A risk budget starts with a single number: the maximum drawdown you are willing to accept over a given period. This is your total risk budget. Everything else flows from dividing that budget across positions, strategies, and asset classes in a way that reflects your views while maintaining the constraint that total risk stays within the budget.
The simplest implementation allocates equal risk to each position. If your risk budget is a 20% maximum drawdown and you hold ten positions, each position gets a 2% risk allocation. Position sizing for each trade is then determined by dividing the risk allocation by the expected worst-case loss for that position. A position with a 20% stop loss would be sized at 10% of the portfolio (2% divided by 20%), while a position with a 50% stop loss would be sized at 4% (2% divided by 50%).
Equal risk allocation is simple but often suboptimal because it ignores the fact that some positions have higher expected returns per unit of risk than others. A more sophisticated approach weights risk allocations by expected Sharpe ratios or by your conviction level. High-conviction, high-Sharpe positions get more risk budget, while lower-quality positions get less.
Correlation between positions affects the total portfolio risk in ways that individual risk budgets do not capture. If all your positions are highly correlated (common in crypto), the portfolio risk is much higher than the sum of individual risk budgets would suggest. You need to either reduce individual risk allocations to account for correlation or explicitly model portfolio-level risk using a covariance matrix.
A practical risk budgeting framework might allocate risk hierarchically. First, divide the total budget across asset classes (40% to crypto, 30% to equities, 20% to commodities, 10% to cash). Then within each asset class, divide among strategies or sectors. Then within each strategy, divide among individual positions. This hierarchical approach ensures that no single level of the portfolio dominates total risk.
Risk budgets should be dynamic. When your portfolio is profitable, the dollar value of your risk budget increases (because 20% of a larger portfolio is more dollars). When the portfolio draws down, the budget shrinks. This naturally leads to reducing position sizes after losses and increasing them after gains, which is the opposite of what most traders do intuitively (doubling down after losses) and the correct approach mathematically.
Tracking risk budget utilization in real time tells you whether you have capacity for new positions. If your current positions are using 90% of your risk budget, adding a new position would push you over the limit unless you reduce existing positions. This constraint forces you to evaluate new opportunities against existing holdings rather than just adding them on top, which is how overconcentration typically develops.
The discipline of maintaining a risk budget is more valuable than the specific numbers you choose. The exact percentage matters less than the habit of consistently sizing positions relative to a total risk limit and never exceeding it. The traders who survive long enough to benefit from their edge are almost always those who have some form of risk budget, explicit or implicit, that prevents catastrophic losses.