Trigger: a session drop of ≤ -5% (earnings-scale) · then a close back above the pre-drop close within 10 sessions · forward returns measured from the gap-fill day · 2005-01-03→2026-07-31
When GOOGL reclaims an earnings-scale gap-down within 10 sessions, forward returns skew positive — median +13.7% at 21 sessions, 60% positive (the still-below-65d cohort, the current analog).
Now: GOOGL 356.13 · -2.7% vs the 65-day MA · -11.5% from ATH · below the 65-day line
You are here. GOOGL dropped -7.1% on 2026-07-23, then closed back above the pre-gap close 6 sessions later (2026-07-31) — still below the 65-day MA, -11.5% from ATH. That is the navy cohort below.
Below-65 · 21d median
+13.7%
60% positive · n=5
vs baseline
+11.8 pts
base +2.0% at 21d
All fills · 63d median
+10.2%
n=10
Completed events
10
5 below-65 · 1 live
Forward returns from the gap-fill day
Median forward return (% positive) at 2 / 5 / 8 / 11 / 15 / 21 / 42 / 63 sessions after the reclaim. The navy cohort — filled while still below the 65-day MA — is the current GOOGL analog. Dashed gold = the unconditional baseline.
all fillsstill below 65d MAbaseline
Horizon
All fillsn=10
Still below 65d MAn=5
Baselineany day
2 sessions
+1.2% 60%
+2.0% 60%
+0.2%
5 sessions
+3.4% 70%
+4.8% 60%
+0.6%
8 sessions
+5.0% 70%
+8.1% 60%
+0.8%
11 sessions
+5.1% 70%
+6.4% 60%
+1.0%
15 sessions
+2.9% 70%
+4.2% 60%
+1.5%
21 sessions
+4.5% 60%
+13.7% 60%
+2.0%
42 sessions
+9.6% 60%
+15.0% 60%
+3.9%
63 sessions
+10.2% 60%
+25.9% 60%
+6.1%
Volatility & the options lens
The directional median is only half the picture — for an options trade what matters is move size vs what the market charges. Below: how this setup behaves in vol/move terms, and what GOOGL options are pricing now.
Metric (21 sessions)
All fillsn=10
Below-65 MAn=5
Baseline
Realized vol (annualized)
38.2%
46.5%
24.2%
Median |21-day move|
14.0%
15.5%
—
Moved > 12%
60.0%
100.0%
—
Direction up
60.0%
60.0%
—
What the options market is pricing · snapshot 2026-07-31
ATM IV (28d, exp 2026-08-28)34.6%
Implied ±1σ move±9.6%
ATM straddle7.1%breakevens ±7.1%
Setup realized vol46.5%median move 15.5%
Implied vs setup-realized: the market is charging 34.6% vol; this setup historically realized 46.5% — options screen 12 vol pts CHEAP. The straddle breaks even at ±7.1%; the setup's median 21-day move was 15.5% (100% of below-65 analogs moved >12%).
What the shape argues for. This is a volatility-expansion setup, not a quiet drift: realized vol runs ~2× baseline and essentially every analog made a large move — but the direction is regime-set (up in uptrends, hard down in bears). That points to owning the move rather than shorting it: a straddle / strangle to harvest the fat tails either way, or a call-tilted risk-reversal / call spread to lean into the 60% up-bias while defining risk. All are long premium — capped, known-at-entry loss, which fits the TEF exit-before-entry rule. Educational structure notes, not a recommendation; size to the risk budget and confirm against a live chain.
Every gap-fill · outcome over time
Each dot is one reclaim at its fill date; height is the forward 21-session return. Navy-ringed = still below the 65-day MA at the fill (the current setup).
The event log
Drop date
Drop
Days to fill
Cohort
% from ATH
Fwd 21d
Fwd 63d
2006-01-20
-8.5%
+2
above 65d
-6.1%
-14.7%
-3.6%
2008-09-08
-5.5%
+9
below 65d
-39.5%
-15.5%
-30.9%
2008-10-07
-6.8%
+4
below 65d
-48.6%
-18.3%
-17.5%
2009-01-20
-5.6%
+1
below 65d
-59.1%
+14.3%
+26.7%
2009-03-09
-5.7%
+2
below 65d
-57.1%
+17.2%
+36.1%
2020-04-01
-5.2%
+3
below 65d
-22.4%
+13.7%
+25.9%
2020-06-26
-5.5%
+3
above 65d
-5.4%
+3.2%
+1.6%
2020-10-28
-5.5%
+2
above 65d
-5.9%
+11.1%
+18.7%
2022-07-22
-5.6%
+5
above 65d
-22.4%
-5.9%
-20.7%
2025-05-07
-7.3%
+5
above 65d
-19.8%
+5.8%
+22.9%
2026-07-23 ◀ now
-7.1%
+6
below 65d
-11.5%
—
—
Read it honestly. Small sample: 10 completed gap-fills in 2005–2026, only 5 of them below the 65-day MA. The below-65 cohort's strength is dominated by 2008–09 and 2020 crisis reclaims off far deeper drawdowns than a single earnings pullback — read it as "constructive," not a base rate. The real swing factor is the broad regime: the same reclaim faded in the 2022 bear but worked in 2020/2025 uptrends. De-clustered by 20 sessions; overlapping forward windows remain. A context gauge, not a signal — cross-check against the TEF mandate before any action.
Computed live from adjusted-close history (yfinance). "Earnings-scale drop" = any single-session decline ≤ -5% (for a mega-cap these are overwhelmingly earnings reactions). Forward return = close-to-close from the gap-fill day. Educational research, not investment advice.