Rebalancing, Risk Budgeting & Position Policy

There is no universally optimal frequency. Research suggests that threshold-based rebalancing, trading only when a position drifts beyond a specified band, generally produces better after-cost outcomes than fixed calendar schedules in most market regimes. A commonly cited starting point is a 5-percentage-point absolute drift band, but the right number depends on your asset-class volatilities, transaction costs, and tax situation.

Position sizing answers "how many dollars do I allocate to this position?" Risk budgeting answers "how much of the portfolio's total volatility or drawdown capacity should this position consume?" A position can have a small dollar weight but a large risk contribution if it is highly volatile or highly correlated with everything else. Risk budgeting makes that contribution explicit and allows you to cap it before you trade.

Rebalancing's primary purpose is to maintain a consistent risk profile, not to generate excess return. The "rebalancing bonus", the idea that systematic selling of winners and buying of losers harvests a premium, depends on mean-reverting asset prices and is not reliable in trending markets. The right frame is: rebalancing keeps the portfolio doing what you designed it to do.

A maximum single-name limit should reflect the maximum loss you can tolerate from a single position going to zero, or its practical equivalent. A common starting rule is 5% of portfolio in any single security for a diversified equity portfolio, but higher limits apply for concentrated strategies. Limits should also be expressed in risk terms: a 5% weight in a 60%-volatility small cap contributes more risk than a 10% weight in a 15%-volatility blue chip.

Volatility targeting means adjusting position sizes (or total portfolio leverage) so that the expected realized volatility stays near a stated annual target, for example, 10% annualized. When measured volatility rises, positions are reduced; when it falls, positions are increased. The approach is most useful for portfolios that include leveraged instruments, futures, or asset classes with wide volatility regimes.

Stop-loss rules and rebalancing rules can conflict: a position that falls to its stop-loss level also appears underweight relative to its target allocation, which would normally trigger a buy under a threshold-based rebalancing policy. The two rules must be sequenced explicitly. The general principle is that a triggered stop-loss takes precedence over a rebalancing buy.

They can, but the measurement step gets harder before the policy step gets easier. Weights, drift and concentration are properties of the combined household portfolio, so each account has to be reconciled to a single position list before any of the rules can be evaluated. Execution then runs the other way, account by account, since tax treatment, available instruments and trading costs differ. Splitting measurement from execution is what keeps a multi-account portfolio governable.

The rules function the same way either way, but an unwritten policy cannot be checked against later. The reason this curriculum treats documentation as part of the method is that the decision most likely to be revised under pressure is the one no record fixes in place. A written target, tolerance and review cadence make it possible to distinguish a deliberate policy change from a reaction to a recent move, which is the distinction the whole discipline rests on.

The tax layer of the decision disappears, which removes the no-trade region created by capital gains and makes tighter tolerance bands cheaper to run. Trading costs, minimum trade sizes and the risk of acting on noisy estimates all remain, so bands do not go to zero. Withdrawal rules replace tax cost as the constraint on moving money, and asset location questions become irrelevant because there is nothing to locate between account types.