6 / 6 · 6 min
When to change size: rules instead of feelings
Position size changes by a rule written in advance, not by mood. The difference between those two is the whole difference between having a system and not having one.
People almost always change size, and almost always at the wrong time. After three wins in a row they feel right and double up — just before a losing run. After three losses they cut — just before the recovery. The feeling of confidence has no link to future results, but it does have one to past results, which is why it is always late.
The only argument for increasing
A growing deposit. With percentage risk this happens by itself: the account grows 20% and positions grow 20%, and you decide nothing. That is the correct way to scale up: slowly, alongside the result, without a single decision taken under an impression.
Increasing after a winning streak
Three wins in a row do not make the fourth more likely — at a 40% hit rate it still comes in 40% of cases. Yet the person raises risk from 1% to 3% precisely on that one, because "it is going well". Then the ordinary losing run arrives, and now it costs not 7% of the deposit but 21%. The size changed, the expectancy did not, and the drawdown tripled.
The price of one "confident" entry
Twenty trades at 1% risk with +0.4R expectancy give roughly +8R, that is +8% of the deposit. One trade where the person put on 6% instead of 1%, and it lost, takes 6% — three quarters of all the work of twenty trades. And this is not a rare case but the typical one: the large bet is almost always placed on the entry the person is "especially confident" about, and confidence is weakly linked to probability.
Reducing is allowed and necessary
A reduction rule is the only size change it is acceptable to make in response to results. It works like this: on reaching a pre-set drawdown, risk per trade halves and returns to normal only after the account has recovered to some level. The point is not to win it back but to survive until the system starts working again.
An example ladder, set in advance
| Drawdown | Risk per trade | What I do |
|---|---|---|
| 0–10% | 1% | normal mode |
| 10–15% | 0.5% | carry on, but half as loud |
| 15–20% | 0.25% | minimum size, review the journal |
| over 20% | 0 | stop, full system review |
Why the return must be slow
The temptation to jump back to the old size after the first profitable trade is enormous — and it is what turns the ladder into nonsense. A sensible rule: step back up only after the account has recovered half the distance it fell. Descend fast, climb slowly: that way the ladder actually protects rather than pretending to.
What is NOT a reason to change size
- "This trade looks especially good" — every trade in your system looked good, otherwise you would not take it.
- "I need to win back yesterday's loss" — yesterday's loss already happened and the market does not know about it.
- "The market is clearly going up today" — if confidence worked, everyone would know about it.
- "I have not traded for a while and want to catch up" — you cannot catch up, only start again at the same risk.
The journal in your account shows the size and result of every trade. Sort by size and look at the largest ones: if your biggest positions performed worse on average than your ordinary ones, you are increasing size at the worst possible moment — and it is visible in your own data.
About the trade journalWrite down your ladder
Build a four-row table: at what drawdown, what risk, and what you do. Write it where you will definitely look: in notes, on paper by the screen, in your strategy description. Not in your head — in your head it will have changed by the third loss. The only job of writing it down is that it exists BEFORE it is needed.
Three winning trades in a row. Increase size on the fourth?
No. Three wins do not change the odds of the fourth: if the system is right 40% of the time, the fourth trade is right 40% of the time regardless of what came before. Size grows on its own alongside the deposit — that is the only growth that is earned. Everything else is a bet on the streak, and streaks do not continue on request.
Drawdown 12%, your ladder says drop to 0.5%. But you see an excellent entry. What now?
Take it at 0.5% risk. The ladder does not forbid trading — it lowers the cost of a mistake during a period when mistakes are arriving more often than usual. An excellent entry at 0.5% pays half as much, but also costs half as much if it turns out to be less excellent than it looked. That case is exactly why the ladder was written in advance.