What is COF's forecasted revenue?

Capital one financial corp is forecasted to generate $14.5B in next-four-quarter revenue, representing +46.6% year-over-year growth.

How accurate is the COF revenue forecast model?

The model has 14.5% training MAPE (Mean Absolute Percentage Error). In a holdout test, it predicted $2.8B versus the actual $2.8B.

Capital one financial corp COF

Revenue Intelligence Report • 57 quarters of SEC filing data • Updated 2026-10-14 • Data through 2026-06-30

Capital one financial corp's latest quarterly revenue is $2.8B for the period ending 2026-06-30. The model projects $14.5B across the next four quarters (+46.6% versus the preceding four quarters). The forecast uses Operating expense proxy data; model selection and estimation exclude the latest quarter, whose retrospective holdout error is 2.0%. These are statistical associations, not estimates of the causal effect of spending.

Investment Thesis

The GAM forecast is based on reported revenue and Operating expense proxy data through 2026-06-30. Training MAPE measures fit to historical data; the separately reported holdout error measures a withheld quarter. Spending coefficients describe associations and do not establish causation.

Next 4 Quarters

$14.5B

+46.6% YoY

Operating expense proxy Elasticity

-0.28x

Training Fit

14.5% MAPE

Retrospective holdout (Q2 FY2026): The model predicted $2.82B vs the actual $2.76B — an error of 2.0%.
Retrospective one-quarter holdout: model fitting and selection exclude this quarter. Historical inputs use the latest available restatements, so this is not a preserved real-time forecast.

Note: The current filings use Operating expense proxy (NoninterestExpense). Spending estimates refer to this expense measure, not separately reported SG&A.

⚠ Model limitation: One or more estimated spending associations are negative. These observational coefficients do not establish a causal return on spending.

Note: Capital one financial corp does not report R&D expenses separately. The model uses the available expense measure without a separate R&D term.

Revenue Forecast

Quarterly Detail

Quarter	Model Forecast	Actual	95% Range	YoY Growth	Status
Q2 FY2026 Period end 2026-06-30	$2.82B	$2.76B	$1.32B–$6.00B	+52.8%	✓ In range
Q3 FY2026 Period end 2026-09-30	$3.21B		$1.28B–$8.09B	+40.8%
Q4 FY2026 Period end 2026-12-31	$3.58B		$1.19B–$10.76B	+46.3%
Q1 FY2027 Period end 2027-03-31	$3.59B		$1.11B–$11.63B	+48.8%
Q2 FY2027 Period end 2027-06-30	$4.13B		$1.02B–$16.80B	+49.6%

Seasonal Factors

Multiplicative seasonal adjustment: These factors capture Capital one financial corp's systematic quarterly revenue patterns relative to the trend model. A factor of 1.05 means that quarter typically runs 5% above the underlying trend; 0.95 means 5% below. Factors are computed as the median of (actual / fitted) across all available quarters.

Fiscal Quarter	Seasonal Factor	vs Trend	Interpretation	Obs.
FQ1 (Jan–Mar)	0.9757	-2.4%	In line with trend	14
FQ2 (Apr–Jun)	1.0308	+3.1%	+3.1% above trend	14
FQ3 (Jul–Sep)	1.0414	+4.1%	+4.1% above trend	14
FQ4 (Oct–Dec)	0.9957	-0.4%	In line with trend	14

Spending and Revenue

Reading this chart: Each line shows the cumulative revenue association with a 1% change in spending over subsequent quarters. These estimates describe historical associations and do not establish causation.

Time-Varying Spending Associations

Time-varying analysis: A penalized spline model (GAM) tracks the association between log spending and log revenue over 57 quarters. Its coefficients describe conditional historical associations, not causal returns on spending.
Current Operating expense proxy elasticity: 0.4963x
The model-term chart shows relative absolute term magnitudes. Its percentages are not shares of revenue or causal contributions to growth.

Enhanced forecast: The time-varying model (GAM) was selected using training fit, before evaluating the withheld quarter. Its holdout error is 2.0%.

Want this analysis for your portfolio?

I build custom revenue intelligence reports for investors and companies using SEC filing data, econometric modeling, and AI-powered insights.

Get in Touch