Why Is 'What Should I Buy?' the Wrong Question?
Two traders can buy exactly the same stocks on exactly the same days and end the year far apart, because entries are the smallest part of a result. Give both a 55% win rate and identical picks: one exits at twice the risk and makes +0.65R a trade, the other cuts at half and holds losers to twice, and loses 0.625R.
What happens when two traders take the same trades?
Hold the picks constant and vary only the exits. Both traders enter the same 100 positions and both are right 55% of the time, because they bought the same things on the same days.
| Trader A | Trader B | |
|---|---|---|
| Entries | Identical | Identical |
| Win rate | 55% | 55% |
| Average win | +2R | +0.5R |
| Average loss | −1R | −2R |
| Expectancy per trade | +0.65R | −0.625R |
| Over 100 trades | +65R | −62.5R |
The gap is 1.275R per trade, or 127.5R over the hundred, and not one unit of it came from stock selection. Trader B took the same opportunities and converted them into a losing year by exiting differently.
Trader B's behaviour is not exotic, either. Taking a quick profit and giving a loser room are the two most natural things a person does in front of an open position: one banks a certainty, the other avoids a defeat. Both feel like judgement at the time.
Which is what shows up whenever anyone measures it. Across 66,465 households from 1991 to 1996, the most active traders earned 11.4% a year while the market returned 17.9%[1] — a gap produced by how they traded, since they were all picking from the same market.
So what is the right question?
Three of them, in order:
- What is my expectancy? Win rate times average win, minus loss rate times average loss. If it is negative, better entries are not the constraint.
- Where does the exit sit, and was it decided before or after entry? A target chosen while the position is open is not a target, it is a feeling with a price on it.
- Can I see what I actually did, rather than what I remember doing? Most traders cannot, and this is the one that has to be fixed first.
Why can't you answer the third one from your own notes?
Because the notes were written by the person being evaluated, in a file that stays editable, and the edits that matter are the ones nobody makes deliberately. The stop you "always intended" is recorded as the stop you always intended. Six months later the journal describes a more disciplined trader than the one who placed the trades, and there is no way to detect the drift from inside.
Does this mean picks do not matter?
No — a strategy with no edge produces nothing to execute well. The claim is narrower and survives the arithmetic above: entries set the ceiling, execution decides how much of it you keep, and traders who cannot make money almost always have more headroom in the second than the first. Check the expectancy before shopping for a new watchlist.
Sharper still: Odean showed that the stocks individual investors sold subsequently outperformed the ones they bought.[2] The selection was not merely unprofitable after costs, it was backwards — which is the strongest available evidence that picking is not where the constraint sits.
kappi is a trade recorder: you commit a trade before the fact, it is sealed on your device for a time-capsuled delay you choose, then kappi publishes it on a Merkle-anchored log. The seal is the whole mechanism — once a stop, a target and a thesis are sealed, revising them is no longer a private edit. $15/month, no free tier.
Sources
Frequently asked questions
Do stock picks matter at all?
They set the ceiling. But two traders with identical entries and a 55% win rate can differ by 1.275R per trade depending only on where their exits sit, so execution usually has more headroom than selection.
What is expectancy?
Win rate times average win, minus loss rate times average loss. A 55% win rate paying +2R and risking 1R is 0.55 x 2 - 0.45 x 1 = +0.65R per trade. The same win rate paying +0.5R against a 2R loss is -0.625R.
Why do traders cut winners and hold losers?
Because banking a small certain gain and avoiding a certain loss are both immediately comfortable. The arithmetic runs the other way, which is why the exit has to be decided before the position is open.
How does kappi help with exits?
The stop and target are part of the commit and are sealed before the fact. Changing them afterwards is no longer a private edit to a file, which removes the option rather than requiring willpower against it.