Setups2 min read
The gap-down filter earned its keep
The single cheapest line in the screen is the one that throws out overnight gaps. Last week it excluded four names and every one of them kept going.
Year Archive
2026
9 posts published in 2026.
Market Notes2 min read
Vol is cheap and nobody is scared
Implied volatility is the one series in this business that reliably reverts. When it is cheap and the tape is calm, the structure of the trade should change before the thesis does.
Method4 min read
The dog and the leash
Why this site exists, and the one metaphor it keeps coming back to: price runs, value walks, and the distance between them is the only thing worth measuring.
Trade Reviews2 min read
The trade that worked for the wrong reason
Closed green, and the review grades it a loss. The thesis was reversion; what actually happened was a sector bid that had nothing to do with the setup.
Method4 min read
What mean reversion actually means
"It fell a lot, so it should bounce" is not mean reversion — it is hope with a chart attached. The real claim is about stationarity, and it is testable.
Three names two sigma below, one worth a look
The screen returned three candidates this morning. Two fail the anchor test on sight. Here is the reasoning on all three, including the ones I'm skipping.
Method4 min read
Building a z-score screen that isn't just a bad-news detector
A naive oversold screen returns the week's worst headlines. Four filters turn it into a list of stretched prices — with the code and the reasoning for each.
Market Notes2 min read
Breadth is narrowing while the index isn't
The index sits near its highs and the average stock inside it does not. That gap is where reversion candidates come from — and it is also how you get run over.
Method4 min read
The three ways a mean-reversion trade dies
High win rates hide the shape of the loss. Here are the three failure modes, what each one looks like on the way in, and the sizing arithmetic that survives them.