Insider Activity Tracker

SEDI (sedi.ca) is the official free source. Search by company or insider name. Commercial services like INK Research, Bloomberg, and Refinitiv also provide cleaned data with analytics.

No, insider buying should confirm, not replace, fundamental analysis. Use it as one input alongside valuation, business quality, and technicals. Insiders can be wrong or early.

Canadian insiders have 5 calendar days to file. So there's up to a week delay between when they trade and when you see it. Some file faster; others use the full 5 days.

Usually not from a single insider. Selling is common for compensation, diversification, and personal needs. Only worry if: multiple insiders selling, unusually large amounts, or selling accompanied by departures.

Generally focus on transactions of $50,000 CAD or more. Smaller amounts may be less meaningful. For large-cap stocks, you might raise this threshold. For small-caps, smaller amounts may still be significant.

Collect SEDI data (manually or via scraping). Filter for open market buys, minimum value ($50K+), recent timeframe (90 days). Count unique insiders per company. Rank by cluster size and total value. Review top candidates.

Check transaction type codes. '10 - Acquisition in public market' is genuine buying. '51 - Exercise of options' followed by '10 - Disposition' is option cashout. Focus on open market acquisitions, not exercises.

Research shows insider buying signal persists for 6-12 months. Plan to hold at least 3-6 months. Exit on thesis change, price target achieved, or if fundamentals deteriorate despite insider buying.

Insider signals are stronger in smaller, less-followed companies where information asymmetry is greater. In large-caps with heavy analyst coverage, insider edge is reduced. Focus on mid and small caps for best insider signal.

Mining and energy (exploration results, commodity timing), financials (credit quality, rates), and small-cap industrials tend to have informative insider activity. Tech is noisier due to compensation structures.

Features: net buy value (30/60/90 day), buyer count, CEO/CFO binary, % of holdings bought. Target: forward 6-month return. Use regression or ML. Walk-forward validation. Combine with other factors for robustness.

Scrape or subscribe to SEC EDGAR data. Match to Canadian companies by ticker/name. Merge with SEDI data, removing duplicates. SEC is faster (2 days vs 5), so may see signal sooner.

Red flags: excessive IR promotion, many small insider buys for optics, insiders quickly selling after buying. Check for registration exemptions (prospectus deals) vs real buying. Verify with business fundamentals.

Research suggests 30-90 days is optimal. Too short misses patterns; too long includes stale signals. Weight recent transactions more heavily. Adjust by sector - fast-moving industries may need shorter lookback.

Measure returns at intervals (1m, 3m, 6m, 12m) post-filing. Signal typically strongest in first 3-6 months, then decays. Build rolling signals that refresh as new buying occurs. Exit if no fresh confirmation.