Backtest: Buy CVX at the close when its 10-day total return underperforms PSX's by more...
Profitable, but badly beaten by a market that barely paused. Over 50 closed trades, this CVX-versus-PSX mean-reversion strategy returned 11% on $100,000 of starting capital — while SPY buy-and-hold gained over 68% across the same stretch. A positive edge did not translate into a good trade.
The premise is elegant: when a refiner like PSX outruns an integrated supermajor like CVX by more than five percentage points over ten days, the market is pricing an overshoot in crack spreads. That gap, the thesis holds, should snap back as flows rotate toward large-cap oil. The exit rule — a one-point outperformance by CVX, a 2% stop, or five days — was meant to catch that snap quickly.
Here is what the 50 trades, the win-loss distribution, and the return-on-capital breakdown actually show.
Buy CVX at the close when its 10-day total return underperforms PSX's by more than 5 percentage points; exit when CVX's 5-day return outperforms PSX's by 1 percentage point, on a 2% stop-loss, or after 5 trading days, whichever comes first. Refiner strength over a supermajor marks a crack-spread spike that overshoots, and the integrated laggard tends to catch up as energy flows rotate back into large-cap oil.
How this was measured
This is a simulated backtest generated from the plain-English strategy below, executed bar-by-bar on historical market data using the price + news data mode with $100,000 starting capital. Strategy: Buy CVX at the close when its 10-day total return underperforms PSX's by more than 5 percentage points; exit when CVX's 5-day return outperforms PSX's by 1 percentage point, on a 2% stop-loss, or after 5 trading days, whichever comes first. Refiner strength over a supermajor marks a crack-spread spike that overshoots, and the integrated laggard tends to catch up as energy flows rotate back into large-cap oil.
The key numbers
The charts
The takeaway
The strategy returned +11.03% on $100,000 starting capital across 50 closed trades with a 46% win rate. Over the same window SPY buy-and-hold returned +68.30%, so the strategy finished trailing the benchmark by 57.27 points. Best single trade +6.39%, worst -3.55%.
The fine print
- Simulated results on historical data — fills, slippage and costs are idealized.
- Past performance does not predict future results.