A1 Β· Post-Earnings Overreaction
Last verified
A1 Β· Post-Earnings Overreaction
A bounce setup: a stock that gapped down hard on an earnings print it actually beat, then started to base.
The short code A1 is display metadata only. The canonical identifier this archetype is keyed by everywhere β config, the screener_candidates.archetype column, the scorer emit β is the descriptive string post_earnings_overreaction.
Thesis
Earnings days produce the yearβs most violent single-day repricings, and the market is not always right in the first hour. A company can print an EPS beat and still gap down double digits because a headline number, a sector-wide de-rating, or pure positioning unwind dominates the tape. When the number was good but the price reaction was punitive, the gap is a candidate for a mechanical overreaction β forced sellers, momentum-following options flow, and stop cascades overshooting fair value. The edge is the divergence between fundamentals (a beat) and price action (a deep gap down), harvested once the selling exhausts and the stock starts to base.
Direction
LONG. This is a recovery / bounce setup β it looks for a beaten-down name turning back up.
What it looks for
The evaluator (evaluate_post_earnings_overreaction) requires five conditions for a primed read; thresholds come from screener_config.json Β§ archetypes.a1:
- Earnings recency β the print was 3 to 10 trading days ago (
earnings_recency_min_days=3,earnings_recency_max_days=10). Fresh enough to still be reverting, stale enough to have stopped falling. - Gap down β the print-day gap was at or below
-8%(gap_floor_pct=-8.0;gap_pctis in percentage points). - Surprise β₯ 0 β EPS surprise (revenue surprise as fallback) was non-negative. The market sold a beat.
- Quality gate pass β a gate failure is hard-disqualifying (
state="none", triggerquality:fail). - Basing confirmed β either RSI recovering out of oversold (RSI-14 in the
(35, 50]band offbasing_rsi_oversold_floor=35.0, or a two-point rise off a sub-35 reading) or price reclaiming its 10- or 20-day SMA (sma_dist_10/20 >= basing_sma_reclaim_tolerance=0.0).
How it scores
The score is a divergence metric: (-gap_pct) Γ max(surprise, 0), so a big drop on a strong beat ranks highest, times a basing_bonus_multiplier=1.15 when stabilization is confirmed. The state machine:
noneβ any core condition unmet (no recent earnings, gap too shallow, missed/unknown surprise, or gate fail). The name is off the board.partialβ all four core conditions hold but basing is not yet confirmed (basing:unconfirmed). The dislocation is real but the turn hasnβt started.primedβ core conditions plus a confirmed base. This is the firing candidate that gets ranked toward S2 promotion.
Pitfalls / when it fails
- The guidance-cut trap. A clean EPS beat can mask a slashed forward outlook β and the selloff is then correct. The public market-data feed carries no guidance data and no BMO/AMC timing, so the screen cannot see it. A high-divergence beat+gap-down may be a justified de-rating, not an overreaction. Always read the guidance commentary before trading.
- Falling knife. Basing is the safeguard against buying a stock still in free-fall, but a single RSI point is a proxy, not proof of a bottom.
- Approximate day-count. When trading-days-since-earnings isnβt pre-supplied, the assembler approximates from calendar days (Γ5/7), so the recency window has soft edges near its bounds.
Related
- The Screener System (vision)
- Screener Pipeline & Lifecycle (operations)
- Mechanical reference cards: Score Β· State Β· Triggers
- Sibling archetypes: A5 Β· Mechanical Reversion Β· A2 Β· Oversold Sympathy