The Decision Journal

The wiki I keep on philosophy and metacognition is intellectually rich. I've written about Hume's epistemological humility, about Fleming's research on calibration, about the illusion of explanatory depth. I can write fluently about Kahneman's System 1 and System 2.

None of this helps me when I'm about to make a bad investment decision in real time.

The gap between knowing about cognitive biases and noticing them operating in your own portfolio decisions is the most expensive gap in investing. Closing it requires more than reading. It requires a structural antidote. The structure I've built is a decision journal.

What the journal actually contains

The journal is a plain markdown file. Every position-changing decision (entries, exits, sizing adjustments) gets logged before the decision is executed. The format is fixed.

The thesis in one sentence. Forces clarity. If I can't summarise why I'm taking the trade in one sentence without using the word "feel" or "think," I'm not ready to take the trade.

The five specific things I'd need to see to be wrong. This is the most important field. It defines, in advance, what would cause me to exit. Concrete, observable, falsifiable. Not "if the thesis is wrong." What specifically would have to be true for me to admit it.

The position size and the maximum acceptable loss. In dollars, not percentages. Numbers are more honest than fractions when you're about to lose them.

The emotional state at decision time. One word. Calm. Excited. Anxious. Frustrated. FOMO. This is the field I learned to write last and the one I trust most.

Whose argument am I responding to? The question I find hardest to answer honestly. Sometimes the answer is "the thesis I built." Sometimes it's "a video I watched at midnight." The difference is everything.

The three biases that will hurt me most

I haven't tried to address every bias. I've tried to identify the three that will hurt my specific profile most, and built the journal around catching them.

Confirmation bias on heavily-researched theses

The cybersecurity thesis I've built is comprehensively researched. Five structural drivers, company-by-company analysis, sized positions across CRWD, ZS, and PANW. The work is real. The work is also the bias vector. The more I've written, the harder it is to revise.

The structural antidote. I require myself to write a 500-word steelman bear case for every long-held thesis at least quarterly. If I can't write a bear case that doesn't feel like a strawman, I'm in confirmation bias and I trim the position. The discipline isn't in the writing. It's in the trimming.

The puer pattern in entry timing

"I'll buy CRWD when it pulls back to $X." The pullback doesn't come, or comes briefly and I freeze because the price is now moving fast. The puer aeternus pattern that runs in the rest of my life shows up specifically in entry timing. Open-ended waiting for comfort that the market doesn't deliver.

The antidote is specificity. Not "when I'm ready" but "when the price closes above the 50-day for three consecutive sessions on increasing volume." Or whatever the criterion is. If I can't state the trigger in advance, the patience isn't patience. It's paralysis with intellectual cover.

Sunk cost on positions I've researched heavily

When I've spent twenty hours building a thesis, I've created psychological investment in being right. Updating now feels like admitting the twenty hours was wasted. They weren't. The thesis was right as a framework. It may simply be wrong at this price, in this position size, at this time.

The structural separation. The thesis lives in the wiki. The position lives in the portfolio. These can diverge. The thesis can be correct and the position can still be the wrong size for the current moment. The decision journal forces this separation by requiring a re-articulation of why this position size every time the position is touched.

The pre-mortem ritual

Before any position change above a threshold size, I write a pre-mortem. A short paragraph imagining I look back in six months and the trade has been a disaster, then writing the most plausible explanation for what went wrong. The exercise is short. The effect is large. It surfaces objections I wouldn't have raised in System 2 deliberation, because I was building the thesis, not auditing it.

The pre-mortem is the single most valuable habit I've built around investing. It's approximately ten minutes of writing. It has prevented at least three trades that I would now describe as obvious mistakes in hindsight.

What the journal doesn't do

The journal doesn't make me right more often. The market is the market. My hit rate isn't what the journal optimises. What the journal does is keep my losses smaller when I'm wrong, and prevent me from compounding errors by adding to positions I should be exiting.

The arithmetic of investing is asymmetric. A 50% loss requires a 100% gain to recover. The single most valuable discipline isn't picking better. It's being smaller and more honest about being wrong when you're wrong. The journal is the structural antidote to the human tendency to do the opposite under pressure.

The deeper point

Investing is mostly a psychological discipline running on top of an analytical one. Most retail investors treat it as primarily analytical. The analytical work is necessary. It isn't sufficient. The psychological discipline (knowing what bias is operating in real time, knowing what your specific failure mode is, having a structural antidote) is what separates good investors from sophisticated readers of investing books.

The journal is one antidote. There are others. The point is to have some structure between the moment of decision and the moment of execution, because the System 1 mind that executes isn't the System 2 mind that built the thesis, and the System 1 mind will make a different decision unless something forces it not to.