Home › Guides › How Q4 GAAP EPS is derived from the 10-K (FY minus Q1–Q3)
How Q4 GAAP EPS is derived from the 10-K (FY minus Q1–Q3)
SEC EDGAR rarely includes an explicit Q4 quarterly EPS fact because most companies file a 10-K for the full fiscal year instead of a separate Q4 10-Q. Aegwah derives Q4 using the industry-standard formula and flags every derived row.
The derivation formula
Q4 EPS = Full-Year EPS − (Q1 EPS + Q2 EPS + Q3 EPS)
This is the same methodology used by Compustat, FactSet, and other institutional data vendors. The full-year value comes from the 10-K XBRL tag; Q1–Q3 come from the three 10-Q filings for that fiscal year.
The _computed_q4 flag
Every derived Q4 record in the Aegwah dataset carries _computed_q4: true. Non-derived quarters (Q1, Q2, Q3, and any quarter with a direct 10-Q filing) have _computed_q4: false or the field is absent.
By default, the REST API excludes computed Q4 rows (include_computed_q4=false). Pass include_computed_q4=true when you need a complete four-quarters-per-year series.
When derivation introduces small differences
Derived Q4 can differ slightly from a standalone Q4 10-Q when:
- The company restated one of Q1–Q3 after the 10-K was filed (the 10-K annual figure may reflect restated quarters).
- Rounding: individual quarters are stored to 4 decimal places; the subtraction can produce a ±$0.01 difference vs. a direct filing.
- Discontinued operations: if the annual EPS concept differs from the quarterly concept (continuing ops vs. total), the subtraction may not reconcile perfectly.
Aegwah's quality framework records derived Q4 as informational issue I2_computed_q4 (0-penalty) — it is expected, not an error.
Quality scoring
Derived Q4 rows do not reduce a ticker's quality score. They are flagged for transparency so you can decide whether to include them in your analysis or filter them out.
See the free sample ZIP for examples of both direct and derived quarters.
Get institutional-quality earnings data
502 tickers · 16+ years of GAAP EPS · ZIP + REST API. From $30/month.
Research / data documentation only. Not investment advice. Aegwah is a dataset product.