Forward Looking
The Next "Bottleneck"
Hello all and happy Sunday. Today I wanted to touch on a few things as we enter the final week of April, which has been nothing short of spectacular for us. Remaining forward looking, we recognize that we just caught some massive moves and rather than being greedy, it’s time to look for other trades, setups, & themes that are poised for their own respective run ups.
If you have been MIA for an extended period of time, you may have missed that we do have a discord for better organizational structure of our daily discussions and trade tracking versus the former of filtering through various substack chats as they can become quite convoluted quickly. If you’d like to join, please follow the link posted below. (Note: this is for paying subscribers only, where we chat all day long & even the weekends about current events, positioning, market sentiment, and various other discussions to provide as much value as possible.)
SPX closed Friday at 7,165, a record high. Nasdaq parked at 24,837, also a record. Russell finally showed up at 2,787, and the Dow lagged at 49,231 weighed down by Merck, IBM, and Verizon. VIX bled to 18.71, four-handle compression on the week. Fourth consecutive weekly gain for the S&P, and the tape is squarely in melt-up posture.
On March 31st, we noticed a major shift in the tape of options flow, which ultimately lead us to believe things were changing in a big way, and that we could have major opportunities immediately. This was the turning point & went over looked by many before surging nearly 12% in 3 weeks time.
The driver Friday was Intel. INTC ripped 24% to a record close, the best single day since 1987, on a Q1 print that destroyed estimates ($13.6B revenue, $0.29 non-GAAP EPS vs. $0.01 expected) and a Q2 guide ($13.8B to $14.8B) that came in nearly $1.4B above consensus. Data Center & AI revenue grew 22%, foundry posted 16%, and Lip-Bu Tan said the quiet part loud on the call: the CPU is reinserting itself as the indispensable foundation of the AI era.
On the macro layer: Iran ‘reopened’ (lol) the Strait of Hormuz, peace talks moved to Islamabad with Pakistan mediating, and Trump ‘extended’ the Israel-Lebanon ceasefire by three weeks. Crude is back to $94 and sliding. The fear premium that defined Q1 has fully rotated into AI infra and CPU bid.
The under-the-radar print this week was the DOE issuing its first $1.6B loan to AEP for a five-state grid upgrade, dropped the same week as the Defense Production Act memo on grid infrastructure. That is the project-level catalyst we said to watch for. The framework is operational. This is the trade we want to be early on across multiple legs.
Sentiment & Breadth
NAAIM Exposure Index printed 94.15 for the week ending April 22, jumping from 79.49 the week prior. One of the largest weekly moves of the cycle, pushing active managers into aggressive bullish positioning. Anything above 90 historically flags chase behavior. Not a sell signal on its own, but it tells you who is doing the buying up here: it is professionals who were underexposed and are now playing catchup.
FactSet has 28% of the S&P reported. Q2 to Q4 2026 EPS growth estimates running 20.6%, 22.7%, and 20.4%. Full-year 2026 at 18.6%. Forward P/E at 20.9 (above the 5-year average of 19.9). The earnings tape is doing the heavy lifting. Valuation is full but defensible while EPS keeps printing.
Translation for our positioning: stop adding fresh longs into strength on the names that have already paid, and start letting setups come to you on names that have not. Tighten stops on runners, take size off the closed-out winners, and rotate dry powder into the new themes that are still undervalued by the tape. We do not chase 94 NAAIM readings. We let them mean-revert into us.
April: The Best Month YTD
April is the standout. We caught the CPU renaissance before it was a consensus call. The thesis was simple: the AI buildout was overpriced into GPUs and underpriced into the silicon that runs inference, the orchestration layer, and the agent workloads. INTC and ARM were the two cleanest expressions. The other layer was AI infrastructure pull-through (CRWV, NBIS, DELL), where unusual flow had been screaming for weeks before price followed.
Closed positions for the month, 27 trades, 22 winners, 5 losers, an 81.5% win rate. Multiple INTC 60C 6/18 lots scaled out at 195%, 286%, and 596% as the position ran into earnings. ARM was the bigger fireworks, with 180C 6/18 lots closing at 219%, 434%, and a 903% top tick on the runners. Below is the breakdown:
Five losses, all sized small relative to the winner book. AMD and META were quick cuts taken without ego. NEE was a utility chase on the wrong leg of the curve, lesson booked. TSEM had one win and one loss inside the same name, which is exactly how foundry exposure trades when the Street is split.
VST is the one that deserves a closer look, because it is the difference between an expression failure and a thesis failure. The May 175C got stopped out for -22% on a timing window that closed too tight. That is the trade. The equity position stayed on the book and the underlying thesis stayed alive. We are not relitigating the loss. We are sizing up the thesis. More on that in the next section.
The structural takeaway: high win rate is nice, but the math that wins the year is winners running 200% to 900% while losers cap at 5% to 45%. We do not need to be right more than 60% of the time when the asymmetry is this stark. That ratio is the entire game. Although, we did, and we were right over 80% of the time.
Let’s dive into some new / fresh ideas presenting similar risk reward ratios going forward.
