The Flaw of Fixed Lot Sizing
Most retail traders trade with an arbitrary fixed lot size—such as 0.1 lots or 1.0 lot—regardless of current market volatility or account equity. This leads to severe risk asymmetry: a 0.5 lot trade on a quiet summer Monday carries vastly different risk than a 0.5 lot trade placed during an FOMC interest rate announcement when ATR expands by 400%.
Dynamic Volatility Sizing via Average True Range (ATR)
To achieve equal monetary risk across varying market regimes, lot sizes must be an inverse function of price volatility:
When market volatility expands, stop distances widen to avoid premature stop-outs, and the algorithm automatically downsizes position volume. Conversely, in low-volatility compression regimes, the algorithm scales up volume while preserving an identical $150 risk limit.
The Kelly Criterion in Continuous Trading
The classical Kelly formula calculates the mathematically optimal fraction of wealth to risk to maximize long-term geometric compounding:
Where W is win rate, L is loss rate (1 - W), and R is the reward-to-risk ratio. However, Full Kelly risk sizing assumes an infinite time horizon and leads to catastrophic drawdowns in high-variance forex regimes. In institutional algo trading, we strictly utilize Quarter-Kelly (0.25 × f*) or Half-Kelly (0.5 × f*).