Evidence Trading
Risk and position size · Chapter 3 of 6

What a losing run actually does to an account

Foundation8 min read · part of Risk and position size

Two facts sit uncomfortably together. Losing runs are ordinary — a strategy that wins half its trades will hand you six losses in a row inside a few hundred trades, and that is expected behaviour, not a malfunction. And a drawdown does not recover symmetrically: down thirty per cent needs about forty-three per cent to get back, down fifty per cent needs a hundred.

Put those together and the case for small risk stops being a matter of temperament. It is arithmetic. At one per cent per trade a run of ten losses costs about a tenth of the account and is recoverable in the ordinary course of trading. At five per cent the same run costs about forty per cent, and now you need a stretch better than anything in your history just to be level.

This is why "risk of ruin" is a more useful frame than expected return. Expected return tells you where you end up if you survive. Risk of ruin tells you the probability you do not. A strategy with a real positive edge and a size that is too big for it has a genuinely high chance of ending at zero, and no amount of edge fixes that — the edge only pays out if you are still there.

The uncomfortable part is that the danger is invisible while things go well. Oversized trading and correctly sized trading look identical during a winning stretch; the only difference is what the eventual losing run does. That is why the size question has to be settled from the distribution rather than from recent results.

The practical takeaway is unglamorous. Pick the risk per trade so that the worst run you can reasonably expect leaves you both solvent and willing to keep following the plan. The second half of that sentence matters as much as the first: a drawdown that technically leaves you trading but breaks your willingness to take the next signal has ruined the strategy just as effectively.

What to take away

  • Losing runs are expected behaviour, not evidence the method broke.
  • Recovery is asymmetric: −50% needs +100%.
  • Size for the worst run you can reasonably expect, not the average one.

Where it goes wrong

  • Reading a normal losing streak as a broken strategy and changing everything.
  • Sizing from the last twenty trades instead of the distribution.
  • Treating survival and expected return as the same question.

This chapter, measured against your own trades

In the app the same chapter ends in your figures rather than an example: how often you did the thing it describes, over your last ninety days. You pick one change to make, and Evidence checks afterwards whether it actually changed — from your journal, arithmetic, no opinion involved. Questions you get wrong come back a week later and again a month after that.

Open the free plan →

Check that it stuck

Answers shown — in the app these are asked before you see them, and the ones you get wrong come back after a week.

An account is down 50%. What return is needed to get back to even?
100% — Halving requires doubling to reverse. This asymmetry is why large drawdowns are so much worse than they look.
Six losses in a row from a method that wins about half its trades means:
Nothing unusual — that run is expected within a few hundred trades — Runs of that length are ordinary in any coin-flip-like sequence. Treating them as a signal is how working methods get abandoned.
Why is risk of ruin a better question than expected return when choosing size?
Expected return only pays out if you are still trading; ruin asks whether you will be — Edge is worthless to an account that hit zero first. Survival is the precondition, not a side issue.
← Sizing backwards from the loss you accept The stop belongs to the idea →