Skip to main content

The Payments Corner

Decoding the Payments Stack
Decoding the Payments Stack
EN
← Back to Markets

Cohorts.

Each layer of the stack, read as a cohort. Are the names moving as one trade — or getting re-underwritten one by one? Lines rebased to 100; cohesion is the average pairwise correlation of daily returns after the S&P is stripped out, so it measures payments-specific co-movement, not the market tide.

AS OF 2026-09-10 · 16:00 ET · each cohort’s window is set independently on its own card

Benchmarkscohesion is market-adjusted (residual to the S&P)

The Tollbooths

· Card Networks
MixedResid corr0.56Mkt β0.48Dispersion10.3pp
1179676100S&P 500

The Tollbooths: mixed — correlation 0.56 over YTD; V +6.0% vs AXP -14.0%. Partial cohesion with individual stories emerging.

The Balance Sheets

· Issuing Banks & Lenders
MixedResid corr0.51Mkt β1.01Dispersion20.7pp
15711165100S&P 500

The Balance Sheets: mixed — correlation 0.51 over YTD; BFH +41.9% vs TBBK -25.7%. Partial cohesion with individual stories emerging.

The Engines

· Processing & Infrastructure
DivergingResid corr0.38Mkt β0.85Dispersion25.0pp
16510239100S&P 500

The Engines have decoupled — correlation just 0.38 over YTD; CPAY +35.9% vs FICO -41.5%. Divergers: CPAY, PAY, GDOT, PGY, FOUR, FICO. Names are being re-underwritten individually.

The Interfaces

· Consumer & Merchant Platforms
DivergingResid corr0.27Mkt β1.85Dispersion33.4pp
30916623100S&P 500

The Interfaces have decoupled — correlation just 0.27 over YTD; SEZL +81.7% vs KLAR -51.7%. Divergers: SEZL, CHYM, XYZ, PAGS, MELI, TOST, PYPL, NU, HAPN, SHOP, STNE, ADYEY, SOFI, UPST, KLAR. Names are being re-underwritten individually.

The Corridors

· Cross-Border Payments
DivergingResid corr0.23Mkt β0.78Dispersion40.5pp
21513657100S&P 500

The Corridors have decoupled — correlation just 0.23 over YTD; RELY +65.9% vs WU -25.4%. Divergers: RELY, FLYW, IMXI, WSE, WU. Names are being re-underwritten individually.

The Alternatives

· Crypto & Alternative Rails
MixedResid corr0.42Mkt β2.55Dispersion16.2pp
16710848100S&P 500

The Alternatives: mixed — correlation 0.42 over YTD; CRCL +8.2% vs COIN -27.2%. Partial cohesion with individual stories emerging.

Methodology: daily-return correlation (not price levels), market-adjusted — the S&P return is regressed out of each name (residual = return − β·S&P) so cohesion reflects co-movement beyond the market (on a risk-on day everything correlates; residuals strip that out). Cohesion = average pairwise correlation of residuals; ≥0.70 = moving together, 0.40–0.70 = mixed, below 0.40 = diverging. Mkt β = the cohort’s average sensitivity to the S&P. Each name’s R² is its fit to the equal-weighted, leave-one-out cohort residual factor (an “out” tag = low fit, a diverger). Charts show actual rebased price paths with the S&P/Nasdaq as toggleable benchmarks. Short windows carry too few trading days to trust the correlation — read the chart, not the number. History depth is ~1 year, so 5-year is not yet offered.