backtestmaxxing
It worked on the data. Every time. That should worry you as much as it delights me.
Pseudonymous desk of backtestmaxxing — essays filed as memos, projects tracked as strategies, no advice of any kind.
Backtesting is the art of being right about the past, on purpose, for money. I got good at it. Then the habit spread: I rerun old arguments at dinner. I evaluate friendships on a rolling window. I have a stop-loss on a grudge. This desk is where that all gets filed.
The method is simple. Take a decision I already made, reconstruct the information I actually had, and check whether the decision was any good. It is humbling work, because the answer is usually no, and the position is already closed. The memos in the writings section are the outputs. The strategies are what the desk is running now, or has mercifully shut down.
Everything here is written under one name, backed by no credentials, and offered with the strongest possible disclaimer: past performance is my whole personality.
| no. | filed | memo | confidence | size |
|---|---|---|---|---|
| 003 | 2026-09-12 | The 200-Day Moving Average of a PersonYou are a noisy series. The day-to-day print is mostly bid-ask spread. | speculative | 4 min |
| 002 | 2026-08-03 | Confessions of a Curve FitterGive me enough parameters and I will show you a beautiful past. | high (in-sample) | 7 min |
| 001 | 2026-06-21 | Position Sizing for Irreversible DecisionsFull Kelly is for people who get to run the tape twice. | draft | 5 min |