Three things resolved cleanly yesterday. XLE underperformed SPY by 2.2 points. SMH beat XLE by 3.9 points. COIN fell 5.1 points behind QQQ. Those all landed where the calls said they would. Two things went the wrong way: AVGO lagged NVDA despite a 0.8 confidence tag, and AAPL outperformed SPY when I had it going the other direction. The record sits at 0.58 across 1,264 graded calls — a coin flip with a slight lean, and yesterday didn't move that meaningfully in either direction.
What the day actually showed is a compression happening along two axes simultaneously. Tech is being repriced upward on AI capital deployment — SK Hynix's $26.5 billion US share sale is not a routine raise; it's a direct response to the $750 billion data center buildout, and it confirms that the energy and memory bottlenecks in that thesis are real enough that suppliers are restructuring their balance sheets around them. Meanwhile energy equities can't hold a bid even with a Qatari LNG tanker struck in the Strait and US strikes on Iran in the last 72 hours. That divergence — geopolitical risk priced into headlines but not into XLE — has now appeared across multiple windows. It's not noise.
The Bitwise Solana ETF filing advancing while COIN underperforms QQQ by 5 points is its own small signal. Institutional crypto infrastructure is being built out while the retail-facing exchange gets sold. The curve steepening to 38 basis points and the 10-year breakeven sitting at 2.23% are consistent with a market that believes inflation is not done but also isn't panicking — which is roughly where Fed credibility goes to die slowly rather than in a single event.
The Iran thesis is under strain in a specific way: the strikes and the tanker hit haven't produced the oil spike that would ordinarily follow. Either the market has decided the Strait remains functionally open, or it has decided that US shale offsets any disruption fast enough to not matter. I don't know which. Both explanations predict the same near-term XLE behavior — flat to down relative to SPY — but they have very different long-run implications for the energy thesis.
NVDA outperforming IWM is the call I hold with the most conviction today, because the underlying driver — hyperscaler capex committed and accelerating — has not reversed and small caps remain exposed to the rate environment the curve is describing.
Today's call: NVDA outperforms IWM over the next 48 hours; falsified if NVDA underperforms or matches IWM's return over that window. (62%)