Position Sizing: The Math That Quietly Decides Your Returns

Ask most investors what they are working on and they will tell you what they are thinking of buying. Ask how much of it they plan to buy, and you will often get a shrug. This is backwards. The question of how much—position sizing—quietly determines your results at least as much as the question of what. You can be right about a company and still be ruined by owning too much of it, and you can be wrong about several and barely feel it if each was small. Getting the size right is where a lot of investing success actually lives, and almost no one talks about it.

I came to appreciate this the way most people do who work with risk and data: by seeing how a single oversized mistake can undo a long string of good decisions. The arithmetic of losses is unforgiving, and position sizing is how you keep that arithmetic on your side.

What position sizing is

Position sizing is simply the decision of how large each holding should be relative to your overall portfolio. If you have $100,000 invested and you put $5,000 into a single stock, that position is 5% of your portfolio. Size it at $30,000 instead and it is 30%—the same idea, the same research, but a vastly different bet. The company has not changed; your exposure to being wrong about it has changed enormously. Position sizing is the dial that controls how much any one decision can help or hurt you.

Why size matters more than being right

The reason sizing dominates is the brutal asymmetry of losses. A position that falls 50% needs to double—a 100% gain—just to get back to where it started. Fall 80%, and you need a 400% gain to recover. Losses and the gains required to undo them are not symmetric, and the deeper the hole, the more disproportionate the climb out. This means that avoiding large losses matters more than capturing large gains, and the surest way to avoid a catastrophic loss is to never let any single position be large enough to inflict one.

Put concretely: if your best idea is 40% of your portfolio and it falls by half, you have lost 20% of everything, and you now need substantial gains across the rest just to break even. If that same idea was 5% of your portfolio, the same 50% collapse costs you just 2.5%—a bad day, not a disaster. Same mistake, wildly different consequence, entirely because of size.

The first rule: never let one position sink you

Every sensible sizing framework starts from a single principle: no individual position should be large enough to seriously damage your financial life if it goes to zero. Companies you were certain about do fail; frauds, disruptions, and plain bad luck happen. The purpose of position sizing is to make sure that when one of your holdings suffers the worst possible outcome, it is a manageable setback rather than a permanent injury. A common guideline for individual stocks is to cap any single name at a modest slice of the portfolio—often no more than around 5%—so that even a total loss is survivable.

Sizing by risk, not just by conviction

It is tempting to size positions purely by how confident you feel: more conviction, bigger bet. But conviction is exactly the feeling that precedes the largest mistakes, because certainty and correctness are not the same thing. A more disciplined approach sizes by risk—how much you can afford to lose if you are wrong—and lets conviction adjust within those limits, never breach them. You might allow a high-conviction idea to sit at the top of your allowed range and a speculative one near the bottom, but the ceiling stays fixed regardless of how sure you feel. The cap protects you from your own certainty.

Don’t forget correlation

A subtle sizing error is treating related positions as if they were independent. If you hold five different companies that all depend on the same industry or the same trend, you have not made five separate 5% bets—you have effectively made one 25% bet, because they will tend to fall together. When you size positions, consider what they have in common, not just their individual weights. True diversification, discussed in a related article, is partly a sizing problem: the goal is to ensure that no single underlying force controls a dangerous share of your portfolio, whether it wears one ticker or several.

The danger of over-betting, even when you are right

There is a branch of mathematics devoted to how much to wager when you have a genuine edge, most famously the Kelly criterion. Its deepest lesson for ordinary investors is counterintuitive: even when the odds are truly in your favor, betting too large is a fast route to ruin, because a run of bad luck against an oversized position can wipe you out before your edge has time to play out. The practical wisdom that falls out of it is to bet smaller than your confidence suggests. Sizing conservatively costs you a little upside in the good cases and saves you from catastrophe in the bad ones—a trade that compounds in your favor over a lifetime of decisions.

The quiet behavioral benefit

Right-sized positions do something beyond limiting losses: they let you behave well. A position small enough that its swings do not frighten you is a position you can hold calmly through the volatility that shakes others out. Oversize a holding and every dip becomes an emotional event that tempts you to sell at the worst moment. In this sense, good position sizing is not only mathematical protection but psychological protection—it keeps your holdings within the range where you can actually stick to your plan.

A simple framework

You do not need complex formulas to size well. Decide in advance a maximum weight for any single position and refuse to exceed it, no matter how compelling the idea. Keep your total exposure to any one theme or industry within a sensible limit, counting correlated holdings together. Let conviction move a position within its allowed band, never past the ceiling. And revisit sizes as positions grow—a winner that swells to an outsized share of your portfolio has quietly become a concentration risk, and trimming it back to your limit is simply good discipline, the same rebalancing that keeps a whole portfolio in line.

A worked example

Picture two investors, each with $200,000, who buy the exact same promising stock and turn out to be wrong: it falls 70% on unexpected bad news. The first sized it as a 5% position, $10,000. Their loss is $7,000—about 3.5% of the portfolio, an unpleasant but forgettable dent. The second, brimming with conviction, made it a 35% position, $70,000. Their loss is $49,000, nearly a quarter of everything, and they now need a punishing gain across the rest of their holdings just to recover. Identical research, identical mistake—one walks it off, the other is set back years. Nothing about the analysis separated them. Only the size did.

Sizing applies to your whole portfolio, not just single stocks

Position sizing is usually discussed in terms of individual stocks, but the same logic governs your biggest allocation decisions. The split between stocks and bonds is itself a sizing choice—how much of your wealth is exposed to market risk at all. How much you hold in any one country, or in a single sector like technology, is a sizing decision too. Even a heavy bet on your own employer’s stock is a sizing problem in disguise. Whenever you decide how much of your total wealth rides on a single outcome, you are sizing a position, and the same rule applies: keep any one bet small enough that being wrong about it is a setback you can absorb, not one that defines your financial life.

Key takeaways

How much you buy matters as much as what you buy. A great idea, oversized, can still ruin you.
Losses are asymmetric. A 50% drop needs a 100% gain to recover—so avoiding big losses beats chasing big wins.
Cap every position. No single holding should be large enough to inflict lasting damage if it goes to zero; count correlated positions together.
Bet smaller than your confidence suggests. Conservative sizing protects you from certainty that turns out to be wrong—and lets you hold calmly.