How much of your allowance should one trade risk?

Take an unusually good strategy and size it at 30% of the account's allowance. Its expected life is about sixty trades. The same strategy at 5% survives essentially forever. Nothing changed but the sizing.

Most accounts are not lost to a bad strategy. They are lost to a reasonable strategy carrying too much of the allowance per trade.

This is the least glamorous number in trading and the one that decides the outcome.

The arithmetic

Take a genuinely good strategy: 60% win rate, winners twice the size of losers. That is better than most things you will ever run. Now size it against a $2,000 allowance.

Simulated — expected life of the account, same strategy:

Risking 30% of the allowance per trade → about 60 trades
Risking 5% of the allowance per trade → on the order of millions

Nothing changed except the sizing.

That is not a rounding difference, and it is not intuition-friendly. Six times the risk per trade does not cost you six times the lifespan. It costs you essentially all of it, because ruin is a compounding path problem, not an average.

Why good traders get this wrong

Because the strategy is fine. Every instinct says look at the edge — the win rate, the reward-to-risk, the equity curve. All of those can be excellent while the account is mathematically doomed. The edge tells you where the equity curve goes. The sizing tells you whether you are still there when it arrives.

There is a second trap underneath it. On a small account the contract is the floor. If one micro contract represents $200 of risk against a $2,000 allowance, your minimum possible position is 10% of everything you have. You cannot choose 5%. The instrument has already chosen for you, and the only lever left is which instrument you trade and how wide the stop is.

What this means in practice

Three consequences, in the order they bite:

The account size determines the strategy, not the other way round. An engine whose ordinary bad month consumes three-quarters of a $2,000 limit is not a bad engine. It is an engine for a bigger account. We publish the drawdown of every engine as a percentage of a $2,000 limit for exactly this reason, and we decline to offer several of our engines on that account size on those grounds.

Two lots is not “a bit more risk”. It is the 30% column. One of our engines sits at roughly a 1.4% chance of breaching a $2,000 limit at one contract and 41.9% at two — while its historical worst drawdown still looks perfectly comfortable, because the historical ordering happened to be lucky.

Adding accounts is safer than adding size. Three accounts at one contract each keeps three separate limits and three separate failure events. One account at three contracts is a single limit taking triple the damage. The money is the same. The survival is not.

The formal version of this — that optimal position size must shrink as you approach a drawdown floor — is Grossman & Zhou (1993), Optimal Investment Strategies for Controlling Drawdowns. It is worth an afternoon.