When to Increase Size After a Drawdown

Restore size on a written trigger decided before the drawdown, not on how the account feels. The two rules that work in practice are a new equity high, which is unambiguous but slow, and a fixed number of trades with expectancy back in line, which is faster but depends on having an honest record.

Most drawdown plans cover the cutting and go quiet here. That gap is where recoveries turn into the next drawdown, because in the absence of a rule size goes back up on confidence, and confidence is the one input with no bearing on whether the edge has returned.

Why the restoration rule is the harder half

Cutting size during a losing run is uncomfortable but obvious: the trigger is a drawdown threshold, the account tells you when it is reached, and the decision is forced.

Going back up has no such moment. Nothing announces that conditions have improved. What usually happens instead is that a few winning trades arrive, the pressure lifts, and size returns to normal somewhere between the third and fifth of them — not by decision but by drift. If those winners were the ordinary variance inside a method that is still out of step with the market, full size has been restored precisely in time for the next run of losses.

The rule exists to take that decision away from the moment it is made in.

Trigger one: a new equity high

Restore one step of size when the account prints a new high-water mark.

What recommends it: it cannot be argued with. The account either made a new high or it did not, there is no estimate involved, and it is impossible to talk yourself into it early.

What it costs: time, and sometimes a lot of it. Recovering a 20% drawdown while trading at half size means the method has to produce roughly twice the gross return to get back, so the rule keeps you small through exactly the period when the edge may already have returned.

This is the right default for anyone without a long, honest performance record, because it needs no inputs that could be wrong.

Trigger two: trades with expectancy back in line

Restore one step after a fixed number of trades — thirty is a reasonable minimum — during which measured expectancy is back inside its normal range.

What recommends it: it measures the thing you actually care about. The question is whether the method is working, and a sample of recent trades answers it more directly than the equity curve, which can be dragged by one outsized loss.

What it costs: it depends on your expectancy figure being real. Thirty trades is a short sample, and if the underlying win rate estimate was wrong to begin with, this trigger will fire on noise. Expectancy and sample size covers how much a figure measured over a short record can move.

Restore by the steps you came down

Whichever trigger you choose, apply it one step at a time.

If risk was halved at a 10% drawdown and halved again at 20%, then recovery runs in reverse: quarter size to half on the first trigger, half to full on the second. Each step earns its own trigger.

The failure mode here is specific and common. An account cut to quarter size has a good fortnight, and size goes straight back to full because the drawdown “feels over”. That single move multiplies exposure by four at the point where the evidence supports multiplying it by two, and it is routinely how the recovery becomes the start of the next decline.

Write it down before you need it

The rule has to be fixed in advance, because its whole purpose is to be binding at a moment when your judgement is predictably poor.

A workable version, written out:

1. Risk halves at 10% from the equity peak.
2. Risk halves again at 20%, and trading stops for a journal review.
3. One step of size returns at each new equity high, or after thirty trades with expectancy inside its usual range, whichever comes first.
4. No step is skipped, regardless of how the recent trades have gone.

Point four is the one that gets broken, which is why it is written as its own line.

Record the trigger, not just the trade

A restoration rule only binds if there is evidence it was followed, which means writing down the moment it fired and what fired it.

Three lines in the journal are enough: the date, which trigger was met, and the size step that resulted. A new equity high on 14 March, quarter to half. Thirty trades closed with expectancy at 0.3R, half to full.

Two things come out of that record over time.

It exposes drift. If size reached full without a matching entry, size went up by feel. That is worth knowing, and it is invisible without the log, because the equity curve looks the same either way until the next drawdown.

It makes the thresholds improvable. After two or three cycles there is a record of how long each recovery actually took. A trigger that has never fired in two years is too strict for the method; one that fires after every good fortnight is too loose. Neither is apparent from a single drawdown, and both are obvious across several.

This is the same argument as keeping a journal at all, applied to the one decision that is hardest to make honestly in the moment. Recording emotions alongside it, as in trading journal: record your emotions, tends to show the pattern before the equity curve does.

When the trigger never fires

If a restoration rule based on measured performance never fires, that is the rule doing its job rather than failing.

A method that has stopped working will not produce a new equity high, and will not produce thirty trades with expectancy in range. Size stays down indefinitely, which is the correct outcome. The alternative — a calendar rule, restoring size after a month regardless — quietly puts full risk back behind an edge that has gone, and the account finds out the expensive way.

The arithmetic of how much that costs, and why position size matters more than any strategy refinement, is in risk of ruin. The short version is that surviving at reduced size leaves the option of trading later open, and nothing else does.