Skip to main content

Fintech & Payments

Updated Jun 24, 2026 at 8:22pm

Research Draft High 1,255 words

"Fintech & payments" is a sprawling sub-sector of Financials that monetizes the movement of money rather than the lending of it. At its core sit the card networks (Visa, Mastercard), which take a tiny toll on a colossal flow of consumer spending; around them orbit merchant acquirers/processors (Adyen, Fiserv, Global Payments, Block, Stripe), digital wallets (PayPal), neobanks, and credit-extending models like buy-now-pay-later (Affirm, Klarna). The central tension for investors is that the franchise quality is extraordinarily bimodal: the networks are among the widest-moat, highest-margin businesses in public markets, while much of the rest of the space is intensely competitive, capital-hungry, and credit- or rate-exposed. Lumping them together as one "fintech" trade is the most common analytical error.

The industry structure

The card economy runs on a four-party model: cardholder → issuing bank → network → acquiring bank/processor → merchant. When a card is swiped, the merchant pays a ~2–3% US merchant discount rate, which splits into three flows to different entities. Interchange (~1.8% on US credit, no federal cap; per Congress.gov it averages ~86% of total merchant cost) goes to the issuing bank, not the network. Scheme/assessment fees (a small fraction of volume) go to the network. The acquirer markup goes to the processor. So Visa and Mastercard collect only the thin assessment-and-data-processing slice — but they collect it on virtually all of it: Visa processed ~$14T in payments volume in FY2025; Mastercard's gross dollar volume is reported around $9–10T depending on source and period (Spark Money cites ~$9.2T; Mastercard's own FY2024 10-K reported $9.8T GDV, and some trade sources cite ~$10.6T for FY2025). With UnionPay, the three networks together handle ~97% of global card transactions (Spark Money, Congress.gov, capitaloneshopping).

Key per-company metrics:

  • TPV / GPV (Total or Gross Payment Volume): the dollar throughput. PayPal ~$1.8T annualized; Stripe ~$1.9T (2025, +34%); Adyen €1.29T (+33%) (Investing.com, Chargeflow, Fintech Wrapup).
  • Take rate: net revenue ÷ volume. Wildly different by model. Adyen's blended take rate was ~16 bps in H2 2024 (enterprise, thin margin per dollar but vast volume); PayPal's take rate is far higher because it owns the consumer relationship; a network's assessment yield is its own number again. Comparing take rates across business models is meaningless without adjusting for what's bundled in.
  • Operating margin is the moat tell: Visa's non-GAAP operating margin runs ~67% and Mastercard's GAAP operating margin sits in the high-50s (~59%) in FY2025 (AInvest, MA 8-K filings) — software-like, near-zero marginal cost per transaction. Acquirers run far lower; lenders carry credit losses.
  • Cross-border volume: the highest-yield line, carrying ~1%+ FX/cross-border fees on top of normal scheme fees; Mastercard's mix skews more international (MA cross-border +13% local-currency in a recent quarter).

How it's used in practice

Analysts decompose growth into volume × take rate × mix, then watch the mix shift toward higher-yield revenue — cross-border travel, e-commerce penetration, and especially value-added services (fraud, data, consulting, tokenization). VAS is now ~40% of Mastercard revenue and ~43% of Visa revenue (American Banker, MatrixBCG) and is the networks' answer to fee pressure: it's recurring, high-margin, and not directly capped by interchange politics. For processors, the watch items are net-new merchant logos, churn, and whether scale is translating into operating leverage (Stripe reaching full-year profitability was the proof point bulls wanted; Adyen sustains a ~50% EBITDA margin). For BNPL/neobanks, the analysis flips to credit: funding cost, charge-off rates, and loan-loss provisioning behave like a lender, not a payments toll-taker.

Adoption, debate & evidence

The networks' wide-moat status is close to consensus among quality-focused investors, and the empirical case is strong: two-sided network effects, operating margins in the high-50s to high-60s percent range, and revenue that, per Wellington Management, tracks nominal consumer spending — which has fallen year-over-year only once since 1960 (the GFC). Wellington frames the cyclicality as roughly linear: a ~400 bps drop in real PCE shaves ~400 bps off payments-company growth, but the base rarely contracts outright, and inflation can help because fees are tied to nominal ticket size.

What is genuinely contested:

  • Regulation. The 2010 Durbin Amendment capped debit interchange; the proposed Credit Card Competition Act (Durbin–Marshall) would force dual-network routing on large issuers' credit cards, with backers claiming >$16B/yr merchant savings (Capstone DC). The DOJ's 2024 antitrust suit targets Visa's debit practices, and Visa/Mastercard agreed to a multi-year settlement temporarily capping posted credit interchange. These are real, recurring overhangs — not existential, but they cap the most political revenue line.
  • Disintermediation. Bill Gurley and others argue stablecoins threaten the 2–3% toll (Fortune, citybiz). The more grounded view (Third Bridge, Payments Dive) is that stablecoins are displacing ACH/SWIFT settlement rather than card front-ends, and the networks are co-opting them — Mastercard agreed to acquire stablecoin-infra firm BVNK for up to $1.8B. Treat "stablecoins kill Visa" as a low-probability, high-impact tail, not a base case.
  • The rest of fintech. Here the evidence is humbling. The 2022–2025 rate cycle exposed unprofitable, cheap-capital-dependent models; BNPL valuations collapsed and fintech stocks sold off hard on consumer-spending and credit fears (CNBC, Mar 2025). PayPal trades around 8x forward earnings — a clear discount to the broader financial sector (Trefis cites a ~15x sector comparison; other screens put the diversified-financials average nearer 11x, but the discount holds either way) — as the market prices erosion of its checkout franchise, a debate that is unresolved.

Strengths & limitations

Where the franchise shines: the networks are toll roads on global commerce — asset-light, inflation-protected, secularly tailwinded by cash-to-card conversion in emerging markets. They tend to compound EPS through cycles with minimal absolute revenue contraction.

Where it fails: (1) Regulatory/political risk concentrates on interchange and routing — slow-moving but real. (2) Cyclicality is real for the credit-bearing and discretionary-exposed names (BNPL, neobanks, processors weighted to travel/retail), even if the networks themselves are resilient. (3) The #1 misuse: treating "fintech" as one homogeneous bet. A wide-moat network and a thinly-capitalized BNPL lender share a sector label and almost nothing else in risk profile, margin structure, or rate sensitivity. Equating their valuations or growth multiples is the classic trap.

Sources

Dispute flags: (a) Stablecoin disruption thesis is genuinely contested — sources split between "existential" (Gurley) and "settlement upgrade, networks co-opt it" (Third Bridge); doc takes the latter as base case. (b) Mastercard's reported volume varies by source and metric/period — ~$9.2T (Spark Money, FY2025), $9.8T (MA FY2024 10-K), ~$10.6T (some FY2025 trade sources); treat as ~$9–10T+. (c) PayPal's "discount vs sector" depends on the comparison set (Trefis ~15x vs screens nearer 11x); the 8x multiple and the discount are firm, the magnitude is a market opinion, not a verdict.