Position Sizing and the Biases That Push It Around

Daniel Kahneman explains why losses register more strongly than equivalent gains and how that asymmetry distorts risk and sizing decisions.

Kahneman on loss aversion, the asymmetry that makes a loss feel heavier than an equivalent gain. This is the mechanism behind sizing up when winning, doubling down when losing, and cutting winners early.

Source: Daniel Kahneman, Nobel laureate in Economic Sciences and author of Thinking, Fast and Slow

Strategy gets the attention. Position sizing does most of the work.

Two traders running the same setup with the same win rate can produce wildly different equity curves based entirely on how they size. A flat 1% per trade, run consistently, will outperform a strategy with twice the edge that drifts between 0.5% and 5% based on conviction. The math is unforgiving on this.

Sizing is also where bias does its most expensive damage. Overconfidence after a streak pushes size up. Loss aversion after a drawdown pulls it down at exactly the moment the strategy is likely to mean-revert. The trader who fixes size in advance and does not renegotiate it during the session takes most of these biases off the table.

Darren Hardy's point in The Compound Effect applies directly here: small, consistent inputs compound, and small, inconsistent inputs do not.

Fixed risk per trade is the most undramatic discipline in this entire course. Run for a year, it produces an equity curve you can actually live with. Skip it on the trades that "feel different," and the variance of your outcomes balloons without your edge improving.

The gap between traders with similar edges is rarely strategy. It is sizing discipline maintained across hundreds of trades, including the ones where the trader felt sure.

The Kelly Criterion gives you the mathematically optimal position size for a known edge: f = (bp - q) / b, where b is the win-to-loss ratio, p is the win probability, and q is one minus that.

Two real-world caveats:

• Your edge is estimated, not known. If you size at full Kelly using an optimistic estimate, you will overbet and the drawdowns will be brutal.
• Full Kelly assumes you can tolerate the volatility of full Kelly. Most operators cannot. The psychological cost is real and shows up as deviation from the plan.

The practical version is fractional Kelly, usually one-quarter to one-half of the full number. That is not a worse math, it is realistic math that accounts for the operator. The point is not to maximize geometric growth, it is to keep sizing inside a range you will actually follow.

Two simple scenarios. Work through the sizing for each and compare against what you would have done by feel.

This is an interactive exercise. The reflection and structured worksheet open in your dashboard tools. Read through the prompt below first, then come back to complete it.

Compare two operators with the same edge over a year.

The first uses fixed risk per trade. The drawdowns are predictable and tolerable. The equity curve is unspectacular but compounds.

The second sizes by feel. Larger after wins, smaller after losses, sometimes much larger on a "high conviction" trade that turns out to be no different statistically. Over the year, the variance of outcomes is much wider. The compound return is usually lower, because the larger losses on oversized trades take longer to recover from than the larger wins added.

The math is on the side of the boring operator. So is the psychology. Boring sizing is also the kind you can actually run on a Tuesday after a tough Monday.