Portfolio & Diversification · Lesson 3 of 5
Rebalancing: What It Is and Common Approaches
Key takeaways
- Market moves continually shift a portfolio’s weights, silently changing its risk level.
- Rebalancing sells what has grown overweight and buys what has fallen behind, restoring the intended allocation.
- Its primary purpose is risk control, not extra return.
- Common methods: calendar-based (e.g., annually), threshold-based (e.g., 5-percentage-point bands), or steering new contributions — the cheapest option of all.
An allocation is a decision; markets immediately begin undoing it. If stocks outperform bonds for a few years — historically the norm — a portfolio built as 60/40 quietly becomes 70/30, then 75/25, carrying meaningfully more risk than its owner chose. Rebalancing is the maintenance activity that periodically restores the intended mix.
The mechanics
Rebalancing means comparing current weights to targets and trading the difference: selling a slice of whatever is overweight, buying whatever is underweight. It feels backwards by design — trimming the asset that has been winning to add to the one that has been losing. That discomfort is the point: the procedure enforces sell-high-buy-low discipline mechanically, replacing the judgment calls that behavioral biases reliably corrupt.
It helps to be clear about purpose. Across most historical periods, an unrebalanced drift toward stocks would have earned slightly more — at the cost of steadily rising risk and a far deeper 2008. Rebalancing is not primarily a return enhancer; it is risk control — keeping the portfolio the one its owner actually chose. (In sideways, mean-reverting markets it has sometimes added return too — a bonus, not the rationale.)
Worked example: the drifting 60/40
A hypothetical $100,000 portfolio is set at 60% stocks ($60,000) / 40% bonds ($40,000). Over three strong years, stocks gain 60% cumulatively and bonds gain 6%:
- Stocks: $96,000. Bonds: $42,400. Total: $138,400 — now 69% / 31%.
- The owner now holds materially more equity risk than chosen. A −35% stock year from here would cost about $33,600 (24% of the portfolio) versus $29,000 (21%) at 60/40.
- Rebalancing: targets are $83,040 stocks / $55,360 bonds — so sell $12,960 of stocks, buy $12,960 of bonds. The portfolio's expected behavior again matches the plan.
Notice what rebalancing did not do: predict anything. It responded to arithmetic, not forecasts.
Common approaches
- Calendar rebalancing. Restore targets on a fixed schedule — annually is common; research finds little benefit to more frequent intervals, and quarterly-or-faster mostly adds costs. Simple, automatic, ignores what markets are doing in between.
- Threshold (band) rebalancing. Act only when a weight strays a set distance from target — commonly 5 percentage points (60% stock target, act at 55% or 65%). Responds to markets rather than the calendar; requires occasional monitoring.
- Hybrid. Check on a schedule, act only if outside bands — a common institutional pattern.
- Contribution steering. Direct new deposits (or withdrawals) toward whichever side is underweight, rebalancing gradually with no selling at all. For accumulating investors, this handles most drift for free.
Rebalancing across the whole household
Targets apply to the household portfolio, not to each account separately. A worker with a 401(k), an IRA, and a taxable account holds one allocation spread across three containers, and the weights that matter are the combined ones. This view opens useful flexibility: if stocks are overweight overall, the correcting trade can happen wherever it is cheapest — typically inside a tax-advantaged account — while each individual account remains lopsided on purpose. It also prevents a common accounting illusion in which every account is "balanced" individually while the household total drifts far from plan. A simple spreadsheet, updated at each scheduled review, is enough bookkeeping for most investors.
Costs and placement
Rebalancing sells winners, and in a taxable account that realizes capital gains. Investors therefore often rebalance inside tax-advantaged accounts (where trades are untaxed — see the accounts lesson), lean on contribution steering in taxable ones, and avoid over-tight bands that trade frequently for trivial risk differences. Many target-date and balanced funds rebalance internally, automating the entire activity.
However it is done, rebalancing works best as a written rule decided in calm times — because the moments it matters most, deep in a crash when it commands buying the asset everyone is fleeing, are precisely the moments improvisation fails. The remaining lessons in this path widen the diversification lens beyond U.S. borders and catalog the mistakes that undo portfolios.
Common misconceptions
“Rebalancing is about boosting returns.”
Its job is keeping risk at the chosen level. Historically, letting stocks drift upward often earned slightly more — while silently transforming a moderate portfolio into an aggressive one. The comparison that matters is against the portfolio you intended to hold.
“Portfolios should be rebalanced constantly.”
Studies find annual or 5-point-band rebalancing captures essentially all the risk-control benefit; more frequent trading mostly generates costs and taxes. Drift accumulates over quarters and years, not days.
“Selling winners to buy losers is throwing good money after bad.”
At the asset-class level, “losers” are cheaper claims on the same long-run engines, not failing companies. Rebalancing systematizes buying asset classes when they are down — the behavior investors praise in hindsight and resist in the moment.