| Purpose | Detect and analyze earnings surprises to identify trading opportunities from post-earnings price movements |
| Core Function | Compares actual earnings results against consensus estimates, calculates surprise magnitude, and generates signals based on historical post-earnings drift patterns |
| Estimate Sources Uk | Bloomberg consensus estimates • LSEG / Refinitiv I/B/E/S consensus estimates • Thin analyst coverage and low liquidity for many small-cap and AIM-listed stocks |
Check: 1) RNS announcements via the London Stock Exchange (or aggregators like Investegate), 2) the company's investor relations 'financial calendar', 3) the FCA National Storage Mechanism for filings, 4) AlgoKing's earnings calendar.
Not necessarily. Options: 1) Hold if you're a long-term investor, 2) Reduce position if earnings are uncertain, 3) Use options to hedge. Consider your investment horizon and risk tolerance.
Initial reactions can be wrong or exaggerated. Reasons: guidance contradicts headline numbers, algorithmic trading creates initial move that humans fade, profit-taking after gap.
Often guidance is more important. A company can beat current quarter EPS but still fall if guidance is lowered. Markets are forward-looking.
Drift typically lasts 1-5 trading days for the main move. Larger surprises can show drift for 20+ days.
Use Standardized Unexpected Earnings (SUE): (Actual - Mean Estimate) / Std Dev of Estimates. This normalizes for different volatilities.
Be aware of IV crush, compare expected move to historical for vol trading edge, consider spreads to reduce IV impact.
This can happen due to guidance disappointment, whisper expectations higher than consensus, or 'buy the rumor, sell the news'. Protect with stops.
Track bellwether results closely, identify specific trends, position in unreported peers with similar exposure, size conservatively.
Fading is risky and not for beginners. Only consider when gap exceeds 2x typical reaction with reversal signals.
Pool data across stocks, use simple models to avoid overfitting, apply regularization, use walk-forward validation, accept wider confidence intervals.
Signals include unusual options volume, IV rising faster than typical, stock price pre-earnings drift. Caveats: much activity is hedging not directional.
Transcribe call, apply financial sentiment model (FinBERT), extract key phrases, quantify tone change from previous quarter, combine with quantitative surprise.
Options include earnings as a factor, earnings timing factor, earnings momentum, or integration with other factors.
Quarterly retrain with 3-5 year lookback recommended, with performance monitoring between. Trigger ad-hoc retrain if accuracy drops significantly.
Full guided lessons, quizzes, and a complete strategy library for the United Kingdom market. One-time purchase. No subscription, ever.
Get United Kingdom access →