The foundations of position sizing

Position sizing is the decision about how much capital to commit to each investment. It is not the same as stock selection, though the two are often conflated: an investor can have an excellent track record of identifying superior businesses but still underperform due to poor position sizing, either because high-conviction positions are too small to matter or because low-conviction positions are large enough to inflict significant damage when they go wrong.

The three primary inputs to any position sizing decision are conviction (how confident is the estimate of edge), risk per share (the difference between entry price and the point at which the thesis is proven wrong, typically the stop-loss or maximum acceptable drawdown), and portfolio risk budget (the maximum percentage of total capital that any single position should be able to cost the portfolio). These three inputs determine both the dollar size of the position and its percentage weight in the overall portfolio.

The expected value framework provides the theoretical foundation. If a trade has a 60% probability of gaining 20% and a 40% probability of losing 10%, the expected value is 60% x 20% minus 40% x 10%, which equals 8% per unit of capital risked. The Kelly criterion translates this expected value into the optimal fraction of capital to deploy: not the fraction that maximizes expected return in one period, but the fraction that maximizes the logarithm of wealth over many periods, which is equivalent to maximizing long-run compound growth while avoiding ruin.

Practical position sizing must also account for correlation. Two positions that appear small individually may produce large combined losses when they are highly correlated and both move adversely at the same time. Diversification within a portfolio requires not just many positions but positions whose risk exposures are genuinely distinct. Sector concentration, factor concentration (all value, all growth, all small-cap), and macro factor concentration (all leveraged to rising rates, all export-dependent) can create correlated losses that make individual position limits insufficient without portfolio-level heat monitoring.

Risk budget allocation across a portfolio

A risk budget treats total portfolio risk as a finite resource to be allocated across positions, sectors, and strategies. The first step is defining the total risk budget: the maximum acceptable drawdown from peak to trough, or the maximum portfolio heat at any given time (the sum of all open positions' maximum defined losses as a percentage of total capital). Common total heat limits range from 10% to 25% of capital, depending on strategy volatility tolerance and investment horizon.

Once the total budget is defined, it is allocated across positions. An equal-risk allocation assigns the same dollar risk to every position, which is not the same as an equal-weight allocation: a highly volatile position (large price range from entry to stop) in equal dollar risk will be smaller in percentage weight than a low-volatility position with the same dollar risk. This is the core insight of risk parity applied at the position level: weight by risk contribution, not by dollar exposure.

Conviction-weighted risk allocation tilts the risk budget toward highest-conviction positions. A position with twice the conviction deserves twice the risk allocation, not necessarily twice the portfolio weight. This framework prevents over-concentrating in volatile positions that consume large amounts of risk budget per dollar invested and under-concentrating in stable positions where conviction is high but volatility-adjusted sizing is small.

Sector and factor constraints supplement individual position limits. Even if each position is within its individual heat limit, concentrating all positions in one sector can produce correlated losses that exhaust the total risk budget simultaneously. Explicit limits on sector heat (for example, no more than 40% of total risk budget in any single sector) and factor heat (no more than 50% of total budget in any single macro factor) prevent correlated drawdowns from individual position discipline.

Every guide in this lab

  • Kelly Criterion: The mathematically optimal position sizing formula, how fractional Kelly reduces drawdown risk, and practical limits on Kelly-based sizing
  • Fixed-Fractional Position Sizing: How to risk a consistent percentage of capital per trade, how stop-loss placement determines position size, and the advantages and limits of this approach
  • Volatility-Adjusted Position Sizing: How to size positions so each contributes equal risk to the portfolio, using ATR and realized volatility to calibrate size
  • Risk Budget Allocation: How to define a total portfolio risk budget and allocate it across positions by conviction and correlation
  • Drawdown-Based Position Limits: How to reduce exposure during losing streaks, position-level loss limits, and step-up protocols for returning to full size

Frequently asked questions

What is position sizing in investing?

Position sizing determines how much capital to allocate to each investment relative to total portfolio assets. The decision links the conviction level in the investment thesis, the risk of loss on that position, and the total risk budget the portfolio can absorb. Correct position sizing prevents a single loss from causing irreparable portfolio damage while still allowing concentrated positions in highest-conviction ideas.

What is a risk budget in portfolio management?

A risk budget is an explicit allocation of total portfolio risk across individual positions and strategies, similar to how a financial budget allocates spending across categories. Total portfolio risk (measured as volatility, value-at-risk, or maximum drawdown tolerance) is the constraint; individual position sizes are calibrated so their combined risk contribution stays within the budget. Risk budgeting ensures that no single position or strategy consumes a disproportionate share of total portfolio risk.

What is the Kelly criterion and when should investors use it?

The Kelly criterion calculates the optimal fraction of capital to bet on a positive-expected-value outcome: f = (bp - q) / b, where b is the net odds received, p is the probability of winning, and q is the probability of losing. Full Kelly maximizes long-run capital growth but produces extreme volatility and large drawdowns. Most practitioners use fractional Kelly (25-50% of full Kelly) to reduce drawdown risk at the cost of lower expected growth. Kelly is most applicable in systematic strategies with well-calibrated win rates and payoff ratios; it is less useful for fundamental equity investors who cannot reliably estimate probabilities.

How does portfolio heat relate to position sizing?

Portfolio heat is the aggregate risk exposure of all open positions, measured as the sum of each position's maximum risk (distance from entry to stop loss, multiplied by position size) as a percentage of total capital. A portfolio heat of 6% means the combined potential loss if all stops are triggered simultaneously equals 6% of capital. Managing portfolio heat prevents the compounding of correlated losses: if all positions are in the same sector and all stop simultaneously, the realized loss equals the heat percentage. Keeping heat below a defined threshold (commonly 15-20%) limits maximum realistic single-event loss.

What is fixed-fractional position sizing?

Fixed-fractional position sizing risks a constant percentage of current account equity on each trade, typically 1-3% per position. If account equity is $100,000 and the risk per trade is 2%, the maximum loss per position is $2,000. Position size in shares equals the dollar risk divided by the per-share risk (entry price minus stop-loss price). Fixed-fractional sizing automatically scales position sizes down after losses (smaller equity base means smaller absolute dollar risk) and up after gains, which has a compounding effect on account growth and a self-limiting effect on drawdowns.