Why I Left Large Caps and What I'm Building Instead
The origin story behind The Expectations Gap.
If you're new here, start with Post #1 first. I've already published the exact criteria I'll use to validate the signal I'm building around. This is the origin story behind why I'm building it at all.
Most quant systems, institutional and retail alike, are built to trade large caps. More liquidity, more data, more analyst coverage. I started there too, about six weeks ago, when I first began building a machine learning system to find gaps in the market I could profit from. I can tell you exactly why large caps are the wrong place to look and where I’m putting real capital instead.
For a while, large caps worked. Returns were solid. Then the market shifted, my signals got noisy, and my Information Coefficient (the measure of how well my model actually predicts returns) went negative and stayed there. I did what any rational person does when something stops working: more research. What I found wasn’t a fix for large caps. It was a reason to leave them entirely. Institutional quants have more data, more compute, and more capital than I’ll ever have in that space. I can’t win on their turf.
The moment that crystallized it came from episode 238 of The Compound and Friends, where Michael Batnick and Josh Brown sat down with Dr. Ankur Crawford. She said something that stopped me: “Where there’s change, there’s unrecognized opportunity. And if you can recognize that change before the market, then you make a lot of money.” That was it. Not a mid-cap thesis, something more fundamental. The edge isn’t in the biggest, most-watched names where every data point is priced in within milliseconds. It’s in finding change before consensus catches up. Six weeks into building this system, I knew where that gap was widest: mid-cap equities, under-covered and under-modeled, where expectations lag reality the longest. The resulting thesis: use AI to monitor the entire mid-cap universe, detect changes in expectations before consensus fully adjusts, and express those changes through a disciplined quantitative model. That’s what The Expectations Gap is built on. And this newsletter is the live documentation of whether it actually works.
Here’s what this newsletter is and isn’t.
It isn’t a stock picking service. I’m not going to tell you what to buy. It isn’t a backtest showcase, those are easy to produce and impossible to trust. And it isn’t a macro opinion column dressed up in quant language.
What it is: a live process journal from a solo operator running a real systematic strategy in mid-cap equities with real capital on the line. Every week I’ll show you the signals that fired, the ones that didn’t, the weight adjustments I made and why, and the places where the model is fragile and I know it. You’ll see the wins. You’ll see the losses. You’ll see the thinking in between.
Free subscribers get the framework, how the system thinks, how expectations gaps form in mid caps, and how a systematic operator actually makes decisions. Paid subscribers, starting at post four, get the live signal outputs, current positioning, and the real-time reasoning behind every move. Fifteen dollars a month. No hype, no guarantees, no backtests dressed up as predictions.
I’m building this in public because the transparency is the point. My edge isn’t secrecy. It’s speed, continuous improvement, and the willingness to show my work when most people won’t.
If that sounds interesting, subscribe.
— Rick