sue zscore drift
What it checks
PEAD using the standardized SUE z-score instead of raw surprise %. SUE is the academic canonical form — it accounts for how noisy each company's analyst consensus is.
Mechanism
Bernard-Thomas 1989 JAE: SUE z-score (actual − consensus)/σ(consensus) outperforms raw surprise % as the PEAD signal because it conditions on the ticker's own consensus-dispersion noise. Original effect: top-decile SUE +6-8% over 60d, bottom-decile -5-7%. We approximate σ(consensus) by 12Q rolling std of own surprise_pct — within-ticker SUE.
Signal rule
Own-history 12Q rolling z of surprise_pct; long top + hold N days, short bottom + hold N days; threshold and hold are harness-tunable.
Data dependencies
earnings_historyWorker data table — see services/worker schema.
daily_pricesAdjusted-close OHLCV for every US-listed ticker; primary price feed.
Expected edge
- Paper alpha
- +6-8% / -5-7% over 60d
- Paper window
- 1974-1986 (in-sample)
Bernard-Thomas 1989: 6-8% (long top) / -5-7% (short bottom) over 60d.
Related families
peadEventPost-Earnings Announcement Drift (Bernard-Thomas 1989): buy after an earnings surprise greater than 1σ, hold 30-60 days. Surprise is computed as (actual − consensus) / |consensus|, with σ taken from a trailing expanding window so prior thresholds don't leak future variance.
pead text confirmationEventVanilla PEAD trades on SUE sign. This family gates by transcript-FinBERT sentiment direction: take position ONLY when sentiment agrees with SUE sign. Beat + bullish call → long; miss + bearish call → short; disagreement (beat-but-cautious, miss-but-optimistic) → skip. Filters out the messy 1/3 of earnings where management contradicts the headline.
strategic friday peadEventManagers schedule bad-news earnings for Fridays after-close. Three-day abnormal returns around the scheduling announcement are significantly negative. PEAD persists up to 2 years for Friday-bad-news.
earnings announcement premiumEventStocks earn an abnormally positive return in the days surrounding scheduled earnings announcements (T−2 to T+1). The original 2007 paper documented ~9bp/day average premium during the window vs the non-announcement baseline; later replications show the effect has compressed to ~5-6bp/day but is still robust. The mechanism is asymmetric resolution of uncertainty plus retail inattention immediately around announcement. Signal: long the (T−2 to T+1) window per announcement, flat otherwise.
Explore sue zscore drift on alphactor.ai
See which tickers this family is currently firing on, with live signals and rankings.