The discipline loop: theses, journals, and honest exits
Quick quiz. Why do you own that position you bought a year and a half ago? Say it out loud. If your answer arrives as a tidy story that conveniently matches how the trade turned out, I have bad news. That is not your reasoning. That is your memory doing PR for you after the fact.
Here is the thing the finance industry would love you to keep believing: that the market is the villain, and if you just had a better forecast you would win. It is a flattering story because none of it is your fault. It is also mostly nonsense. Most investors do not get beaten by the market. They get beaten by their own unexamined process, repeating the same mistakes on a loop because nobody ever wrote the mistakes down. The good news, and there is genuinely good news here, is that a process is the one thing you can actually fix.
Write down why you own the thing
Radical idea: before you buy something, say why, in writing, where you cannot later pretend you said something cleverer. A thesis in Strata is a living document stapled to a ticker. The drivers. The key assumptions. How big the position is and why. And the part everyone skips because it stings, the invalidation rules: the exact conditions under which you agree, in advance, that the idea is broken and you were wrong.
Each thesis carries a conviction score you revisit as the world changes, and every change you make gets logged to a timeline. No quiet edits. No tiptoeing back to rewrite history so past-you looks like a genius. The receipts stay. This is the same reason accountants do not use erasers, and it is exactly the standard your own money deserves. We are stubborn about this on purpose: Strata is manual by design.
Signals keep the thesis honest so you do not have to
Writing it down is step one. Step two is admitting that future-you is a slippery character who will talk themselves into holding a losing idea. So you do not leave it to willpower. Each thesis carries signals, a plain metric condition like gross margin staying above 70 per cent, each one showing its current value and a state: OK, WATCH, or BREACH. When a signal breaches, the thesis flips itself to AT-RISK. No nudge required from you.
Read that again, because it is the whole trick. The system nags you at precisely the moment your past, calmer, more honest self said you would want to be nagged. Not when you happen to glance at the screen. Not when the loss finally gets loud enough to ignore. The moment the line you drew gets crossed. These are real metric conditions, not vibes, the same transparent inputs behind factors without black boxes.
The journal closes the loop
Now the bit that turns all this from a nice diary into an actual edge. Before the trade, you log the plan: entry, stop, target, size. After, you log what really happened. The journal lines them up side by side, plan versus actual, and adds the numbers your memory is far too kind to keep honestly: MFE and MAE (how far it ran your way, how far it ran against you), your realised R, and an emotion scorecard for conviction, patience, discipline and that old gremlin, FOMO.
Do this enough times and your win rate and your expectancy stop being feelings and start being facts. You learn, with evidence, that you are great at one kind of setup and a menace to your own wealth at another. That is worth more than any forecast.
And notice what none of this needed. No crystal ball. No prediction about next quarter. Just a place where your reasoning lives, ages, and gets politely confronted with what actually happened. That is the discipline loop, and it is the quiet engine of the whole workbench.