There is a hierarchy of investment decisions that most people get backwards. The popular version goes: find the right asset, buy it, and hold it. The real version is: decide how much of your capital to put at risk, then decide where to put it. The first decision, the position size, matters far more than the second.

Here is why. Even a correct thesis can lose you money if the position is too large. An investor who puts fifty percent of their portfolio into a single idea that falls forty percent before recovering has lost twenty percent of their total capital. That loss requires a twenty-five percent gain just to break even. If the drawdown is larger, or if it forces a liquidation, the position size has turned a good thesis into a permanent loss.

Conversely, even a mediocre thesis does limited damage if the position is appropriately sized. A two percent allocation to an idea that goes to zero costs you two percent. That is recoverable. You live to invest another day.

The key variable is not conviction. It is the relationship between your position size and the maximum plausible drawdown. For volatile assets, this means smaller positions. For concentrated portfolios, it means accepting that your path will be rougher and sizing accordingly. For assets with non-linear payoff profiles, like options or early-stage positions, it means keeping each bet small enough that a total loss is tolerable.

This sounds obvious, but in practice it is the discipline that most investors fail at. The more excited you are about an idea, the more you want to overweight it. The better an asset has performed recently, the more it feels safe to own in size. Both impulses lead you toward exactly the position sizes that will do the most damage if the thesis is wrong or the timing is off.

The professional edge is not better ideas. It is better sizing. Hedge funds, trading desks, and institutional allocators obsess over position sizing, correlation, and portfolio-level risk far more than they obsess over individual picks. The reason is arithmetic. Over a career, the investor who sizes correctly and survives compounds far more wealth than the investor who sizes aggressively and blows up, even once.

Size the position for the drawdown, not for the upside. That single rule will do more for your returns than any research insight.

The Kelly Criterion and Why Nobody Uses It Raw

There is a mathematically optimal answer to position sizing, and almost nobody uses it directly. The Kelly criterion computes the bet size that maximises long-run compound growth given an edge and the odds, and its output is famously aggressive: full Kelly sizing produces gut-wrenching drawdowns on the way to its theoretical optimum, because maximising growth and minimising pain are different objectives, and Kelly only cares about the first.

The practical adaptation is fractional Kelly, sizing at a half or a quarter of the formula’s output, which sacrifices a modest amount of expected growth in exchange for a dramatic reduction in drawdown depth. The deeper lesson survives the dilution: position size should be a function of edge and uncertainty, not conviction or excitement, and the sizing error that ruins portfolios is almost never sizing too small. An investor who is wrong at 2 percent of the portfolio has a lesson. An investor who is wrong at 40 percent has a new financial life, and the arithmetic of recovering from deep losses, needing a 100 percent gain to repair a 50 percent loss, is the asymmetry the entire discipline exists to respect.

The Kelly Criterion and Why Nobody Uses It at Full Strength

There is a mathematically optimal answer to position sizing, the Kelly criterion, which maximises long-run compound growth by sizing each position to the edge and odds of the bet. The formula is unforgiving in both directions: bet less than Kelly and you leave growth on the table, bet more and you guarantee eventual ruin even with a genuine edge, because oversized bets turn a winning strategy into a losing one through volatility alone.

In practice almost every serious practitioner sizes at a fraction of Kelly, commonly a quarter to a half, and the reason is epistemic humility rather than timidity. Full Kelly assumes you know your edge precisely, and nobody does. Overestimating a 55 percent win rate as 60 percent, a small-sounding error, pushes full-Kelly sizing well past the ruin threshold. Fractional Kelly buys enormous protection against being wrong about your own edge at a modest cost to theoretical growth, which is the same trade-off, robustness over optimality, that runs through every durable approach to markets.