$ why did I buy this?
Your portfolio remembers what you bought.
Nothing remembers why.
A journal for what you own — and what you are still deciding about. Write the thesis, score how sure you are, and record what would prove you wrong. Months later it brings all three back, before hindsight rewrites the story.
what would prove me wrong
Active buyer numbers decline for two consecutive quarters once comparisons normalize.
- 2024-09-20thesisconv 7/10The habit formed during lockdown outlasts the lockdown. Handmade is a category, not a moment.conv 7/10
- 2025-05-18updateconv 5/10Buyers fell one quarter. That is half my falsifier. Watching the next print rather than deciding now.conv 5/10
- 2025-08-06updateconv 3/10Second consecutive decline. The falsifier I wrote down ten months ago has fired exactly as specified.conv 3/10
- 2025-08-10exitconv 2/10Sold. Not because it fell, but because the thing I said would prove me wrong happened. Worth noticing how strongly I wanted to reinterpret my own sentence.conv 2/10
write the next one
the journal this came from →A journal that pushes back
You could write all of this in a notes app. What you could not do is make it hard to lie to yourself later.
You have to say what would prove you wrong
A note lets you write only the bull case. Here it is its own field, and it is the question you get asked at review time — not a paragraph you can quietly skip.
Your old score is hidden while you review
Seeing that you once said 8 anchors whatever you write next. The previous conviction stays withheld until you commit a new one.
The review comes and finds you
A note sits where you left it. Set a date and the thesis arrives by email when it falls due — and positions you have simply stopped writing about resurface on their own.
No recommendations, no signals, no feed. No live prices — closes only, at least a trading day old, so nothing moves while you read. What it does and does not do →
Does conviction predict returns?
For every buy in the demo journal, the conviction recorded before it, weighted by money committed, against what happened since. Nothing here is a projection — it is what already happened.
Your lower-conviction buys have done better — 5–6 leads 9–10 by 45.8 points. Worth asking what the confident ones had in common.
- 5–6on the fence6 buys+22.71%
- 7–8confident4 buys-6.26%
- 9–10all in3 buys-23.14%
Write the next one before you know how it ends
The record only works if it goes in before the outcome does.
No card. First 10 positions free — see pricing →