Rebalance on a threshold, not on a calendar
Quarterly rebalancing is a scheduling convention, not a risk decision. Bands are what actually keep a portfolio inside the policy you signed — and they cost less to run.
The date is not the risk
Ask most private clients when their portfolio is rebalanced and you will hear a date: quarter-end, year-end, “when we review it in March”. Ask why that date, and the honest answer is that it is when somebody has time.
A calendar tells you when the accounting period closes. It tells you nothing about whether your equity sleeve is eight points above the weight you agreed to. Those two facts drift apart quickly, and they drift furthest in exactly the markets where the gap matters — a portfolio that ran up through a quarter is a portfolio carrying more risk than its owner signed for, for as long as the calendar says to wait.
Rebalancing because it is March
Calendar rebalancing trades a portfolio that has barely moved and leaves one that has moved a long way sitting out of policy until the date arrives. It is a schedule for the adviser’s workload, not a control on the client’s risk.
What a band actually does
A band replaces the date with a condition. Each sleeve of the policy carries a tolerance — an absolute one, expressed in percentage points of the total, and a relative one, expressed as a share of the sleeve’s own target. When either is breached, the trade is generated. When neither is breached, nothing happens, and nothing needs to.
At Sightline every mandate runs the same pair: ±3.0 points absolute and ±20% relative, checked weekly against custodian positions. A 40% equity target is therefore in policy anywhere between 37.0% and 43.0%; a 5% real-assets sleeve, where three points would be most of the position, is bounded by the relative band at 4.0–6.0% instead. The absolute band governs the big sleeves, the relative band governs the small ones, and whichever binds first is the one that fires.
Write both bands into the policy statement, state which sleeves each one binds first, and record the date and the reason of every breach. A rebalance you cannot explain two years later is indistinguishable from a hunch.
Fewer trades, not more
The objection we hear is that a threshold means constant trading. In practice it means the opposite. A quarterly calendar generates four rebalancing events a year whether or not the portfolio needed any of them. A band generates an event only when the portfolio has actually moved — which, for a diversified balanced mandate, has historically been closer to one or two a year, concentrated in the periods when being out of policy was most expensive.
The cost saving is real but secondary. The point is that the portfolio spends far more of its life inside the policy you signed, and that the times it is furthest outside are measured in days rather than in months.
If your equity weight is six points above target today, what is the mechanism that corrects it — and what is the longest it can stay there before something happens?
Where bands go wrong
Three ways, all avoidable, all of which we have had to unwind on an inherited portfolio.
Bands set too tight. A ±1 point band on a volatile sleeve is a trading strategy with a policy document stapled to it. The transaction and tax costs will exceed anything the discipline earns back.
Bands with no tax budget. In a taxable account the rebalance is a realisation event. The band tells you that to trade; a tax budget, agreed in advance, tells you how — which lots, in which account, and what to do instead when the cost of the trade exceeds the risk it removes.
Bands nobody measures. A threshold checked once a quarter is a calendar wearing a threshold’s clothes. If the drift is not measured on a schedule far tighter than the drift itself, the band is decorative.
Concentrated positions are not a sleeve A single line worth half your net worth does not belong inside a band. It gets its own unwind schedule, its own tax budget and its own horizon — and until it is down to the policy ceiling, the rest of the portfolio is built around it rather than beside it.
Putting bands into a policy statement
- Set the targets first. Bands are meaningless without a weight to be a tolerance around, and the weight comes from your spending, not from a peer group.
- Choose both tolerances. ±3.0 points absolute and ±20% relative is our house pair; the right one for you depends on sleeve sizes and tax position.
- Name the measurement interval. Weekly, from custodian data, not from an internal book — and write down who looks at it.
- Agree the tax budget before the first breach, in currency, for the year.
- Record the exception path. Who can override a band, in writing, and what the reason has to be. Ours requires committee sign-off and a written note.
Before you change how you rebalance
- Every sleeve has a target weight and both tolerances written down
- Drift is measured at least weekly, from third-party custodian positions
- A tax budget for the year exists in currency, not in principle
- Concentrated single lines are excluded and scheduled separately
- Every past breach is logged with its date, its size and its resolution
What this looks like on a statement
A mandate run on bands produces a duller quarterly document than one run on a calendar, and that is the point. Most quarters read: measured, in policy, no action. The quarters that do not read that way say precisely which sleeve breached, on which date, what was traded, what it cost in tax, and where the weight sits now.
That is the whole argument for the method. Not that it earns more — over a full cycle the difference is modest and honest people disagree about its size — but that at any moment you can answer the question is this portfolio inside the policy I agreed to, and answer it with a number rather than with a date.
See your own drift, measured
A confidential portfolio review takes forty minutes. We measure your current weights against the policy you have — or write the first cut of one, if you do not have it in writing yet.
- How a Sightline mandate is written and run
- The composite, net of all fees, including the years we lost
- Sightline Capital, Policy Statement Template v4.2 — bands, tax budget and exception path, issued to every client at onboarding.