Earnings Surprise Detector

TMX website has an earnings calendar. Yahoo Finance shows dates on stock pages. Company investor relations pages announce dates. Earnings Whispers and Seeking Alpha also cover Canadian stocks.

Depends on risk tolerance. Earnings are binary events - you can win big or lose big. If you have conviction, you can hold with appropriate size. If unsure, reduce position before or use options for defined risk.

Stocks are valued on future earnings, not past. Guidance tells you what future earnings will be. A beat today with lower guidance means future is worse than thought. A miss with raised guidance means future is better. Future dominates.

Research shows drift can last 60+ trading days. The first few days are usually the strongest. Consider holding post-earnings momentum trades for at least a week, potentially to the next earnings.

GAAP is official accounting earnings. Adjusted/Non-GAAP excludes 'one-time' items like restructuring. Many analysts focus on adjusted. Understand what's excluded - sometimes companies adjust out recurring items, which is misleading.

Implied move = ATM straddle / stock price. It shows expected move. If implied is 5% and stock historically moves 7%, straddle may be cheap. If historical is 3%, straddle may be expensive. Compare implied to realized.

Rising estimates in weeks before earnings often predict beats - analysts have new positive information. Falling estimates predict misses. Revision breadth (% revising up vs down) is also predictive. Follow the revisions.

The whisper number is the unofficial expectation, often higher than published consensus. It represents what sophisticated investors actually expect. If consensus is $1.00 but whisper is $1.10, beating $1.05 may still disappoint.

Enter after initial reaction settles (15-60 min post-earnings). Trade in direction of significant surprise. Hold for days to weeks. Exit before next earnings or on momentum exhaustion. Academic research confirms drift is tradeable.

When early reporter (e.g., RY) releases, identify sector-relevant information (mortgage volumes, provisions). Assess if positive/negative for peers. Position in peers (TD, BMO) before their reports. Be aware each company has idiosyncratic factors.

Features: estimate revisions (30-day, 60-day), revision breadth, historical beat rate, sector earnings momentum, macro indicators. Target: beat/miss or surprise magnitude. Use logistic regression or gradient boosting. Walk-forward validation critical.

Calculate accruals (earnings - operating cash flow). High accruals = low quality. Check revenue recognition (DSO changes). Check inventory (days increasing = demand concern). Compare GAAP to adjusted - widening gap is red flag.

Depends on view: Long straddle if expect move > implied. Iron condor if expect move < implied. Calendar spread if expect IV crush but some move. Direction + magnitude view → call/put spreads. No single best strategy.

HFTs react in milliseconds. As retail/systematic: you can't compete on speed. Wait 15-60 minutes for initial reaction to settle. Trade the drift, not the spike. If you must trade immediately, use limits not market orders.

Build calendar database. Generate pre-earnings signals (positioning). Generate post-earnings signals (surprise-based). Track drift signals. Use earnings as filter for other strategies (reduce exposure pre-earnings). Track performance by signal type.