Imperial Oil's Record Quarter vs. Mixed Energy Quant Signals

This week's energy headlines are hard to ignore. Imperial Oil reported record net income of $2,190 million in Q2 2026, up sharply from a year ago, on higher commodity prices and strong cash flow. The fundamental story looks bulletproof. But the quant research published on trades.run this week tells a more complicated story: energy earnings may be booming, but most systematic timing signals in the sector are not keeping pace with the S&P 500.

One Signal That Stood Out

The clear exception is a backtest on APA. Buying the stock at the close whenever it lands in the bottom 20% of its 20-day closing range generated a +156.10% return on $100,000 across 37 closed trades, with a 70% win rate. Over the same window, SPY buy-and-hold returned +68.30%, so the strategy beat the benchmark by 87.80 points. The best single trade was +14.03%, the worst -12.28%. That is a decisive edge, but it is also idiosyncratic: it is one stock, one specific dip-buying trigger, and a small sample of 37 trades. It suggests that selective, drawdown-based entry points can work, even if the broader sector does not reward simple timing.

The Sector-Wide Signals Faded

Everything else is a different story. A backtest that buys XLE when its daily high-low range is in the bottom decile of its 20-day range returned just +12.66% across 50 trades, trailing SPY by 55.64 points. A VLO strategy based on outperforming XLE over five days scratched out +22.86% over 62 trades, but still trailed by 45.44 points. An EQT oversold signal tied to Brent crude managed +13.15% over 11 trades, trailing by 55.15 points. Even the macro question—whether XOM beats SPY in CPI-surprise months—the data basically calls a coin flip: XOM beat SPY in only 7 of 13 surprise months, and the mean excess return gap ( +0.38% vs -1.69%) has a p-value of 0.54. None of these are robust edges.

The Volatility Connection Is Weaker Than It Looks

One finding cuts against the prevailing narrative that oil volatility makes energy names more sensitive. For BKR, the daily-return beta to Brent crude was 0.263 in high-volatility months versus 0.255 in normal months—a difference of just 0.009, with a t-stat of 0.46. That is economically trivial and statistically indistinguishable from noise. So the instinct to overweight energy names when oil gets choppy has no support in the data. The relationship is stable, which means the recent earnings-driven excitement may not translate into additional sensitivity to crude price swings.

The data leans toward a clear conclusion: strong fundamentals are real, but the market may have already priced them in. The APA result shows that selective, contrarian signals can still beat the S&P by a wide margin, but the broader set of sector-level timing strategies—range compression, relative strength, oversold bounces—are all lagging. For systematic research, the takeaway is not that energy is a bad sector; it is that the easy signals are crowded and the edge, where it exists, is narrow and stock-specific. News flow may be bullish, but the quant evidence says: show me the trade, not just the headline.