Here is a fact that surprises most people the first time they meet it: two investors can pick exactly the same stocks and end the decade with wildly different results, purely because they sized them differently. Picture two analysts who, over ten years, buy the identical twenty ideas: same names, same entry and exit dates. Analyst A puts 15% of her capital into her three highest-conviction ideas and 2–3% into the rest; Analyst B equal-weights everything at 5%. If the high-conviction ideas happened to be the big winners, A crushes B; if two of A's 15% positions were the losers, B crushes A. Same research, same picks, different outcome, and the only difference was sizing. Selection decides what you own; sizing decides how much each one matters, and "how much each one matters" is at least half of the return and nearly all of the risk. The great fund manager Stanley Druckenmiller has said the single most important lesson he learned was that it is not whether you are right or wrong that matters, but how much you make when you are right and how much you lose when you are wrong, which is a statement about sizing, not selection.
Advertisement
Why is position sizing under-taught?
So why is sizing chronically under-taught relative to selection? Three reasons, worth naming so you can resist them:
- Selection is where the story is. Stock-picking has narrative: the moat, the misunderstood turnaround, the cheap compounder. Sizing is arithmetic and restraint; it makes no dinner-party story. So the entire ecosystem of books, media, and idea-dinners over-weights selection and barely mentions sizing.
- Sizing feels like it should be obvious, and isn't. "Put more in your best ideas" sounds like the whole answer until you ask how much more, measured against what, and what stops one mistake from ruining you, at which point you need the machinery.
- The feedback on sizing is even slower and noisier than on selection. You find out whether a pick was good in a few years; you find out whether your sizing discipline was good only across many positions and a full cycle, when the oversized bet you got away with nine times finally isn't. Slow, noisy feedback means the lesson is rarely learned before it is expensive.
How do you make sizing a written step?
The cure is to make sizing a computed, written step of every decision, not a gut feeling formed at the moment of purchase when you are most in love with the idea. The next sections give you three sizing tools in increasing order of practicality: Kelly (the theoretically-correct ceiling, see Kelly criterion and fractional Kelly), the conviction × downside heuristic (the practical day-to-day method, see conviction and downside sizing), and the anti-ruin overlay (the non-negotiable caps, see the anti-ruin overlay), plus the crucial hidden-correlation correction (hidden correlation) that stops you from accidentally making one giant bet while believing you made thirty small ones.