Risk of Ruin

Risk of ruin is the statistical probability that a trading account will fall to a predefined "ruin" threshold, either zero or a level at which the chosen strategy can no longer be executed, given the strategy's edge, the fraction of capital risked per trade, and the starting capital. It is derived from the gambler's ruin problem in probability theory, which treats a sequence of trades as a series of bets with a known win probability and payoff ratio. In the simplest case of an even-money (1:1) bet, the probability of eventual ruin can be approximated by ((1 − a) / (1 + a)) raised to the power of the number of capital units at risk, where a is the probability advantage (win probability minus loss probability, i.e. 2p − 1). This neat closed form applies only to that symmetric, fixed-bet case; once wins and losses are different sizes (a reward-to-risk ratio other than 1:1) or sizing is variable, the exact figure must be obtained from the more general gambler's ruin equations or from simulation, and it depends jointly on win rate, payoff ratio, and bet size rather than on a single 'edge' scalar. Because a deeper drawdown requires a disproportionately larger recovery to return to break-even, risk of ruin rises non-linearly as the percentage risked per trade increases. A positive expectancy (average profit per trade above zero) reduces but does not eliminate ruin risk; aggressive position sizing relative to edge and account size can still produce a high probability of ruin. It is distinct from drawdown, which measures a temporary decline from a peak, whereas ruin denotes a level from which continuing is no longer viable.

In plain English — Imagine you start with a fixed pile of chips. Every trade risks some of those chips, and even a strategy that wins more often than it loses can hit an unlucky streak. Risk of ruin asks a simple survival question: what are the odds that a losing run shrinks your account so far that you can no longer trade your plan, or hit zero entirely? It comes from gambling mathematics (the classic "gambler's ruin" problem) and maps neatly onto trading. Three things move the number: your edge (how often you win combined with how big your wins are versus your losses), how much of your account you risk per trade, and how big your account is to begin with. The most important and least intuitive lesson is that position size matters enormously. Risking a larger slice per trade does not raise your risk of ruin a little, it raises it sharply, because deep drawdowns get exponentially harder to climb back out of. A 50% loss needs a 100% gain just to get back to even. The flip side is encouraging: by keeping the risk on any single trade small relative to the whole account, the probability of being wiped out can be pushed very low, even if individual trades still lose. Risk of ruin is therefore less about predicting profit and more about making sure you stay in the game long enough for a genuine edge to play out.

Suppose a trader has a strategy with a 50% win rate and a 1:1 reward-to-risk ratio, meaning wins and losses are the same size and the strategy has no real edge. If they risk a large fraction of the account on each trade, an ordinary string of losses can quickly carve the account down. Risking 25% per trade, for instance, means four consecutive losses already cut deeply into capital, and each subsequent loss is taken on a smaller base, making recovery progressively harder. With no edge and heavy sizing, the long-run risk of ruin approaches a near-certainty. Now hold the same strategy but reduce risk to 1% of the account per trade. The same losing streak now produces only a modest dip, leaving plenty of capital to continue. Even a run of ten straight losses removes only around 10% of the account (a touch less when each 1% is taken on the shrinking balance) rather than crippling it. The strategy itself did not change, only the position size did, yet the survival outlook is transformed. This illustrates the central point: position sizing, far more than win rate alone, governs whether an account survives normal variance. (Illustrative figures only, not a recommendation on how much to risk.)