Credit Card Balances 90 Days Late Have Nearly Doubled Since 2022. New Delinquencies Have Not Moved.

Capital One’s Q2 2026 30-day delinquency rate fell to 3.13% (down YoY and below Q1), while Synchrony declined to 4.16% and Bread Financial to 5.25%—suggesting current credit performance is holding up despite hot inflation. The Fed (NY Liberty Street Economics) notes 90+ delinquencies rose from 7.6% (2022:Q3) to 12.8% (2026:Q1), but the “flow” delinquency rate has been relatively stable since ~2024, implying new delinquency pressure is elevated yet not accelerating materially. Net: credit risk is concentrated among already-struggling borrowers, with delinquency metrics still warranting close monitoring.

Analysis

The market is likely to over-read the headline delinquency level and under-read the vintage mix. If new flow into delinquency is flat, then the incremental reserve shock should stay contained even if reported 90+ day balances remain noisy; that matters because valuation risk for COF/SYF/BFH is more about future credit-cost acceleration than about where the stock of delinquencies sits today. In that setup, the immediate downside from “credit scare” headlines is probably more about sentiment than fundamentals unless we see a fresh turn in early-stage delinquency.

Second-order, the real loser would be lenders with weaker underwriting and higher private-label exposure, because they have less room to absorb even a modest uptick in net charge-offs without a step-up in provision expense. BFH is the most exposed on that dimension; COF should be the cleanest barometer of whether the consumer is actually weakening because it has scale and more transparent credit metrics. If consumer stress were truly broadening, the first place it would show up is in 30-day flow and revolving utilization, not in lagged 90+ day stock data.

Over the next 1-3 months, the catalyst is earnings-season reserve guidance rather than macro headlines: watch for management commentary on vintage performance, payment rates, and reserve build assumptions. Over 6-18 months, sustained inflation or a labor-market rollover would finally matter, but absent that, the current setup argues for skepticism toward bear cases built only on elevated delinquency stocks. The contrarian view is that credit is bad enough to keep lenders defensive, but not bad enough to justify a wholesale multiple de-rating; that supports relative-value, not outright sector bearishness.

Key Decisions for Investors

• Long COF / short BFH as a relative-value pair into the next earnings cycle; thesis is that the market should reward the cleaner credit profile while BFH remains more exposed to reserve pressure. Use a 1-3 month horizon and cut if BFH management shows stabilizing vintages or if COF provisions step up materially.
• Avoid chasing outright shorts in SYF or COF on this headline alone; wait for a real inflection in 30-day flow or a two-quarter rise in net charge-offs before betting on a consumer-credit break. Falsifier: sequential deterioration in early-stage delinquencies plus higher payment-rate pressure.
• If you want consumer-credit beta, prefer COF over BFH/SYF on any drawdown driven by macro fear. The setup favors the better-underwritten name because the bad-news path is likely already reflected in reserve assumptions, while weaker issuers still face multiple compression risk.
• Watch Q3/Q4 earnings for provision guidance and vintage seasoning. If reserve builds stay flat while delinquencies remain stable, rotate off defensive credit shorts and into the better-quality card names; if not, reassess for a broader consumer downgrade.
• No high-conviction options trade here unless you see a fresh acceleration in early-stage delinquencies; the signal is currently too mixed. For now, the cleaner expression is relative value rather than directional convexity.