Position Sizing: The Decision That Comes Before the Trade
Ask ten investors what makes a good trade and nine will talk about direction: where the market is going, and why. It is the glamorous half of investing, the part that fills interviews and headlines. Yet when portfolios fail, the autopsy rarely points to a wrong opinion about direction. It points to a defensible opinion held in indefensible size.
Position sizing is the decision that comes before the trade: how much capital a single idea is allowed to put at risk, defined before the position exists and independent of how confident the idea feels. In professional currency management it is not an afterthought of the analysis. It is a policy, and in our experience it separates portfolios that survive their mistakes from portfolios that are defined by them.
Why sizing errors hurt more than direction errors
Every manager is wrong about direction regularly. That is not a flaw in the process; it is a property of markets. Decades of evidence show that even excellent currency strategies are wrong on a large share of their individual positions, and we have built our entire approach around that fact, as we described in why process beats prediction.
Being wrong is survivable when the loss is bounded. What is not survivable is the combination of a wrong view and an oversized position, because losses do not scale linearly with recovery effort. A portfolio that loses ten percent needs roughly eleven percent to recover. One that loses forty percent needs sixty-seven. The arithmetic of drawdowns, which we examined in detail in our article on drawdown management, is the reason sizing is a survival discipline rather than an optimization exercise: the damage of being wrong grows faster than the benefit of being right.
This is also why sizing cannot be left to conviction. Conviction is highest exactly where caution matters most, at the moments when a view feels obvious, the evidence seems overwhelming and the temptation to scale up is strongest. A sizing rule written in calm conditions exists precisely to overrule that feeling, because the feeling is not information about the future. It is information about the trader.
What a sizing framework actually decides
A disciplined framework answers three questions before any position is opened.
The first is the maximum loss a single idea may inflict on the portfolio. Not the expected loss, and not the loss if things go mildly wrong: the loss if the position is stopped out at its predefined exit. That number is set as a fraction of capital small enough that no individual mistake, or short sequence of mistakes, threatens the portfolio’s ability to keep operating.
The second is how positions combine. Two trades that look independent on a screen can be the same bet wearing different clothes: long one dollar pair and short another may both be, in substance, a single view on the dollar. Sizing limits that ignore correlation are limits in name only, which is why we treat this as part of the same discipline we described in our work on correlation risk: the unit being sized is the exposure, not the ticket.
The third is how size responds to conditions. Volatility changes what a position of a given notional actually risks: the same size that is conservative in a quiet market is aggressive in a stressed one. A framework that holds risk constant must therefore let notional vary, trimming positions as volatility expands rather than discovering, after the fact, that the portfolio had grown without anyone adding a trade.
Size as policy, not opinion
The common thread in all three answers is that none of them depends on how good the idea looks. That is deliberate. Direction is an opinion, and opinions should compete, be challenged and sometimes be wrong. Size is a policy, and policies exist to be applied uniformly, especially when applying them is uncomfortable.
In practice this means the sizing rule is decided at the level of the investment process, documented, and applied to every position without exception: not renegotiated trade by trade, not suspended for high-conviction ideas, not loosened after a good run. The moments when a manager most wants to make an exception are reliably the moments when the exception is most expensive, because overconfidence and risk concentration arrive together.
There is a quieter benefit as well. When size is fixed by rule, the quality of decisions becomes measurable. A portfolio where every idea risks a comparable fraction of capital produces a track record that reflects judgment rather than courage, and errors become data instead of events. Over the many decisions that make up a year of currency management, that feedback loop is worth more than any single trade.
The question that protects the next hundred decisions
The practical test of a sizing discipline is a single question, asked before every trade: how much is this position allowed to lose if it is wrong? Not how much it should make, and not how likely it is to work. If the answer is a number fixed in advance by policy, sizing is doing its job. If the answer depends on how the idea feels, the portfolio is one streak of confidence away from a problem.
Direction decides whether a trade makes money. Size decides whether the portfolio survives being wrong long enough for the process to work. Of the two decisions, only one can be fully controlled in advance, and professional management is, above all, the discipline of controlling what can be controlled.
This article is provided for informational and educational purposes only. It does not constitute investment advice, an offer, or a solicitation to invest in any product or service. Income Capital Management s.r.o. is registered under ยง15 of the Czech Act on Investment Companies and Investment Funds (ZISIF). Past performance is not indicative of future results. Investing involves risk, including the possible loss of capital.
