In the sample (2016 Jan through now), strong up days (>1%) were followed by slightly negative average next-day returns with higher volatility; flat/weak days had small positive average next-day returns.
Data below -- SPY next-day return stats by current-day (bucket by bucket count, mean, and std):
2. weak_trend_within1pct (±1%): ROI +92.16%, CAGR 29.70%, Sharpe 1.79, N=633
3. strong_uptrend_gt1pct (>+1%): ROI -37.48%, CAGR -30.68%, Sharpe -1.44, N=323
4. other: ROI +38.04%, CAGR 34.51%, Sharpe 1.14, N=274
So here's how I'd interpret this: trade next day only when today is converging_near0 or weak_trend_within1pct; assume flat otherwise. Obviously there's probably some overfitting going on here as well.
What does '4. other' mean?
What are the SPY backtest results with those rules, Strong uptrend: > +1%, Weak trend: ±0.5% and Converging trend: near 0%?
In the sample (2016 Jan through now), strong up days (>1%) were followed by slightly negative average next-day returns with higher volatility; flat/weak days had small positive average next-day returns.
Data below -- SPY next-day return stats by current-day (bucket by bucket count, mean, and std):
1. converging_near0 (±0.5%): ROI +99.35%, CAGR 14.90% (252/len window), Sharpe 1.11,
N=1252
2. weak_trend_within1pct (±1%): ROI +92.16%, CAGR 29.70%, Sharpe 1.79, N=633
3. strong_uptrend_gt1pct (>+1%): ROI -37.48%, CAGR -30.68%, Sharpe -1.44, N=323
4. other: ROI +38.04%, CAGR 34.51%, Sharpe 1.14, N=274
So here's how I'd interpret this: trade next day only when today is converging_near0 or weak_trend_within1pct; assume flat otherwise. Obviously there's probably some overfitting going on here as well.