bet-sizing

Optimal position sizing is calculated using the Kelly criterion and risk budgeting frameworks. This process computes optimal capital fractions for discrete bets and continuous returns by mathematically balancing expected excess return against volatility.

Fractional Kelly applies a conservative fraction—such as half, third, or quarter—to the full Kelly bet size. It is used to mitigate risk from estimation errors, preventing over-concentration while still targeting risk-adjusted growth.

Volatility scaling adjusts position sizes inversely with asset volatility to maintain consistent risk exposure. This ensures that highly volatile assets receive smaller capital allocations within the overall portfolio construction process.

Yes, numpy supports the precise mathematical computations required for Kelly criterion calculations. It handles the array operations needed to process continuous investment returns, volatility scaling, and conviction weighting efficiently.

Risk budgeting allocates capital by assigning specific risk limits, such as Value at Risk or tracking error, to individual positions. This framework distributes risk exposure across the portfolio rather than allocating strictly by capital amount.

Full Kelly sizing can suggest highly leveraged allocations, such as a 200% position, which introduces severe over-concentration risk. Estimation errors in expected return or volatility can drastically amplify losses when using the full fraction.