Cutting Losses and Sizing Positions: The Math Beginners Skip

Start with the arithmetic that makes everything else obvious

Before any philosophy, one table. It explains why experienced traders sound almost obsessive about small losses, and why beginners think they are exaggerating.

The relationship is not linear, and that is the entire point. Small losses are recoverable through ordinary performance. Large ones require exceptional performance, arriving at exactly the moment your confidence and capital are both depleted. A 10% loss is a bad week. A 70% loss is a different life.

What a 1923 memoir still gets right

Reminiscences of a Stock Operator is a lightly fictionalised account of the speculator Jesse Livermore, written by journalist Edwin Lefèvre. It is a hundred years old, describes a market with no computers and few rules, and remains on nearly every trading reading list — for two reasons that have nothing to do with its trading techniques.

The first is its honesty about why traders hold losers. The narrator loses money repeatedly in the same way: a position moves against him, he decides the market is wrong, and he waits for it to agree with him. He knows better each time. He does it anyway. Reading a skilled professional make the identical mistake across a career is more instructive than any rule stated in the abstract, because it removes the excuse that this is a beginner's problem.

The second is its treatment of patience as a skill. The book's most quoted lesson is that the money is made in the waiting rather than in the trading — that being right about direction is common and staying with a correct position is rare. Most beginners are the mirror image: they hold losers patiently and take gains quickly, which inverts the arithmetic of the table above.

Position sizing: how the rule becomes a number

"Cut your losses" is advice everyone agrees with and few implement, because it is stated as a sentiment rather than a calculation. Here is the calculation. It is three lines of arithmetic and it is the single most useful thing on this page.

1. 1Decide the maximum percentage of your account you will lose on one idea. Many traders use 1%; for a long-term investor a larger figure can be reasonable. What matters is that the number exists before the trade.
2. 2Decide where the idea is wrong — a price level, or better, an event that would invalidate your reason for owning it.
3. 3Divide. Position size = (account × risk %) ÷ (distance from entry to that invalidation point).

Read the last two rows carefully, because they contain the insight. You never chose to put 10% of your account into this position. You chose how much you were willing to lose and where you would be wrong, and the position size fell out of those two decisions. That is the correct direction of reasoning, and it is the opposite of how most people size a trade.

It also produces the right behaviour automatically. A position whose invalidation point sits far away gets a smaller size. A tight, well-defined idea gets a larger one. You are sized by the quality of your definition of being wrong — which is exactly what a written trading system is meant to enforce, and what a running commentary never will.

Where mechanical stop-losses go wrong

Cutting losses is not the same thing as putting a stop-loss 8% below every purchase and calling it risk management. Three failure modes are common enough to name.

• A stop at a round number is a stop at everyone's number. Levels chosen for tidiness rather than meaning are dense with resting orders and get taken out on noise.
• A stop too tight for the instrument's normal movement is a guaranteed loss with extra steps. If a stock routinely moves 4% a day, a 3% stop is not risk control — it is a fee.
• Moving a stop down is the whole problem. The moment you widen a stop because the price approached it, you have removed the only mechanical protection you had, and you have done it under exactly the emotional conditions the rule was written to survive.

The deeper version of the rule is not about price at all. Exit when your reason for owning it is no longer true. A price level is a crude proxy for that, useful mainly because it is enforceable when you are not thinking clearly. If you can state the invalidating fact instead — a lost contract, a broken margin trend, a thesis that assumed something that did not happen — you have a better exit than any percentage.

Why you will not do this, and what actually helps

Everything above is simple and almost no one follows it, which should tell you the obstacle is not comprehension. Selling at a loss is the act that converts a story you are still telling yourself into a fact you have to record. Holding, by contrast, keeps the outcome pleasantly undetermined.

Two things reliably help, and neither is willpower:

1. 1Decide before you are exposed. The invalidation level written before entry is chosen by a different person than the one who will be staring at a 12% drawdown. Write it down at the moment of least pressure.
2. 2Make the record automatic. If exits are logged with the reason attached, you can eventually see your own pattern — that your losers were, on average, held three times longer than your winners. That pattern is invisible from memory and obvious from a replayed record.