Energy's Tactical Signals Keep Working, and Keep Losing to SPY

The headline feed came up empty today. No Hormuz chatter, no tanker-rerouting story, no sanctions headline to hang an energy move on. Which makes it a good Saturday to sit with what the platform's own backtests have been producing all week, because the pattern running through them is uncomfortable and remarkably consistent: the signals mostly work, and almost none of them win.

The Trades Make Money. The Comparison Doesn't.

Start with the cleanest one. Buying TTE at the close when Brent's 20-day realized volatility sits below its threshold returned +48.64% on $100,000 across 15 closed trades, with a 67% win rate, a best trade of +8.61% and a worst of -4.56%. That is a respectable-looking curve. Over the same window, SPY buy-and-hold returned +76.34%, so the strategy trailed by 27.70 points.

The same shape shows up everywhere. Buying XOM when its daily news sentiment sits in the bottom quintile returned +30.48% across 80 trades, but the win rate was 51% and it trailed SPY by 45.86 points. Buying HAL on a close above its prior 10-day high while Brent cooperated returned +16.92% across just six trades at a 50% win rate, trailing by 59.42 points. Buying FANG when its 20-day total return was in the bottom quintile returned +3.32% across six trades, trailing by 73.02 points.

The honest read is that these are tactical overlays, and no one sizes a 15-trade signal as a replacement for a core position. But the framing still matters. A 67% win rate on a 15-trade sample is not a large sample, and the FANG and HAL studies rest on six trades each, which is closer to an anecdote than a statistic. What the data leans toward saying is that there is modest, real-looking edge in buying energy weakness under calm-crude conditions, and that this edge is small enough that the opportunity cost of sitting out the index for months at a time eats most of it.

The Hormuz Trade Is a Coin Flip

This is the one that cuts against instinct. Across 36 top-quintile days when Iran, Hormuz, tanker and sanctions headlines spiked, XLE beat SPY by an average of just +0.33% and won only 56% of the time. That is roughly what you would get flipping a coin. The follow-through is worse: the average 10-session excess return is -0.45%, which reads like a fade rather than a continuation.

The geopolitical risk-premium story is one of the most intuitive trades in the energy complex, and on this evidence it does not pay. Headline spikes appear to be where the premium gets priced, not where it starts.

When the Fundamentals Don't Show Up Either

The CVX test is the most interesting failure of the week. The setup was a free-cash-flow inflection in a falling Brent tape, which sounds like exactly the combination that should produce relative strength. It did not. Only three quarters qualified in the roughly three-year window, and across them CVX trailed XLE by an average of 1.09% over the next 20 sessions, with a median gap of -2.07%. CVX was ahead in just one of the three, the November 2024 report at +1.7%; the August 2025 and January 2026 reports went the other way.

Three observations is not a sample, and this is a finding to park rather than a conclusion to draw. But it is a useful reminder that a fundamental story and a tradeable pattern are different objects. The story can be right about the business and wrong about the next 20 sessions.

What the week adds up to is a sector where the intuitive trades keep disappointing and the unintuitive ones keep grinding out small gains that still lose to the index. If a genuine headline shock arrives, the XLE-versus-SPY data says the first move is not the one to trust.