CAIA® rebalancing: calendar, bands or risk triggers?

On this page

Choose the rebalancing rule by asking what must stay close to target, how quickly it can drift, and what a corrective trade costs. A calendar date is simple, a tolerance band responds to weight drift, and a risk based allocation may need a trigger tied to risk contribution. For a CAIA® Level II candidate, the useful answer is conditional: the appropriate rule depends on the allocation approach and the constraint that matters.

First identify the target

A strategic portfolio starts with policy weights intended to express a long term mix. Market returns then alter those weights even if the manager places no trade. A tactical allocation may deliberately change a target for a defined purpose. A core and satellite structure may have a stable core and a more actively managed satellite. Risk parity seeks balance in contribution to portfolio risk rather than identical capital weights. Treating all four as though they shared one permanent percentage target obscures the decision the rebalancing rule must make.

The CAIA Association curriculum overview lists other asset allocation approaches and rebalancing strategies within Level II asset allocation. The point of comparing them is to reason from the portfolio’s objective to the monitoring variable. In a strategic mix, the live weight relative to policy weight is often the natural starting point. In a risk contribution framework, a weight can look unchanged while estimated volatility or correlation shifts the distribution of risk. The reverse can also happen: a capital weight moves, yet its contribution to overall risk remains near the desired level.

Rebalancing restores a chosen exposure. It is not automatically an attempt to forecast which asset will outperform next. Selling a sleeve that has grown beyond policy can look contrary to recent momentum, but its purpose is to restore the agreed risk mix. A tactical tilt, by contrast, changes the intended mix. Before calculating any trade, the candidate should say which target is in force and whether the apparent drift is unwanted or planned.

What each trigger actually does

A calendar rule checks or resets the allocation on scheduled review dates. Its chief advantage is governance: the committee and dealing team know when the assessment occurs, and records are easy to organise. Its weakness is that market moves are indifferent to the calendar. A quiet period can lead to trades that deliver little risk correction, while a sharp move shortly after a review can leave exposures far from target until the next date. A review date need not compel a trade; the policy should say whether the date triggers a measurement or a reset.

A tolerance band sets a permitted range around the target. Once the monitored exposure crosses the boundary, the manager considers a trade under the rule. An absolute band is expressed in percentage points; a relative band scales with the target weight. These are different for a small sleeve. A broad absolute band may permit a modest allocation to change dramatically as a proportion of its intended size. A relative band tightens the permitted point movement for that sleeve. Neither convention is universally best: liquidity, estimation noise and the risk of the sleeve all affect the sensible width.

A dynamic or optimisation based process estimates the benefit of reducing drift against expected trading costs, tax effects where relevant and any portfolio constraints. It can choose a partial rebalance instead of a full reset. The price of this flexibility is dependence on forecasts, cost estimates, and governance that makes the recommendation intelligible. A mathematically detailed recommendation is weak if nobody can explain which assumptions caused a trade. The comparison should therefore include operational auditability, not just theoretical efficiency.

Work through a small portfolio

Use invented, illustrative values. A portfolio starts with 100 units of value: 60 in a liquid core sleeve, 30 in bonds and 10 in an alternative sleeve. The target weights are therefore 60%, 30% and 10%. Suppose their values change to 66, 29 and 11 before a scheduled review. Total value is 106. The live weights are about 62.26%, 27.36% and 10.38%. The deviations from target are approximately plus 2.26, minus 2.64 and plus 0.38 percentage points. These figures are arithmetic for this example, not observed returns or a recommended allocation.

Under an illustrative absolute band of three percentage points, no sleeve breaches its own boundary. A band policy that trades only on a breach would hold. A calendar policy that resets at this review would trade despite the modest drift. To restore the original mix at the new total, the target values would be 63.6, 31.8 and 10.6. The portfolio would sell 2.4 units of the core and 0.4 of the alternative sleeve, and buy 2.8 units of bonds, before trading costs. The amounts balance. The arithmetic shows why a full reset can create several trades even when each drift is individually small.

Now change only the band convention. Suppose the illustrative relative band is 20% of each target weight. Its permitted drift becomes twelve percentage points for the 60% target, six points for the 30% target and two points for the 10% target. No sleeve breaches here either, but the alternative sleeve is being watched much more closely in percentage point terms than under the same twelve point absolute allowance that would apply to the core. The example is a check on interpretation: “20% band” means nothing until one says whether it refers to the target weight or to percentage points.

The candidate should also ask whether the band triggers a review, a partial trade back inside the band, or a complete return to target. These policies have different turnover. If the trading cost is high, returning exactly to target each time may invite repeated small trades near the boundary. A policy can instead specify a destination inside the band. The trigger, destination and monitoring frequency are separate choices and should not be collapsed into the word “rebalance”.

Match the rule to the approach

For a strategic allocation, the chosen policy weights form the baseline. A calendar review works where simplicity and a predictable governance cycle matter, provided the investor accepts interim drift. Bands give a direct limit on deviation if monitoring and execution are feasible. The policy should account for cash flows: an incoming contribution or distribution may let the manager reduce drift with less trading than selling a holding. The candidate can describe that option without assuming it is always available.

For a tactical approach, use the current authorised target, not an outdated strategic target, when deciding whether a position has drifted accidentally. A trade that restores strategic weights could erase a deliberate tactical view. The control question becomes who approved the temporary target, when it expires, and how it will be reviewed. In a core and satellite structure, a slower core process and a more responsive satellite process can be coherent because their purposes and frictions differ. The answer should explain the difference rather than prescribe identical bands to both.

For risk parity, inspect contribution to risk. The CAIA Association discussion of portfolio objectives describes risk parity in terms of balancing each asset class’s contribution to portfolio risk. That definition explains why a capital weight trigger alone can miss the relevant change. Rising volatility in one sleeve can increase its risk contribution without a large change in its market value. The manager still needs reliable estimates and a trading process; frequent model revisions can themselves generate turnover. A risk trigger needs governance over the data used to calculate it.

Cost, illiquidity and false precision

Trading is not free. A liquid holding may be easy to adjust; an illiquid fund interest may have notice periods, transfer restrictions or a valuation that arrives less frequently. The candidate should not apply a continuous daily band as if every sleeve could be traded and measured at a firm price on demand. A practical policy can rebalance liquid sleeves while recognising that the illiquid weight is only an estimate. It should state how uncertainty is handled rather than pretend a stale mark has the precision of a live quote.

There is also a cost to remaining away from target. A rule with no trades can allow concentration, liquidity needs or risk contributions to become inconsistent with the original policy. The comparison is between those risks and the costs of correcting them. Widening a band saves some trading only under particular paths; it does not mechanically improve outcomes. Likewise, a complex optimiser does not produce a better decision if its inputs are wrong. The examiner is testing whether the candidate can identify which exposure the rule controls and which friction it accepts.

The CAIA programme page locates this study area in a wider preparation path, while Omni’s how it works page explains the study setting. For the allocation judgement itself, use the official curriculum as the anchor and make the assumptions in each worked case visible.

A decision checklist for the exam

State the approved target and whether it is strategic, tactical, sleeve specific or risk based. Calculate the current exposure using values that sum to the same portfolio total. Compare it with the correct threshold, including whether the threshold is absolute or relative. Identify when the rule checks and where a triggered trade would take the portfolio. Then mention the relevant transaction, tax, liquidity or estimation cost. This sequence answers the question more clearly than naming a preferred rule without conditions. The main mistake is applying a capital weight band to every portfolio, especially one whose intended balance is measured in risk contribution.