← All forecasts · October 19

COF / Q3 2026

Capital One Financial Corporation

Reported: October 20, 2026

Published forecast

2026-10-20T00:01:52.291814+00:00

Opus 5.5 High in Shortcut

Shortcut

5.63

Shortcut normalized EPS

+2.5% vs Street

Off by $1.06 per share

Street

5.50

Street normalized EPS

Frozen October 18 snapshot

Off by $1.19 per share

Real

6.69

Issuer adjusted diluted EPS

Shortcut win

Reported result · October 20 ↗ · Adjusted EPS excludes the issuer’s disclosed adjusting items; GAAP diluted EPS was $5.61.

USD per per diluted common share. Percentages use unrounded forecasts.

Model download

Explore the assumptions and calculations behind the forecast.

Shortcut’s reasoning.

The forecast combines the Credit Card, Consumer Banking and Commercial Banking builds for net interest income, non-interest revenue, credit provisions and operating expenses. Consolidated net interest income equals the segment forecast through a residual allocation; the consolidated interest-rate lines are not an independent revenue forecast.

The model bridges GAAP net income available to common shareholders of $2,771.736 million to adjusted net income of $3,426.846 million by adding $870 million of adjustments and a negative $214.890 million tax effect.

Dividing adjusted net income of $3,426.84628778907 million by 608.15 million diluted shares gives adjusted diluted EPS of 5.634870159975451, versus the frozen Street estimate of $5.50.

The complete public export preserves all 188 cells in the scored EPS dependency chain. The separate 67.14% beat probability is a heuristic and does not multiply forecast EPS. The tracked cash-flow forecast has a narrower definition than issuer historical operating cash flow and is not scored.

On the record.

Reporting basis
adjusted normalized diluted EPS to common
Forecast completed
2026-10-19T21:55:49.584951+00:00
Published
2026-10-20T00:01:52.291814+00:00
Model assumptions and review notes
  • MEDIUM: Consolidated net interest income equals the segment forecast through a residual. Consolidated rates and interest allocations therefore do not independently validate earnings. Cash is a balancing item, and internal PASS and linkage checks are not independent validation of forecast assumptions.
  • MEDIUM: The September-end Credit Card balance of $262,400 million and Consumer Banking balance of $92,900 million are forecast assumptions anchored to August data, not reported September actuals. The cached August Credit Card and auto balances were $260,642 million and $91,886 million.
  • MEDIUM: Tracked operating cash flow of $7,333.305 million includes a narrow $805 million depreciation/amortization add-back, versus $1,585 million on the broader latest historical issuer line. Deferred taxes and some other operating items are zero; gross securities flows are not separately forecast. This is not a comparable issuer-basis CFO forecast. An issuer-defined free-cash-flow metric was not identified in the reviewed cache.
  • MEDIUM: Twenty-four growth diagnostic rows repeat the earlier parent-metric calculations beneath intervening driver or monthly-input rows. They refer to the parent metric, not the immediately preceding monthly driver. They do not feed scored EPS, but their placement can mislead.
  • The midpoint timing assumption for share repurchases and issuance is approximate. Applied to past quarters, it produces 620.5 million versus 622.5 million reported basic shares in Q1 and 614.7 million versus 620.5 million in Q2. These are timing diagnostics, not proved future errors. Adding 2 million diluted shares to the forecast denominator would reduce EPS to about $5.6164.
  • Credit Card average loans use a simple monthly mean rather than day weights. The conditional day-weighted alternative would lower EPS by about $0.0009045. The original monthly-mean assumption remains fixed. Independent average cash, debt and deposit balances are also assumptions.
  • The CapIQ net-income consensus basis is unconfirmed, so Cover net-income variances may not compare the same common-shareholder/preferred-dividend basis. The scored EPS comparison remains separate. Preferred cash dividends are assumed at $73 million; a prior-declaration timing alternative differs by $16 million in cash.
  • The 8.26% to 8.36% NIM range is a historical proxy rather than issuer guidance. One-decimal display formatting hides basis-point precision; full underlying values are preserved. Some wrapped raw history headers may clip.
  • MEDIUM: Post-cutoff retrieval and displayed publication dates do not prove immutable source pages as of the information cutoff. Original cached source bytes were reverified; no new financial research, data or assumptions were introduced during this review.
  • The frozen 0.78 prior, eight pseudo-observations and net negative 0.06 judgment give a 67.14% beat probability from four strict beats in six quarters. These are heuristics. Large-company market capitalization qualifies independently of an unverified current index-membership headline.
Research sources