The Inverted Regulatory Perimeter in Singapore Buy-Now-Pay-Later

The Inverted Regulatory Perimeter in Singapore Buy-Now-Pay-Later | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article The Inverted Regulatory Perimeter in Singapore Buy-Now-Pay-Later Boon Chuan Lim This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9581198/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 7 You are reading this latest preprint version Abstract Singapore Buy-Now-Pay-Later (BNPL) products operate outside the rate-cap architecture that governs every other consumer credit instrument in the jurisdiction. Licensed moneylenders are capped at 4 percent monthly interest under the Moneylenders Act, and credit card issuers face MAS conduct rules and a 12-times-monthly-income aggregate underwriting limit. BNPL faces neither. I compute the effective monthly rate implied by the late fee schedules of the three accredited Singapore BNPL providers (Atome, Grab PayLater, SeaMoney SPayLater) across seventeen ticket-size scenarios and benchmark these against the Moneylenders Act ceiling. On per-incident absolute fees, BNPL charges fall below the moneylender maximum. On effective rates the picture inverts: BNPL produces monthly rates of 1 to 75 percent on the deferred amount. The same S$15 Atome late fee implies a 75 percent rate on a S$20 missed instalment, 15 percent on S$100, and 4.5 percent on S$333. The statutory rate cap binds uniformly on rate; the BNPL flat fee binds uniformly on dollar amount. I calibrate the analysis to MAS data and to the CFPB 2025 BNPL studies, which show high loan stacking and concentration in subprime borrowers. The structural patterns hold across jurisdictions. BNPL escapes the rate-cap regime through legal classification (deferred payment, not loan) rather than through superior consumer protection. I propose extending Moneylenders Act discipline through a two-component fee structure (capped flat administrative fee plus percentage rate cap) and mandating full Credit Bureau Singapore reporting of BNPL exposure, building on rather than replacing the current industry Code of Conduct. JEL Codes: G18, G28, G51, K23, D18. Buy-Now-Pay-Later consumer credit regulation Moneylenders Act regulatory arbitrage Singapore fintech 1. Introduction Singapore regulates consumer credit through a coherent statutory architecture. Credit card issuers operate under MAS Notice 635, which sets a S $ 30,000 minimum annual income floor for cardholders and caps aggregate unsecured borrowing at 12 times monthly income across all financial institutions. Licensed moneylenders operate under the Moneylenders Act 2008, which caps monthly interest at 4 percent on outstanding principal and on overdue amounts, caps late fees at S $ 60 per month, and limits cumulative charges to the principal of the loan. 1 Both regimes are statutory, both are enforced by a designated regulator, and both impose binding rate or fee ceilings. The architecture is conservative by design: it is designed to prevent extreme effective rates regardless of contract size or borrower segment. Buy-Now-Pay-Later operates outside this architecture. Atome, Grab PayLater, and SeaMoney SPayLater offer instalment payment products to Singapore consumers without income verification, without aggregate exposure caps that bind across other credit instruments, and without statutory regulation. The Singapore FinTech Association issued a Buy-Now-Pay-Later Code of Conduct in October 2022 and accredited five providers in April 2024. The Code caps a customer’s outstanding payments at S $ 2,000 with each individual provider before an additional credit assessment is required. The cap binds per-provider rather than across the BNPL sector, so a user can in principle accumulate S $ 2,000 each at Atome, Grab, and SeaMoney concurrently without any provider triggering the assessment. 2 The Code is industry self-regulation. It does not set a statutory rate cap, does not impose income verification, and does not feed BNPL exposure into the Credit Bureau Singapore that banks consult for card and loan underwriting. Before stating the central claims of the paper, two scope clarifications are useful. First, BNPL is genuinely interest-free for the modal user. A customer who pays each instalment on the due date pays exactly the purchase price, with no interest, fees, or other charges. The product is more attractive than a credit card on this dimension. The paper is concerned with the conditional cost given a missed payment, not with the average user-level cost across the population. Second, the late fees discussed below apply to instalment plans of three to four monthly payments. SPayLater also offers longer tenures of six to twenty-four months on which monthly interest is charged explicitly; these are analytically closer to bank instalment plans and outside the central scope of the paper. The architectural argument focuses on the pay-in-three or pay-in-four products that are the consumer-facing default at Atome, Grab, and SeaMoney. This paper makes three claims. First, that BNPL fee structures generate effective monthly rates that would be statutorily prohibited under the Moneylenders Act if BNPL were classified as a loan. Second, that the rate variability is a direct consequence of flat-fee architecture: a fixed dollar fee implies a high effective rate on small principals and a low effective rate on large principals, an outcome that statutory rate caps were designed to prevent. Third, that the regulatory arbitrage operates on architecture, not absolute fee levels. On a per-incident basis, BNPL late fees are below what a licensed moneylender could lawfully charge. The arbitrage is the choice of cap structure (none vs rate-based vs flat fee), not the choice of fee level. I treat these claims as empirical and verify them against the published fee schedules of the three active accredited Singapore BNPL providers as of April 2026, the statutory ceilings published by MinLaw under the Moneylenders Rules 2015, and the BNPL Code of Conduct administered by the Singapore FinTech Association. 3 The contribution is the synthesis: identifying that a coherent rate-cap regulatory hierarchy exists in Singapore consumer credit, demonstrating that BNPL sits outside it, and quantifying the asymmetry. The paper proceeds as follows. Section 2 reviews the BNPL literature, with emphasis on regulatory arbitrage and effective-cost measurement. Section 3 sets out the three Singapore consumer credit regimes side by side. Section 4 presents the analytical core: a worked example, a cost grid across providers and ticket sizes, and the comparison to the Moneylenders Act ceilings. Section 5 calibrates the analysis to empirical data, drawing on aggregate market statistics, provider-level financial disclosures, MAS-disclosed Singapore user statistics, and the user-level evidence on BNPL borrowers from the CFPB’s 2025 studies. Section 6 considers counter-arguments. Section 7 sketches a policy proposal. Section 8 sets out limitations and scope. Section 9 concludes. 2. Related Literature 2.1 Effective cost of BNPL deHaan et al. ( 2024 ) examine the effects of BNPL adoption on US consumer financial health using banking data for 10.6 million US consumers. They find that new BNPL users experience rapid increases in bank overdraft charges and credit card interest and fees compared with non-users. They use an instrumental variable based on consumers’ pre-BNPL shopping habits to argue for a causal interpretation, and conclude that BNPL plausibly facilitates overborrowing. The paper anchors the borrower-side empirical literature on BNPL and frames the welfare question as one of total cost across all of a user’s credit instruments, not just BNPL fees in isolation. Cornelli et al. ( 2023 ) take the cross-country adoption view in a BIS Quarterly Review article. They document that BNPL adoption is stronger in countries with higher e-commerce penetration, higher inflation, less efficient banking sectors, and looser financial regulation. They explicitly raise the regulatory arbitrage question: whether BNPL grows fastest where the regulatory perimeter is weakest. They do not resolve the question empirically but treat it as the leading hypothesis. Berg et al. ( 2025 ) provide the merchant-side analysis. They develop a structural model in which BNPL functions as a price-discrimination device: merchants accept BNPL because it expands sales among low-creditworthiness customers, and the merchant discount rate (3 to 6 percent) is the primary revenue source for BNPL providers, not late fees on consumers. 11 The Berg et al. result complicates the simple “BNPL extracts surplus from consumers” narrative: most extraction is from merchants, who pass through some portion via prices. But the consumer-cost question remains relevant for the late-fee tail of users who fail to pay on time. The CFPB (2022) market report on BNPL is the most cited regulatory analysis. The report identifies three concerns: accumulated debt (loan stacking across providers), regulatory arbitrage (BNPL avoiding TILA disclosures and Reg Z protections), and data harvesting (BNPL providers monetising user data). The CFPB issued an interpretive rule in May 2024 applying Regulation Z to BNPL providers, which was withdrawn in 2025. 8 The US regulatory record demonstrates the difficulty of bringing BNPL inside the consumer credit perimeter once it is already established as a payment product rather than a loan product. 2.2 Singapore-specific work I find no peer-reviewed empirical paper on Singapore BNPL effective costs or regulatory perimeter as of April 2026. The available material consists of MAS and SFA announcements, Lexology and law-firm commentaries on the Code of Conduct, and Straits Times coverage of provider accreditation. The 2024 trustmark accreditation of Atome, Grab, SeaMoney, LatitudePay (since departed), and Abnk.ai received industry press but no academic treatment. The Experian-administered BNPL credit bureau is operational since 2024 but has not produced public empirical reports on the user distribution. This is the gap the present paper fills. The contribution is not new theory: it is the application of existing effective-cost methodology, adapted from the CFPB and deHaan et al., to the Singapore regulatory environment, where a coherent rate-cap regime for moneylending exists and provides a natural benchmark. 2.3 The moneylender comparison The US literature compares BNPL to credit cards and to payday loans. Both comparisons are imperfect for Singapore. Credit cards are heavily underwritten and serve a different population. Payday lending in the US sense barely exists in Singapore: short-term lending to unbanked consumers is regulated under the Moneylenders Act with hard caps. The relevant comparator in Singapore is the licensed moneylender, which serves a similar borrower segment to BNPL (consumers without ready access to bank credit) but operates under heavier regulation, not lighter. I have not found prior work that draws this comparison directly. The comparison is fruitful because it inverts the usual framing. In the US, BNPL is sometimes positioned as a more responsible alternative to predatory payday lending. In Singapore, BNPL operates under lighter rules than the licensed moneylender regime that itself replaced predatory lending. The regulatory hierarchy runs in the opposite direction of what the BNPL marketing narrative suggests. 3. The Three Singapore Consumer Credit Regimes Singapore operates three statutory or quasi-statutory consumer credit regimes. I summarise each below and then present them in tabular form to make the comparison explicit. 3.1 Credit cards (MAS) Credit card issuance is regulated by MAS under the Banking Act and MAS Notice 635 (Unsecured Credit Rules). 4 The minimum annual income for cardholders is S $ 30,000 for citizens and permanent residents and S $ 40,000 for foreigners. Aggregate unsecured borrowing across all financial institutions is capped at 12 times monthly income, enforced through the Credit Bureau Singapore. Issuers conduct hard credit checks; payment history reports back to CBS. There is no statutory cap on interest rates or fees, but issuers operate under MAS conduct rules and competitive pressure. Effective annual rates on revolving balances typically run 24 to 28 percent. 3.2 Licensed moneylenders (MinLaw) Licensed moneylending is regulated under the Moneylenders Act 2008 with implementing rules from the Moneylenders Rules 2015. 5 The maximum interest rate is 4 percent per month on outstanding principal, calculated on a reducing-balance basis. The maximum late interest is also 4 percent per month, applied only to overdue amounts. The maximum administrative fee is 10 percent of principal, charged once at loan origination. The maximum late fee is S $ 60 per month of late repayment. The Act’s principal-ceiling rule is critical: the cumulative total of all charges over the life of the loan, including interest, late interest, the upfront administrative fee, and all monthly late fees, cannot exceed the principal of the loan. 13 This applies to charges across the whole loan, not to any individual month or instalment. Unsecured borrowing is capped at S $ 3,000 across all licensed moneylenders for borrowers with annual income below S $ 20,000, and at six times monthly income for borrowers above that threshold. Lender data flows into the Moneylenders Credit Bureau. Borrowers may apply for self-exclusion through the MLCB. 10 The regime is enforced by MinLaw’s Registry of Moneylenders. Sanctions for non-compliance include licence revocation and criminal penalties. 3.3 BNPL (industry Code of Conduct) Buy-Now-Pay-Later providers in Singapore operate outside any specific statutory regime. The Buy-Now-Pay-Later Code of Conduct, launched in October 2022 by the Singapore FinTech Association under MAS guidance, is industry self-regulation. 2 The Code’s principal substantive provision caps customer outstanding BNPL exposure at S $ 2,000 across all accredited providers without an additional credit assessment. 12 The Code requires fee disclosure, prohibits compounding of fees, and allows voluntary self-exclusion. It does not set a rate cap, does not require income verification, and does not require BNPL data to flow into the Credit Bureau Singapore. Five providers were accredited under the trustmark in April 2024: Atome, SeaMoney, Grab, LatitudePay, and Abnk.ai. LatitudePay subsequently exited the Singapore market. As of April 2026, the three active providers offering general-purpose BNPL services are Atome, Grab PayLater, and SeaMoney SPayLater. 9 The Grab PayLater Terms of Service (last modified 21 November 2025) include a consumer advisory stating that the service is a buy-now-pay-later product and does not constitute a loan, credit card or charge card facility, and that BNPL services are not regulated by the Monetary Authority of Singapore. 6 3.4 The architecture, side by side Table 1 summarises the three regimes across eight regulatory dimensions. The asymmetry is most visible on the rate cap, total charge ceiling, statutory regulator, and credit reporting rows. Table 1 Singapore consumer credit regimes compared Dimension Credit cards (MAS) Licensed moneylenders (MinLaw) BNPL (SFA Code) Underwriting threshold Min. annual income S $ 30k (citizen/PR), bureau check, aggregate cap 12x monthly income Income-based borrowing tiers (S $ 3,000 cap below S $ 20k income, 6x monthly above); MLCB check; self-exclusion available No income verification; sign-up via NRIC and linked card Interest rate cap No statutory cap (MAS conduct rules) 4% per month on principal; 4% per month on overdue No cap; nominally interest-free if paid on time Late fee cap Issuer-specific; subject to MAS conduct rules S $ 60 per month of late repayment (statutory) No statutory cap; provider-specific (e.g. Atome S $ 60/transaction) Total charge ceiling No statutory ceiling on cumulative charges Total charges ≤ principal of loan No cumulative charge ceiling Aggregate exposure cap 12x monthly income across all unsecured (MAS rule) Tiered by income (S $ 3,000 / 6x monthly income) S $ 2,000 per provider before additional assessment (Code, not statute); no inter-provider aggregation rule Credit reporting Full bureau (CBS) reporting MLCB reporting (since 2016) Intra-BNPL bureau only (Experian, since 2024); not visible to bank lenders Statutory regulator MAS MinLaw / Registry of Moneylenders None (industry-led, MAS guidance only) Sanctions for breach Licence action by MAS Licence revocation, criminal penalties Loss of trustmark accreditation Notes. Compiled from MAS Notice 635, Moneylenders Act 2008 with Moneylenders Rules 2015, and the BNPL Code of Conduct (SFA, October 2022). Sources cited in footnotes. The pattern across Table 1 is a coherent regulatory hierarchy with one clear exception. Credit cards have heaviest income-based underwriting and lightest rate caps, on the theory that underwriting controls the borrower population. Moneylenders have lighter income-based underwriting (a tiered borrowing cap rather than a hard income floor) and heaviest rate caps, on the theory that rate caps protect a population that the underwriting cannot effectively screen. Both regimes use income to determine how much a borrower can access, and both bind across providers within the regime through their respective bureaus. BNPL uses income for nothing. The Code of Conduct’s S $ 2,000 outstanding cap applies per provider rather than across all BNPL providers, so cross-provider stacking is not constrained by the cap itself. The cap is also invariant to the borrower’s income, employment status, or existing debt elsewhere. The Code is enforced by industry self-regulation rather than statute. 4. The Effective Cost Comparison 4.1 Method I compute the effective monthly rate on a missed BNPL instalment by dividing the late fee by the deferred amount and treating the result as a monthly rate. The deferred amount is the value of the missed instalment, which is the credit actually extended past the due date. This follows the convention used by the CFPB (2022) and is the economically correct denominator for an APR calculation: it answers the question “what rate is being charged for the use of these funds for this period?” 14 Two alternative denominators warrant comment because they appear in industry materials and would yield different effective rates. The first is the full transaction value at point of purchase. This denominator amortises the late fee over the entire ticket regardless of how much of it has been repaid, and so understates the cost of credit on the residual amount actually deferred. For an Atome 3-instalment plan on a S $ 300 transaction, the customer has already paid S $ 100 at checkout when the first instalment is due. Charging the S $ 15 missed-instalment fee against the original S $ 300 produces a 5 percent monthly rate; against the S $ 100 actually overdue, 15 percent. The latter measures what the customer is paying for the use of credit they still have, which is the standard APR question. The second alternative is the residual outstanding balance immediately after the missed instalment, which for the same example is S $ 200 (the missed S $ 100 plus the not-yet-due S $ 100). Charging S $ 15 against S $ 200 produces 7.5 percent. This denominator captures total balance at risk but conflates funds the borrower owes immediately with funds not yet due. The CFPB-convention denominator (the deferred instalment itself) is the economically cleanest for the question of effective rate on a single missed payment. Table 2 below uses this convention. Section 8 acknowledges the sensitivity, and an alternative-denominator robustness column is reported in the worked example. Three rate measures follow naturally. The simple APR is the monthly rate multiplied by 12. The effective annual rate (EAR) is the monthly rate compounded over 12 months. Both annualisations are reported, with the simple APR used for headline comparisons and the EAR for substantive interpretation. The Moneylenders Act benchmark for one month of overdue is the sum of two components: (a) 4 percent monthly interest on the overdue amount, and (b) up to S $ 60 in late fees per month. The per-month maximum therefore varies with the missed amount and equals (missed amount × 0.04) + S $ 60. For a S $ 100 missed instalment this is S $ 64.00; for S $ 33.33 (a missed instalment on a S $ 100 BNPL transaction split three ways) this is S $ 61.33; for S $ 500 it is S $ 80.00. Cumulative charges across the loan life are capped separately at 100 percent of principal, but this cumulative cap does not bind in any single-month scenario considered here. I compute both an absolute fee comparison (BNPL fee versus MA per-month maximum) and a rate comparison (BNPL effective monthly rate versus MA 4 percent rate cap). 4.2 A worked example Consider an Atome standard 3-instalment plan on a S $ 300 transaction. The borrower pays S $ 100 at checkout, S $ 100 in month 1, and S $ 100 in month 2. Suppose the borrower misses the month-1 instalment and pays it 30 days late. BNPL effective cost. Atome charges a S $ 15 admin fee for orders below S $ 1,000. Effective monthly rate (deferred-amount denominator) = S$15 / S$100 = 15.0% Simple APR = 15.0% × 12 = 180% Effective annual rate = (1 + 0.15)¹² − 1 = 435% For comparison, the alternative denominators discussed in Section 4.1 yield: Against full transaction value (S $ 300): S $ 15 / S $ 300 = 5.0% monthly (60% simple APR, 80% EAR) Against residual outstanding balance (S $ 200, includes not-yet-due S $ 100): S $ 15 / S $ 200 = 7.5% monthly (90% simple APR, 138% EAR) All three denominators produce monthly rates above the 4 percent Moneylenders Act ceiling on this transaction. The deferred-amount denominator is used as the headline measure throughout the paper because it is the standard APR convention for measuring the effective rate on credit actually extended past its due date. The architectural finding does not depend on the choice. Moneylenders Act benchmark. On a S $ 100 overdue for one month: Maximum interest = 4% × S $ 100 = S$4 Maximum late fee = S$60 (statutory) MA total per-month maximum = S$64 Two comparisons follow. On absolute fees, the Atome S $ 15 charge is below the MA per-month maximum of S $ 64. Per-incident, BNPL is cheaper than what a moneylender could legally charge on the same overdue amount . I emphasise this because an earlier framing of the question, popular in policy commentary, has BNPL fees “exceeding” moneylender ceilings on a per-incident basis. That framing does not survive the arithmetic. On effective rates, the picture inverts. The Atome 15.0 percent monthly rate is 3.75 times the 4 percent statutory rate cap that applies to every other consumer credit instrument in Singapore. The Moneylenders Act is structured to prevent precisely this outcome : by capping the rate at 4 percent rather than the absolute fee at S $ 60, the Act limits effective cost-of-credit regardless of the size of the contract. BNPL’s flat-fee structure produces effective rates that scale inversely with the missed amount, generating high rates on small principals. 4.3 The cost grid Table 2 extends the worked example to seventeen scenarios across the three providers and a range of ticket sizes. Each row reports the missed-instalment amount, the late fee charged by the provider, the implied effective monthly rate on the deferred amount, the simple APR and EAR, and two comparison columns: the BNPL fee as a percentage of the MA per-month maximum, and the BNPL effective rate as a multiple of the 4 percent MA statutory rate cap. Table 2 Effective cost of BNPL late fees vs Moneylenders Act ceilings Provider Plan Ticket (S $ ) Missed (S $ ) Late fee (S $ ) Monthly rate Simple APR EAR BNPL/MA max Rate / 4% cap Atome 3 instalments 60 20.00 15 75.0% 900% 82,401% 23% 18.8x Atome 3 instalments 100 33.33 15 45.0% 540% 8,538% 23% 11.2x Atome 3 instalments 300 100.00 15 15.0% 180% 435% 23% 3.8x Atome 3 instalments 600 200.00 15 7.5% 90% 138% 23% 1.9x Atome 3 instalments 999 333.00 15 4.5% 54% 70% 23% 1.1x Atome 3 instalments 1,500 500.00 30 6.0% 72% 101% 38% 1.5x Atome 3 instalments 2,000 666.67 30 4.5% 54% 70% 36% 1.1x Grab PayLater Pay next month 100 100.00 15 15.0% 180% 435% 23% 3.8x Grab PayLater Pay next month 300 300.00 15 5.0% 60% 80% 21% 1.2x Grab PayLater Pay next month 1,000 1,000.00 15 1.5% 18% 20% 15% 0.4x Grab PayLater 4 instalments 400 100.00 10 10.0% 120% 214% 16% 2.5x Grab PayLater 4 instalments 1,000 250.00 10 4.0% 48% 60% 14% 1.0x Grab PayLater 4 instalments 2,000 500.00 10 2.0% 24% 27% 12% 0.5x SeaMoney SPayLater 3 monthly bills 150 50.00 5 10.0% 120% 214% 8% 2.5x SeaMoney SPayLater 3 monthly bills 300 100.00 5 5.0% 60% 80% 8% 1.3x SeaMoney SPayLater 3 monthly bills 600 200.00 5 2.5% 30% 34% 7% 0.6x SeaMoney SPayLater 3 monthly bills 1,500 500.00 5 1.0% 12% 13% 6% 0.2x Notes. Effective monthly rate = late fee / missed instalment. Simple APR = monthly rate × 12. EAR = (1 + monthly rate)¹² − 1. BNPL/MA max compares the BNPL late fee to the moneylender per-month maximum, computed as (missed instalment × 0.04) + S$60 and varying row-by-row with the missed amount. Rate / 4 percent cap compares the BNPL effective monthly rate to the statutory rate ceiling; the displayed multiple is rounded to one decimal place using round-half-up. Ticket sizes are illustrative; provider fee schedules verified from Atome, Grab, and Shopee/SeaMoney published terms as of 30 April 2026. 4.4 Three patterns Three patterns emerge from Table 2 . First, BNPL late fees are below the Moneylenders Act per-month maximum in every row. The BNPL/MA max column ranges from 6 percent (SeaMoney on a S $ 1,500 ticket) to 38 percent (Atome on a S $ 1,500 ticket). On no row does a BNPL fee exceed what a licensed moneylender could lawfully charge for the same overdue amount in the same month. The simple framing “BNPL costs more than moneylending” is therefore false at the per-incident level. This matters because that framing has appeared in policy discussion and would not survive scrutiny. Second, BNPL effective rates exceed the 4 percent statutory rate cap on the small-ticket end of the distribution. Atome on a S $ 60 ticket implies a 75.0 percent monthly rate, which is 18.8 times the moneylender rate cap. Atome on a S $ 300 ticket implies a 15.0 percent monthly rate, 3.8 times the cap. The pattern is consistent across providers: SeaMoney on a S $ 150 ticket is 2.5 times the cap and on a S $ 300 ticket is 1.25 times the cap; Grab PayLater Pay-Next-Month on a S $ 100 ticket is 3.8 times the cap. The effective rate falls below the 4 percent ceiling only at ticket sizes above approximately S $ 1,000 for Atome and Grab, and above S $ 300 for SeaMoney. Third, the rate variability is large. Across the seventeen scenarios, the implied effective monthly rate ranges from 1.0 percent (SeaMoney, S $ 1,500 ticket) to 75.0 percent (Atome, S $ 60 ticket). This is the architectural finding. Statutory rate caps are designed to deliver uniform effective rates regardless of contract size. Flat-fee structures by definition do not. The same dollar fee implies very different effective rates depending on the size of the missed amount, with the highest rates concentrated on the smallest tickets. 4.5 Why ticket-size sensitivity matters The BNPL user distribution is concentrated in small-ticket transactions. Atome’s default plan splits payments into three instalments, encouraging use on purchases in the S $ 100–S $ 500 range. Grab PayLater’s in-app payment limit and SeaMoney SPayLater’s integration with Shopee likewise drive small-ticket usage. The user demographic is younger, lower-income, and more financially fragile than card holders, consistent with international evidence on US BNPL borrowers (deHaan et al., 2024 ) and US late-payment incidence (LendingTree, 2026). The combination is consequential. The product is most heavily used in the ticket range where the flat-fee structure produces the highest effective rates, and the user population is the segment for which statutory rate caps would normally provide the most protection. A regulatory regime that imposes a 4 percent monthly cap on licensed moneylenders, who serve the same borrower segment, but does not impose any rate constraint on BNPL is asymmetric in a way that is hard to justify on first principles. 5. Empirical Calibration Section 4 established the comparative-rate finding analytically. This section calibrates the analysis to the available empirical data. I draw on four sources: aggregate market statistics published by the Monetary Authority of Singapore and from secondary industry research, provider-level financial disclosures from listed and quasi-listed parents, MAS-disclosed Singapore user-level statistics from a February 2026 parliamentary reply, and the user-level evidence base from the Consumer Financial Protection Bureau’s 2025 studies of US BNPL borrowers. The aggregate data establishes the scale of the BNPL market relative to the comparator credit regimes. The provider data establishes that BNPL providers have reached scale and profitability, and that consumer late fees are not the primary revenue source. The MAS Singapore data and CFPB US data jointly establish the structural patterns of BNPL usage at the user level: small typical outstanding, dominant loan stacking, concentration in subprime and deep-subprime borrowers, and elevated balances on other unsecured credit. These patterns hold across jurisdictions and bear directly on the architectural argument. 5.1 Aggregate market scale Singapore BNPL gross merchandise value reached an estimated US $ 1.19 billion (approximately S $ 1.6 billion) in 2024 and is projected at US $ 1.32 billion in 2025, growing at 11.1 percent annually. 17 For comparison, total credit card billings in Singapore averaged S $ 23 to S $ 24 billion per quarter in 2024, equivalent to roughly S $ 95 billion annualised. Card rollover balances (the share carried at interest) reached S $ 8.3 billion at end-2024. 18 The relative scale matters for the policy framing. MAS confirmed in February 2026 that BNPL transactions remain less than 2 percent of the value of total credit and debit card payments. The aggregate market is small enough that BNPL does not threaten financial stability. Whether it threatens consumer protection at the user level is a separate question, addressed below. Table 3 sets out the aggregate scale comparison. Table 3 Aggregate scale of Singapore consumer credit instruments Instrument Annual flow (S $ bn) Outstanding (S $ bn) Reference period Source Credit cards (billings) ~ 95 ~ 8.3 (rollover) 2024 full year MAS Stat. Bulletin I.17A BNPL (gross merchandise) ~ 1.6 n.d. 2024 full year Industry research; MAS confirms < 2% of card billings Licensed moneylenders n.d. ~ 2.0 (estimated) 2024 (most recent published) MLCB; MinLaw Notes. BNPL gross merchandise value is the total value of transactions financed via BNPL plans, not the outstanding amount on those plans at any point in time. Card billings are the equivalent gross flow for credit cards. Card outstanding is the rollover balance (the share carried beyond the due date and subject to interest). Public quarterly disclosures of BNPL outstanding are not published by MAS or SFA. Licensed moneylender outstanding is approximate, based on MLCB administrative data; precise figures are not in regular public release. The pattern across Table 3 is that BNPL is the smallest of the three regimes by transaction value, but the only one without rate-cap regulation. The asymmetry is therefore not about the size of the market: a comprehensive consumer credit architecture would be expected to apply rate-cap discipline regardless of segment size, and Singapore in fact does this for licensed moneylending despite that segment being smaller still in outstanding terms than the credit card market. 5.2 Provider-level revenue mix Atome’s parent, Atome Financial (part of Singapore-headquartered Advance Intelligence Group), reported regional gross merchandise volume of US $ 2 billion in 2024, up 50 percent year-on-year, with annualised revenue surpassing US $ 500 million by Q2 2025. Atome Financial reported full-year operating income of US $ 236 million in 2024 and turned profitable in Q1 2024. 19 The disclosures emphasise growth in the Atome PayLater Anywhere Card (1.5 million cards issued in the Philippines by mid-2025), expansion into insurance, savings, and lending, and partnership-based merchant integration. Late fees on consumer BNPL are not separately disclosed as a revenue line. Singapore-specific figures are not separately reported. SeaMoney, the digital financial services arm of Sea Limited (NYSE: SE), is reported alongside parent group financials in SEC filings. SeaMoney consumer and SME loans principal outstanding reached US $ 5.1 billion at end-2024, up 63.9 percent year-on-year. Non-performing loans past due more than 90 days were 1.2 percent of principal outstanding. Sea Limited’s digital financial services GAAP revenue was US $ 2.4 billion in 2024. 20 SPayLater is the consumer-facing BNPL product within this segment. As with Atome, late fees are not separately disclosed. SeaMoney operates in eight markets across Southeast Asia, Taiwan, and Brazil; Singapore is one of these but not the largest by revenue. Two findings follow from the provider data. First, BNPL providers in Southeast Asia have reached scale, profitability, and institutional capital backing. The argument that BNPL is a fragile fintech sector requiring regulatory forbearance no longer applies. Second, late fees do not appear as a separately disclosed revenue line in either provider’s public filings, consistent with the merchant discount rate being the primary revenue source. This supports the Section 6.4 counter-argument that late fees are not the main profit centre. It also supports the policy implication: replacing flat-fee late charges with rate-based late charges would not materially affect provider economics, because late fees are a small share of revenue. 5.3 User-level evidence Section 5.1 established the aggregate market scale and Section 5.2 established the provider economics. Neither reaches the level of individual borrower behaviour, which is where the architectural argument has to land if the policy proposal is to be defensible. Singapore-specific user-level statistics are limited to what MAS has disclosed in parliamentary replies. The most rigorous user-level evidence available globally is from the United States Consumer Financial Protection Bureau, which has published two studies in 2025 drawing on aggregate data from six large pay-in-four lenders covering approximately 40 percent of the US point-of-sale financing market. I summarise the relevant US findings below and compare them to what is known about Singapore. The two jurisdictions differ in regulatory framework, provider mix, and consumer protections, so the US data should be read as evidence about the structure of BNPL usage rather than as a direct claim about Singapore. 23 That said, the structural patterns are unlikely to be jurisdiction-specific. Table 4 BNPL market and borrower metrics from CFPB data (six large US lenders) Metric 2021 2022 2023 Source Loan originations (millions) 196.6 273.8 335.8 CFPB Dec 2025, Table 1 Dollar originations (US $ billions, CPI-adj.) 29.5 35.8 45.2 CFPB Dec 2025, Table 1 Average loan size (US $ , CPI-adj.) 150 131 135 CFPB Dec 2025, Table 1 Unique users (millions) n.r. 48.0 53.6 CFPB Dec 2025, Table 2 Average loans per user per year n.r. 5.7 6.3 CFPB Dec 2025, Table 2 Average annual amount per user (US $ ) n.r. 745 848 CFPB Dec 2025, Table 2 Loan-level late fee assessment rate 7.5% 5.2% 4.1% CFPB Dec 2025, Table 3 Late fees collected as % of GMV 0.32% 0.24% 0.18% CFPB Dec 2025, Table 3 Average late fee assessed (US $ , CPI-adj.) 11.81 10.72 9.99 CFPB Dec 2025, Table 3 Loan charge-off rate 2.44% 2.63% 1.83% CFPB Dec 2025, Table 4 Notes. n.r. = not reported in source. CFPB sample comprises six large pay-in-four lenders: Affirm, Cash App Afterpay, Klarna, PayPal, Sezzle, and Zip. Late fee metrics in Table 3 cover the four firms that charged late fees during the survey period. The sample represents approximately 40 percent of the US point-of-sale financing market in 2023. Dollar figures are inflation-adjusted to 2024 dollars where indicated. Several patterns from Table 4 inform the policy framing. First, late-fee incidence has fallen, not risen, over the survey period (4.1 percent of loans in 2023, down from 7.5 percent in 2021). Late-fee revenue as a share of gross merchandise volume halved from 0.32 percent to 0.18 percent. The CFPB attributes this decline partly to industry self-correction in response to regulatory scrutiny and partly to tightening underwriting. The Singapore Code of Conduct, in force since October 2022, would be expected to produce similar dampening effects. Second, the size of the typical late fee charged is small. The CFPB reports an average late fee assessed of US $ 9.99 in 2023 and an average late fee collected of US $ 5.87 (the gap reflects discretionary waivers and uncollected fees). The Singapore equivalents — Atome S $ 15, Grab S $ 10-S $ 15, SeaMoney S $ 5 — are in the same range. The bounded-harm argument has support in absolute fee terms. Where the data is more concerning is the user-level distribution of usage. Table 5 BNPL borrower characteristics from matched CFPB credit-records data Finding Value Reference Share of US consumers (with credit record) who used BNPL in 2022 21% CFPB Jan 2025, Table 2 Share of borrowers characterised as heavy users (> 1 loan/month avg.) 20% CFPB Jan 2025, Exec. Summary Share of borrowers with simultaneous loans (any firm) in 2022 63% CFPB Jan 2025, Section V Share of borrowers with simultaneous loans across multiple firms 33% CFPB Jan 2025, Section V Share of originations to deep subprime borrowers (FICO 300–579) 45% CFPB Jan 2025, Table 3 Share of originations to subprime borrowers (FICO 580–619) 16% CFPB Jan 2025, Table 3 Share of originations to no-FICO-score borrowers 4% CFPB Jan 2025, Table 3 BNPL default rate among users (2019–2022 average) 2% CFPB Jan 2025, Section V Credit-card default rate, same users, same period 10% CFPB Jan 2025, Section V Additional credit card debt of BNPL users vs non-users (matched) +US $ 871 CFPB Jan 2025, Section VI Additional personal loan debt of BNPL users vs non-users +US $ 453 CFPB Jan 2025, Section VI BNPL share of total unsecured debt for borrowers aged 18–24 28% CFPB Jan 2025, Section VI BNPL share of total unsecured debt across all age groups 17% CFPB Jan 2025, Section VI Avg. credit card utilisation, BNPL users (2020–2023) 60–66% CFPB Jan 2025, Section VI Avg. credit card utilisation, never-BNPL users 34% CFPB Jan 2025, Section VI Notes. All figures from the CFPB January 2025 study, which uses a matched sample of de-identified BNPL applications and originations from six large pay-in-four lenders linked to credit records from one of the three nationwide consumer reporting companies. The matched-sample design corrects for the fact that BNPL exposure is not generally reported to credit bureaus, allowing the CFPB to measure stacking and to compare BNPL users to credit-record matched non-users. Dollar comparisons in the lower panel are differences in mean balance, controlling for age and FICO score. The default rate reported in row 8 is the share of originated loans charged off as bad debt (typically loans more than 120 days past due in the CFPB definition), not the share of loans on which any late fee was assessed. Late-fee assessment rates are higher (4.1 percent of loans in 2023, per Table 4 above), reflecting the much larger share of loans that incur a single missed instalment without progressing to charge-off. Three substantive patterns emerge from Table 5 , all of which bear on the architectural argument. Loan stacking is the dominant usage pattern, not the exception. Sixty-three percent of US BNPL borrowers held simultaneous loans within 2022. One-third of borrowers held simultaneous loans across multiple firms. The CFPB explicitly notes that these stacked obligations are largely invisible to non-BNPL lenders because BNPL exposure is not generally reported to the three nationwide consumer reporting companies. The Singapore framework addresses this through the Experian-administered BNPL bureau, which shares data among accredited BNPL providers from April 2024 onward. 7 It does not, however, share BNPL data with the Credit Bureau Singapore that banks consult for card and loan underwriting. The intra-BNPL stacking concern is partially addressed; the cross-product invisibility concern is not. BNPL users are concentrated in lower credit-score bands. In the CFPB matched sample, 45 percent of originations went to deep-subprime borrowers (FICO 300–579) and a further 16 percent to subprime (580–619). Borrowers with no FICO score account for 4 percent of originations. These shares correspond to the population that, in Singapore, would not generally qualify for credit cards under MAS Notice 635 (S $ 30,000 income floor, aggregate cap of 12 times monthly income). Singapore-specific borrower distributions are not publicly available, but the structural argument holds: BNPL serves the population the credit-card regime is designed to exclude. The Moneylenders Act regime, by contrast, is designed to serve precisely this population — with the rate-cap protections that BNPL does not provide. BNPL users hold materially higher balances on other unsecured credit, even controlling for age and credit score. The CFPB matched-sample analysis finds that BNPL users hold US $ 871 more in credit card debt and US $ 453 more in personal loan debt than non-BNPL users with the same age and FICO score. Average credit card utilisation among BNPL users is 60–66 percent, against 34 percent among never-BNPL users. The interpretation in deHaan et al. ( 2024 ) and replicated by the CFPB is that BNPL adoption is associated with increasing financial stress on other instruments rather than substituting away from them. Among the youngest age group (18–24), BNPL accounts for 28 percent of total unsecured debt during borrowing months. The under-21 share of Singapore BNPL users (1.5 percent per MAS February 2026) is lower than this US distribution suggests, but the mechanism is the same. 5.4 What this implies for Singapore The Singapore-specific evidence is consistent with the US patterns where it can be observed. MAS confirmed in February 2026 that the median Singapore BNPL user with missed payments holds aggregate outstanding below S $ 400, with the vast majority below S $ 1,000. 15 This figure is consistent with the modest typical outstanding the CFPB documents for US BNPL users when measured per-firm. It is also consistent with the small absolute fee level (S $ 5-S $ 15 per missed instalment) that the Singapore providers charge. What the Singapore aggregate disclosures do not yet reach is the user-level distribution of stacking, the share of users with simultaneous plans across providers, or the cumulative annual fee burden on heavy users. The CFPB borrower-level patterns suggest these distributions are highly concentrated: 20 percent of US borrowers are heavy users (more than one loan per month), 33 percent stack across firms, and the borrower population skews to subprime. If similar patterns hold in Singapore, the architectural finding from Section 4 has its strongest welfare implications precisely for the heavy-user subset, not for the median user. Two pieces of Singapore-specific evidence would resolve the open empirical questions. The first is publication of aggregate user-level statistics from the Experian-administered BNPL bureau, which holds the necessary data. The MLCB and CBS publish equivalent statistics for licensed moneylenders and credit cards respectively; the BNPL bureau does not. The second is publication of per-provider Singapore-specific revenue breakdowns by source (merchant discount fees versus consumer late fees), which currently are reported only at the regional or group level. Both releases would convert the present analysis from a structural argument supplemented by international evidence into a directly empirical Singapore study. 6. Counter-Arguments Three counter-arguments are common in policy discussion of BNPL regulation. I address each in turn. 6.1 “BNPL is interest-free” The accuracy of this claim is conditional on no late payment. The published Atome, Grab, and SeaMoney terms charge no interest if the borrower pays each instalment on the due date. The relevant question is whether this conditional claim represents the realised cost across the user distribution. International evidence indicates that a non-trivial fraction of users incur late fees. The CFPB (2022) reports late-fee revenue at around 13 percent of BNPL provider revenues. LendingTree (2026) reports that 41 percent of US BNPL users paid late at least once in the previous year, up from 34 percent the previous year, and that 25 percent used BNPL to buy groceries, up from 14 percent the year prior. I do not have comparable Singapore figures, but the structural drivers (multiple due dates per month across stacked plans, automatic deduction failures from insufficient funds) operate identically. The “interest-free” framing describes the modal contract, not the realised cost across users. 6.2 “Tickets are small, harm is bounded” MAS confirmed in February 2026 that BNPL accounts for less than 2 percent of the value of total credit and debit card payments in Singapore, and that the median user with missed payments has outstanding amounts below S $ 400 across all providers, with the vast majority below S $ 1,000. 15 The aggregate market is therefore small, and the bounded-harm argument has some force at the systemic level. The Singapore BNPL market does not pose financial stability risk. The harm calculation at the user level is different. If the median user with a missed payment has an aggregate outstanding balance below S $ 400, a single S $ 15 BNPL late fee represents close to 4 percent of that user’s total BNPL exposure in a single month. For a user whose cash flow is constrained enough to miss a payment, a 4 percent hit to their outstanding balance is material. Repeat occurrences within the year compound this. The economic logic is similar to a user who carries a S $ 400 credit card balance and incurs late fees and interest of comparable percentage at an effective rate that statutory rate caps would otherwise prevent. The harm is also not bounded relative to the user’s cash flow more broadly. A user living paycheck-to-paycheck who pays S $ 15 on a S $ 100 missed instalment is paying 15 percent of the missed amount as a fee. For a user with sufficient cash flow to absorb the fee, the absolute amount is small. For a user without that cushion, the fee can trigger further missed payments, including on rent, utilities, or other obligations. The bounded-harm argument requires assumptions about user cash flow that are stronger than the BNPL marketing material acknowledges. The aggregate-share statistic is the wrong instrument for measuring this: a small share of total payments can still translate into significant per-user cost concentration in a small subset of financially fragile users. 6.3 “Industry self-regulation works” The April 2024 trustmark accreditation demonstrated that the SFA Code can be enforced as an industry standard, with PwC conducting independent audits of the four accredited firms. The Code does several things well: it caps total BNPL outstanding at S $ 2,000, requires fee disclosure, prohibits compounding, and requires hardship assistance. These are real provisions, and the Code’s implementation should be credited. What the Code does not do is impose a rate cap or a structurally analogous discipline. The S $ 2,000 outstanding cap is a quantity constraint on the loan size. It does not constrain the price (the implied rate) on overdue amounts. The Moneylenders Act does both: a quantity cap on borrowing relative to income, and a price cap on interest and fees. The Singapore regulatory tradition for consumer credit serving comparable populations is to cap both quantity and price. The Code of Conduct adopts only the first half. The argument that industry self-regulation is sufficient is therefore conditional on the view that price discipline is not necessary, which is precisely the point at issue. 6.4 “Late fees are not the main revenue source” BNPL providers earn most of their revenue from merchant discount fees, not from consumer late fees. Industry estimates put the merchant discount rate at 3 to 8 percent of the purchase price, well above the 1.5 to 3 percent typical of credit card networks (CFPB, 2022; Berg et al., 2025 ). Berg et al. ( 2025 ) document this as the central feature of the BNPL business model: merchants pay for expanded sales among low-creditworthiness customers. I accept this point but it does not displace the late-fee analysis. The fact that merchant fees dominate provider revenue means that late fees are not designed to be a profit centre, but they remain a real cost to the subset of users who incur them. The relevant question for consumer protection is not what fraction of provider revenue late fees represent in aggregate, but what cost they impose on the affected users at the margin. The architectural argument in this paper concerns the latter: even if late fees are a small share of provider revenue, the rate they imply on small-ticket overdue amounts is structurally disconnected from the rate-cap framework that governs all other Singapore consumer credit. Acknowledging the merchant-rebate revenue model strengthens, rather than weakens, the case that BNPL late fees should not be designed primarily as profit-maximising charges. They could be calibrated as cost-recovery charges on a percentage basis without disrupting provider economics. 7. A Rate-Cap Extension The natural policy implication is to extend Moneylenders Act rate-cap discipline to BNPL fee structures. I sketch the principal elements below, noting that the design space is large and the proposal is illustrative rather than fully specified. 7.1 Re-architect late fees as percentage caps with a flat-fee floor The natural-sounding proposal is to replace flat-dollar late fees with the Moneylenders Act 4 percent rate cap. Mechanically applied, this would impose late fees of S $ 0.80 on a S $ 20 missed instalment and S $ 4.00 on a S $ 100 missed instalment. At those levels, the deterrence and cost-recovery functions of late fees collapse: a missed payment becomes effectively costless, and the BNPL business model that relies on partial fee revenue from delinquent users would not survive the transition. The naive proposal is unworkable. A workable proposal needs more structure. The right design is two-component: a percentage rate cap on the rolling cost of credit, plus a capped flat administrative fee that reflects the genuine collection cost a provider incurs on a missed payment. The Moneylenders Act regime itself is built this way. The 4 percent monthly rate caps interest on the overdue principal, and the S $ 60 per-month late fee permits a flat administrative charge above the rate cap, recognising that overdue collection has fixed costs. A BNPL-adapted version would set the rate cap at a level calibrated to the per-incident absolute fees Singapore providers currently charge (S $ 5-S $ 15) but would impose the percentage discipline on small tickets that currently produce 75 percent monthly rates. A specific worked specification illustrates the principle. Suppose the rule is: a flat administrative fee of up to S $ 15 per missed instalment, plus an additional rate-based component capped at 4 percent of the overdue amount per month, with the total of both components not exceeding 30 percent of the missed instalment in any single month. On a S $ 100 missed instalment, the cap would be the lesser of (S $ 15 + S $ 4) = S $ 19 or (30% × S $ 100) = S $ 30; the binding constraint is S $ 19, which is the Atome fee plus a small interest charge. On a S $ 20 missed instalment, the cap would be the lesser of (S $ 15 + S $ 0.80) = S $ 15.80 or (30% × S $ 20) = S $ 6; the binding constraint is the percentage limit at S $ 6. Provider revenue is preserved on standard tickets where the flat fee dominates. The architectural pathology of S $ 15 fees on S $ 20 tickets is eliminated. The 30 percent monthly rate is high relative to the Moneylenders Act 4 percent cap, but acknowledges that BNPL late fees combine an interest component and a fixed administrative cost. Finer calibration is possible; the principle is that any cap with a percentage component closes the small-ticket arbitrage even if it sets the overall ceiling well above the moneylender benchmark. A simpler alternative is a percentage-only cap with a more generous rate, say 10 percent monthly on the overdue amount, set at a level designed to preserve provider economics on the modal ticket while eliminating the small-ticket distortion. The key design principle in either variant is that the cap should bind on the rate, not on the absolute fee, so that effective costs do not vary by an order of magnitude across ticket sizes. 7.2 Apply the principal-ceiling rule The Moneylenders Act’s rule that cumulative charges cannot exceed principal is a robust backstop. 13 For BNPL, the analogous rule would be that cumulative late fees and interest on a transaction cannot exceed the transaction value. Atome’s current S $ 60 per-transaction admin fee cap operates similarly, but only applies on a single transaction. The principal-ceiling rule is more general: it applies to the loan as a unit, regardless of how the lender structures fees. Extending this rule to BNPL would prevent the corner cases where stacked late fees exceed the transaction value. 7.3 Mandate full credit bureau reporting BNPL exposure is currently reported only within the Experian-administered BNPL bureau, not into the Credit Bureau Singapore that banks consult for card and loan underwriting. MAS has confirmed that financial institutions are not required to include outstanding BNPL amounts when computing the Total Debt Servicing Ratio. 16 A borrower can therefore hold simultaneous BNPL plans up to S $ 2,000 across providers and have those exposures invisible to a bank making credit decisions on a separate card or loan. The asymmetry is a quiet form of off-balance-sheet leverage at the household level. Mandating BNPL reporting into CBS would close this gap. It would also bring BNPL into alignment with the moneylender regime, which feeds the MLCB, and with the credit card regime, which feeds CBS. The argument for keeping BNPL reporting siloed rests on the legal classification of BNPL as a deferred payment arrangement rather than a loan. This classification is formalistic: it follows the contractual structure rather than the economic substance. Deferred payment for consumption financed by a third party that bears the credit risk is functionally a loan, regardless of how the contract is named. The Singapore regulatory tradition for credit broadly tracks economic substance: hire-purchase agreements are regulated under their own statute, even though they share the deferred-payment label that BNPL also uses. The case for treating BNPL as a payment product for regulatory purposes is therefore weak relative to its credit characteristics. Bringing BNPL into CBS reporting is the natural first step in aligning the regulatory perimeter with economic substance. 7.4 Income verification at higher exposure thresholds The S $ 2,000 outstanding cap currently applies per provider, not aggregated across providers. A user can accumulate S $ 2,000 with Atome, Grab, and SeaMoney concurrently without any provider triggering the additional credit assessment that the Code requires above the cap. The Experian-administered BNPL bureau in principle enables cross-provider visibility, but the Code does not currently mandate that the assessment threshold bind on aggregate exposure. A statutory regime would either harmonise the cap on aggregate exposure (the analogue of the MAS Notice 635 12x-monthly-income aggregate ceiling) or specify a structured higher cap with income verification (the analogue of the moneylender tiered cap). The principle that quantity caps should bind across the regime in proportion to income is well-established in Singapore consumer credit regulation. There is no obvious reason BNPL should be exempt. 8. Limitations and Scope Five limitations of the present analysis are worth stating explicitly. First, the cost grid models a single-incident scenario: one missed instalment for one month. Cumulative late fees across multiple missed instalments, multiple transactions, or multiple providers are not modelled. Atome’s S $ 60 per-transaction admin fee cap implies that within a single transaction, cumulative fees are bounded; across transactions and across providers there is no equivalent ceiling. The Moneylenders Act’s principal-ceiling rule applies to the loan as a unit and provides a more comprehensive backstop. A cumulative-fee analysis would strengthen the architectural argument but requires user-level data that is not publicly available. Second, the effective rates reported are sensitive to the choice of denominator. I use the missed instalment (the deferred amount), following the CFPB convention. An alternative is to use the full transaction value, which produces lower effective rates because the fee is amortised over the entire purchase rather than the deferred fraction. The transaction-value framing is the provider-friendly alternative; it understates the cost of credit because most of the transaction value has already been paid at the point of late incidence. The Excel companion model reports both denominators. Third, the analysis treats the Moneylenders Act per-month maximum as the relevant comparison benchmark. An alternative is to compare BNPL fees to the cumulative-charge ceiling, which equals 100 percent of principal across the loan life. On a one-month horizon the cumulative ceiling does not bind. On a multi-month horizon with repeated late fees, BNPL’s lack of an equivalent cumulative ceiling becomes more consequential. The single-month framing in Section 4 understates this dimension of the asymmetry. Fourth, provider fee schedules are subject to revision. The schedules used here are dated as of 30 April 2026 and were verified against current published terms. The April 2024 trustmark accreditation prompted some fee restructuring across providers; further revision is possible. The architectural argument is robust to fee-level revisions because it concerns the choice of cap structure rather than the level of any specific fee. The numerical illustrations would need updating if fee schedules change materially. Fifth, behavioural channels are not modelled. The international literature documents that BNPL increases spending (deHaan et al., 2024 ; Berg et al., 2025 ) and that users disproportionately stack plans across providers. These behavioural channels likely amplify the welfare cost of BNPL relative to the static fee analysis here. The present paper deliberately limits its claims to the statutory architecture; the behavioural extension is a separate research question and would require Singapore-specific user-level data. These limitations do not affect the central finding. The rate-cap architecture for licensed moneylending exists, BNPL operates outside it, and the flat-fee structure produces effective rates that the rate-cap structure would not permit. The limitations affect the magnitude and policy specificity of the recommendations rather than the qualitative finding itself. The most actionable gap is data. The Singapore FinTech Association, through the Experian-administered BNPL credit bureau, holds user-level transaction and late-fee data sufficient to answer the empirical questions the present paper has to leave open: the distribution of users by ticket size, the incidence rate of late fees, the cumulative late fees per user per year, and the extent of stacking across providers. The Moneylenders Credit Bureau publishes monthly aggregate statistics on borrower counts, loan totals, and overdue exposure. The Credit Bureau Singapore publishes regular indicators on credit card balances and delinquencies. The BNPL bureau, by contrast, does not currently publish equivalent aggregate data. A monthly aggregate release analogous to MLCB and CBS publications would advance the regulatory discussion considerably. In the absence of such data, the architectural argument has to stand on the regulatory analysis alone. 9. Conclusion Singapore has built a coherent statutory architecture for consumer credit. Credit cards are heavily underwritten and reported. Licensed moneylenders are heavily rate-capped and reported. Buy-Now-Pay-Later is neither. The 2022 Code of Conduct provides quantity caps and disclosure rules but does not impose rate-cap discipline on per-transaction fee structures. The result is that BNPL late fees, while modest in absolute terms, generate effective monthly rates that vary from 1 percent to 75 percent across ticket sizes, exceeding the 4 percent moneylender rate cap on transactions below approximately S $ 1,000. The contribution of this paper is to make the comparison explicit and to identify the architectural rather than the absolute-fee nature of the asymmetry. The point is not that BNPL providers charge predatory fees: per-incident, they charge less than what a licensed moneylender could lawfully charge for the same overdue amount. The point is that BNPL operates outside the rate-cap regulatory framework that governs every other consumer credit instrument in the jurisdiction, and that the flat-fee structure produces precisely the rate variability that statutory rate caps were designed to prevent. I do not argue that BNPL should be brought under the Moneylenders Act. That would be a category error: BNPL is a payment product with credit characteristics, not a moneylending product. I argue that the regulatory perimeter for BNPL should incorporate the rate-cap discipline that Singapore already applies to comparable credit, with adaptation for the specifics of the BNPL business model. The 2022 Code of Conduct is a useful first step. The next step is statutory backing and rate-cap architecture. The questions left open by this paper are empirical. What share of Singapore BNPL users incur late fees in a given year? How concentrated is BNPL exposure in the small-ticket range? How does cross-provider stacking interact with the per-provider S $ 2,000 outstanding cap? The Experian BNPL bureau holds the data needed to answer these questions. Public release of aggregate statistics, analogous to those that the MLCB and CBS produce, would advance the policy discussion considerably. In the absence of such data, the architectural argument has to stand on its own. The Singapore regulatory tradition is conservative for good reason. Rate caps and quantity caps are blunt instruments, but they have served the jurisdiction well in keeping consumer credit costs predictable across business cycles and borrower segments. Extending that tradition to BNPL is not radical; it is consistent with how Singapore has chosen to regulate every other consumer credit instrument since 2008. Declarations Replication. All effective-rate calculations in Table 2 (Section 4.3 ) follow directly from the inputs in Tables A1 – A3 and the formulas given in Section 4.1 . Effective monthly rate = late fee / missed instalment. Simple APR = monthly rate × 12. Effective annual rate = (1 + monthly rate)¹² − 1. The Moneylenders Act per-month maximum is computed as (missed instalment × 0.04) + S $ 60. No further inputs are required to reproduce the cost grid. Competing interests. The author has no competing interests to declare that are relevant to the content of this article. Funding. The author received no specific funding, grant, or financial support from any agency in the public, commercial, or not-for-profit sectors for the research, authorship, or publication of this article. Author Contribution B.C.L. is the sole author. B.C.L. conceptualised and designed the study, conducted the regulatory and effective-cost analysis, performed all calculations, and wrote and revised the manuscript. Data Availability All data used in this study are publicly available secondary sources. No new data were generated or collected by the author.The primary inputs are: (1) the late-fee schedules of three Singapore BNPL providers (Atome, Grab PayLater, SeaMoney SPayLater), retrieved from the providers' published Help Centre and Terms of Service pages and verified on 30 April 2026; (2) statutory parameters from the Singapore Moneylenders Act 2008 and Moneylenders Rules 2015, available from the Ministry of Law (https://rom.mlaw.gov.sg/); (3) MAS Notice 635 unsecured credit rules and the MAS February 2026 parliamentary reply on BNPL transactions (https://www.mas.gov.sg/); (4) the BNPL Code of Conduct, available from the Singapore FinTech Association (https://singaporefintech.org/); and (5) aggregate statistics from CFPB reports of January 2025 and December 2025, available at https://files.consumerfinance.gov/. All inputs are compiled with citations in Appendix A of the manuscript.The cost grid in Table 2 is deterministic computation from these inputs using the formulas given in Section 4.1 (effective monthly rate = late fee / missed instalment; simple APR = monthly rate × 12; effective annual rate = (1 + monthly rate)^12 - 1). No proprietary or restricted data are used, and no statistical estimation requires replication code. References Atome (2026). Help Centre: What happens when I make a late payment? Retrieved 30 April 2026, from https://help.atome.sg/hc/en-gb/articles/4403114483225 Atome Financial (2025, August 26). Atome Financial posts record US $ 236M operating income, marks full-year profit in 2024 [Press release]. Berg, T., Burg, V., Keil, J., & Puri, M. (2025). The economics of Buy Now, Pay Later: A merchant’s perspective. Journal of Financial Economics , 171 , 104093. https://doi.org/10.1016/j.jfineco.2025.104093 Consumer Financial Protection Bureau. (2022). Buy now, pay later: Market trends and consumer impacts . CFPB. Consumer Financial Protection Bureau (2024). Truth in Lending (Regulation Z); Use of digital user accounts to access buy now, pay later loans. 89 Fed. Reg. 47,068 (May 31, 2024). Formally withdrawn 12 May 2025 (90 Fed. Reg. 20084). Consumer Financial Protection Bureau. (2025a). Consumer use of buy now, pay later and other unsecured debt . CFPB Office of Research. https://files.consumerfinance.gov/f/documents/cfpb_BNPL_Report_2025_01.pdf Consumer Financial Protection Bureau. (2025b). The buy now, pay later market: Data spotlight . CFPB Office of Markets. https://files.consumerfinance.gov/f/documents/cfpb_bnpl-market-report_2025-12.pdf Cornelli, G., Gambacorta, L., & Pancotto, L. (2023). December). Buy now, pay later: A cross-country analysis . BIS Quarterly Review. deHaan, E., Kim, J., Lourie, B., & Zhu, C. (2024). Buy now pay (pain?) later. Management Science , 70 (8), 5586–5598. https://doi.org/10.1287/mnsc.2022.03266 Grab, H. (2025, November 21). PayLater by Grab Singapore: Terms of service. https://www.grab.com/sg/terms-policies/paylater/ LendingTree (2026, February). LendingTree BNPL tracker. Ministry of Law (2026). Registry of Moneylenders: FAQs on borrowing from licensed moneylenders. Retrieved 30 April 2026, from https://rom.mlaw.gov.sg/information-for-borrowers/guide-to-borrowing-from-licensed-moneylenders-english/ Monetary Authority of Singapore (2024). MAS Notice 635: Unsecured credit rules. Monetary Authority of Singapore. (2025a). February 28). Oral reply to parliamentary question on the rising use of buy now . pay later schemes. Monetary Authority of Singapore. (2025b). Monthly statistical bulletin, Table I.17A . Credit and charge card statistics. Monetary Authority of Singapore (2026, February 4). Written reply to parliamentary questions on buy now, pay later transactions and the maximum purchase limit for those below 21 years old. https://www.mas.gov.sg/news/parliamentary-replies/2026/written-reply-to-parliamentary-questions-on-buy-now-pay-later-transactions Moneylenders Act (2008 (Singapore)). as amended, with Moneylenders Rules 2015 (effective 1 October 2015). ResearchAndMarkets (2025). Singapore buy now pay later business and investment opportunities databook — Q3 2025 update [Industry research report]. Sea, L. (2025, March 4). Form 6-K filed with the U.S. Securities and Exchange Commission, full-year 2024 results. Shopee Singapore (2026). SPayLater terms of service and help centre. Retrieved 30 April 2026, from https://help.shopee.sg/4/article/76485 Singapore FinTech Association (2022, October). Buy now, pay later code of conduct. https://singaporefintech.org/news/buy-now-pay-later-bnpl-working-group-launches-bnpl-code-of-conduct-for-singapore/ Singapore FinTech Association (2024, April). BNPL trustmark accreditation announcement. The Straits Times (2024, March 29). 5 buy now, pay later players to get accredited in April after minor shake-up in industry. Appendix Appendix A. Primary-source compilation of regulatory inputs and fee schedules This appendix consolidates the verified regulatory parameters and provider fee schedules used throughout the paper. Each input is given with its primary source. Where a source URL is included, it was accessed and verified on 30 April 2026. Table A1 Moneylenders Act regime parameters (statutory, MinLaw) Parameter Value Authority Maximum monthly interest rate 4% per month, on outstanding principal Moneylenders Rules 2015 (eff. 1 Oct 2015) Maximum monthly late interest rate 4% per month, on overdue amount only Moneylenders Rules 2015 Maximum late fee S $ 60 per month of late repayment Moneylenders Rules 2015 Maximum administrative/approval fee 10% of principal, one-off upfront Moneylenders Rules 2015 Total charges ceiling (whole-of-loan) Cumulative charges ≤ principal of the loan Moneylenders Act 2008, s. 23(1)(b) Unsecured loan cap (annual income < S $ 20,000) S $ 3,000 across all licensed moneylenders Moneylenders Rules Unsecured loan cap (annual income ≥ S $ 20,000) 6 × monthly income across all moneylenders Moneylenders Rules Statutory regulator MinLaw / Registry of Moneylenders Moneylenders Act 2008 Credit reporting MLCB (Moneylenders Credit Bureau, since 2016) MinLaw; CBS (operator since 2021) Notes. Source: Ministry of Law, Registry of Moneylenders, FAQs on Borrowing From Licensed Moneylenders, last updated 7 April 2026, available at https://rom.mlaw.gov.sg/information-for-borrowers/guide-to-borrowing-from-licensed-moneylenders-english/ . Table A2 BNPL Code of Conduct parameters (industry self-regulation, SFA) Parameter Value Authority Outstanding cap (no additional credit assessment) S $ 2,000 per accredited BNPL provider; no aggregation rule across providers BNPL Code of Conduct, Section 4 (SFA, October 2022) Statutory rate cap None No statute applies Compounding of fees Prohibited BNPL Code of Conduct MAS direct regulation No (MAS guidance only) Grab T&C, 21 Nov 2025; MAS Parl. Reply, Feb 2025 Income verification at sign-up No Atome / Grab / SeaMoney sign-up flows Account suspension on default Mandatory BNPL Code of Conduct Compounding interest on overdue amounts Prohibited BNPL Code of Conduct Credit reporting Intra-BNPL bureau only (Experian, since April 2024) SFA / Experian; not visible to CBS Statutory regulator None (industry-led) SFA accreditation; PwC independent audits Sanction for breach Loss of trustmark accreditation SFA Code Notes. Sources: Singapore FinTech Association, BNPL Code of Conduct, October 2022 , https://singaporefintech.org/news/buy-now-pay-later-bnpl-working-group-launches-bnpl-code-of-conduct-for-singapore/ ; trustmark accreditation announcement April 2024 , https://fintechnews.sg/95130/lending/abnk-atome-grab-and-seamoney-earn-singapore-bnpl-compliance-trustmark/ ; MAS Parliamentary Replies, 28 February 2025 and 4 February 2026. Table A3 Provider late fee schedules (verified from primary sources) Provider / plan Late fee structure Per-transaction cap Effective date Source Atome (orders < S $ 1,000) S $ 15 per missed instalment S $ 60 per transaction 1 May 2023 help.atome.sg/hc/en-gb/articles/4403114483225 Atome (orders ≥ S $ 1,000) S $ 30 per missed instalment S $ 60 per transaction 1 May 2023 help.atome.sg/hc/en-gb/articles/4403114483225 Grab PayLater (Pay Next Month) S $ 15 per missed payment No published cap 1 April 2023 grab.com/sg/consumer/finance/paylater/ Grab PayLater Instalments S $ 10 per missed instalment S $ 30 per transaction Current grab.com/sg/terms-policies/paylater/ SeaMoney SPayLater S $ 5 per overdue monthly bill No published cap Current help.shopee.sg/4/article/76485 Notes. All fee schedules accessed and verified on 30 April 2026 against the providers’ published Help Centre or Terms of Service pages. Where multiple plan structures exist, only the consumer-facing default plans are reported. Atome does not currently distinguish between online and offline merchant fees. Grab PayLater Instalments and Pay Next Month are separate products with distinct fee schedules. SeaMoney SPayLater is operated by Monee (Credit) Singapore Private Limited. Table A4 Singapore credit-card regulatory parameters (MAS, for comparison) Parameter Value Authority Minimum annual income (citizens / PRs) S $ 30,000 MAS Notice 635 Minimum annual income (foreigners) S $ 40,000 MAS Notice 635 Aggregate unsecured borrowing cap 12 × monthly income across all FIs MAS Notice 635 Statutory interest rate cap None (subject to MAS conduct rules) MAS Notice 635 Bureau reporting Full reporting to Credit Bureau Singapore (CBS) CBS Act Total Debt Servicing Ratio inclusion Yes (BNPL is excluded; cards are included) MAS Parliamentary Reply, 28 Feb 2025 Statutory regulator MAS Banking Act Sanction for breach Licence action by MAS MAS rule-making powers Notes. Source: Monetary Authority of Singapore, MAS Notice 635 (Unsecured Credit Rules); MAS Parliamentary Replies, 28 February 2025 and 4 February 2026. Card-issuer specific underwriting standards (e.g. higher income thresholds for premium cards) are layered on top of the MAS minimums and are not reported here. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 14 May, 2026 Reviewers agreed at journal 11 May, 2026 Reviewers agreed at journal 06 May, 2026 Reviewers invited by journal 06 May, 2026 Editor assigned by journal 05 May, 2026 Submission checks completed at journal 01 May, 2026 First submitted to journal 30 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9581198","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":640340875,"identity":"138ade7c-345a-4955-8add-3c9398fe6743","order_by":0,"name":"Boon Chuan Lim","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6klEQVRIiWNgGAWjYBACAxDB2AAk2BsYDpCohecAkhZ8mhFaJBKQhPFpMWdvPvbg547D8uaSzx8eLvhll8cvffgA84cK3Fose46lG/aeOWy4c3aOweGZfcnFkn1pCQwHzuBx2I0cM2nGtsOMG27nMBzm7WFO3HCGx4DhYBthLfYbbh5/ANRSD9TC/4EoLYkbbjAYHOb5cRhkCwNeLUC/pEn2tqUnbzgD9Atvw/FiyR42gwNn8PgFFGISP9usbTccP/74M8+f6jx+HuaHDyrwhBgqYGxjSADRB4jVAAR/IFpGwSgYBaNgFCADAFcoXEhc4/nwAAAAAElFTkSuQmCC","orcid":"","institution":"","correspondingAuthor":true,"prefix":"","firstName":"Boon","middleName":"Chuan","lastName":"Lim","suffix":""}],"badges":[],"createdAt":"2026-04-30 21:08:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9581198/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9581198/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109375547,"identity":"03417ad9-a659-4c33-be99-eb67894a08fc","added_by":"auto","created_at":"2026-05-16 14:26:01","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":508520,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9581198/v1/e2ab05ca-f03e-4629-a66c-b4a9dfeb3112.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"The Inverted Regulatory Perimeter in Singapore Buy-Now-Pay-Later","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eSingapore regulates consumer credit through a coherent statutory architecture. Credit card issuers operate under MAS Notice 635, which sets a S\u003cspan\u003e$\u003c/span\u003e30,000 minimum annual income floor for cardholders and caps aggregate unsecured borrowing at 12 times monthly income across all financial institutions.\u003c/p\u003e \u003cp\u003eLicensed moneylenders operate under the Moneylenders Act 2008, which caps monthly interest at 4 percent on outstanding principal and on overdue amounts, caps late fees at S\u003cspan\u003e$\u003c/span\u003e60 per month, and limits cumulative charges to the principal of the loan. \u003csup\u003e1\u003c/sup\u003e Both regimes are statutory, both are enforced by a designated regulator, and both impose binding rate or fee ceilings. The architecture is conservative by design: it is designed to prevent extreme effective rates regardless of contract size or borrower segment.\u003c/p\u003e \u003cp\u003eBuy-Now-Pay-Later operates outside this architecture. Atome, Grab PayLater, and SeaMoney SPayLater offer instalment payment products to Singapore consumers without income verification, without aggregate exposure caps that bind across other credit instruments, and without statutory regulation. The Singapore FinTech Association issued a Buy-Now-Pay-Later Code of Conduct in October 2022 and accredited five providers in April 2024. The Code caps a customer\u0026rsquo;s outstanding payments at S\u003cspan\u003e$\u003c/span\u003e2,000 with each individual provider before an additional credit assessment is required. The cap binds per-provider rather than across the BNPL sector, so a user can in principle accumulate S\u003cspan\u003e$\u003c/span\u003e2,000 each at Atome, Grab, and SeaMoney concurrently without any provider triggering the assessment.\u003csup\u003e2\u003c/sup\u003e The Code is industry self-regulation. It does not set a statutory rate cap, does not impose income verification, and does not feed BNPL exposure into the Credit Bureau Singapore that banks consult for card and loan underwriting.\u003c/p\u003e \u003cp\u003eBefore stating the central claims of the paper, two scope clarifications are useful. First, BNPL is genuinely interest-free for the modal user. A customer who pays each instalment on the due date pays exactly the purchase price, with no interest, fees, or other charges. The product is more attractive than a credit card on this dimension. The paper is concerned with the conditional cost given a missed payment, not with the average user-level cost across the population. Second, the late fees discussed below apply to instalment plans of three to four monthly payments. SPayLater also offers longer tenures of six to twenty-four months on which monthly interest is charged explicitly; these are analytically closer to bank instalment plans and outside the central scope of the paper. The architectural argument focuses on the pay-in-three or pay-in-four products that are the consumer-facing default at Atome, Grab, and SeaMoney.\u003c/p\u003e \u003cp\u003eThis paper makes three claims. First, that BNPL fee structures generate effective monthly rates that would be statutorily prohibited under the Moneylenders Act if BNPL were classified as a loan. Second, that the rate variability is a direct consequence of flat-fee architecture: a fixed dollar fee implies a high effective rate on small principals and a low effective rate on large principals, an outcome that statutory rate caps were designed to prevent. Third, that the regulatory arbitrage operates on architecture, not absolute fee levels. On a per-incident basis, BNPL late fees are below what a licensed moneylender could lawfully charge. The arbitrage is the choice of cap structure (none vs rate-based vs flat fee), not the choice of fee level.\u003c/p\u003e \u003cp\u003eI treat these claims as empirical and verify them against the published fee schedules of the three active accredited Singapore BNPL providers as of April 2026, the statutory ceilings published by MinLaw under the Moneylenders Rules 2015, and the BNPL Code of Conduct administered by the Singapore FinTech Association.\u003csup\u003e3\u003c/sup\u003e The contribution is the synthesis: identifying that a coherent rate-cap regulatory hierarchy exists in Singapore consumer credit, demonstrating that BNPL sits outside it, and quantifying the asymmetry.\u003c/p\u003e \u003cp\u003eThe paper proceeds as follows. Section \u003cspan refid=\"Sec2\" class=\"InternalRef\"\u003e2\u003c/span\u003e reviews the BNPL literature, with emphasis on regulatory arbitrage and effective-cost measurement. Section \u003cspan refid=\"Sec6\" class=\"InternalRef\"\u003e3\u003c/span\u003e sets out the three Singapore consumer credit regimes side by side. Section \u003cspan refid=\"Sec11\" class=\"InternalRef\"\u003e4\u003c/span\u003e presents the analytical core: a worked example, a cost grid across providers and ticket sizes, and the comparison to the Moneylenders Act ceilings. Section \u003cspan refid=\"Sec17\" class=\"InternalRef\"\u003e5\u003c/span\u003e calibrates the analysis to empirical data, drawing on aggregate market statistics, provider-level financial disclosures, MAS-disclosed Singapore user statistics, and the user-level evidence on BNPL borrowers from the CFPB\u0026rsquo;s 2025 studies. Section \u003cspan refid=\"Sec22\" class=\"InternalRef\"\u003e6\u003c/span\u003e considers counter-arguments. Section \u003cspan refid=\"Sec27\" class=\"InternalRef\"\u003e7\u003c/span\u003e sketches a policy proposal. Section \u003cspan refid=\"Sec32\" class=\"InternalRef\"\u003e8\u003c/span\u003e sets out limitations and scope. Section \u003cspan refid=\"Sec33\" class=\"InternalRef\"\u003e9\u003c/span\u003e concludes.\u003c/p\u003e"},{"header":"2. Related Literature","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Effective cost of BNPL\u003c/h2\u003e \u003cp\u003edeHaan et al. (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) examine the effects of BNPL adoption on US consumer financial health using banking data for 10.6\u0026nbsp;million US consumers. They find that new BNPL users experience rapid increases in bank overdraft charges and credit card interest and fees compared with non-users. They use an instrumental variable based on consumers\u0026rsquo; pre-BNPL shopping habits to argue for a causal interpretation, and conclude that BNPL plausibly facilitates overborrowing. The paper anchors the borrower-side empirical literature on BNPL and frames the welfare question as one of total cost across all of a user\u0026rsquo;s credit instruments, not just BNPL fees in isolation.\u003c/p\u003e \u003cp\u003eCornelli et al. (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) take the cross-country adoption view in a BIS Quarterly Review article. They document that BNPL adoption is stronger in countries with higher e-commerce penetration, higher inflation, less efficient banking sectors, and looser financial regulation. They explicitly raise the regulatory arbitrage question: whether BNPL grows fastest where the regulatory perimeter is weakest. They do not resolve the question empirically but treat it as the leading hypothesis.\u003c/p\u003e \u003cp\u003eBerg et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) provide the merchant-side analysis. They develop a structural model in which BNPL functions as a price-discrimination device: merchants accept BNPL because it expands sales among low-creditworthiness customers, and the merchant discount rate (3 to 6 percent) is the primary revenue source for BNPL providers, not late fees on consumers.\u003csup\u003e11\u003c/sup\u003e The Berg et al. result complicates the simple \u0026ldquo;BNPL extracts surplus from consumers\u0026rdquo; narrative: most extraction is from merchants, who pass through some portion via prices. But the consumer-cost question remains relevant for the late-fee tail of users who fail to pay on time.\u003c/p\u003e \u003cp\u003eThe CFPB (2022) market report on BNPL is the most cited regulatory analysis. The report identifies three concerns: accumulated debt (loan stacking across providers), regulatory arbitrage (BNPL avoiding TILA disclosures and Reg Z protections), and data harvesting (BNPL providers monetising user data). The CFPB issued an interpretive rule in May 2024 applying Regulation Z to BNPL providers, which was withdrawn in 2025.\u003csup\u003e8\u003c/sup\u003e The US regulatory record demonstrates the difficulty of bringing BNPL inside the consumer credit perimeter once it is already established as a payment product rather than a loan product.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Singapore-specific work\u003c/h2\u003e \u003cp\u003eI find no peer-reviewed empirical paper on Singapore BNPL effective costs or regulatory perimeter as of April 2026. The available material consists of MAS and SFA announcements, Lexology and law-firm commentaries on the Code of Conduct, and Straits Times coverage of provider accreditation. The 2024 trustmark accreditation of Atome, Grab, SeaMoney, LatitudePay (since departed), and Abnk.ai received industry press but no academic treatment. The Experian-administered BNPL credit bureau is operational since 2024 but has not produced public empirical reports on the user distribution.\u003c/p\u003e \u003cp\u003eThis is the gap the present paper fills. The contribution is not new theory: it is the application of existing effective-cost methodology, adapted from the CFPB and deHaan et al., to the Singapore regulatory environment, where a coherent rate-cap regime for moneylending exists and provides a natural benchmark.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 The moneylender comparison\u003c/h2\u003e \u003cp\u003eThe US literature compares BNPL to credit cards and to payday loans. Both comparisons are imperfect for Singapore. Credit cards are heavily underwritten and serve a different population. Payday lending in the US sense barely exists in Singapore: short-term lending to unbanked consumers is regulated under the Moneylenders Act with hard caps. The relevant comparator in Singapore is the licensed moneylender, which serves a similar borrower segment to BNPL (consumers without ready access to bank credit) but operates under heavier regulation, not lighter.\u003c/p\u003e \u003cp\u003eI have not found prior work that draws this comparison directly. The comparison is fruitful because it inverts the usual framing. In the US, BNPL is sometimes positioned as a more responsible alternative to predatory payday lending. In Singapore, BNPL operates under lighter rules than the licensed moneylender regime that itself replaced predatory lending. The regulatory hierarchy runs in the opposite direction of what the BNPL marketing narrative suggests.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. The Three Singapore Consumer Credit Regimes","content":"\u003cp\u003eSingapore operates three statutory or quasi-statutory consumer credit regimes. I summarise each below and then present them in tabular form to make the comparison explicit.\u003c/p\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Credit cards (MAS)\u003c/h2\u003e \u003cp\u003eCredit card issuance is regulated by MAS under the Banking Act and MAS Notice 635 (Unsecured Credit Rules).\u003csup\u003e4\u003c/sup\u003e The minimum annual income for cardholders is S\u003cspan\u003e$\u003c/span\u003e30,000 for citizens and permanent residents and S\u003cspan\u003e$\u003c/span\u003e40,000 for foreigners. Aggregate unsecured borrowing across all financial institutions is capped at 12 times monthly income, enforced through the Credit Bureau Singapore. Issuers conduct hard credit checks; payment history reports back to CBS. There is no statutory cap on interest rates or fees, but issuers operate under MAS conduct rules and competitive pressure. Effective annual rates on revolving balances typically run 24 to 28 percent.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Licensed moneylenders (MinLaw)\u003c/h2\u003e \u003cp\u003eLicensed moneylending is regulated under the Moneylenders Act 2008 with implementing rules from the Moneylenders Rules 2015.\u003csup\u003e5\u003c/sup\u003e The maximum interest rate is 4 percent per month on outstanding principal, calculated on a reducing-balance basis. The maximum late interest is also 4 percent per month, applied only to overdue amounts. The maximum administrative fee is 10 percent of principal, charged once at loan origination. The maximum late fee is S\u003cspan\u003e$\u003c/span\u003e60 per month of late repayment. The Act\u0026rsquo;s principal-ceiling rule is critical: the cumulative total of all charges over the life of the loan, including interest, late interest, the upfront administrative fee, and all monthly late fees, cannot exceed the principal of the loan.\u003csup\u003e13\u003c/sup\u003e This applies to charges across the whole loan, not to any individual month or instalment.\u003c/p\u003e \u003cp\u003eUnsecured borrowing is capped at S\u003cspan\u003e$\u003c/span\u003e3,000 across all licensed moneylenders for borrowers with annual income below S\u003cspan\u003e$\u003c/span\u003e20,000, and at six times monthly income for borrowers above that threshold. Lender data flows into the Moneylenders Credit Bureau. Borrowers may apply for self-exclusion through the MLCB.\u003csup\u003e10\u003c/sup\u003e The regime is enforced by MinLaw\u0026rsquo;s Registry of Moneylenders. Sanctions for non-compliance include licence revocation and criminal penalties.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.3 BNPL (industry Code of Conduct)\u003c/h2\u003e \u003cp\u003eBuy-Now-Pay-Later providers in Singapore operate outside any specific statutory regime. The Buy-Now-Pay-Later Code of Conduct, launched in October 2022 by the Singapore FinTech Association under MAS guidance, is industry self-regulation.\u003csup\u003e2\u003c/sup\u003e The Code\u0026rsquo;s principal substantive provision caps customer outstanding BNPL exposure at S\u003cspan\u003e$\u003c/span\u003e2,000 across all accredited providers without an additional credit assessment.\u003csup\u003e12\u003c/sup\u003e The Code requires fee disclosure, prohibits compounding of fees, and allows voluntary self-exclusion. It does not set a rate cap, does not require income verification, and does not require BNPL data to flow into the Credit Bureau Singapore.\u003c/p\u003e \u003cp\u003eFive providers were accredited under the trustmark in April 2024: Atome, SeaMoney, Grab, LatitudePay, and Abnk.ai. LatitudePay subsequently exited the Singapore market. As of April 2026, the three active providers offering general-purpose BNPL services are Atome, Grab PayLater, and SeaMoney SPayLater.\u003csup\u003e9\u003c/sup\u003e The Grab PayLater Terms of Service (last modified 21 November 2025) include a consumer advisory stating that the service is a buy-now-pay-later product and does not constitute a loan, credit card or charge card facility, and that BNPL services are not regulated by the Monetary Authority of Singapore.\u003csup\u003e6\u003c/sup\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e3.4 The architecture, side by side\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e summarises the three regimes across eight regulatory dimensions. The asymmetry is most visible on the rate cap, total charge ceiling, statutory regulator, and credit reporting rows.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSingapore consumer credit regimes compared\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDimension\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCredit cards (MAS)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLicensed moneylenders (MinLaw)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBNPL (SFA Code)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUnderwriting threshold\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMin. annual income S\u003cspan\u003e$\u003c/span\u003e30k (citizen/PR), bureau check, aggregate cap 12x monthly income\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIncome-based borrowing tiers (S\u003cspan\u003e$\u003c/span\u003e3,000 cap below S\u003cspan\u003e$\u003c/span\u003e20k income, 6x monthly above); MLCB check; self-exclusion available\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNo income verification; sign-up via NRIC and linked card\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInterest rate cap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo statutory cap (MAS conduct rules)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4% per month on principal; 4% per month on overdue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNo cap; nominally interest-free if paid on time\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLate fee cap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIssuer-specific; subject to MAS conduct rules\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e60 per month of late repayment (statutory)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNo statutory cap; provider-specific (e.g. Atome S\u003cspan\u003e$\u003c/span\u003e60/transaction)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal charge ceiling\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo statutory ceiling on cumulative charges\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTotal charges\u0026thinsp;\u0026le;\u0026thinsp;principal of loan\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNo cumulative charge ceiling\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAggregate exposure cap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12x monthly income across all unsecured (MAS rule)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTiered by income (S\u003cspan\u003e$\u003c/span\u003e3,000 / 6x monthly income)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e2,000 per provider before additional assessment (Code, not statute); no inter-provider aggregation rule\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCredit reporting\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFull bureau (CBS) reporting\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMLCB reporting (since 2016)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eIntra-BNPL bureau only (Experian, since 2024); not visible to bank lenders\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStatutory regulator\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMAS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMinLaw / Registry of Moneylenders\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNone (industry-led, MAS guidance only)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSanctions for breach\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLicence action by MAS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLicence revocation, criminal penalties\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLoss of trustmark accreditation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eNotes. Compiled from MAS Notice 635, Moneylenders Act 2008 with Moneylenders Rules 2015, and the BNPL Code of Conduct (SFA, October 2022). Sources cited in footnotes.\u003c/em\u003e \u003c/p\u003e \u003cp\u003eThe pattern across Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e is a coherent regulatory hierarchy with one clear exception. Credit cards have heaviest income-based underwriting and lightest rate caps, on the theory that underwriting controls the borrower population. Moneylenders have lighter income-based underwriting (a tiered borrowing cap rather than a hard income floor) and heaviest rate caps, on the theory that rate caps protect a population that the underwriting cannot effectively screen. Both regimes use income to determine how much a borrower can access, and both bind across providers within the regime through their respective bureaus. BNPL uses income for nothing. The Code of Conduct\u0026rsquo;s S\u003cspan\u003e$\u003c/span\u003e2,000 outstanding cap applies per provider rather than across all BNPL providers, so cross-provider stacking is not constrained by the cap itself. The cap is also invariant to the borrower\u0026rsquo;s income, employment status, or existing debt elsewhere. The Code is enforced by industry self-regulation rather than statute.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. The Effective Cost Comparison","content":"\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Method\u003c/h2\u003e \u003cp\u003eI compute the effective monthly rate on a missed BNPL instalment by dividing the late fee by the deferred amount and treating the result as a monthly rate. The deferred amount is the value of the missed instalment, which is the credit actually extended past the due date. This follows the convention used by the CFPB (2022) and is the economically correct denominator for an APR calculation: it answers the question \u0026ldquo;what rate is being charged for the use of these funds for this period?\u0026rdquo;\u003csup\u003e14\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eTwo alternative denominators warrant comment because they appear in industry materials and would yield different effective rates. The first is the full transaction value at point of purchase. This denominator amortises the late fee over the entire ticket regardless of how much of it has been repaid, and so understates the cost of credit on the residual amount actually deferred. For an Atome 3-instalment plan on a S\u003cspan\u003e$\u003c/span\u003e300 transaction, the customer has already paid S\u003cspan\u003e$\u003c/span\u003e100 at checkout when the first instalment is due. Charging the S\u003cspan\u003e$\u003c/span\u003e15 missed-instalment fee against the original S\u003cspan\u003e$\u003c/span\u003e300 produces a 5 percent monthly rate; against the S\u003cspan\u003e$\u003c/span\u003e100 actually overdue, 15 percent. The latter measures what the customer is paying for the use of credit they still have, which is the standard APR question. The second alternative is the residual outstanding balance immediately after the missed instalment, which for the same example is S\u003cspan\u003e$\u003c/span\u003e200 (the missed S\u003cspan\u003e$\u003c/span\u003e100 plus the not-yet-due S\u003cspan\u003e$\u003c/span\u003e100). Charging S\u003cspan\u003e$\u003c/span\u003e15 against S\u003cspan\u003e$\u003c/span\u003e200 produces 7.5 percent. This denominator captures total balance at risk but conflates funds the borrower owes immediately with funds not yet due. The CFPB-convention denominator (the deferred instalment itself) is the economically cleanest for the question of effective rate on a single missed payment. Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e below uses this convention. Section \u003cspan refid=\"Sec32\" class=\"InternalRef\"\u003e8\u003c/span\u003e acknowledges the sensitivity, and an alternative-denominator robustness column is reported in the worked example.\u003c/p\u003e \u003cp\u003eThree rate measures follow naturally. The simple APR is the monthly rate multiplied by 12. The effective annual rate (EAR) is the monthly rate compounded over 12 months. Both annualisations are reported, with the simple APR used for headline comparisons and the EAR for substantive interpretation.\u003c/p\u003e \u003cp\u003eThe Moneylenders Act benchmark for one month of overdue is the sum of two components: (a) 4 percent monthly interest on the overdue amount, and (b) up to S\u003cspan\u003e$\u003c/span\u003e60 in late fees per month. The per-month maximum therefore varies with the missed amount and equals (missed amount \u0026times; 0.04)\u0026thinsp;+\u0026thinsp;S\u003cspan\u003e$\u003c/span\u003e60. For a S\u003cspan\u003e$\u003c/span\u003e100 missed instalment this is S\u003cspan\u003e$\u003c/span\u003e64.00; for S\u003cspan\u003e$\u003c/span\u003e33.33 (a missed instalment on a S\u003cspan\u003e$\u003c/span\u003e100 BNPL transaction split three ways) this is S\u003cspan\u003e$\u003c/span\u003e61.33; for S\u003cspan\u003e$\u003c/span\u003e500 it is S\u003cspan\u003e$\u003c/span\u003e80.00. Cumulative charges across the loan life are capped separately at 100 percent of principal, but this cumulative cap does not bind in any single-month scenario considered here. I compute both an absolute fee comparison (BNPL fee versus MA per-month maximum) and a rate comparison (BNPL effective monthly rate versus MA 4 percent rate cap).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e4.2 A worked example\u003c/h2\u003e \u003cp\u003eConsider an Atome standard 3-instalment plan on a S\u003cspan\u003e$\u003c/span\u003e300 transaction. The borrower pays S\u003cspan\u003e$\u003c/span\u003e100 at checkout, S\u003cspan\u003e$\u003c/span\u003e100 in month 1, and S\u003cspan\u003e$\u003c/span\u003e100 in month 2. Suppose the borrower misses the month-1 instalment and pays it 30 days late.\u003c/p\u003e \u003cp\u003eBNPL effective cost. Atome charges a S\u003cspan\u003e$\u003c/span\u003e15 admin fee for orders below S\u003cspan\u003e$\u003c/span\u003e1,000.\u003c/p\u003e \u003cp\u003eEffective monthly rate (deferred-amount denominator)\u0026thinsp;=\u0026thinsp;\u003cb\u003eS$15 / S$100\u0026thinsp;=\u0026thinsp;15.0%\u003c/b\u003e\u003c/p\u003e \u003cp\u003eSimple APR\u0026thinsp;=\u0026thinsp;15.0% \u0026times; 12\u0026thinsp;=\u0026thinsp;\u003cb\u003e180%\u003c/b\u003e\u003c/p\u003e \u003cp\u003eEffective annual rate = (1\u0026thinsp;+\u0026thinsp;0.15)\u0026sup1;\u0026sup2; \u0026minus; 1\u0026thinsp;=\u0026thinsp;\u003cb\u003e435%\u003c/b\u003e\u003c/p\u003e \u003cp\u003eFor comparison, the alternative denominators discussed in Section \u003cspan refid=\"Sec12\" class=\"InternalRef\"\u003e4.1\u003c/span\u003e yield:\u003c/p\u003e \u003cp\u003eAgainst full transaction value (S\u003cspan\u003e$\u003c/span\u003e300): S\u003cspan\u003e$\u003c/span\u003e15 / S\u003cspan\u003e$\u003c/span\u003e300\u0026thinsp;=\u0026thinsp;\u003cb\u003e5.0% monthly\u003c/b\u003e (60% simple APR, 80% EAR)\u003c/p\u003e \u003cp\u003eAgainst residual outstanding balance (S\u003cspan\u003e$\u003c/span\u003e200, includes not-yet-due S\u003cspan\u003e$\u003c/span\u003e100): S\u003cspan\u003e$\u003c/span\u003e15 / S\u003cspan\u003e$\u003c/span\u003e200\u0026thinsp;=\u0026thinsp;\u003cb\u003e7.5% monthly\u003c/b\u003e (90% simple APR, 138% EAR)\u003c/p\u003e \u003cp\u003eAll three denominators produce monthly rates above the 4 percent Moneylenders Act ceiling on this transaction. The deferred-amount denominator is used as the headline measure throughout the paper because it is the standard APR convention for measuring the effective rate on credit actually extended past its due date. The architectural finding does not depend on the choice.\u003c/p\u003e \u003cp\u003eMoneylenders Act benchmark. On a S\u003cspan\u003e$\u003c/span\u003e100 overdue for one month:\u003c/p\u003e \u003cp\u003eMaximum interest\u0026thinsp;=\u0026thinsp;4% \u0026times; S\u003cspan\u003e$\u003c/span\u003e100\u0026thinsp;=\u0026thinsp;\u003cb\u003eS$4\u003c/b\u003e\u003c/p\u003e \u003cp\u003eMaximum late fee\u0026thinsp;=\u0026thinsp;\u003cb\u003eS$60\u003c/b\u003e (statutory)\u003c/p\u003e \u003cp\u003eMA total per-month maximum\u0026thinsp;=\u0026thinsp;\u003cb\u003eS$64\u003c/b\u003e\u003c/p\u003e \u003cp\u003eTwo comparisons follow.\u003c/p\u003e \u003cp\u003eOn absolute fees, the Atome S\u003cspan\u003e$\u003c/span\u003e15 charge is below the MA per-month maximum of S\u003cspan\u003e$\u003c/span\u003e64. \u003cem\u003ePer-incident, BNPL is cheaper than what a moneylender could legally charge on the same overdue amount\u003c/em\u003e. I emphasise this because an earlier framing of the question, popular in policy commentary, has BNPL fees \u0026ldquo;exceeding\u0026rdquo; moneylender ceilings on a per-incident basis. That framing does not survive the arithmetic.\u003c/p\u003e \u003cp\u003eOn effective rates, the picture inverts. The Atome 15.0 percent monthly rate is 3.75 times the 4 percent statutory rate cap that applies to every other consumer credit instrument in Singapore. \u003cem\u003eThe Moneylenders Act is structured to prevent precisely this outcome\u003c/em\u003e: by capping the rate at 4 percent rather than the absolute fee at S\u003cspan\u003e$\u003c/span\u003e60, the Act limits effective cost-of-credit regardless of the size of the contract. BNPL\u0026rsquo;s flat-fee structure produces effective rates that scale inversely with the missed amount, generating high rates on small principals.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e4.3 The cost grid\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e extends the worked example to seventeen scenarios across the three providers and a range of ticket sizes. Each row reports the missed-instalment amount, the late fee charged by the provider, the implied effective monthly rate on the deferred amount, the simple APR and EAR, and two comparison columns: the BNPL fee as a percentage of the MA per-month maximum, and the BNPL effective rate as a multiple of the 4 percent MA statutory rate cap.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eEffective cost of BNPL late fees vs Moneylenders Act ceilings\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProvider\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePlan\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTicket (S\u003cspan\u003e$\u003c/span\u003e)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMissed (S\u003cspan\u003e$\u003c/span\u003e)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLate fee (S\u003cspan\u003e$\u003c/span\u003e)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eMonthly rate\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSimple APR\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eEAR\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eBNPL/MA max\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003eRate / 4% cap\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAtome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 instalments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e20.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e75.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e900%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e82,401%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e23%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e18.8x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAtome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 instalments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e33.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e45.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e540%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e8,538%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e23%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e11.2x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAtome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 instalments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e100.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e15.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e180%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e435%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e23%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e3.8x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAtome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 instalments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e600\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e200.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e7.5%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e90%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e138%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e23%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e1.9x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAtome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 instalments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e999\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e333.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4.5%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e54%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e70%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e23%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e1.1x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAtome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 instalments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e500.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e6.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e72%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e101%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e38%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e1.5x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAtome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 instalments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e666.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4.5%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e54%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e70%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e36%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e1.1x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrab PayLater\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePay next month\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e100.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e15.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e180%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e435%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e23%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e3.8x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrab PayLater\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePay next month\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e300.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e5.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e60%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e80%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e21%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e1.2x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrab PayLater\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePay next month\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1,000.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.5%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e18%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e20%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e15%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.4x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrab PayLater\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4 instalments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e100.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e10.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e120%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e214%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e16%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e2.5x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrab PayLater\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4 instalments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e250.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e48%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e60%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e14%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e1.0x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrab PayLater\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4 instalments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e500.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e24%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e27%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e12%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.5x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSeaMoney SPayLater\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 monthly bills\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e150\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e50.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e10.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e120%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e214%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e8%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e2.5x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSeaMoney SPayLater\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 monthly bills\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e100.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e5.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e60%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e80%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e8%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e1.3x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSeaMoney SPayLater\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 monthly bills\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e600\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e200.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2.5%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e30%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e34%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e7%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.6x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSeaMoney SPayLater\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 monthly bills\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e500.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e12%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e13%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e6%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.2x\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eNotes. Effective monthly rate\u0026thinsp;=\u0026thinsp;late fee / missed instalment. Simple APR\u0026thinsp;=\u0026thinsp;monthly rate \u0026times; 12. EAR = (1\u0026thinsp;+\u0026thinsp;monthly rate)\u0026sup1;\u0026sup2; \u0026minus; 1. BNPL/MA max compares the BNPL late fee to the moneylender per-month maximum, computed as (missed instalment \u0026times; 0.04)\u0026thinsp;+\u0026thinsp;S$60 and varying row-by-row with the missed amount. Rate / 4 percent cap compares the BNPL effective monthly rate to the statutory rate ceiling; the displayed multiple is rounded to one decimal place using round-half-up. Ticket sizes are illustrative; provider fee schedules verified from Atome, Grab, and Shopee/SeaMoney published terms as of 30 April 2026.\u003c/em\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Three patterns\u003c/h2\u003e \u003cp\u003eThree patterns emerge from Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eFirst, BNPL late fees are below the Moneylenders Act per-month maximum in every row. The BNPL/MA max column ranges from 6 percent (SeaMoney on a S\u003cspan\u003e$\u003c/span\u003e1,500 ticket) to 38 percent (Atome on a S\u003cspan\u003e$\u003c/span\u003e1,500 ticket). On no row does a BNPL fee exceed what a licensed moneylender could lawfully charge for the same overdue amount in the same month. The simple framing \u0026ldquo;BNPL costs more than moneylending\u0026rdquo; is therefore false at the per-incident level. This matters because that framing has appeared in policy discussion and would not survive scrutiny.\u003c/p\u003e \u003cp\u003eSecond, BNPL effective rates exceed the 4 percent statutory rate cap on the small-ticket end of the distribution. Atome on a S\u003cspan\u003e$\u003c/span\u003e60 ticket implies a 75.0 percent monthly rate, which is 18.8 times the moneylender rate cap. Atome on a S\u003cspan\u003e$\u003c/span\u003e300 ticket implies a 15.0 percent monthly rate, 3.8 times the cap. The pattern is consistent across providers: SeaMoney on a S\u003cspan\u003e$\u003c/span\u003e150 ticket is 2.5 times the cap and on a S\u003cspan\u003e$\u003c/span\u003e300 ticket is 1.25 times the cap; Grab PayLater Pay-Next-Month on a S\u003cspan\u003e$\u003c/span\u003e100 ticket is 3.8 times the cap. The effective rate falls below the 4 percent ceiling only at ticket sizes above approximately S\u003cspan\u003e$\u003c/span\u003e1,000 for Atome and Grab, and above S\u003cspan\u003e$\u003c/span\u003e300 for SeaMoney.\u003c/p\u003e \u003cp\u003eThird, the rate variability is large. Across the seventeen scenarios, the implied effective monthly rate ranges from 1.0 percent (SeaMoney, S\u003cspan\u003e$\u003c/span\u003e1,500 ticket) to 75.0 percent (Atome, S\u003cspan\u003e$\u003c/span\u003e60 ticket). This is the architectural finding. Statutory rate caps are designed to deliver uniform effective rates regardless of contract size. Flat-fee structures by definition do not. The same dollar fee implies very different effective rates depending on the size of the missed amount, with the highest rates concentrated on the smallest tickets.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e4.5 Why ticket-size sensitivity matters\u003c/h2\u003e \u003cp\u003eThe BNPL user distribution is concentrated in small-ticket transactions. Atome\u0026rsquo;s default plan splits payments into three instalments, encouraging use on purchases in the S\u003cspan\u003e$\u003c/span\u003e100\u0026ndash;S\u003cspan\u003e$\u003c/span\u003e500 range. Grab PayLater\u0026rsquo;s in-app payment limit and SeaMoney SPayLater\u0026rsquo;s integration with Shopee likewise drive small-ticket usage. The user demographic is younger, lower-income, and more financially fragile than card holders, consistent with international evidence on US BNPL borrowers (deHaan et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and US late-payment incidence (LendingTree, 2026).\u003c/p\u003e \u003cp\u003eThe combination is consequential. The product is most heavily used in the ticket range where the flat-fee structure produces the highest effective rates, and the user population is the segment for which statutory rate caps would normally provide the most protection. A regulatory regime that imposes a 4 percent monthly cap on licensed moneylenders, who serve the same borrower segment, but does not impose any rate constraint on BNPL is asymmetric in a way that is hard to justify on first principles.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Empirical Calibration","content":"\u003cp\u003eSection \u003cspan refid=\"Sec11\" class=\"InternalRef\"\u003e4\u003c/span\u003e established the comparative-rate finding analytically. This section calibrates the analysis to the available empirical data. I draw on four sources: aggregate market statistics published by the Monetary Authority of Singapore and from secondary industry research, provider-level financial disclosures from listed and quasi-listed parents, MAS-disclosed Singapore user-level statistics from a February 2026 parliamentary reply, and the user-level evidence base from the Consumer Financial Protection Bureau\u0026rsquo;s 2025 studies of US BNPL borrowers. The aggregate data establishes the scale of the BNPL market relative to the comparator credit regimes. The provider data establishes that BNPL providers have reached scale and profitability, and that consumer late fees are not the primary revenue source. The MAS Singapore data and CFPB US data jointly establish the structural patterns of BNPL usage at the user level: small typical outstanding, dominant loan stacking, concentration in subprime and deep-subprime borrowers, and elevated balances on other unsecured credit. These patterns hold across jurisdictions and bear directly on the architectural argument.\u003c/p\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e5.1 Aggregate market scale\u003c/h2\u003e \u003cp\u003eSingapore BNPL gross merchandise value reached an estimated US\u003cspan\u003e$\u003c/span\u003e1.19\u0026nbsp;billion (approximately S\u003cspan\u003e$\u003c/span\u003e1.6\u0026nbsp;billion) in 2024 and is projected at US\u003cspan\u003e$\u003c/span\u003e1.32\u0026nbsp;billion in 2025, growing at 11.1 percent annually.\u003csup\u003e17\u003c/sup\u003e For comparison, total credit card billings in Singapore averaged S\u003cspan\u003e$\u003c/span\u003e23 to S\u003cspan\u003e$\u003c/span\u003e24\u0026nbsp;billion per quarter in 2024, equivalent to roughly S\u003cspan\u003e$\u003c/span\u003e95\u0026nbsp;billion annualised. Card rollover balances (the share carried at interest) reached S\u003cspan\u003e$\u003c/span\u003e8.3\u0026nbsp;billion at end-2024.\u003csup\u003e18\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eThe relative scale matters for the policy framing. MAS confirmed in February 2026 that BNPL transactions remain less than 2 percent of the value of total credit and debit card payments. The aggregate market is small enough that BNPL does not threaten financial stability. Whether it threatens consumer protection at the user level is a separate question, addressed below.\u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e sets out the aggregate scale comparison.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eAggregate scale of Singapore consumer credit instruments\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInstrument\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAnnual flow (S\u003cspan\u003e$\u003c/span\u003ebn)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOutstanding (S\u003cspan\u003e$\u003c/span\u003ebn)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eReference period\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSource\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCredit cards (billings)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e~\u0026thinsp;95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e~\u0026thinsp;8.3 (rollover)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2024 full year\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMAS Stat. Bulletin I.17A\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBNPL (gross merchandise)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e~\u0026thinsp;1.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003en.d.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2024 full year\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eIndustry research; MAS confirms\u0026thinsp;\u0026lt;\u0026thinsp;2% of card billings\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLicensed moneylenders\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003en.d.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e~\u0026thinsp;2.0 (estimated)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2024 (most recent published)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMLCB; MinLaw\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eNotes. BNPL gross merchandise value is the total value of transactions financed via BNPL plans, not the outstanding amount on those plans at any point in time. Card billings are the equivalent gross flow for credit cards. Card outstanding is the rollover balance (the share carried beyond the due date and subject to interest). Public quarterly disclosures of BNPL outstanding are not published by MAS or SFA. Licensed moneylender outstanding is approximate, based on MLCB administrative data; precise figures are not in regular public release.\u003c/em\u003e \u003c/p\u003e \u003cp\u003eThe pattern across Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e is that BNPL is the smallest of the three regimes by transaction value, but the only one without rate-cap regulation. The asymmetry is therefore not about the size of the market: a comprehensive consumer credit architecture would be expected to apply rate-cap discipline regardless of segment size, and Singapore in fact does this for licensed moneylending despite that segment being smaller still in outstanding terms than the credit card market.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e5.2 Provider-level revenue mix\u003c/h2\u003e \u003cp\u003eAtome\u0026rsquo;s parent, Atome Financial (part of Singapore-headquartered Advance Intelligence Group), reported regional gross merchandise volume of US\u003cspan\u003e$\u003c/span\u003e2\u0026nbsp;billion in 2024, up 50 percent year-on-year, with annualised revenue surpassing US\u003cspan\u003e$\u003c/span\u003e500\u0026nbsp;million by Q2 2025. Atome Financial reported full-year operating income of US\u003cspan\u003e$\u003c/span\u003e236\u0026nbsp;million in 2024 and turned profitable in Q1 2024.\u003csup\u003e19\u003c/sup\u003e The disclosures emphasise growth in the Atome PayLater Anywhere Card (1.5\u0026nbsp;million cards issued in the Philippines by mid-2025), expansion into insurance, savings, and lending, and partnership-based merchant integration. Late fees on consumer BNPL are not separately disclosed as a revenue line. Singapore-specific figures are not separately reported.\u003c/p\u003e \u003cp\u003eSeaMoney, the digital financial services arm of Sea Limited (NYSE: SE), is reported alongside parent group financials in SEC filings. SeaMoney consumer and SME loans principal outstanding reached US\u003cspan\u003e$\u003c/span\u003e5.1\u0026nbsp;billion at end-2024, up 63.9 percent year-on-year. Non-performing loans past due more than 90 days were 1.2 percent of principal outstanding. Sea Limited\u0026rsquo;s digital financial services GAAP revenue was US\u003cspan\u003e$\u003c/span\u003e2.4\u0026nbsp;billion in 2024.\u003csup\u003e20\u003c/sup\u003e SPayLater is the consumer-facing BNPL product within this segment. As with Atome, late fees are not separately disclosed. SeaMoney operates in eight markets across Southeast Asia, Taiwan, and Brazil; Singapore is one of these but not the largest by revenue.\u003c/p\u003e \u003cp\u003eTwo findings follow from the provider data. First, BNPL providers in Southeast Asia have reached scale, profitability, and institutional capital backing. The argument that BNPL is a fragile fintech sector requiring regulatory forbearance no longer applies. Second, late fees do not appear as a separately disclosed revenue line in either provider\u0026rsquo;s public filings, consistent with the merchant discount rate being the primary revenue source. This supports the Section \u003cspan refid=\"Sec26\" class=\"InternalRef\"\u003e6.4\u003c/span\u003e counter-argument that late fees are not the main profit centre. It also supports the policy implication: replacing flat-fee late charges with rate-based late charges would not materially affect provider economics, because late fees are a small share of revenue.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e5.3 User-level evidence\u003c/h2\u003e \u003cp\u003eSection \u003cspan refid=\"Sec18\" class=\"InternalRef\"\u003e5.1\u003c/span\u003e established the aggregate market scale and Section \u003cspan refid=\"Sec19\" class=\"InternalRef\"\u003e5.2\u003c/span\u003e established the provider economics. Neither reaches the level of individual borrower behaviour, which is where the architectural argument has to land if the policy proposal is to be defensible. Singapore-specific user-level statistics are limited to what MAS has disclosed in parliamentary replies. The most rigorous user-level evidence available globally is from the United States Consumer Financial Protection Bureau, which has published two studies in 2025 drawing on aggregate data from six large pay-in-four lenders covering approximately 40 percent of the US point-of-sale financing market.\u003c/p\u003e \u003cp\u003eI summarise the relevant US findings below and compare them to what is known about Singapore. The two jurisdictions differ in regulatory framework, provider mix, and consumer protections, so the US data should be read as evidence about the structure of BNPL usage rather than as a direct claim about Singapore.\u003csup\u003e23\u003c/sup\u003e That said, the structural patterns are unlikely to be jurisdiction-specific.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBNPL market and borrower metrics from CFPB data (six large US lenders)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMetric\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2021\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2022\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2023\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSource\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLoan originations (millions)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e196.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e273.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e335.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCFPB Dec 2025, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDollar originations (US\u003cspan\u003e$\u003c/span\u003e billions, CPI-adj.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e29.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e35.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e45.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCFPB Dec 2025, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage loan size (US\u003cspan\u003e$\u003c/span\u003e, CPI-adj.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e150\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e131\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e135\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCFPB Dec 2025, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUnique users (millions)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003en.r.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e48.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e53.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCFPB Dec 2025, Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage loans per user per year\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003en.r.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCFPB Dec 2025, Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage annual amount per user (US\u003cspan\u003e$\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003en.r.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e745\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e848\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCFPB Dec 2025, Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLoan-level late fee assessment rate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7.5%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.2%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.1%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCFPB Dec 2025, Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLate fees collected as % of GMV\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.32%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.24%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.18%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCFPB Dec 2025, Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage late fee assessed (US\u003cspan\u003e$\u003c/span\u003e, CPI-adj.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e10.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e9.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCFPB Dec 2025, Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLoan charge-off rate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.44%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.63%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.83%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCFPB Dec 2025, Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eNotes. n.r. = not reported in source. CFPB sample comprises six large pay-in-four lenders: Affirm, Cash App Afterpay, Klarna, PayPal, Sezzle, and Zip. Late fee metrics in\u003c/em\u003e Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e \u003cem\u003ecover the four firms that charged late fees during the survey period. The sample represents approximately 40 percent of the US point-of-sale financing market in 2023. Dollar figures are inflation-adjusted to 2024 dollars where indicated.\u003c/em\u003e\u003c/p\u003e \u003cp\u003eSeveral patterns from Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e inform the policy framing. First, late-fee incidence has fallen, not risen, over the survey period (4.1 percent of loans in 2023, down from 7.5 percent in 2021). Late-fee revenue as a share of gross merchandise volume halved from 0.32 percent to 0.18 percent. The CFPB attributes this decline partly to industry self-correction in response to regulatory scrutiny and partly to tightening underwriting. The Singapore Code of Conduct, in force since October 2022, would be expected to produce similar dampening effects.\u003c/p\u003e \u003cp\u003eSecond, the size of the typical late fee charged is small. The CFPB reports an average late fee assessed of US\u003cspan\u003e$\u003c/span\u003e9.99 in 2023 and an average late fee collected of US\u003cspan\u003e$\u003c/span\u003e5.87 (the gap reflects discretionary waivers and uncollected fees). The Singapore equivalents \u0026mdash; Atome S\u003cspan\u003e$\u003c/span\u003e15, Grab S\u003cspan\u003e$\u003c/span\u003e10-S\u003cspan\u003e$\u003c/span\u003e15, SeaMoney S\u003cspan\u003e$\u003c/span\u003e5 \u0026mdash; are in the same range. The bounded-harm argument has support in absolute fee terms.\u003c/p\u003e \u003cp\u003eWhere the data is more concerning is the user-level distribution of usage.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBNPL borrower characteristics from matched CFPB credit-records data\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFinding\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eValue\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eReference\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShare of US consumers (with credit record) who used BNPL in 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShare of borrowers characterised as heavy users (\u0026gt;\u0026thinsp;1 loan/month avg.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Exec. Summary\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShare of borrowers with simultaneous loans (any firm) in 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e63%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Section V\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShare of borrowers with simultaneous loans across multiple firms\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e33%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Section V\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShare of originations to deep subprime borrowers (FICO 300\u0026ndash;579)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e45%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShare of originations to subprime borrowers (FICO 580\u0026ndash;619)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShare of originations to no-FICO-score borrowers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBNPL default rate among users (2019\u0026ndash;2022 average)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Section V\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCredit-card default rate, same users, same period\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Section V\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAdditional credit card debt of BNPL users vs non-users (matched)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e+US\u003cspan\u003e$\u003c/span\u003e871\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Section VI\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAdditional personal loan debt of BNPL users vs non-users\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e+US\u003cspan\u003e$\u003c/span\u003e453\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Section VI\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBNPL share of total unsecured debt for borrowers aged 18\u0026ndash;24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e28%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Section VI\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBNPL share of total unsecured debt across all age groups\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e17%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Section VI\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAvg. credit card utilisation, BNPL users (2020\u0026ndash;2023)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e60\u0026ndash;66%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Section VI\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAvg. credit card utilisation, never-BNPL users\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e34%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFPB Jan 2025, Section VI\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eNotes. All figures from the CFPB January 2025 study, which uses a matched sample of de-identified BNPL applications and originations from six large pay-in-four lenders linked to credit records from one of the three nationwide consumer reporting companies. The matched-sample design corrects for the fact that BNPL exposure is not generally reported to credit bureaus, allowing the CFPB to measure stacking and to compare BNPL users to credit-record matched non-users. Dollar comparisons in the lower panel are differences in mean balance, controlling for age and FICO score. The default rate reported in row 8 is the share of originated loans charged off as bad debt (typically loans more than 120 days past due in the CFPB definition), not the share of loans on which any late fee was assessed. Late-fee assessment rates are higher (4.1 percent of loans in 2023, per\u003c/em\u003e Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e \u003cem\u003eabove), reflecting the much larger share of loans that incur a single missed instalment without progressing to charge-off.\u003c/em\u003e\u003c/p\u003e \u003cp\u003eThree substantive patterns emerge from Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, all of which bear on the architectural argument.\u003c/p\u003e \u003cp\u003eLoan stacking is the dominant usage pattern, not the exception. Sixty-three percent of US BNPL borrowers held simultaneous loans within 2022. One-third of borrowers held simultaneous loans across multiple firms. The CFPB explicitly notes that these stacked obligations are largely invisible to non-BNPL lenders because BNPL exposure is not generally reported to the three nationwide consumer reporting companies. The Singapore framework addresses this through the Experian-administered BNPL bureau, which shares data among accredited BNPL providers from April 2024 onward.\u003csup\u003e7\u003c/sup\u003e It does not, however, share BNPL data with the Credit Bureau Singapore that banks consult for card and loan underwriting. The intra-BNPL stacking concern is partially addressed; the cross-product invisibility concern is not.\u003c/p\u003e \u003cp\u003eBNPL users are concentrated in lower credit-score bands. In the CFPB matched sample, 45 percent of originations went to deep-subprime borrowers (FICO 300\u0026ndash;579) and a further 16 percent to subprime (580\u0026ndash;619). Borrowers with no FICO score account for 4 percent of originations. These shares correspond to the population that, in Singapore, would not generally qualify for credit cards under MAS Notice 635 (S\u003cspan\u003e$\u003c/span\u003e30,000 income floor, aggregate cap of 12 times monthly income). Singapore-specific borrower distributions are not publicly available, but the structural argument holds: BNPL serves the population the credit-card regime is designed to exclude. The Moneylenders Act regime, by contrast, is designed to serve precisely this population \u0026mdash; with the rate-cap protections that BNPL does not provide.\u003c/p\u003e \u003cp\u003eBNPL users hold materially higher balances on other unsecured credit, even controlling for age and credit score. The CFPB matched-sample analysis finds that BNPL users hold US\u003cspan\u003e$\u003c/span\u003e871 more in credit card debt and US\u003cspan\u003e$\u003c/span\u003e453 more in personal loan debt than non-BNPL users with the same age and FICO score. Average credit card utilisation among BNPL users is 60\u0026ndash;66 percent, against 34 percent among never-BNPL users. The interpretation in deHaan et al. (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and replicated by the CFPB is that BNPL adoption is associated with increasing financial stress on other instruments rather than substituting away from them. Among the youngest age group (18\u0026ndash;24), BNPL accounts for 28 percent of total unsecured debt during borrowing months. The under-21 share of Singapore BNPL users (1.5 percent per MAS February 2026) is lower than this US distribution suggests, but the mechanism is the same.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e5.4 What this implies for Singapore\u003c/h2\u003e \u003cp\u003eThe Singapore-specific evidence is consistent with the US patterns where it can be observed. MAS confirmed in February 2026 that the median Singapore BNPL user with missed payments holds aggregate outstanding below S\u003cspan\u003e$\u003c/span\u003e400, with the vast majority below S\u003cspan\u003e$\u003c/span\u003e1,000.\u003csup\u003e15\u003c/sup\u003e This figure is consistent with the modest typical outstanding the CFPB documents for US BNPL users when measured per-firm. It is also consistent with the small absolute fee level (S\u003cspan\u003e$\u003c/span\u003e5-S\u003cspan\u003e$\u003c/span\u003e15 per missed instalment) that the Singapore providers charge.\u003c/p\u003e \u003cp\u003eWhat the Singapore aggregate disclosures do not yet reach is the user-level distribution of stacking, the share of users with simultaneous plans across providers, or the cumulative annual fee burden on heavy users. The CFPB borrower-level patterns suggest these distributions are highly concentrated: 20 percent of US borrowers are heavy users (more than one loan per month), 33 percent stack across firms, and the borrower population skews to subprime. If similar patterns hold in Singapore, the architectural finding from Section \u003cspan refid=\"Sec11\" class=\"InternalRef\"\u003e4\u003c/span\u003e has its strongest welfare implications precisely for the heavy-user subset, not for the median user.\u003c/p\u003e \u003cp\u003eTwo pieces of Singapore-specific evidence would resolve the open empirical questions. The first is publication of aggregate user-level statistics from the Experian-administered BNPL bureau, which holds the necessary data. The MLCB and CBS publish equivalent statistics for licensed moneylenders and credit cards respectively; the BNPL bureau does not. The second is publication of per-provider Singapore-specific revenue breakdowns by source (merchant discount fees versus consumer late fees), which currently are reported only at the regional or group level. Both releases would convert the present analysis from a structural argument supplemented by international evidence into a directly empirical Singapore study.\u003c/p\u003e \u003c/div\u003e"},{"header":"6. Counter-Arguments","content":"\u003cp\u003eThree counter-arguments are common in policy discussion of BNPL regulation. I address each in turn.\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section2\"\u003e \u003ch2\u003e6.1 \u0026ldquo;BNPL is interest-free\u0026rdquo;\u003c/h2\u003e \u003cp\u003eThe accuracy of this claim is conditional on no late payment. The published Atome, Grab, and SeaMoney terms charge no interest if the borrower pays each instalment on the due date. The relevant question is whether this conditional claim represents the realised cost across the user distribution.\u003c/p\u003e \u003cp\u003eInternational evidence indicates that a non-trivial fraction of users incur late fees. The CFPB (2022) reports late-fee revenue at around 13 percent of BNPL provider revenues. LendingTree (2026) reports that 41 percent of US BNPL users paid late at least once in the previous year, up from 34 percent the previous year, and that 25 percent used BNPL to buy groceries, up from 14 percent the year prior. I do not have comparable Singapore figures, but the structural drivers (multiple due dates per month across stacked plans, automatic deduction failures from insufficient funds) operate identically. The \u0026ldquo;interest-free\u0026rdquo; framing describes the modal contract, not the realised cost across users.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003e6.2 \u0026ldquo;Tickets are small, harm is bounded\u0026rdquo;\u003c/h2\u003e \u003cp\u003eMAS confirmed in February 2026 that BNPL accounts for less than 2 percent of the value of total credit and debit card payments in Singapore, and that the median user with missed payments has outstanding amounts below S\u003cspan\u003e$\u003c/span\u003e400 across all providers, with the vast majority below S\u003cspan\u003e$\u003c/span\u003e1,000.\u003csup\u003e15\u003c/sup\u003e The aggregate market is therefore small, and the bounded-harm argument has some force at the systemic level. The Singapore BNPL market does not pose financial stability risk.\u003c/p\u003e \u003cp\u003eThe harm calculation at the user level is different. If the median user with a missed payment has an aggregate outstanding balance below S\u003cspan\u003e$\u003c/span\u003e400, a single S\u003cspan\u003e$\u003c/span\u003e15 BNPL late fee represents close to 4 percent of that user\u0026rsquo;s total BNPL exposure in a single month. For a user whose cash flow is constrained enough to miss a payment, a 4 percent hit to their outstanding balance is material. Repeat occurrences within the year compound this. The economic logic is similar to a user who carries a S\u003cspan\u003e$\u003c/span\u003e400 credit card balance and incurs late fees and interest of comparable percentage at an effective rate that statutory rate caps would otherwise prevent.\u003c/p\u003e \u003cp\u003eThe harm is also not bounded relative to the user\u0026rsquo;s cash flow more broadly. A user living paycheck-to-paycheck who pays S\u003cspan\u003e$\u003c/span\u003e15 on a S\u003cspan\u003e$\u003c/span\u003e100 missed instalment is paying 15 percent of the missed amount as a fee. For a user with sufficient cash flow to absorb the fee, the absolute amount is small. For a user without that cushion, the fee can trigger further missed payments, including on rent, utilities, or other obligations. The bounded-harm argument requires assumptions about user cash flow that are stronger than the BNPL marketing material acknowledges. The aggregate-share statistic is the wrong instrument for measuring this: a small share of total payments can still translate into significant per-user cost concentration in a small subset of financially fragile users.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec25\" class=\"Section2\"\u003e \u003ch2\u003e6.3 \u0026ldquo;Industry self-regulation works\u0026rdquo;\u003c/h2\u003e \u003cp\u003eThe April 2024 trustmark accreditation demonstrated that the SFA Code can be enforced as an industry standard, with PwC conducting independent audits of the four accredited firms. The Code does several things well: it caps total BNPL outstanding at S\u003cspan\u003e$\u003c/span\u003e2,000, requires fee disclosure, prohibits compounding, and requires hardship assistance. These are real provisions, and the Code\u0026rsquo;s implementation should be credited.\u003c/p\u003e \u003cp\u003eWhat the Code does not do is impose a rate cap or a structurally analogous discipline. The S\u003cspan\u003e$\u003c/span\u003e2,000 outstanding cap is a quantity constraint on the loan size. It does not constrain the price (the implied rate) on overdue amounts. The Moneylenders Act does both: a quantity cap on borrowing relative to income, and a price cap on interest and fees. The Singapore regulatory tradition for consumer credit serving comparable populations is to cap both quantity and price. The Code of Conduct adopts only the first half. The argument that industry self-regulation is sufficient is therefore conditional on the view that price discipline is not necessary, which is precisely the point at issue.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section2\"\u003e \u003ch2\u003e6.4 \u0026ldquo;Late fees are not the main revenue source\u0026rdquo;\u003c/h2\u003e \u003cp\u003eBNPL providers earn most of their revenue from merchant discount fees, not from consumer late fees. Industry estimates put the merchant discount rate at 3 to 8 percent of the purchase price, well above the 1.5 to 3 percent typical of credit card networks (CFPB, 2022; Berg et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Berg et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) document this as the central feature of the BNPL business model: merchants pay for expanded sales among low-creditworthiness customers.\u003c/p\u003e \u003cp\u003eI accept this point but it does not displace the late-fee analysis. The fact that merchant fees dominate provider revenue means that late fees are not designed to be a profit centre, but they remain a real cost to the subset of users who incur them. The relevant question for consumer protection is not what fraction of provider revenue late fees represent in aggregate, but what cost they impose on the affected users at the margin. The architectural argument in this paper concerns the latter: even if late fees are a small share of provider revenue, the rate they imply on small-ticket overdue amounts is structurally disconnected from the rate-cap framework that governs all other Singapore consumer credit. Acknowledging the merchant-rebate revenue model strengthens, rather than weakens, the case that BNPL late fees should not be designed primarily as profit-maximising charges. They could be calibrated as cost-recovery charges on a percentage basis without disrupting provider economics.\u003c/p\u003e \u003c/div\u003e"},{"header":"7. A Rate-Cap Extension","content":"\u003cp\u003eThe natural policy implication is to extend Moneylenders Act rate-cap discipline to BNPL fee structures. I sketch the principal elements below, noting that the design space is large and the proposal is illustrative rather than fully specified.\u003c/p\u003e \u003cdiv id=\"Sec28\" class=\"Section2\"\u003e \u003ch2\u003e7.1 Re-architect late fees as percentage caps with a flat-fee floor\u003c/h2\u003e \u003cp\u003eThe natural-sounding proposal is to replace flat-dollar late fees with the Moneylenders Act 4 percent rate cap. Mechanically applied, this would impose late fees of S\u003cspan\u003e$\u003c/span\u003e0.80 on a S\u003cspan\u003e$\u003c/span\u003e20 missed instalment and S\u003cspan\u003e$\u003c/span\u003e4.00 on a S\u003cspan\u003e$\u003c/span\u003e100 missed instalment. At those levels, the deterrence and cost-recovery functions of late fees collapse: a missed payment becomes effectively costless, and the BNPL business model that relies on partial fee revenue from delinquent users would not survive the transition. The naive proposal is unworkable. A workable proposal needs more structure.\u003c/p\u003e \u003cp\u003eThe right design is two-component: a percentage rate cap on the rolling cost of credit, plus a capped flat administrative fee that reflects the genuine collection cost a provider incurs on a missed payment. The Moneylenders Act regime itself is built this way. The 4 percent monthly rate caps interest on the overdue principal, and the S\u003cspan\u003e$\u003c/span\u003e60 per-month late fee permits a flat administrative charge above the rate cap, recognising that overdue collection has fixed costs. A BNPL-adapted version would set the rate cap at a level calibrated to the per-incident absolute fees Singapore providers currently charge (S\u003cspan\u003e$\u003c/span\u003e5-S\u003cspan\u003e$\u003c/span\u003e15) but would impose the percentage discipline on small tickets that currently produce 75 percent monthly rates.\u003c/p\u003e \u003cp\u003eA specific worked specification illustrates the principle. Suppose the rule is: a flat administrative fee of up to S\u003cspan\u003e$\u003c/span\u003e15 per missed instalment, plus an additional rate-based component capped at 4 percent of the overdue amount per month, with the total of both components not exceeding 30 percent of the missed instalment in any single month. On a S\u003cspan\u003e$\u003c/span\u003e100 missed instalment, the cap would be the lesser of (S\u003cspan\u003e$\u003c/span\u003e15\u0026thinsp;+\u0026thinsp;S\u003cspan\u003e$\u003c/span\u003e4)\u0026thinsp;=\u0026thinsp;S\u003cspan\u003e$\u003c/span\u003e19 or (30% \u0026times; S\u003cspan\u003e$\u003c/span\u003e100)\u0026thinsp;=\u0026thinsp;S\u003cspan\u003e$\u003c/span\u003e30; the binding constraint is S\u003cspan\u003e$\u003c/span\u003e19, which is the Atome fee plus a small interest charge. On a S\u003cspan\u003e$\u003c/span\u003e20 missed instalment, the cap would be the lesser of (S\u003cspan\u003e$\u003c/span\u003e15\u0026thinsp;+\u0026thinsp;S\u003cspan\u003e$\u003c/span\u003e0.80)\u0026thinsp;=\u0026thinsp;S\u003cspan\u003e$\u003c/span\u003e15.80 or (30% \u0026times; S\u003cspan\u003e$\u003c/span\u003e20)\u0026thinsp;=\u0026thinsp;S\u003cspan\u003e$\u003c/span\u003e6; the binding constraint is the percentage limit at S\u003cspan\u003e$\u003c/span\u003e6. Provider revenue is preserved on standard tickets where the flat fee dominates. The architectural pathology of S\u003cspan\u003e$\u003c/span\u003e15 fees on S\u003cspan\u003e$\u003c/span\u003e20 tickets is eliminated. The 30 percent monthly rate is high relative to the Moneylenders Act 4 percent cap, but acknowledges that BNPL late fees combine an interest component and a fixed administrative cost. Finer calibration is possible; the principle is that any cap with a percentage component closes the small-ticket arbitrage even if it sets the overall ceiling well above the moneylender benchmark.\u003c/p\u003e \u003cp\u003eA simpler alternative is a percentage-only cap with a more generous rate, say 10 percent monthly on the overdue amount, set at a level designed to preserve provider economics on the modal ticket while eliminating the small-ticket distortion. The key design principle in either variant is that the cap should bind on the rate, not on the absolute fee, so that effective costs do not vary by an order of magnitude across ticket sizes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec29\" class=\"Section2\"\u003e \u003ch2\u003e7.2 Apply the principal-ceiling rule\u003c/h2\u003e \u003cp\u003eThe Moneylenders Act\u0026rsquo;s rule that cumulative charges cannot exceed principal is a robust backstop. \u003csup\u003e13\u003c/sup\u003e For BNPL, the analogous rule would be that cumulative late fees and interest on a transaction cannot exceed the transaction value. Atome\u0026rsquo;s current S\u003cspan\u003e$\u003c/span\u003e60 per-transaction admin fee cap operates similarly, but only applies on a single transaction. The principal-ceiling rule is more general: it applies to the loan as a unit, regardless of how the lender structures fees. Extending this rule to BNPL would prevent the corner cases where stacked late fees exceed the transaction value.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec30\" class=\"Section2\"\u003e \u003ch2\u003e7.3 Mandate full credit bureau reporting\u003c/h2\u003e \u003cp\u003eBNPL exposure is currently reported only within the Experian-administered BNPL bureau, not into the Credit Bureau Singapore that banks consult for card and loan underwriting. MAS has confirmed that financial institutions are not required to include outstanding BNPL amounts when computing the Total Debt Servicing Ratio.\u003csup\u003e16\u003c/sup\u003e A borrower can therefore hold simultaneous BNPL plans up to S\u003cspan\u003e$\u003c/span\u003e2,000 across providers and have those exposures invisible to a bank making credit decisions on a separate card or loan. The asymmetry is a quiet form of off-balance-sheet leverage at the household level.\u003c/p\u003e \u003cp\u003eMandating BNPL reporting into CBS would close this gap. It would also bring BNPL into alignment with the moneylender regime, which feeds the MLCB, and with the credit card regime, which feeds CBS. The argument for keeping BNPL reporting siloed rests on the legal classification of BNPL as a deferred payment arrangement rather than a loan. This classification is formalistic: it follows the contractual structure rather than the economic substance. Deferred payment for consumption financed by a third party that bears the credit risk is functionally a loan, regardless of how the contract is named. The Singapore regulatory tradition for credit broadly tracks economic substance: hire-purchase agreements are regulated under their own statute, even though they share the deferred-payment label that BNPL also uses. The case for treating BNPL as a payment product for regulatory purposes is therefore weak relative to its credit characteristics. Bringing BNPL into CBS reporting is the natural first step in aligning the regulatory perimeter with economic substance.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec31\" class=\"Section2\"\u003e \u003ch2\u003e7.4 Income verification at higher exposure thresholds\u003c/h2\u003e \u003cp\u003eThe S\u003cspan\u003e$\u003c/span\u003e2,000 outstanding cap currently applies per provider, not aggregated across providers. A user can accumulate S\u003cspan\u003e$\u003c/span\u003e2,000 with Atome, Grab, and SeaMoney concurrently without any provider triggering the additional credit assessment that the Code requires above the cap. The Experian-administered BNPL bureau in principle enables cross-provider visibility, but the Code does not currently mandate that the assessment threshold bind on aggregate exposure. A statutory regime would either harmonise the cap on aggregate exposure (the analogue of the MAS Notice 635 12x-monthly-income aggregate ceiling) or specify a structured higher cap with income verification (the analogue of the moneylender tiered cap). The principle that quantity caps should bind across the regime in proportion to income is well-established in Singapore consumer credit regulation. There is no obvious reason BNPL should be exempt.\u003c/p\u003e \u003c/div\u003e"},{"header":"8. Limitations and Scope","content":"\u003cp\u003eFive limitations of the present analysis are worth stating explicitly.\u003c/p\u003e \u003cp\u003eFirst, the cost grid models a single-incident scenario: one missed instalment for one month. Cumulative late fees across multiple missed instalments, multiple transactions, or multiple providers are not modelled. Atome\u0026rsquo;s S\u003cspan\u003e$\u003c/span\u003e60 per-transaction admin fee cap implies that within a single transaction, cumulative fees are bounded; across transactions and across providers there is no equivalent ceiling. The Moneylenders Act\u0026rsquo;s principal-ceiling rule applies to the loan as a unit and provides a more comprehensive backstop. A cumulative-fee analysis would strengthen the architectural argument but requires user-level data that is not publicly available.\u003c/p\u003e \u003cp\u003eSecond, the effective rates reported are sensitive to the choice of denominator. I use the missed instalment (the deferred amount), following the CFPB convention. An alternative is to use the full transaction value, which produces lower effective rates because the fee is amortised over the entire purchase rather than the deferred fraction. The transaction-value framing is the provider-friendly alternative; it understates the cost of credit because most of the transaction value has already been paid at the point of late incidence. The Excel companion model reports both denominators.\u003c/p\u003e \u003cp\u003eThird, the analysis treats the Moneylenders Act per-month maximum as the relevant comparison benchmark. An alternative is to compare BNPL fees to the cumulative-charge ceiling, which equals 100 percent of principal across the loan life. On a one-month horizon the cumulative ceiling does not bind. On a multi-month horizon with repeated late fees, BNPL\u0026rsquo;s lack of an equivalent cumulative ceiling becomes more consequential. The single-month framing in Section \u003cspan refid=\"Sec11\" class=\"InternalRef\"\u003e4\u003c/span\u003e understates this dimension of the asymmetry.\u003c/p\u003e \u003cp\u003eFourth, provider fee schedules are subject to revision. The schedules used here are dated as of 30 April 2026 and were verified against current published terms. The April 2024 trustmark accreditation prompted some fee restructuring across providers; further revision is possible. The architectural argument is robust to fee-level revisions because it concerns the choice of cap structure rather than the level of any specific fee. The numerical illustrations would need updating if fee schedules change materially.\u003c/p\u003e \u003cp\u003eFifth, behavioural channels are not modelled. The international literature documents that BNPL increases spending (deHaan et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Berg et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and that users disproportionately stack plans across providers. These behavioural channels likely amplify the welfare cost of BNPL relative to the static fee analysis here. The present paper deliberately limits its claims to the statutory architecture; the behavioural extension is a separate research question and would require Singapore-specific user-level data.\u003c/p\u003e \u003cp\u003eThese limitations do not affect the central finding. The rate-cap architecture for licensed moneylending exists, BNPL operates outside it, and the flat-fee structure produces effective rates that the rate-cap structure would not permit. The limitations affect the magnitude and policy specificity of the recommendations rather than the qualitative finding itself.\u003c/p\u003e \u003cp\u003eThe most actionable gap is data. The Singapore FinTech Association, through the Experian-administered BNPL credit bureau, holds user-level transaction and late-fee data sufficient to answer the empirical questions the present paper has to leave open: the distribution of users by ticket size, the incidence rate of late fees, the cumulative late fees per user per year, and the extent of stacking across providers. The Moneylenders Credit Bureau publishes monthly aggregate statistics on borrower counts, loan totals, and overdue exposure. The Credit Bureau Singapore publishes regular indicators on credit card balances and delinquencies. The BNPL bureau, by contrast, does not currently publish equivalent aggregate data. A monthly aggregate release analogous to MLCB and CBS publications would advance the regulatory discussion considerably. In the absence of such data, the architectural argument has to stand on the regulatory analysis alone.\u003c/p\u003e"},{"header":"9. Conclusion","content":"\u003cp\u003eSingapore has built a coherent statutory architecture for consumer credit. Credit cards are heavily underwritten and reported. Licensed moneylenders are heavily rate-capped and reported. Buy-Now-Pay-Later is neither. The 2022 Code of Conduct provides quantity caps and disclosure rules but does not impose rate-cap discipline on per-transaction fee structures. The result is that BNPL late fees, while modest in absolute terms, generate effective monthly rates that vary from 1 percent to 75 percent across ticket sizes, exceeding the 4 percent moneylender rate cap on transactions below approximately S\u003cspan\u003e$\u003c/span\u003e1,000.\u003c/p\u003e \u003cp\u003eThe contribution of this paper is to make the comparison explicit and to identify the architectural rather than the absolute-fee nature of the asymmetry. The point is not that BNPL providers charge predatory fees: per-incident, they charge less than what a licensed moneylender could lawfully charge for the same overdue amount. The point is that BNPL operates outside the rate-cap regulatory framework that governs every other consumer credit instrument in the jurisdiction, and that the flat-fee structure produces precisely the rate variability that statutory rate caps were designed to prevent.\u003c/p\u003e \u003cp\u003eI do not argue that BNPL should be brought under the Moneylenders Act. That would be a category error: BNPL is a payment product with credit characteristics, not a moneylending product. I argue that the regulatory perimeter for BNPL should incorporate the rate-cap discipline that Singapore already applies to comparable credit, with adaptation for the specifics of the BNPL business model. The 2022 Code of Conduct is a useful first step. The next step is statutory backing and rate-cap architecture.\u003c/p\u003e \u003cp\u003eThe questions left open by this paper are empirical. What share of Singapore BNPL users incur late fees in a given year? How concentrated is BNPL exposure in the small-ticket range? How does cross-provider stacking interact with the per-provider S\u003cspan\u003e$\u003c/span\u003e2,000 outstanding cap? The Experian BNPL bureau holds the data needed to answer these questions. Public release of aggregate statistics, analogous to those that the MLCB and CBS produce, would advance the policy discussion considerably. In the absence of such data, the architectural argument has to stand on its own.\u003c/p\u003e \u003cp\u003eThe Singapore regulatory tradition is conservative for good reason. Rate caps and quantity caps are blunt instruments, but they have served the jurisdiction well in keeping consumer credit costs predictable across business cycles and borrower segments. Extending that tradition to BNPL is not radical; it is consistent with how Singapore has chosen to regulate every other consumer credit instrument since 2008.\u003c/p\u003e"},{"header":"Declarations","content":" \u003cp\u003e \u003cb\u003eReplication.\u003c/b\u003e All effective-rate calculations in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e (Section \u003cspan refid=\"Sec14\" class=\"InternalRef\"\u003e4.3\u003c/span\u003e) follow directly from the inputs in Tables \u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003eA1\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Tab8\" class=\"InternalRef\"\u003eA3\u003c/span\u003e and the formulas given in Section \u003cspan refid=\"Sec12\" class=\"InternalRef\"\u003e4.1\u003c/span\u003e. Effective monthly rate\u0026thinsp;=\u0026thinsp;late fee / missed instalment. Simple APR\u0026thinsp;=\u0026thinsp;monthly rate \u0026times; 12. Effective annual rate = (1\u0026thinsp;+\u0026thinsp;monthly rate)\u0026sup1;\u0026sup2; \u0026minus; 1. The Moneylenders Act per-month maximum is computed as (missed instalment \u0026times; 0.04)\u0026thinsp;+\u0026thinsp;S\u003cspan\u003e$\u003c/span\u003e60. No further inputs are required to reproduce the cost grid.\u003c/p\u003e \u003cp\u003e \u003ch2\u003eCompeting interests.\u003c/h2\u003e \u003cp\u003eThe author has no competing interests to declare that are relevant to the content of this article.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding.\u003c/h2\u003e \u003cp\u003eThe author received no specific funding, grant, or financial support from any agency in the public, commercial, or not-for-profit sectors for the research, authorship, or publication of this article.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eB.C.L. is the sole author. B.C.L. conceptualised and designed the study, conducted the regulatory and effective-cost analysis, performed all calculations, and wrote and revised the manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eAll data used in this study are publicly available secondary sources. No new data were generated or collected by the author.The primary inputs are: (1) the late-fee schedules of three Singapore BNPL providers (Atome, Grab PayLater, SeaMoney SPayLater), retrieved from the providers' published Help Centre and Terms of Service pages and verified on 30 April 2026; (2) statutory parameters from the Singapore Moneylenders Act 2008 and Moneylenders Rules 2015, available from the Ministry of Law (https://rom.mlaw.gov.sg/); (3) MAS Notice 635 unsecured credit rules and the MAS February 2026 parliamentary reply on BNPL transactions (https://www.mas.gov.sg/); (4) the BNPL Code of Conduct, available from the Singapore FinTech Association (https://singaporefintech.org/); and (5) aggregate statistics from CFPB reports of January 2025 and December 2025, available at https://files.consumerfinance.gov/. All inputs are compiled with citations in Appendix A of the manuscript.The cost grid in Table 2 is deterministic computation from these inputs using the formulas given in Section 4.1 (effective monthly rate = late fee / missed instalment; simple APR = monthly rate \u0026times; 12; effective annual rate = (1 + monthly rate)^12 - 1). No proprietary or restricted data are used, and no statistical estimation requires replication code.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAtome (2026). Help Centre: What happens when I make a late payment? Retrieved 30 April 2026, from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://help.atome.sg/hc/en-gb/articles/4403114483225\u003c/span\u003e\u003cspan address=\"https://help.atome.sg/hc/en-gb/articles/4403114483225\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAtome Financial (2025, August 26). Atome Financial posts record US\u003cspan\u003e$\u003c/span\u003e236M operating income, marks full-year profit in 2024 [Press release].\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBerg, T., Burg, V., Keil, J., \u0026amp; Puri, M. (2025). The economics of Buy Now, Pay Later: A merchant\u0026rsquo;s perspective. \u003cem\u003eJournal of Financial Economics\u003c/em\u003e, \u003cem\u003e171\u003c/em\u003e, 104093. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jfineco.2025.104093\u003c/span\u003e\u003cspan address=\"10.1016/j.jfineco.2025.104093\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConsumer Financial Protection Bureau. (2022). \u003cem\u003eBuy now, pay later: Market trends and consumer impacts\u003c/em\u003e. CFPB.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConsumer Financial Protection Bureau (2024). Truth in Lending (Regulation Z); Use of digital user accounts to access buy now, pay later loans. 89 Fed. Reg. 47,068 (May 31, 2024). Formally withdrawn 12 May 2025 (90 Fed. Reg. 20084).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConsumer Financial Protection Bureau. (2025a). \u003cem\u003eConsumer use of buy now, pay later and other unsecured debt\u003c/em\u003e. CFPB Office of Research. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://files.consumerfinance.gov/f/documents/cfpb_BNPL_Report_2025_01.pdf\u003c/span\u003e\u003cspan address=\"https://files.consumerfinance.gov/f/documents/cfpb_BNPL_Report_2025_01.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConsumer Financial Protection Bureau. (2025b). \u003cem\u003eThe buy now, pay later market: Data spotlight\u003c/em\u003e. CFPB Office of Markets. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://files.consumerfinance.gov/f/documents/cfpb_bnpl-market-report_2025-12.pdf\u003c/span\u003e\u003cspan address=\"https://files.consumerfinance.gov/f/documents/cfpb_bnpl-market-report_2025-12.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCornelli, G., Gambacorta, L., \u0026amp; Pancotto, L. (2023). \u003cem\u003eDecember). Buy now, pay later: A cross-country analysis\u003c/em\u003e. BIS Quarterly Review.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003edeHaan, E., Kim, J., Lourie, B., \u0026amp; Zhu, C. (2024). Buy now pay (pain?) later. \u003cem\u003eManagement Science\u003c/em\u003e, \u003cem\u003e70\u003c/em\u003e(8), 5586\u0026ndash;5598. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1287/mnsc.2022.03266\u003c/span\u003e\u003cspan address=\"10.1287/mnsc.2022.03266\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrab, H. (2025, November 21). PayLater by Grab Singapore: Terms of service. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.grab.com/sg/terms-policies/paylater/\u003c/span\u003e\u003cspan address=\"https://www.grab.com/sg/terms-policies/paylater/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLendingTree (2026, February). LendingTree BNPL tracker.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMinistry of Law (2026). Registry of Moneylenders: FAQs on borrowing from licensed moneylenders. Retrieved 30 April 2026, from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://rom.mlaw.gov.sg/information-for-borrowers/guide-to-borrowing-from-licensed-moneylenders-english/\u003c/span\u003e\u003cspan address=\"https://rom.mlaw.gov.sg/information-for-borrowers/guide-to-borrowing-from-licensed-moneylenders-english/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMonetary Authority of Singapore (2024). MAS Notice 635: Unsecured credit rules.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMonetary Authority of Singapore. (2025a). \u003cem\u003eFebruary 28). Oral reply to parliamentary question on the rising use of buy now\u003c/em\u003e. pay later schemes.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMonetary Authority of Singapore. (2025b). \u003cem\u003eMonthly statistical bulletin, Table I.17A\u003c/em\u003e. Credit and charge card statistics.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMonetary Authority of Singapore (2026, February 4). Written reply to parliamentary questions on buy now, pay later transactions and the maximum purchase limit for those below 21 years old. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.mas.gov.sg/news/parliamentary-replies/2026/written-reply-to-parliamentary-questions-on-buy-now-pay-later-transactions\u003c/span\u003e\u003cspan address=\"https://www.mas.gov.sg/news/parliamentary-replies/2026/written-reply-to-parliamentary-questions-on-buy-now-pay-later-transactions\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoneylenders Act (2008 (Singapore)). as amended, with Moneylenders Rules 2015 (effective 1 October 2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eResearchAndMarkets (2025). Singapore buy now pay later business and investment opportunities databook \u0026mdash; Q3 2025 update [Industry research report].\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSea, L. (2025, March 4). Form 6-K filed with the U.S. Securities and Exchange Commission, full-year 2024 results.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShopee Singapore (2026). SPayLater terms of service and help centre. Retrieved 30 April 2026, from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://help.shopee.sg/4/article/76485\u003c/span\u003e\u003cspan address=\"https://help.shopee.sg/4/article/76485\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSingapore FinTech Association (2022, October). Buy now, pay later code of conduct. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://singaporefintech.org/news/buy-now-pay-later-bnpl-working-group-launches-bnpl-code-of-conduct-for-singapore/\u003c/span\u003e\u003cspan address=\"https://singaporefintech.org/news/buy-now-pay-later-bnpl-working-group-launches-bnpl-code-of-conduct-for-singapore/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSingapore FinTech Association (2024, April). BNPL trustmark accreditation announcement.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThe Straits Times (2024, March 29). 5 buy now, pay later players to get accredited in April after minor shake-up in industry.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Appendix","content":"\u003cp\u003eAppendix A. Primary-source compilation of regulatory inputs and fee schedules\u003c/p\u003e\u003cp\u003eThis appendix consolidates the verified regulatory parameters and provider fee schedules used throughout the paper. Each input is given with its primary source. Where a source URL is included, it was accessed and verified on 30 April 2026.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tab6\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable A1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMoneylenders Act regime parameters (statutory, MinLaw)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eParameter\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eValue\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAuthority\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMaximum monthly interest rate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4% per month, on outstanding principal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMoneylenders Rules 2015 (eff. 1 Oct 2015)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMaximum monthly late interest rate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4% per month, on overdue amount only\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMoneylenders Rules 2015\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMaximum late fee\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e60 per month of late repayment\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMoneylenders Rules 2015\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMaximum administrative/approval fee\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10% of principal, one-off upfront\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMoneylenders Rules 2015\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal charges ceiling (whole-of-loan)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCumulative charges\u0026thinsp;\u0026le;\u0026thinsp;principal of the loan\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMoneylenders Act 2008, s. 23(1)(b)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUnsecured loan cap (annual income\u0026thinsp;\u0026lt;\u0026thinsp;S\u003cspan\u003e$\u003c/span\u003e20,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e3,000 across all licensed moneylenders\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMoneylenders Rules\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUnsecured loan cap (annual income\u0026thinsp;\u0026ge;\u0026thinsp;S\u003cspan\u003e$\u003c/span\u003e20,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6 \u0026times; monthly income across all moneylenders\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMoneylenders Rules\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStatutory regulator\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMinLaw / Registry of Moneylenders\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMoneylenders Act 2008\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCredit reporting\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMLCB (Moneylenders Credit Bureau, since 2016)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMinLaw; CBS (operator since 2021)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eNotes. Source: Ministry of Law, Registry of Moneylenders, FAQs on Borrowing From Licensed Moneylenders, last updated 7 April 2026, available at\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://rom.mlaw.gov.sg/information-for-borrowers/guide-to-borrowing-from-licensed-moneylenders-english/\u003c/span\u003e\u003cspan address=\"https://rom.mlaw.gov.sg/information-for-borrowers/guide-to-borrowing-from-licensed-moneylenders-english/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tab7\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable A2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBNPL Code of Conduct parameters (industry self-regulation, SFA)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eParameter\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eValue\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAuthority\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOutstanding cap (no additional credit assessment)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e2,000 per accredited BNPL provider; no aggregation rule across providers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBNPL Code of Conduct, Section \u003cspan refid=\"Sec11\" class=\"InternalRef\"\u003e4\u003c/span\u003e (SFA, October 2022)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStatutory rate cap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNo statute applies\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCompounding of fees\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eProhibited\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBNPL Code of Conduct\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMAS direct regulation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo (MAS guidance only)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGrab T\u0026amp;C, 21 Nov 2025; MAS Parl. Reply, Feb 2025\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIncome verification at sign-up\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAtome / Grab / SeaMoney sign-up flows\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAccount suspension on default\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMandatory\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBNPL Code of Conduct\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCompounding interest on overdue amounts\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eProhibited\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBNPL Code of Conduct\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCredit reporting\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIntra-BNPL bureau only (Experian, since April 2024)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSFA / Experian; not visible to CBS\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStatutory regulator\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNone (industry-led)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSFA accreditation; PwC independent audits\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSanction for breach\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLoss of trustmark accreditation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSFA Code\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eNotes. Sources: Singapore FinTech Association, BNPL Code of Conduct, October 2022\u003c/em\u003e, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://singaporefintech.org/news/buy-now-pay-later-bnpl-working-group-launches-bnpl-code-of-conduct-for-singapore/\u003c/span\u003e\u003cspan address=\"https://singaporefintech.org/news/buy-now-pay-later-bnpl-working-group-launches-bnpl-code-of-conduct-for-singapore/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e; \u003cem\u003etrustmark accreditation announcement April 2024\u003c/em\u003e, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://fintechnews.sg/95130/lending/abnk-atome-grab-and-seamoney-earn-singapore-bnpl-compliance-trustmark/\u003c/span\u003e\u003cspan address=\"https://fintechnews.sg/95130/lending/abnk-atome-grab-and-seamoney-earn-singapore-bnpl-compliance-trustmark/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e; \u003cem\u003eMAS Parliamentary Replies, 28 February 2025 and 4 February 2026.\u003c/em\u003e\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tab8\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable A3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eProvider late fee schedules (verified from primary sources)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProvider / plan\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLate fee structure\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePer-transaction cap\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEffective date\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSource\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAtome (orders\u0026thinsp;\u0026lt;\u0026thinsp;S\u003cspan\u003e$\u003c/span\u003e1,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e15 per missed instalment\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e60 per transaction\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1 May 2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ehelp.atome.sg/hc/en-gb/articles/4403114483225\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAtome (orders\u0026thinsp;\u0026ge;\u0026thinsp;S\u003cspan\u003e$\u003c/span\u003e1,000)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e30 per missed instalment\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e60 per transaction\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1 May 2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ehelp.atome.sg/hc/en-gb/articles/4403114483225\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrab PayLater (Pay Next Month)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e15 per missed payment\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNo published cap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1 April 2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003egrab.com/sg/consumer/finance/paylater/\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrab PayLater Instalments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e10 per missed instalment\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e30 per transaction\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCurrent\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003egrab.com/sg/terms-policies/paylater/\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSeaMoney SPayLater\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e5 per overdue monthly bill\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNo published cap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCurrent\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ehelp.shopee.sg/4/article/76485\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eNotes. All fee schedules accessed and verified on 30 April 2026 against the providers\u0026rsquo; published Help Centre or Terms of Service pages. Where multiple plan structures exist, only the consumer-facing default plans are reported. Atome does not currently distinguish between online and offline merchant fees. Grab PayLater Instalments and Pay Next Month are separate products with distinct fee schedules. SeaMoney SPayLater is operated by Monee (Credit) Singapore Private Limited.\u003c/em\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tab9\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable A4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSingapore credit-card regulatory parameters (MAS, for comparison)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eParameter\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eValue\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAuthority\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMinimum annual income (citizens / PRs)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e30,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMAS Notice 635\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMinimum annual income (foreigners)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eS\u003cspan\u003e$\u003c/span\u003e40,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMAS Notice 635\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAggregate unsecured borrowing cap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12 \u0026times; monthly income across all FIs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMAS Notice 635\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStatutory interest rate cap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNone (subject to MAS conduct rules)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMAS Notice 635\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBureau reporting\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFull reporting to Credit Bureau Singapore (CBS)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCBS Act\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal Debt Servicing Ratio inclusion\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eYes (BNPL is excluded; cards are included)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMAS Parliamentary Reply, 28 Feb 2025\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStatutory regulator\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMAS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBanking Act\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSanction for breach\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLicence action by MAS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMAS rule-making powers\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eNotes. Source: Monetary Authority of Singapore, MAS Notice 635 (Unsecured Credit Rules); MAS Parliamentary Replies, 28 February 2025 and 4 February 2026. Card-issuer specific underwriting standards (e.g. higher income thresholds for premium cards) are layered on top of the MAS minimums and are not reported here.\u003c/em\u003e \u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"journal-of-consumer-policy","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"copo","sideBox":"Learn more about [Journal of Consumer Policy](http://link.springer.com/journal/10603)","snPcode":"10603","submissionUrl":"https://submission.springernature.com/new-submission/10603/3","title":"Journal of Consumer Policy","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Buy-Now-Pay-Later, consumer credit regulation, Moneylenders Act, regulatory arbitrage, Singapore, fintech","lastPublishedDoi":"10.21203/rs.3.rs-9581198/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9581198/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSingapore Buy-Now-Pay-Later (BNPL) products operate outside the rate-cap architecture that governs every other consumer credit instrument in the jurisdiction. Licensed moneylenders are capped at 4 percent monthly interest under the Moneylenders Act, and credit card issuers face MAS conduct rules and a 12-times-monthly-income aggregate underwriting limit. BNPL faces neither. I compute the effective monthly rate implied by the late fee schedules of the three accredited Singapore BNPL providers (Atome, Grab PayLater, SeaMoney SPayLater) across seventeen ticket-size scenarios and benchmark these against the Moneylenders Act ceiling. On per-incident absolute fees, BNPL charges fall below the moneylender maximum. On effective rates the picture inverts: BNPL produces monthly rates of 1 to 75 percent on the deferred amount. The same S$15 Atome late fee implies a 75 percent rate on a S$20 missed instalment, 15 percent on S$100, and 4.5 percent on S$333. The statutory rate cap binds uniformly on rate; the BNPL flat fee binds uniformly on dollar amount. I calibrate the analysis to MAS data and to the CFPB 2025 BNPL studies, which show high loan stacking and concentration in subprime borrowers. The structural patterns hold across jurisdictions. BNPL escapes the rate-cap regime through legal classification (deferred payment, not loan) rather than through superior consumer protection. I propose extending Moneylenders Act discipline through a two-component fee structure (capped flat administrative fee plus percentage rate cap) and mandating full Credit Bureau Singapore reporting of BNPL exposure, building on rather than replacing the current industry Code of Conduct.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eJEL Codes: \u003c/strong\u003eG18, G28, G51, K23, D18.\u003c/p\u003e","manuscriptTitle":"The Inverted Regulatory Perimeter in Singapore Buy-Now-Pay-Later","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-16 14:25:55","doi":"10.21203/rs.3.rs-9581198/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-05-14T13:29:16+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"281600350120938357789549447489178577398","date":"2026-05-11T14:26:35+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"277431784634837303073059876116165843561","date":"2026-05-06T13:18:41+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-05-06T13:10:41+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-05-05T17:33:46+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-05-01T07:03:38+00:00","index":"","fulltext":""},{"type":"submitted","content":"Journal of Consumer Policy","date":"2026-04-30T20:52:55+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"journal-of-consumer-policy","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"copo","sideBox":"Learn more about [Journal of Consumer Policy](http://link.springer.com/journal/10603)","snPcode":"10603","submissionUrl":"https://submission.springernature.com/new-submission/10603/3","title":"Journal of Consumer Policy","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"7cabaf0b-aa76-4100-9fb7-38099287aa04","owner":[],"postedDate":"May 16th, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-14T13:29:16+00:00","index":13,"fulltext":""},{"type":"reviewerAgreed","content":"281600350120938357789549447489178577398","date":"2026-05-11T14:26:35+00:00","index":11,"fulltext":""},{"type":"reviewerAgreed","content":"277431784634837303073059876116165843561","date":"2026-05-06T13:18:41+00:00","index":10,"fulltext":""},{"type":"reviewersInvited","content":"4","date":"2026-05-06T13:10:41+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-05-05T17:33:46+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-05-01T07:03:38+00:00","index":"","fulltext":""},{"type":"submitted","content":"Journal of Consumer Policy","date":"2026-04-30T20:52:55+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-16T14:25:55+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-16 14:25:55","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9581198","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9581198","identity":"rs-9581198","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}