The BaselineWatch Drift Index is at 53 — its highest reading since the 2008 Financial Crisis. The VIX is at 26. Nearly half of all SEC filings in our coverage universe are showing elevated or critical language changes. Companies are hedging their words at a rate we haven’t seen in 18 years.
But here’s what most people miss: not every stock that drops in a selloff deserves to drop.
We score every 10-K and 10-Q filed with the SEC for how much the language changed from the prior filing. When management starts adding risk disclaimers, hedging forward guidance, or expanding their “factors that could affect results” sections — we see it. When the language stays stable and confident — we see that too.
What 30 Years of Data Says About Stressed Markets
During periods when the aggregate drift index is at or above 50 and the market sells off, we observe a clear pattern in 60-day forward returns:
| Group | n | Raw 60d | Alpha | Win Rate |
|---|---|---|---|---|
| LOW drift (≤ 25) | 873 | +4.98% | +11.27% | 54% |
| HIGH drift (≥ 60) | 2,203 | +4.77% | +11.25% | 53% |
| MID | 1,548 | +3.69% | +9.72% | 50% |
Low-drift companies — the ones whose filings say “business as usual” — generate the same alpha as high-drift companies in a market bounce. But with fundamentally cleaner filings. Their management isn’t flagging problems. They got dragged down by the market, not by their own deterioration.
The Dip-Buy List: Low Drift, Clean Filings
These companies have filed recent 10-Ks or 10-Qs with minimal language changes. Management is not hedging. Disclosures are stable. If these names sell off with the broader market, the filings say it’s contagion — not fundamentals.
| Ticker | Score | Form | Filed |
|---|---|---|---|
| MSFT | 0 | 10-Q | Jan 28 |
| CSCO | 0 | 10-Q | Feb 17 |
| LRCX | 0 | 10-Q | Jan 29 |
| VRTX | 0 | 10-K | Feb 13 |
| WM | 0 | 10-K | Feb 9 |
| ORCL | 0 | 10-Q | Dec 11 |
| GEV | 0 | 10-K | Jan 29 |
| GOLD | 0 | 10-Q | Feb 6 |
| PANW | 1 | 10-Q | Feb 18 |
| WMT | 6 | 10-Q | Dec 3 |
| BA | 10 | 10-K | Jan 30 |
| JNJ | 10 | 10-K | Feb 11 |
| V | 10 | 10-Q | Jan 30 |
| DELL | 11 | 10-Q | Dec 9 |
| OXY | 13 | 10-K | Feb 18 |
| SMCI | 15 | 10-Q | Feb 6 |
| UNH | 15 | 10-Q | Oct 28 |
| PFE | 16 | 10-Q | Nov 4 |
| LOW | 17 | 10-Q | Nov 26 |
| HD | 19 | 10-Q | Nov 25 |
| INTC | 19 | 10-K | Jan 23 |
| RTX | 23 | 10-K | Feb 6 |
| PAYX | 23 | 10-Q | Dec 22 |
A score of zero means the filing language barely changed from the prior period. These are companies showing management confidence, stable operations, and no material new risk disclosures.
The Avoid List: High Drift, Real Problems
These names have drift scores at or above 75. Their filings show significant language changes — expanded disclaimers, hedged guidance, new risk factors. If these drop in a selloff, you cannot tell whether the decline is market contagion or the market finally pricing in what the filing already disclosed. That ambiguity is risk.
| Ticker | Score | Form | Filed |
|---|---|---|---|
| NVDA | 100 | 10-K | Feb 25 |
| ADBE | 100 | 10-K | Jan 15 |
| GOOG | 100 | 10-K | Feb 5 |
| AVGO | 100 | 10-Q | Mar 11 |
| LLY | 100 | 10-K | Feb 12 |
| TGT | 100 | 10-K | Mar 11 |
| TMUS | 100 | 10-K | Feb 11 |
| FCX | 100 | 10-K | Feb 13 |
| DIS | 98 | 10-Q | Feb 2 |
| NFLX | 94 | 10-K | Jan 23 |
| MCD | 91 | 10-K | Feb 24 |
| QCOM | 90 | 10-Q | Feb 4 |
| HLT | 87 | 10-K | Feb 11 |
| LMT | 87 | 10-K | Jan 29 |
| IBM | 85 | 10-K | Feb 24 |
| UPS | 85 | 10-K | Feb 17 |
| BKNG | 85 | 10-K | Feb 18 |
The Framework
This is not a prediction. It is a conditional setup with clearly defined trigger conditions:
1. Aggregate drift index at or above 50. Currently 53. ✓
2. VIX elevated. Currently 26. ✓
3. Wait for a broad selloff.
4. Buy low-drift names — they bounced with +11% alpha historically.
5. Avoid high-drift names — you don’t know what you’re catching.
When the market is up during stressed periods, nothing works — negative alpha across the board. The signal is specifically useful when fear creates a broad selloff and you need to separate the companies that got hit by contagion from the ones that actually have problems.
Methodology
BaselineWatch analyzes over 37,000 scored filings across 6,700+ U.S. public companies, with data spanning 30 years of SEC filing history. Returns are calculated on a 60-trading-day forward basis with Fama-French 5-factor controls. Alpha is measured against the expected return given market, size, value, profitability, and investment factor exposures. The methodology has passed 19 of 24 rigorous statistical validation tests.
We publish validation results. We protect the methodology.