What it is
Post-earnings announcement drift is the well-documented tendency for stocks that gap up on a positive earnings surprise to keep drifting higher over the following weeks, and for negative surprises to keep drifting lower. This playbook trades the drift, not the gap: you let the gap day happen, wait for the first pullback that holds above the gap, and enter with a stop below the gap-day low. Holds are 2 to 8 weeks.
The logic
Analysts, funds and index-tracking flows do not update instantly. Estimates are revised over days, position sizes are rebalanced over weeks, and retail attention follows price rather than leading it. That slow incorporation of information is the drift. The gap itself is the fast part of the repricing and is fought over by algorithms; the drift is the slow part and is accessible to a swing trader.
On the other side are holders taking profit after the gap, short sellers betting on a fill, and traders who did not believe the print. When the surprise is real and guidance was raised, those participants are slowly proven wrong. When the surprise was one-off (a tax item, an accounting change), they are right and the gap fades.
Setup rules
- Market: US stocks with average daily dollar volume above $50 million and market cap above $2 billion; the effect exists in smaller names but costs and gaps make it hard to capture.
- Timeframe: daily.
- Gap conditions: an earnings-day gap of at least 5 percent that holds; the stock closes the gap day in the top third of its range and above the pre-earnings high; volume at least 3x average; the company raised guidance or beat on both revenue and eps, not just one.
- Pullback conditions: within 2 to 10 days after the gap, price pulls back on declining volume without closing below the gap-day midpoint; the first day that closes above the prior day's high is the trigger.
- Disqualifiers: the gap was a one-time item; the company is in a sector that is currently in a strong downtrend; the stock is a heavily shorted name whose gap was mostly a short-squeeze.
- Surprise check: the gap should exceed the move the options market had priced before the report (the at-the-money straddle). A gap smaller than the priced move is not a surprise, whatever the headline says, and drift after it is weaker.
Entry, stop, target
Enter on the trigger close. Stop below the gap-day low (or the gap-day midpoint for a tighter version). Target 1 is the gap-day high plus the gap size; target 2 is trailed on a 10-day low close or the 20-day EMA. Exit before the next earnings report regardless.
| Item | Level | Notes |
|---|---|---|
| Pre-earnings close | 80.00 | |
| Gap-day range | 88.00 to 92.00 | Gap 10 percent, closed at 91.50 |
| Pullback low | 89.20 | Day 4, light volume |
| Entry | 90.80 | Close above prior day's high |
| Stop | 87.50 | Below gap-day low, risk 3.30 |
| Target 1 | 100.00 | Gap high plus gap size, reward 9.20, 2.8R |
| Trailed exit | 20-day EMA | Often 3 to 5R on strong drifts |
Position sizing and risk
Size from the stop at /tools/position-size and risk 0.5 to 1 percent of equity per trade. Earnings drifts cluster during earnings season, when a portfolio of five drift trades in the same sector can behave like one trade; cap sector exposure and total heat per /learn/risk-management. The gap risk between earnings is lower than during them, which is why exiting before the next report is a rule, not a suggestion.
What breaks it
- Gap fade. A meaningful share of earnings gaps fully reverse within 20 days, especially those driven by a single line item or a squeeze. The pullback entry avoids the immediate fade but not the slow one.
- Market regime. Drift is weak or negative in a falling market, where even good reports get sold.
- Edge decay. Academic drift has shrunk materially since the 1990s as more capital targets it, and the remaining effect is concentrated in smaller, less-followed names where costs are higher. In large caps the drift is modest; in mega caps it is close to zero.
- Costs. Low per trade, but the strategy's expectancy is small enough that a 0.2 percent round-trip cost matters.
- Drawdowns. Losses come in clusters at the end of earnings seasons and during corrections.
How to test it
Build a dataset of earnings dates and gaps for at least 8 years and a universe that includes delisted stocks. Record the 20 and 60-day return of every qualifying gap, then the return under your pullback entry and exit rules with costs. Separate results by market cap tercile and by whether guidance was raised; expect the effect to be strongest in the smallest liquid tercile and with raised guidance. Minimum 300 trades. Paper-trade one full earnings season before going live; the pace of signals during season is hard to manage and the test will show whether you can.
Variations
- Negative drift shorts: the mirror trade on large gap-downs with lowered guidance; borrow costs and squeezes make it harder.
- Gap-day entry: enter on the gap day's close instead of the pullback; more trades, worse R:R, more exposure to fades.
- Options version: a long call spread after the iv-crush instead of stock; see earnings-iv-crush.
Further reading
earnings-report, guidance, eps, gap, iv-crush, short-squeeze, short-interest, pullback, survivorship-bias, swing-trading.
Related playbooks: gap-and-go-gap-fill, earnings-iv-crush, news-event-trading-process, gap-fill-swing