# Fintech Concept Card: [Your name, elective or company, and date]

> An elective-level concept card for fintech: how payments, lending and digital asset businesses make money, where
> the risk sits, and which rules govern each activity. It is for MBA students in a fintech, technology strategy or
> financial institutions elective and for candidates preparing for product, strategy or investment roles at payment
> companies, digital lenders, banks and crypto-asset firms. It teaches the unit economics of three business models
> (merchant acquiring, buy now pay later lending and a payment stablecoin issuer), a regulation map organised by
> activity rather than by company, and the questions an investor or regulator asks about each line. Worked example A
> solves the economics of Varnelle Pay, a fictional multi-line fintech; Worked example B is a half-solved credit stress
> case; Exercise C asks you to reprice the payments line for a different market, find the break-even revenue share of
> a stablecoin and draft a regulation map for a launch. Source basis: Rochet and Tirole (2003), Diamond and Dybvig
> (1983), Philippon (2016), Buchak, Matvos, Piskorski and Seru (2018), Fuster, Plosser, Schnabl and Vickery (2019),
> Berg, Burg, Gombovic and Puri (2020) and Gorton and Zhang (2021), written up in original words with fictional
> numbers. The regulation map is a dated table checked on 2026-10-30. Replace every [bracketed] value and every row
> marked *Example*. This is a learning aid, not legal, regulatory, investment or financial advice; check the current
> official text and a qualified professional before you rely on any rule.

<!--
Guidance: How to use this template
- Read "The idea in plain language" and "Formulas and notation" first. Fintech interviews test whether you can say,
  in one line, where a business earns its margin and what could take that margin away.
- Rebuild Worked example A in a blank workbook from the inputs alone, one business line at a time, before you look
  at the solved tables.
- Treat the regulation map as a dated snapshot. Laws in payments and digital assets change every year; open each
  official link and update the "checked" date before you use the map in a memo or a case write-up.
- Fill every [ ] in Worked example B, then do Exercise C without looking at the answer key.
- Every number was computed in Python and is shown to two decimals; Excel formulas sit next to the results so you can
  check your own build.
- Related templates: mba/mba-platform-strategy-card, mba/mba-technology-strategy-guide,
  mba/mba-digital-transformation-card, mba/mba-ai-risk-governance-card, mba/mba-customer-lifetime-value-card,
  mba/mba-pricing-card, mba/mba-fixed-income-card, mba/mba-venture-capital-guide, mba/mba-business-law-guide,
  mba/mba-nonmarket-strategy-card, finance/unit-economics-analysis, finance/credit-memo, finance/ecl-methodology,
  industry/open-banking-api-design, industry/card-acquiring-payments-design, industry/iso20022-payments-design,
  system-design/design-digital-wallet, system-design/design-payment-ledger, system-design/design-fraud-detection.
-->

## Document control

<!-- Guidance: Record when you last checked the regulation map. A fintech analysis with an undated rule table is out of date the day it is written. -->

| Field | Value |
|---|---|
| Learner | [Your name] |
| Elective or company | [Fintech elective, interview preparation, company name] |
| Business lines analysed | [Payments, lending, digital assets, other] |
| Markets in scope | [EU, UK, US, other] |
| Regulation map last checked | YYYY-MM-DD |
| Reviewed by | [Learning team member, teaching assistant or industry mentor] |
| Version | [0.1] |
| Last updated | YYYY-MM-DD |

## Contents

- [When to use it](#when-to-use-it)
- [The idea in plain language](#the-idea-in-plain-language)
- [Formulas and notation](#formulas-and-notation)
- [Assumptions and limits](#assumptions-and-limits)
- [Regulation map by activity (dated)](#regulation-map-by-activity-dated)
- [How investors, operators and regulators use the card](#how-investors-operators-and-regulators-use-the-card)
- [Worked example A (fully solved): Varnelle Pay](#worked-example-a-fully-solved-varnelle-pay)
- [Worked example B (half solved): a credit stress case](#worked-example-b-half-solved-a-credit-stress-case)
- [Exercise C (unsolved): a new market and a launch map](#exercise-c-unsolved-a-new-market-and-a-launch-map)
- [Connections to other cards](#connections-to-other-cards)
- [Common mistakes and misconceptions](#common-mistakes-and-misconceptions)
- [Self-test questions](#self-test-questions)
- [My notes and next steps](#my-notes-and-next-steps)
- [Sources and further reading](#sources-and-further-reading)
- [Answer key](#answer-key)

## When to use it

<!-- Guidance: Use the card whenever a case or interview asks "how does this company make money, and what could stop it". The three models cover most fintech revenue. -->

Use this card when:

- a case, interview or investment memo asks you to explain the revenue and cost of a payment company, a digital
  lender or a crypto-asset business, and to say which line of the income statement is most exposed;
- you must compare a fintech with a bank doing the same activity, and decide whether its advantage comes from
  technology, from a lighter regulatory burden or from cheaper funding;
- a product team plans to launch a payment, credit or digital asset feature in a new market and needs a first map of
  the licences and rules it will meet;
- you are valuing a fintech and need to separate revenue that scales with volume from revenue that depends on interest
  rates or credit conditions, because the two deserve different multiples.

Do not use the card as a compliance checklist. It teaches the economics and the shape of regulation; a launch needs
the current official texts and advice from qualified lawyers in each market.

## The idea in plain language

<!-- Guidance: Fintech is banking activity rebuilt around software. The economics of the activity do not change; only who performs it, at what cost, and under which rules. -->

A bank bundles several activities: it holds deposits, moves money, lends, and manages risk, all inside one licensed
balance sheet. Fintech firms usually unbundle these activities and do one of them with better software, a better
customer experience or a different distribution channel. The useful question is never "is this a fintech?" but
"which banking activity is this firm doing, and how does it earn a margin on it?"

**Payments** earn a small percentage of a very large flow. A merchant acquirer or payment facilitator charges the
merchant a merchant service charge (MSC) on each card payment. Most of that charge is passed on: the interchange fee
goes to the card issuer and scheme fees go to the card network. What is left, the net take rate, must cover
processing, fraud and chargeback losses, and customer acquisition. Payment networks are two-sided markets in the
sense of Rochet and Tirole (2003): the network sets a price structure across merchants and cardholders, and
interchange is the tool that shifts cost from one side to the other. That is why interchange is regulated in many
markets. Payments businesses scale well because cost per transaction falls with volume, but the net take rate is
under constant pressure from competition and regulation.

**Lending** earns a spread and takes credit risk. A digital lender, including a buy now pay later (BNPL) provider,
earns interest, fees or a merchant discount; it pays for funding and loses money when borrowers default. The
lender's edge is usually in underwriting and distribution: Berg, Burg, Gombovic and Puri (2020) show that simple
digital footprint variables (such as device type and email provider) predict default about as well as a credit
bureau score in their sample, and Fuster, Plosser, Schnabl and Vickery (2019) find that technology-based mortgage
lenders process applications faster without higher default rates. The risk is that a lender funded by warehouse
lines or securitisation, not by deposits, faces a funding squeeze exactly when losses rise. Buchak, Matvos, Piskorski
and Seru (2018) attribute much of the growth of shadow-bank and fintech mortgage lending to regulatory differences
and part to technology, which is a reminder that a cost advantage can be a regulatory gap that closes.

**Digital assets** cover crypto-asset trading, custody and payment stablecoins. A payment stablecoin issuer takes
money from holders, invests it in safe short-term assets and earns the yield on those reserves, while promising to
redeem each token at par. The model is old: Gorton and Zhang (2021) compare stablecoins with the private bank notes of
the US free banking era and argue that par redemption is only credible if reserves are safe and liquid. The economics
are simple and highly sensitive to interest rates, because holders are not paid interest under the main regimes
(see the regulation map) and the issuer keeps the yield, often sharing it with the platforms that distribute the
token. A run is the main risk: if holders doubt the reserves, they all redeem at once, the same mechanism Diamond and
Dybvig (1983) describe for banks.

The common thread is that each activity has a natural owner of risk and a natural regulator. Philippon (2016) notes
that the unit cost of financial intermediation in the US stayed roughly flat for a century despite advances in
technology, and argues that fintech entry and regulation that favours competition could lower it. Regulation follows
the activity: a firm that holds client money, lends to consumers or issues a token meets the rules for that activity,
whatever it calls itself.

## Formulas and notation

<!-- Guidance: Six relations cover the three business models. Learn which input each line is most sensitive to: volume for payments, losses and funding for lending, rates for stablecoins. -->

Payments: net revenue and contribution on total payment volume (TPV):

$$
NR = TPV \cdot (m - ic - sf) \qquad
C_{pay} = NR - c_{tx} \cdot N - \lambda \cdot TPV \qquad N = \frac{TPV}{\bar{t}}
$$

Net take rate as a share of volume, and the break-even merchant service charge:

$$
\text{net take rate} = \frac{C_{pay}}{TPV} \qquad m^{*} = ic + sf + \frac{c_{tx} \cdot N}{TPV} + \lambda
$$

Lending (short-term instalment credit funded by debt): average receivables, contribution and the break-even loss rate:

$$
\bar{R} = G \cdot \frac{w}{52} \qquad
C_{lend} = f \cdot G - r_f \cdot \bar{R} - \ell \cdot G - s \cdot G \qquad
\ell^{*} = \frac{f \cdot G - r_f \cdot \bar{R} - s \cdot G}{G}
$$

Expected credit loss for a single exposure, the building block behind $\ell$:

$$
EL = PD \cdot LGD \cdot EAD
$$

Payment stablecoin issuer profit, with outstanding tokens equal to reserves $R$:

$$
\Pi = (1 - \sigma) \cdot y \cdot R - K \qquad y^{*} = \frac{K}{(1 - \sigma) \cdot R}
$$

| Symbol | Meaning | Note |
|---|---|---|
| $TPV$, $N$, $\bar{t}$ | Total payment volume, number of transactions, average ticket | Volume drives most payment revenue |
| $m$ | Merchant service charge, as a share of volume | Blended across card types and merchant sizes |
| $ic$, $sf$ | Interchange fee and scheme fees, as a share of volume | Passed through to issuers and networks |
| $c_{tx}$ | Processing cost per transaction | Falls with scale |
| $\lambda$ | Fraud and chargeback losses, as a share of volume | Rises with card-not-present and new merchants |
| $NR$, $C_{pay}$ | Net revenue and contribution of the payments line | Before customer acquisition and overheads |
| $G$, $w$ | Credit volume originated a year and average weeks outstanding | Pay-in-four over six weeks gives $w = 6$ |
| $\bar{R}$ | Average receivables on the balance sheet | Turns per year $= G / \bar{R}$ |
| $f$ | Merchant fee or other revenue, as a share of volume | Consumer interest would be a second revenue line |
| $r_f$ | Cost of funds on receivables | Warehouse, securitisation or deposit funding |
| $\ell$, $s$ | Credit losses and servicing cost, as a share of volume | $\ell$ is net of recoveries |
| $PD$, $LGD$, $EAD$ | Probability of default, loss given default, exposure at default | See finance/ecl-methodology |
| $R$, $y$ | Reserves (equal to tokens outstanding) and reserve yield | Short-term government assets in the main regimes |
| $\sigma$, $K$ | Share of reserve income paid to distribution partners, fixed operating and compliance cost | Holders receive no interest |

Excel: with named cells, payments contribution is `=TPV*(MSC-IC-SF)-Cost_tx*TPV/Ticket-Fraud*TPV`; BNPL
contribution is `=Fee*GMV-CoF*GMV*Weeks/52-Loss*GMV-Serv*GMV`; break-even loss rate is
`=(Fee*GMV-CoF*GMV*Weeks/52-Serv*GMV)/GMV`; stablecoin profit is `=(1-Share)*Yield*Reserves-Fixed_cost`. Use a data
table (Data, What-If Analysis, Data Table) to show profit against the yield or the loss rate.

## Assumptions and limits

<!-- Guidance: State each simplification in your memo. Most fintech valuation errors come from treating rate-driven or credit-driven income as if it were fee income. -->

- **Contribution, not profit.** Each line is shown before customer acquisition cost, technology development and
  head-office overheads. A fintech can have positive contribution and still lose money; link to
  mba/mba-customer-lifetime-value-card for acquisition economics.
- **Blended rates.** Real merchant pricing varies by merchant size, card type, region and channel. A blended rate is
  enough for a case but hides mix effects: a shift towards credit cards or cross-border cards raises interchange.
- **Interchange as a pass-through.** The card-acquiring model assumes "interchange plus" pricing is not used, so the
  acquirer bears mix risk. Under interchange-plus pricing, the merchant bears it and the acquirer's margin is the
  markup alone.
- **Short-term credit only.** The lending formula fits instalment credit repaid within weeks. Longer loans need
  amortisation schedules, interest income and loss curves by vintage; finance/credit-memo and finance/ecl-methodology
  show the fuller treatment. Accounting provisions (expected credit loss under IFRS 9 or CECL in the US) recognise
  losses earlier than the cash loss rate used here.
- **Funding cost on all receivables.** The example charges the cost of funds on the whole receivable book. In
  practice an equity tranche funds part of it; the return on that equity is the real test of the model.
- **Stablecoin reserves at a single yield.** Reserve income follows short-term rates almost one for one. The yields in
  Worked example A are illustrative assumptions, not forecasts; use the current short-term government rate in your
  own work.
- **No tax and no regulatory capital.** A bank or licensed lender also holds capital against its assets, which lowers
  the return on equity; payment and e-money institutions hold own funds and safeguard client money. Both change the
  comparison between a fintech and a bank.

## Regulation map by activity (dated)

<!-- Guidance: Organise rules by activity, not by company. For each row, open the official link and update the status before you rely on it. Do not copy this table into fixed guidance: it is a dated snapshot. -->

Regulation attaches to what a firm does. The table below lists, for each activity, the main instrument in the EU,
the UK and the US as examples, with the status found on the official page on 2026-10-30. It is a learning map, not a
complete list: anti-money laundering, data protection, consumer protection and conduct rules apply on top of every
row, and many countries outside these three have their own regimes. This is not legal advice; check the current
official text and a qualified professional.

| ID | Activity | Jurisdiction | Instrument and status (checked 2026-10-30) | What it means for the business model | Official source |
|---|---|---|---|---|---|
| REG-01 | Payment services and e-money | EU | Payment services are licensed under the second Payment Services Directive (PSD2) and e-money under the E-Money Directive. The proposed Payment Services Regulation and PSD3 (procedures 2023/0210(COD) and 2023/0209(COD)) are not yet law: the Parliament committee approved the negotiated text on 5 May 2026 (early second reading), the Council's first-reading position was pending, and a plenary sitting was listed as indicative for 14 December 2026 | A payment or e-money licence, safeguarding of client funds, strong customer authentication; fraud liability rules may change when the new texts apply | https://oeil.europarl.europa.eu/oeil/en/procedure-file?reference=2023/0210(COD) |
| REG-02 | Instant euro payments | EU | Regulation (EU) 2024/886: euro-area payment service providers must receive instant euro credit transfers by 9 January 2025 and send them by 9 October 2025; providers outside the euro area by 9 January 2027 and 9 July 2027 | Instant transfers at no higher charge than standard transfers, payee verification, a cheaper alternative to cards for some payments | https://publications.europa.eu/resource/celex/32024R0886 |
| REG-03 | Card interchange | EU | Regulation (EU) 2015/751 caps interchange on consumer debit cards at 0.2% and on consumer credit cards at 0.3% of the transaction value; member states may set lower domestic caps | Low interchange keeps merchant fees low, so acquirers compete on processing cost and services | https://publications.europa.eu/resource/celex/32015R0751 |
| REG-04 | Authorised push payment fraud | UK | Mandatory reimbursement of authorised push payment fraud for payments on or after 7 October 2024, up to GBP 85,000 per claim, with an optional excess of up to GBP 100 (not for vulnerable consumers), claims within 13 months | Sending and receiving firms share fraud losses, so fraud controls become a cost line in the payments model | https://www.psr.org.uk/information-for-consumers/app-fraud-reimbursement-protections/ |
| REG-05 | Open banking and data access | US | The CFPB issued its personal financial data rights final rule on 22 October 2024 and an advance notice of proposed rulemaking on reconsideration on 22 August 2025; a federal court stayed the compliance dates on 29 October 2025 (as recorded in industry/open-banking-api-design); the CFPB page did not state a 2026 outcome when checked, so check later court and agency steps directly | Access to bank data underpins account-to-account payments and cash-flow underwriting; the rule's final shape decides who may charge for data | https://www.consumerfinance.gov/personal-financial-data-rights/ |
| REG-06 | Consumer credit, including many BNPL products | EU | Directive (EU) 2023/2225 (second Consumer Credit Directive): national laws adopted by 20 November 2025 and applied from 20 November 2026; it applies to credit agreements up to EUR 100,000, with listed exclusions in Article 2 | Creditworthiness assessment, pre-contractual information and other consumer duties extend to many short-term and interest-free products | https://publications.europa.eu/resource/celex/32023L2225 |
| REG-07 | Crypto-asset services and e-money and asset-referenced tokens | EU | Regulation (EU) 2023/1114 (MiCA): the stablecoin titles (III and IV) applied from 30 June 2024 and the rest from 30 December 2024; firms operating under national law could continue until 1 July 2026 at the latest, or until authorisation was granted or refused (Article 143(3)); issuers and service providers may not grant interest on e-money tokens (Article 50) | Authorisation for issuers and service providers, reserve and redemption rules, no interest paid to holders | https://publications.europa.eu/resource/celex/32023R1114 and https://www.esma.europa.eu/esmas-activities/digital-finance-and-innovation/markets-crypto-assets-regulation-mica |
| REG-08 | Payment stablecoins | US | GENIUS Act, Public Law 119-27, approved 18 July 2025: at least one-to-one reserves in listed safe assets (section 4(a)(1)), no interest or yield paid to holders solely for holding the coin (section 4(a)(11)), a state regime option for issuers with up to USD 10 billion outstanding (section 4(c)); effective on the earlier of 18 months after enactment or 120 days after final implementing regulations (section 20) | The issuer earns the reserve yield and cannot share it with holders as interest; distribution partners are paid from it | https://www.govinfo.gov/app/details/PLAW-119publ27 |
| REG-09 | Digital operational resilience | EU | Regulation (EU) 2022/2554 (DORA) has applied since 17 January 2025 to most financial entities, including payment, e-money and crypto-asset service providers | ICT risk management, incident reporting, resilience testing and oversight of critical third-party providers become a fixed cost | https://www.eiopa.europa.eu/digital-operational-resilience-act-dora_en |

```mermaid
flowchart TD
    A["Describe the product in one sentence"] --> B{"Does the firm hold or move client money?"}
    B -->|"Yes"| C["Payment or e-money rules: REG-01 to REG-05"]
    B -->|"No"| D{"Does the firm extend credit to consumers?"}
    C --> D
    D -->|"Yes"| E["Consumer credit rules: REG-06 and local equivalents"]
    D -->|"No"| F{"Does the firm issue, hold or trade crypto-assets or stablecoins?"}
    E --> F
    F -->|"Yes"| G["Crypto-asset and stablecoin rules: REG-07 and REG-08"]
    F -->|"No"| H["Check whether any licence is needed at all"]
    G --> I["Add cross-cutting rules: resilience REG-09, anti-money laundering, data protection, conduct"]
    H --> I
```

Questions to ask for each activity in a launch map: which licence, or which licensed partner, carries the activity;
who holds the client money and how it is safeguarded; who bears fraud and credit losses; which consumer disclosures
apply; and which date in the table could change the economics in the next two years.

## How investors, operators and regulators use the card

<!-- Guidance: Each reader looks at a different line of the same unit economics. Know which line matters to whom. -->

| Reader | Question | What they look at |
|---|---|---|
| Venture or growth investor | Does the margin survive scale and competition? | Net take rate trend, contribution per transaction, share of revenue that depends on interest rates |
| Credit or warehouse lender | Will receivables repay if losses rise? | Loss rate by vintage, break-even loss rate, advance rate and excess spread |
| Product and strategy team | Which product or market to launch next? | Contribution by line, licences needed, time to launch |
| Supervisor | Are client money and consumers protected? | Safeguarding, reserves, redemption, complaints, fraud, operational resilience |
| Bank partner | Does the partnership carry risk the bank cannot see? | Compliance controls, concentration, the fintech's own funding and liquidity |

Fintech valuations moved a lot in the past decade because markets repriced the same revenue once its source became
clear. Fee revenue from payments tends to grow with volume and to be valued like software or processing. Revenue that
depends on short-term rates (float and reserve income) or on credit performance (lending spreads) is valued more like
a bank, because it falls when rates fall or losses rise. A good analysis splits the income statement along these
lines before choosing a multiple; see mba/mba-dcf-valuation-card and mba/mba-venture-capital-guide.

## Worked example A (fully solved): Varnelle Pay

<!-- Guidance: Solve each line on its own first. Then compare the three lines: which one is most exposed to competition, to credit, and to interest rates? -->

Varnelle Pay is a fictional European fintech with three business lines. All figures are *Example* inputs for one
year, in EUR m for payments and lending and USD m for the stablecoin line, which is run by a separate fictional US
issuer, Varnelle Digital.

### Line 1: merchant acquiring (EUR m)

*Example* inputs: TPV 2,000; average ticket EUR 40, so 50 million transactions; blended merchant service charge
1.20%; consumer card mix 70% debit at the 0.2% interchange cap and 30% credit at the 0.3% cap, so blended interchange
0.23%; scheme fees 0.12%; processing cost EUR 0.03 a transaction; fraud and chargeback losses 0.05% of TPV.

| Item | Computation | Result | Excel formula |
|---|---|---|---|
| Transactions (million) | 2,000 / 40 | 50.00 | `=2000/40` |
| Blended interchange | 70% x 0.2% + 30% x 0.3% | 0.23% | `=0.7*0.2%+0.3*0.3%` |
| Net revenue | 2,000 x (1.20% - 0.23% - 0.12%) | 17.00 | `=2000*(1.2%-0.23%-0.12%)` |
| Processing cost | 0.03 x 50 | 1.50 | `=0.03*50` |
| Fraud and chargebacks | 0.05% x 2,000 | 1.00 | `=0.05%*2000` |
| Contribution | 17.00 - 1.50 - 1.00 | 14.50 | `=17-1.5-1` |
| Net take rate | 14.50 / 2,000 | 0.725% | `=14.5/2000` |
| Contribution per transaction (EUR) | 14.50 / 50 | 0.29 | `=14.5/50` |
| Break-even merchant service charge | 0.23% + 0.12% + 1.50 / 2,000 + 0.05% | 0.475% | `=0.23%+0.12%+1.5/2000+0.05%` |

The gross charge of 1.20% looks large, but only 0.725% of volume is left after pass-through costs, processing and
fraud. A competitor that cuts the merchant service charge by 10 basis points takes 2.00 out of contribution
(`=2000*0.1%`), or 14% of it.

### Line 2: buy now pay later (EUR m)

*Example* inputs: merchant-funded pay-in-four product; gross merchandise value financed 500; merchant fee 4.0%; six
weeks outstanding on average; cost of funds 6.0% a year on average receivables; net credit losses 1.0% of volume;
servicing and collections 0.5% of volume; no consumer interest or late fees.

| Item | Computation | Result | Excel formula |
|---|---|---|---|
| Average receivables | 500 x 6 / 52 | 57.69 | `=500*6/52` |
| Receivable turns a year | 500 / 57.69 | 8.67 | `=500/57.69` |
| Merchant fee revenue | 4.0% x 500 | 20.00 | `=4%*500` |
| Funding cost | 6.0% x 57.69 | 3.46 | `=6%*57.69` |
| Credit losses | 1.0% x 500 | 5.00 | `=1%*500` |
| Servicing and collections | 0.5% x 500 | 2.50 | `=0.5%*500` |
| Contribution | 20.00 - 3.46 - 5.00 - 2.50 | 9.04 | `=20-3.46-5-2.5` |
| Contribution as a share of volume | 9.04 / 500 | 1.81% | `=9.04/500` |
| Contribution on average receivables | 9.04 / 57.69 | 15.67% | `=9.04/57.69` |
| Break-even loss rate | (20.00 - 3.46 - 2.50) / 500 | 2.81% | `=(20-3.46-2.5)/500` |

Short duration is the engine of this model: each euro of receivables turns over almost nine times a year, so a 6%
funding rate costs only 0.69% of volume (`=3.46/500`). Credit losses are the largest cost, and the loss rate can rise
to 2.81% before contribution reaches zero. That buffer is the number a warehouse lender asks for first.

### Line 3: payment stablecoin issuer (USD m)

*Example* inputs for Varnelle Digital: tokens outstanding and reserves 1,000; reserves held in short-term government
assets yielding 4.0% a year (an illustrative assumption, not a forecast); 50% of reserve income paid to distribution
partners; fixed operating, attestation and compliance cost 8; no interest paid to holders, as required under REG-07
and REG-08.

| Reserve yield | Gross reserve income | Paid to partners | Issuer profit | Excel formula for profit |
|---|---|---|---|---|
| 1.0% | 10.00 | 5.00 | -3.00 | `=(1-50%)*1%*1000-8` |
| 2.0% | 20.00 | 10.00 | 2.00 | `=(1-50%)*2%*1000-8` |
| 3.0% | 30.00 | 15.00 | 7.00 | `=(1-50%)*3%*1000-8` |
| 4.0% (base) | 40.00 | 20.00 | 12.00 | `=(1-50%)*4%*1000-8` |
| 5.0% | 50.00 | 25.00 | 17.00 | `=(1-50%)*5%*1000-8` |

Break-even yield: `=8/((1-50%)*1000)` = 1.60%. Every percentage point of short-term rates is worth 5.00 of profit to
this issuer, so a fall in rates from 4% to 2% cuts profit by 83% with no change in the business.

```vega-lite
{
  "$schema": "https://vega.github.io/schema/vega-lite/v5.json",
  "description": "Varnelle Digital: issuer profit against reserve yield, USD m, computed example",
  "width": 420,
  "height": 240,
  "data": {
    "values": [
      {"yield": 1.0, "profit": -3.0}, {"yield": 1.5, "profit": -0.5}, {"yield": 2.0, "profit": 2.0},
      {"yield": 2.5, "profit": 4.5}, {"yield": 3.0, "profit": 7.0}, {"yield": 3.5, "profit": 9.5},
      {"yield": 4.0, "profit": 12.0}, {"yield": 4.5, "profit": 14.5}, {"yield": 5.0, "profit": 17.0}
    ]
  },
  "mark": {"type": "line", "point": true},
  "encoding": {
    "x": {"field": "yield", "type": "quantitative", "title": "Reserve yield (%)"},
    "y": {"field": "profit", "type": "quantitative", "title": "Issuer profit (USD m)"}
  }
}
```

### Comparing the three lines

| Line | Contribution | Main driver | Main threat | Bank-like or software-like revenue? |
|---|---|---|---|---|
| Merchant acquiring | EUR 14.50m | Volume and net take rate | Price competition, interchange and fraud rules | Software and processing-like |
| BNPL | EUR 9.04m | Merchant fee and turnover speed | Credit losses and funding cost in a downturn | Lender-like |
| Stablecoin issuer | USD 12.00m | Short-term interest rates | Falling rates, runs, reserve rules | Bank-like (float income) |

The judgement to state: the payments line is the one an investor would pay the highest multiple for, because it is
least exposed to rates and credit; the stablecoin line looks the most profitable per unit of fixed cost today, but it
is a bet on short-term rates and on trust in the reserves.

## Worked example B (half solved): a credit stress case

<!-- Guidance: Change only the loss rate and the cost of funds. Then write the two sentences a warehouse lender would want to read. -->

Same BNPL line as Worked example A. A downturn doubles net credit losses to 2.0% of volume and raises the cost of
funds to 8.0% a year, because the warehouse lender reprices the facility. Volume, merchant fee, duration and servicing
cost are unchanged.

| Item | Computation | Result |
|---|---|---|
| Average receivables | 500 x 6 / 52 | 57.69 |
| Merchant fee revenue | 4.0% x 500 | 20.00 |
| Funding cost | `=8%*57.69` | [ ] |
| Credit losses | `=2%*500` | [ ] |
| Servicing and collections | 0.5% x 500 | 2.50 |
| Contribution | 20.00 - [funding] - [losses] - 2.50 | [ ] |
| Contribution as a share of volume | [contribution] / 500 | [ ] |
| New break-even loss rate | (20.00 - [funding] - 2.50) / 500 | [ ] |

Questions to answer: by how much does contribution fall, and which of the two shocks causes most of the fall? Why is
the break-even loss rate almost unchanged even though funding is a third more expensive? What would you ask the
merchant partners to change first? [ ]

## Exercise C (unsolved): a new market and a launch map

<!-- Guidance: C1 and C2 are numerical; C3 is a short written map. Use the formulas section, not the solved tables. -->

C1. Varnelle Pay considers a market without interchange caps. Keep TPV, ticket size, processing cost and fraud rate
from Worked example A. The merchant service charge is 2.50%, blended interchange 1.75% and scheme fees 0.14%. Find net
revenue, contribution and the net take rate. Explain why the gross charge more than doubles while contribution falls.

C2. Short-term rates fall to 1.5%. Keeping the fixed cost of 8 and reserves of 1,000, what is the highest share of
reserve income Varnelle Digital can pay its distribution partners and still break even? What does that tell you about
the bargaining position of large distribution platforms?

C3. Varnelle Pay plans to offer its BNPL product to consumers in one EU member state and a stablecoin wallet to US
users. In no more than 150 words, list the activities involved, the row of the regulation map each one touches, two
questions for counsel in each market, and the dates that could change the economics in the next two years.

## Connections to other cards

<!-- Guidance: Fintech sits between strategy, finance and law. Use these links to go deeper on any line of the unit economics. -->

| Card or template | Connection |
|---|---|
| mba/mba-platform-strategy-card | Two-sided markets, network effects and pricing across sides, the theory behind interchange |
| mba/mba-technology-strategy-guide, mba/mba-digital-transformation-card | Build, partner or buy choices for banks facing fintech entrants |
| mba/mba-customer-lifetime-value-card | Acquisition cost and retention, the lines below contribution |
| mba/mba-fixed-income-card | Short-term rates that drive reserve and float income |
| mba/mba-ai-risk-governance-card | Model risk and fairness in automated credit decisions |
| mba/mba-venture-capital-guide | How fintech start-ups are funded and valued |
| mba/mba-business-law-guide, mba/mba-nonmarket-strategy-card | Licences, regulation and engagement with regulators |
| finance/unit-economics-analysis, finance/credit-memo, finance/ecl-methodology | Fuller unit economics, credit analysis and loss provisioning |
| industry/card-acquiring-payments-design, industry/open-banking-api-design, industry/iso20022-payments-design | How the payment rails and data access work in practice |
| system-design/design-digital-wallet, system-design/design-payment-ledger, system-design/design-fraud-detection | The systems behind the business lines |

## Common mistakes and misconceptions

<!-- Guidance: Check your own analysis against this list before you present it. -->

1. **Treating the merchant service charge as revenue.** Most of it passes through to issuers and networks. Quote
   net revenue or the net take rate when you compare payment companies.
2. **Calling float or reserve income "fee income".** It is rate income and falls when central banks cut rates; value
   it accordingly.
3. **Ignoring duration in short-term lending.** The same annual funding rate costs very different amounts per unit
   of volume for a six-week product and a two-year loan.
4. **Using the loss rate of a growing book in good times as a through-the-cycle rate.** Young vintages have not yet
   defaulted; look at losses by vintage.
5. **Assuming a fintech has no balance-sheet risk because it has no banking licence.** Lenders funded by warehouse
   lines can be cut off in a downturn, exactly when losses rise.
6. **Assuming stablecoin holders earn interest.** Under the EU and US regimes in the map, issuers may not pay interest
   to holders; returns offered by third parties raise separate legal questions.
7. **Mapping regulation by label.** Calling a product "payments", "credit" or "rewards" does not decide which rules
   apply; the activity does.
8. **Using an undated regulation table.** Several rows in the map changed within the last two years; every row needs
   a checked date and a link.
9. **Confusing a regulatory cost advantage with a technology advantage.** If the gap comes from lighter rules, it may
   close when rules change (Buchak et al. 2018).
10. **Forgetting fraud as a cost line.** Reimbursement rules such as REG-04 move fraud losses onto payment firms.

## Self-test questions

<!-- Guidance: Answer from memory, then check the answer key. Repeat after two days. -->

1. Name the three pass-through or direct cost items between the merchant service charge and payments contribution in
   Worked example A.
2. What is Varnelle Pay's net take rate, and how much contribution does a 10 basis point price cut remove?
3. Why does a 6% funding rate cost the BNPL line only 0.69% of volume?
4. Write the formula for the break-even loss rate and compute it for Worked example A.
5. What is the issuer's break-even reserve yield in Worked example A, and why does it depend on the partner share?
6. Why is stablecoin reserve income closer to bank income than to software income?
7. Give one reason in Rochet and Tirole's framework why card networks charge interchange.
8. According to Berg et al. (2020), what kind of information can predict default about as well as a credit bureau
   score in their sample?
9. What is the main source of run risk for a payment stablecoin, and which paper describes the same mechanism for
   banks?
10. Which EU regulation caps consumer card interchange, and at what levels?
11. On the checked date, what was the status of the EU Payment Services Regulation and PSD3?
12. Under the GENIUS Act, when does the law take effect?

## My notes and next steps

<!-- Guidance: Log each business you analyse with the line you found most exposed. Patterns across companies are what interviewers look for. -->

| Date | Company or case | Business lines | Most exposed line and why | Next step |
|---|---|---|---|---|
| *Example:* YYYY-MM-DD | Varnelle Pay (Worked example A) | Acquiring, BNPL, stablecoin | Stablecoin: profit falls 83% if rates fall to 2% | Split revenue by driver in my next case write-up |
| [YYYY-MM-DD] | [ ] | [ ] | [ ] | [ ] |

## Sources and further reading

<!-- Guidance: Primary works first, then official pages for dated facts, then textbooks by name only. -->

Primary works:

- Diamond, D. W. and Dybvig, P. H. (1983). Bank runs, deposit insurance, and liquidity. *Journal of Political
  Economy*, 91(3), 401-419.
- Rochet, J.-C. and Tirole, J. (2003). Platform competition in two-sided markets. *Journal of the European Economic
  Association*, 1(4), 990-1029.
- Philippon, T. (2016). The FinTech opportunity. NBER Working Paper 22476, https://doi.org/10.3386/w22476.
- Buchak, G., Matvos, G., Piskorski, T. and Seru, A. (2018). Fintech, regulatory arbitrage, and the rise of shadow
  banks. *Journal of Financial Economics*, 130(3), 453-483.
- Fuster, A., Plosser, M., Schnabl, P. and Vickery, J. (2019). The role of technology in mortgage lending. *The Review
  of Financial Studies*, 32(5), 1854-1899.
- Berg, T., Burg, V., Gombovic, A. and Puri, M. (2020). On the rise of FinTechs: credit scoring using digital
  footprints. *The Review of Financial Studies*, 33(7), 2845-2897.
- Gorton, G. B. and Zhang, J. (2021). Taming wildcat stablecoins. Working paper, SSRN,
  https://doi.org/10.2139/ssrn.3888752.

Official pages for dated facts (checked 2026-10-30):

- European Parliament Legislative Observatory, procedure 2023/0210(COD), Payment Services Regulation,
  https://oeil.europarl.europa.eu/oeil/en/procedure-file?reference=2023/0210(COD).
- Regulation (EU) 2024/886 on instant credit transfers in euro, https://publications.europa.eu/resource/celex/32024R0886.
- Regulation (EU) 2015/751 on interchange fees for card-based payment transactions, Articles 3 and 4,
  https://publications.europa.eu/resource/celex/32015R0751.
- Payment Systems Regulator, APP fraud reimbursement protections,
  https://www.psr.org.uk/information-for-consumers/app-fraud-reimbursement-protections/.
- Consumer Financial Protection Bureau, Personal financial data rights,
  https://www.consumerfinance.gov/personal-financial-data-rights/.
- Directive (EU) 2023/2225 on credit agreements for consumers, Articles 2 and 48,
  https://publications.europa.eu/resource/celex/32023L2225.
- Regulation (EU) 2023/1114 on markets in crypto-assets, Articles 50, 143 and 149,
  https://publications.europa.eu/resource/celex/32023R1114; ESMA MiCA page,
  https://www.esma.europa.eu/esmas-activities/digital-finance-and-innovation/markets-crypto-assets-regulation-mica.
- GENIUS Act, Public Law 119-27, sections 4 and 20, https://www.govinfo.gov/app/details/PLAW-119publ27.
- EIOPA, Digital Operational Resilience Act, https://www.eiopa.europa.eu/digital-operational-resilience-act-dora_en.

Textbooks (further reading, by name only): Mishkin and Eakins, *Financial Markets and Institutions*; Saunders and
Cornett, *Financial Institutions Management*.

Open resource: OpenStax, *Principles of Finance* (CC BY-NC-SA 4.0), https://openstax.org.

## Answer key

<!-- Guidance: Every number below comes from the calculation script. Check your workbook against these to two decimals. -->

Worked example B: funding cost 4.62 (`=8%*57.69`); credit losses 10.00; contribution 2.88 (`=20-4.62-10-2.5`), or
0.58% of volume; new break-even loss rate 2.58% (`=(20-4.62-2.5)/500`). Contribution falls by 6.15, from 9.04 to
2.88: 5.00 of the fall comes from losses and 1.15 from funding. The break-even loss rate barely moves (2.81% to 2.58%)
because the six-week duration keeps funding cost small relative to volume; the real damage is that actual losses of
2.0% now sit close to the 2.58% break-even, so a further small rise wipes out the line. Ask merchants for a higher fee
on high-risk categories or tighten approval rates first; a warehouse lender will want to see vintage loss curves and
an early amortisation trigger.

Exercise C:

- C1: net revenue `=2000*(2.5%-1.75%-0.14%)` = 12.20; contribution `=12.2-1.5-1` = 9.70; net take rate 0.485%. The
  gross charge is higher because interchange is higher, but interchange and scheme fees pass through, so the acquirer
  keeps less (0.61% of volume before costs, against 0.85% in Worked example A).
- C2: break even requires `(1-Share)*1.5%*1000 >= 8`, so the maximum share is `=1-8/(1.5%*1000)` = 46.67%. At low rates
  the issuer cannot afford to pay large platforms the 50% they earned at higher rates; platforms with many users can
  demand most of the reserve income, or issue their own coin.
- C3: a good answer names consumer credit (REG-06 and the member state's transposing law, applied from 20 November
  2026), payment services and safeguarding for collecting repayments (REG-01), stablecoin issuance or distribution
  in the US (REG-08, effective date in section 20), operational resilience in the EU (REG-09) and anti-money
  laundering rules in both markets. Questions for counsel: does our product fall within an Article 2 exclusion; which
  licence or partner carries the credit; who is the issuer and who holds reserves; can our rewards be treated as
  interest; what changes when the PSR and PSD3 apply.

Self-test:

1. Interchange and scheme fees (pass-through), then processing cost and fraud and chargeback losses.
2. 0.725% (`=14.5/2000`); a 10 basis point cut removes 2.00 (`=2000*0.1%`), about 14% of contribution.
3. Average receivables are only six fifty-seconds of annual volume (57.69 on 500), so funding cost is
   `=6%*57.69/500` = 0.69% of volume.
4. Break-even loss rate equals fee revenue minus funding and servicing, divided by volume:
   `=(20-3.46-2.5)/500` = 2.81%.
5. `=8/((1-50%)*1000)` = 1.60%; the issuer keeps only the share of reserve income not paid to partners, so a higher
   partner share raises the break-even yield.
6. It is earned by investing money held for customers in interest-bearing assets and moves with short-term rates,
   like a bank's deposit margin, not with transaction volume.
7. In a two-sided market, the network chooses how to split the total price between merchants and cardholders;
   interchange shifts cost to the merchant side to attract cardholders and usage.
8. Digital footprint variables, such as the device type or email provider used when ordering online.
9. Doubt about whether reserves can meet redemptions at par, which leads holders to redeem at once; Diamond and
   Dybvig (1983) model the same self-fulfilling run for banks.
10. Regulation (EU) 2015/751: 0.2% of the transaction value for consumer debit cards and 0.3% for consumer credit
    cards (Articles 3 and 4).
11. Not law: the Parliament committee approved the negotiated text on 5 May 2026, the plenary vote was listed as
    indicative for 14 December 2026, and the texts were not in the Official Journal.
12. On the earlier of 18 months after enactment (the Act was approved on 18 July 2025) or 120 days after the
    regulators issue final implementing regulations (section 20).

## Revision history

<!-- Guidance: Add a row each time you update the regulation map or rebuild the unit economics. -->

| Version | Date | Author | Change summary |
|---|---|---|---|
| 0.1 | YYYY-MM-DD | [Your name] | First build of Worked example A and regulation map checked |
| 0.2 | YYYY-MM-DD | [Your name] | *Example:* Updated REG-01 after the plenary vote and added a UK credit row |

## Standard and licence note

Based on standard payments, lending and financial intermediation economics, the primary works listed above, and the
public legal texts and regulator pages named in the regulation map, checked on 2026-10-30. The text, the unit
economics, the numbers and the exercises are original; no textbook, course material or published case is copied or
paraphrased, and Varnelle Pay, Varnelle Digital and every number in this card are fictional. Laws are summarised in
original words with article and section numbers; they are not quoted at length. The template is independent and not
affiliated with or endorsed by any regulator, payment network, bank, business school or publisher. Payments and
digital asset rules change often: check the current official text and a qualified professional before you rely on
any row. This is not legal, regulatory, investment or financial advice.
Template licence: MIT.