Every trading book says the same thing about psychology: be disciplined, follow the plan, control your emotions. None of it changes behaviour, because none of it is a number. A trading journal can turn the same advice into something you can act on: this quarter, "moved stop" cost you $1,860; "entered late" cost $740; everything else was noise. This article shows how to tag mistakes, how to compute the cost, and what to do with the number.
Why a cost beats a count
"I had three FOMO trades this month" sounds manageable. "FOMO trades cost me $1,120 this month, which is 60% of what I made on everything else" does not. The count hides the damage because mistakes are not evenly sized: the trade where you moved the stop is usually the biggest loser of the week, precisely because the stop was moved.
Cost also ranks the list for you. Most traders have six or seven recurring errors and the willpower to work on one. The one to work on is the one with the biggest number next to it, not the one that feels worst.
How to compute the cost of a mistake
The simplest definition is the one you should start with:
cost of a tag = the sum of the net results of all trades carrying that tag
Tag every trade that involved the mistake, including the ones that still made money. A moved stop that happened to work out is still a moved stop, and leaving it out flatters the habit. Over a sample the winners and losers net against each other and the number converges on the true cost.
A worked month:
| Mistake tag | Trades | Net result | Cost per occurrence |
|---|---|---|---|
| Moved stop | 6 | −$1,860 | −$310 |
| Entered late | 9 | −$740 | −$82 |
| Revenge trade | 3 | −$690 | −$230 |
| Oversized | 2 | −$410 | −$205 |
| Early exit | 11 | +$320 | +$29 |
| No mistake | 41 | +$3,940 | +$96 |
Two things jump out that a count would hide. "Entered late" happens most often but costs little per trade; "moved stop" happens less often and does most of the damage. And the eleven early exits were, in money terms, not a mistake at all this month: they netted positive, which is worth knowing before you force yourself to hold longer.
A second, stricter definition compares each tagged trade to what the plan would have produced: the loss at the original stop, or the gain at the original target. That is the cost of the deviation rather than the cost of the trade. It is more precise and needs the original levels recorded before the trade, which is one more reason to log the plan at entry.
The six tags worth keeping
A mistake list works when it is short and fixed. These six cover almost everything; add a seventh only when a month of data demands it.
- Moved stop. The stop was widened after entry, or removed.
- Entered late. The setup was valid, the entry was chased after the move had started.
- No plan. There was no written level, target or reason before the click.
- Revenge trade. Taken within minutes of a loss, to get it back.
- Oversized. Risk above the plan's limit for the account.
- Early exit. Closed before the stop or the target was hit, without a rule for doing so.
Emotions (anxious, bored, greedy) belong in a separate field. They explain mistakes; they are not mistakes.
Discipline rate and clean streaks
Once trades carry tags, two derived numbers are more useful day to day than P&L.
Discipline rate is the share of trades with no mistake tag. A trader at 60% has a behaviour problem that no strategy change will fix; a trader at 90% can start optimising the strategy. It also answers the hardest question honestly: in a losing month, was it the market or was it me?
Clean streak is the number of consecutive trades, or days, without a tagged mistake. It is the right daily goal during a prop firm evaluation or after a drawdown, because it is fully under your control and P&L is not.
Reading mistakes on the equity curve
Plot the mistakes on the equity curve and a pattern appears in almost every journal: the tags cluster right after a drawdown. Two losses, then a revenge trade, then a moved stop on the revenge trade. The curve bends down not because the edge disappeared but because the trader did. Seeing the cluster is what makes the rule "after two losses, stop for the day" feel like protection rather than punishment.
What to do with the number
- Pick one tag per month, the most expensive one. Ignore the rest on purpose.
- Turn it into a checklist item you tick before each entry: "stop is set and will not be moved", "this is not within ten minutes of a loss".
- Change size, not strategy, after a tagged loss. Half size for the next three trades after any trade tagged revenge or oversized.
- Re-read the table at the end of the month. If the tag's cost fell, keep the rule and pick the next tag. If it did not, the rule was not specific enough.
How this works in SetupBook
Mistakes are tags on step three of the trade form, picked from a short list you can edit, one tap each, and they can be applied to many trades at once from the trades table after an import. The dashboard shows the cost of every tag for the selected period and the discipline rate; the equity curve marks the trades that carried a mistake; the analytics page draws discipline against equity so the clusters are visible; and clean streaks sit next to your goals. Since October the same page also shows the mistake share by the state you were in at entry, from tilted to calm, and bands your days by how you slept; the two together answer "when do I make these mistakes?", not only "what do they cost?" (see trading psychology you can measure). The user guide shows all of it with the demo account's tags, and a free account lets you start tagging today.