Contrarian tool - extreme sentiment often precedes reversals
| Strategy Type | Market Psychology and Sentiment Measurement Framework |
| Market Outlook | Contrarian tool - extreme sentiment often precedes reversals |
| Risk Profile | Analytical tool - improves timing and risk assessment |
| Reward Profile | Identifies crowded trades and potential turning points |
| Time Horizon | Daily to weekly sentiment assessment |
| Iv Environment | High fear sentiment often correlates with elevated IV |
| Breakeven | N/A - analytical framework for sentiment measurement |
| Sentiment Indicators | S&P/TSX 60 VIX - Canadian fear gauge • Montreal Exchange options data • Advance-decline, new highs-lows • Canadian ETF and mutual fund flows |
| Canadian Factors | Energy and mining sentiment affects TSX • Housing market sentiment impacts banks • Rate expectations and policy sentiment |
| Data Sources | TSX market data and breadth • Montreal Exchange options data • Canadian investor surveys (limited) • Canadian ETF flow data |
| Cross Border | US VIX, AAII often leads Canadian sentiment • TSX sentiment closely tied to US • US indicators as proxy when Canadian unavailable |
Sentiment is often a coincident or slightly leading indicator at extremes. It doesn't predict exact timing but extreme readings often precede reversals by days to weeks. Use it with price confirmation.
Sentiment alone is not enough. Use it as a filter or confirmation. Combine with trend analysis and price action. Sentiment tells you when conditions are ripe for reversal, but price action tells you when reversal is happening.
Use VIXC (TSX VIX) when available, TSX breadth data, and Montreal Exchange put/call data. US VIX and AAII are useful proxies since TSX correlates highly with US markets.
Daily check of VIX and put/call is good. Weekly for surveys (AAII). Major sentiment shifts are relatively rare, so you don't need constant monitoring. Set alerts for extreme readings.
Extreme fear means selling is exhausted. When everyone who wants to sell has sold, selling pressure ends. Additionally, extreme fear means most are in cash - they're future buyers when they return. Supply/demand shifts in favor of buyers.
Look for: multiple indicators at extremes (not just one), sentiment divergence from price, price confirmation (reversal candle, higher low). Also consider if fundamental catalyst could justify sentiment. Not all extremes reverse immediately.
In uptrend: extreme fear readings are buy-the-dip opportunities. In downtrend: extreme greed readings are sell-the-rally opportunities. Sentiment extremes against the trend are higher probability. Sentiment extremes with the trend need more confirmation.
Common weighting: VIX 25-30%, put/call 15-20%, breadth 20-25%, surveys 15-20%, flows 10-15%. Adjust based on availability and backtested predictive power. Start equal-weighted and refine.
When indicators conflict, rely on those at extremes. If VIX is at extreme but surveys are neutral, weight VIX more. If nothing is extreme, sentiment is neutral - don't force a contrarian trade. Wait for clearer signal.
Yes, but different indicators: short interest, analyst ratings consensus, options unusual activity, social media mentions. Extreme short interest can lead to squeezes; universal 'buy' ratings often precede disappointments.
Use pre-trained models like FinBERT (HuggingFace). Feed headlines or text, get sentiment classification. Aggregate across sources with volume and time weighting. Backtest against forward returns. Monitor for model drift.
Calculate 25-delta put IV minus 25-delta call IV. Positive = puts more expensive = fear premium. Track z-score vs history. Extreme positive skew = fear extreme. Negative skew = call FOMO. Use as component in composite sentiment.
Architecture: Data feeds (market data, news APIs, social APIs) → Stream processing (Kafka, Spark) → NLP/scoring → Aggregation → Dashboard (Grafana, custom). Store historical for percentile calculations. Alert on extremes.
Backtest: predict forward returns (1-day, 5-day, 30-day) from sentiment. Metrics: correlation, hit rate, conditional returns (returns following extreme sentiment). Walk-forward testing essential. Compare to simple benchmarks.
Use as filter: only take signals when sentiment agrees. Or as position sizer: reduce size at extreme greed, increase at extreme fear. Or as regime indicator: different strategies by sentiment regime. Backtest integration before live deployment.
Full guided lessons, quizzes, and a complete strategy library for the Canada market. One-time purchase. No subscription, ever.
Get Canada access →