Using 10-Ks, 10-Qs, and 8-Ks Effectively

Difficulty: Intermediate Published: 2025-12-28

The practical point: you're trying to convert 10–15 hours per holding per year of reading into one binary decision—size the position or don't—using documents that arrive on predictable clocks (60/75/90 days, 40/45 days, and 4 business days) and move prices in measurable windows (3 days, 20 trading days, 12 months). The point is to treat filings as an operational process, not a one-time research project.

Why SEC Filings Matter

SEC filings matter because they contain audited (10-K) and systematically updated (10-Q) baselines plus event-triggered (8-K) surprises, and each has empirically measured information-delay effects. In one study, more complex 10-Ks were followed by 2.3% excess returns over the next 20 trading days versus 0.4% for simpler filings—an explicit "processing delay" you can exploit only if you read fast and consistently (You & Zhang, 2009). In another, 8-K disclosures produced 3.2% average absolute abnormal returns in a 3-day window, with Item 2.02 averaging 4.1%—a timing advantage you lose if you wait for the next quarterly report (Lerman & Livnat, 2010).

Filing Types: what you get, when you get it, and what you do with it

10-K (annual, 1× per year): your baseline model

What it is (1 number that matters): the only filing that gives you audited annual financial statements plus full-year narrative context (1 baseline per year).

When it arrives (deadline math):

• Large accelerated filer (public float > $700M): 60 days after year-end
• Accelerated filer ($75M–$700M): 75 days
• Non-accelerated (< $75M): 90 days

What you read first (4 sections, in order):

1. Item 1A Risk Factors
2. Item 7 MD&A
3. Item 8 Financials + Notes
4. Item 9A Controls and Procedures

How you quantify "the narrative is getting worse" (2 thresholds):

• Risk-factor expansion: a >15% year-over-year word-count increase is your "conditions deteriorating" tripwire.
• Language shock: >30% textual change in Risk Factors is linked to 4.1% underperformance over 12 months, and a systematic long–short based on 10-K/10-Q textual change produced 8.4% annual alpha (Cohen, Malloy & Nguyen, 2020).

How you quantify "they're hiding the ball" (Fog Index):

• MD&A Fog Index > 19.4 is associated with 4.7% lower earnings persistence versus filings with Fog Index < 16.8 (Li, 2008).

The point is not literary criticism; it's a numeric proxy for the probability you'll underwrite a fading earnings stream by mistake.

Footnote red flags (3 mechanical checks):

• Goodwill cushion: if fair value exceeds carrying value by <10%, impairment risk is "elevated" on your checklist.
• Debt maturity wall: if >25% of long-term debt matures within 24 months, you force a refinancing scenario (rates, covenants, liquidity).
• Related-party transactions: you don't label them "bad"; you count dollars and pages.

10-Q (quarterly, 3× per year): your change detector

What it is (1 framing): an unaudited delta against the 10-K, filed for Q1/Q2/Q3 (3 filings; Q4 is covered by the 10-K).

When it arrives (days):

• Large accelerated / accelerated: 40 days after quarter-end
• Non-accelerated: 45 days

What you read first (3 sections):

• Item 1 Financials + Notes
• Item 1A Risk Factors (updates only)
• Item 2 MD&A

Timing as signal (3 timing metrics):

• After-hours: filings after 4:00 PM ET show 1.8% more negative earnings surprises, with 23% higher short interest pre-filing (Bao et al., 2019).
• End-loading: filing in the final 5 days of the deadline window correlates with 2.1% more negative surprises.
• Friday effect: Friday filings have a 12% higher likelihood of adverse information.

The point is that timing is measurable behavior, not vibes: you log timestamp, day-of-week, and days-after-period-end as numeric features.

8-K (event-driven): your "material change" alert

What it is (1 clock): a current report due within 4 business days of the triggering event.

Highest-impact items (4 items you treat as immediate reads):

• Item 2.02 Results of Operations (4.1% average abnormal return in a 3-day window) (Lerman & Livnat, 2010)
• Item 4.01 Auditor Changes
• Item 4.02 Non-Reliance on Prior Financials
• Item 5.02 Director/Officer Departures

Auditor-change edge case (1 statistic): auditor-change 8-Ks (Item 4.01) precede restatements 34% more frequently than non-changers, so you treat them as a "stop-and-verify" event.

Red Flags You Can Operationalize (numbers, not adjectives)

Use this as a threshold table you can apply in 15–30 minutes per 8-K, 1–2 hours per 10-Q, and 4–6 hours per 10-K:

• Risk Factors: >15% YoY word-count increase; >30% textual change (expect -4.1% 12-month underperformance signal).
• Readability (MD&A): Fog Index >19.4 (associate with -4.7% earnings persistence).
• Filing behavior: after 4:00 PM ET (+1.8% negative surprise frequency); final 5 days (+2.1% more negative surprises); Friday (+12% adverse likelihood).
• Accounting complexity (XBRL): >850 unique XBRL tags implies 37% higher subsequent restatement likelihood vs <450 tags (Hoitash & Hoitash, 2018).
• Balance-sheet stress: >25% of long-term debt due within 24 months; goodwill cushion <10%.

Worked Example: you analyze filings like a process (6 steps, 300 minutes)

Scenario (numbers first): you evaluate a mid-cap industrial company with $8.2B market cap for a 4% portfolio position ($32,800 out of $820,000) with 6–8 hours available.

1. You download the 10-K and prior-year 10-K (45 minutes). You diff Item 1A and you count words. You find 3 new supply-chain risks, 2 removed risks, and a 12% increase in Risk Factors length (4,892–5,479 words). The point is that you've turned "tone" into one numeric delta you can track.

2. You extract MD&A quant (60 minutes). You pull every percentage change management mentions into a spreadsheet. You document 7 segments with revenue growth from -4.2% to +11.3%, and you flag 2 segments where margin compression exceeds 200 bps. The point is to force the narrative to produce numbers you can reconcile.

3. You audit the footnotes that carry blow-up risk (90 minutes). You prioritize revenue recognition, debt maturities, goodwill testing, and related-party transactions. You find a $340M debt maturity in 18 months, and goodwill of $2.1B tested at only 8% above fair value—below the 10% cushion threshold, so you tag impairment risk as "high."

4. You use the 10-Q as a working-capital lie detector (45 minutes). You compute turnover changes: DSO moves 42–51 days quarter-over-quarter; inventory turnover slows 5.2×–4.7× annualized. The point is to catch deterioration 1 quarter earlier than the annual report.

5. You review the last 12 months of 8-Ks (30 minutes). You log the event types and dates: a CFO departure appears in Month 8 under Item 5.02, and a restructuring charge of $47M appears in Month 5 under Item 2.05. The point is that "management stability" becomes countable events.

6. You score timing and readability (30 minutes). You calculate the 10-K filing delay: 58 days after year-end (inside a 60-day accelerated-filer deadline) versus 52 days last year, and you compute an MD&A Fog Index of 18.7 (below the 19.4 red-flag threshold). You now have a filing-behavior time series with 2 annual observations you can extend.

Decision framing (probabilities and payoff ranges):

• 40% probability: stable case with 8–12% expected annualized return.
• 25% probability: mispriced growth case with 15–22% annualized return over 2–3 quarters.
• 35% probability: avoidable loss case of -15% to -30% if you ignore the red flags; $0 loss if you pass.

Historical Examples (dates + outcomes you can't ignore)

• Enron (10-K filed 2001-04-02; bankruptcy 2001-12-02): Footnote disclosures included $1.2B in related-party transactions across 67 pages; SPE mentions rose 8–47 (+488%). Investors who identified $3.9B off-balance-sheet obligations avoided a 99.7% stock collapse ($83.13–$0.26) over 12 months (SEC File No. 001-13159; Form 10-K FY2000).
• GE (10-K filed 2017-02-24; 8-K dated 2018-01-16): the 2016 10-K disclosed $15.1B insurance reserves with "significant uncertainty," while critical audit matters expanded 2.0–4.5 pages and loss ratio rose 76%–89%. The later 8-K announced a $15B reserve charge; investors who acted on the 10-K warning avoided a 44% drop ($29.93–$16.72) (SEC File No. 001-00035).
• SVB (10-K filed 2023-02-24; failure 2023-03-10): the 10-K disclosed $91.3B HTM securities against $173.1B assets (52.7%), and $15.1B unrealized HTM losses versus $11.5B equity, with 6.2-year average duration. Analysts who marked equity to market had 14 days to act; the stock fell 60% in the final session before seizure (SEC File No. 001-39154).

Common Implementation Mistakes (and what they cost you)

1. You read the earnings press release instead of the 10-Q. You miss footnotes because press releases omit 73% of footnote disclosures on average, and you can miss estimate changes that affected 23% of S&P 500 companies' reported earnings in 2023. Fix: you read the full 10-Q within 48 hours and cover Items 1, 1A, 2 plus all notes.

2. You ignore 8-Ks between quarters. You fall behind by 45–90 days because the average S&P 500 company files 12.4 8-Ks per year. Fix: you set EDGAR alerts and review each 8-K within 24 hours, logging Item number + date + dollar impact.

3. You don't compare the current 10-K to the prior-year 10-K. You forfeit a documented signal: >30% Risk Factor language changes predict -4.1% 12-month underperformance, and the average warning lead time is 2.7 quarters before deterioration. Fix: you diff Item 1A and store a change log with additions, deletions, and word counts.

Implementation Checklist (tiered by ROI)

Tier 1 (highest ROI, 60–120 minutes setup):

• Track deadlines: 60/75/90 (10-K), 40/45 (10-Q), 4 business days (8-K) and log actual days-to-file.
• Build a 4-section 10-K first-read template: Item 1A, 7, 8, 9A with 1 numeric takeaway per section.
• Set alerts and a review SLA: 24 hours for 8-Ks, 48 hours for 10-Qs.

Tier 2 (high ROI, 2–4 hours per year per holding):

• Diff Item 1A and record: word-count change (%) and text-change (%); flag >15% expansion and >30% change.
• Add 3 balance-sheet tests: debt >25% due in 24 months, goodwill cushion <10%, working-capital deltas (DSO, inventory turns).

Tier 3 (situational ROI, 1–2 hours when triggered):

• Compute MD&A Fog Index; treat >19.4 as a "read deeper" threshold (Li, 2008).
• Count XBRL tag complexity; treat >850 unique tags as elevated restatement risk (+37%) (Hoitash & Hoitash, 2018).
• Treat auditor changes (8-K Item 4.01) as a halt condition because restatements are 34% more frequent after changes.

The durable lesson

The durable lesson: you win by being mechanical—turn each filing into a timestamped dataset (days-to-file, after-hours vs before-hours, Friday vs non-Friday), a text delta (Risk Factor +15% and +30% thresholds), and a balance-sheet stress test (<10% goodwill cushion, >25% maturity wall). Your edge is not reading "more"; it's applying the same 12–15 quantified checks on every 10-K, every 10-Q, and every 8-K, and letting the numbers force the decision.