Rebalancing: selling winners on purpose

What rebalancing is, mechanically

You chose a mix. Say 60 per cent growth assets and 40 per cent defensive assets. Markets then move, and they do not move together. A year later the growth sleeve has run and the split is 68 to 32. You never made a decision, but you now hold a meaningfully more volatile portfolio than the one you signed up for.

Rebalancing is the act of moving it back. You sell some of what grew, or you direct new money into what lagged, until the portfolio matches the target again.

That is the entire mechanism. There is nothing clever in it. The difficulty is not technical, it is that the trade always feels backwards. You are selling the holding that has made you money and buying the one that has disappointed you, on purpose, using a rule you wrote when you were calm.

It is worth being precise about what rebalancing is for, because the reason is widely misdescribed.

Rebalancing is a risk control, not a return strategy

You will see rebalancing sold as a way to "buy low and sell high automatically". Sometimes it does add return. Often it does not. Over long periods in which one asset class simply outperforms another for a decade, rebalancing away from the winner reduces your return compared with leaving it alone. That is not a flaw. It is the trade you are making.

The honest description is this: rebalancing keeps your portfolio's risk near the level you chose. Any return effect is a by-product, and its sign depends on whether markets in that period trended or reverted.

If you never rebalance, three things happen slowly:

1. Your equity weight drifts upward over time, because growth assets usually grow faster over long stretches. Your portfolio gets riskier the longer you hold it.
2. Your portfolio becomes concentrated in whatever recently worked, which is precisely the exposure most likely to be expensively priced.
3. The drift is invisible. You do not feel the risk increase until the market delivers it in one go.

So the case for rebalancing is a case about control, not about cleverness.

Why it feels wrong every single time

The discomfort is structural, and knowing why helps you execute anyway.

When you rebalance, the asset you are trimming has a recent record of going up, and the asset you are adding has a recent record of going nowhere or down. Human judgement extrapolates recent trends. So the trade contradicts your own forecast at the moment you make it.

It also creates a clean regret story. If the winner keeps winning after you trim it, you can point at the exact amount you gave up. If the laggard keeps lagging, you can point at the exact amount you wasted. Rebalancing produces vivid, attributable regret in both directions, whereas doing nothing produces a diffuse risk you cannot picture. People systematically prefer diffuse risk to vivid regret.

The remedy is not willpower. It is pre-commitment — deciding the rule in advance and reducing the moment of action to an administrative task. Written rules and pre-commitment are a recurring theme in financial capability work for exactly this reasonSource^source.

Three rebalancing rules and what each one costs

There is no single correct method. There are three families, and each trades off cost, effort and tracking accuracy differently.

Calendar rebalancing

Rebalance on a fixed date. Once a year, or twice a year, you check the portfolio and restore the target weights.

• **Advantages.** Trivially simple. Impossible to argue with yourself about. Bounded number of transactions, so bounded cost. Easy to automate as a diary entry.
• **Disadvantages.** The date is arbitrary. A market can move violently in March and be back to normal by your December review, so you rebalance nothing when it mattered and something when it did not. Between reviews, drift can get large.

Annual is the common choice, and for most people it is sufficient. More frequent calendar rebalancing mainly adds cost without adding control.

Threshold or band rebalancing

Rebalance when a weight drifts beyond a set band, regardless of the date. For example, act whenever any asset class is more than 5 percentage points away from its target.

• **Advantages.** Responds to what actually happened rather than to the calendar. Does nothing in quiet periods. Acts promptly after large moves, which is when drift is largest.
• **Disadvantages.** Requires monitoring, which for many people means looking at the portfolio more often than is good for them. In volatile periods it can trigger repeatedly, generating costs. Needs a rule for what counts as a band breach — absolute percentage points or a relative percentage of the target weight.

Note the difference between absolute and relative bands. A 5-percentage-point absolute band around a 60 per cent target means acting at 55 or 65. The same absolute band around a 5 per cent target is meaningless, because the holding can go to zero without breaching it. For small sleeves, use relative bands — for example, act when a holding is 25 per cent away from its target weight.

Cash-flow rebalancing

Do not sell anything. Direct new contributions, dividends and interest into whichever asset is furthest below target. Do the same in reverse with withdrawals, taking money from whatever is furthest above target.

• **Advantages.** No selling means no realised gains, minimal transaction costs, and no psychological trade against your own instinct. It is the cheapest method by a wide margin.
• **Disadvantages.** It only works while contributions are large relative to the portfolio. Once your pot is big, a monthly contribution cannot correct a 10-point drift. It also cannot fix a large gap quickly.

The ordering rule

These are not exclusive. The practical policy is an ordering:

1. Rebalance with contributions and income first, always.
2. If a band is still breached after that, sell to correct it.
3. Use the calendar only as a backstop — a fixed annual date to check whether the bands are working and whether the target itself still fits your life.

This ordering is the part most people miss. They set a calendar rule, then rebalance by selling in January while simultaneously making monthly contributions into whatever they already hold most of. The contributions are doing the opposite of the policy.

A worked example over three hypothetical years

Suppose you hold 200,000 with a target of 60 per cent growth and 40 per cent defensive, and you add 3,000 a month. All figures below are invented to show the mechanics.

**Year one.** Growth assets rise 22 per cent, defensive assets rise 2 per cent. Ignoring contributions for a moment, the 120,000 growth sleeve becomes 146,400 and the 80,000 defensive sleeve becomes 81,600. Total 228,000. Growth is now 64.2 per cent — inside a 5-point band, so no sale required. Meanwhile your 36,000 of annual contributions went entirely into the defensive sleeve under the ordering rule, which pulls the weight back toward target without a single sell trade.

**Year two.** Growth assets fall 28 per cent. Now the growth sleeve is well below target. Contributions redirect into growth. This is the year the policy earns its keep, because it tells you to keep buying the thing that just fell, at a moment when every instinct says to stop. Note carefully what it does not tell you: it does not tell you the fall is over, and it does not tell you to add extra money beyond your normal contribution.

**Year three.** Growth recovers 20 per cent. Because you kept buying through the fall, you own more units of the growth sleeve than you would have had you paused contributions. Your recovery is correspondingly larger. This is the mechanical benefit of the policy — not a forecast, just the arithmetic of having kept buying at lower prices.

Now the counter-case, which is equally instructive. If growth assets had simply fallen for four consecutive years, the policy would have moved more money into a falling asset the whole way down, and your portfolio would be worse than if you had stopped. That outcome is real and it is not a failure of the rule. It is the risk you accepted when you chose the allocation.

Rebalancing cannot know whether an asset is cheap or broken. It responds to weight, not to value. Any story you tell yourself about it "buying the dip" is a story you added.

The friction nobody models

Textbook rebalancing assumes trading is free and instant. It is not.

Costs of trading

Every sell and buy carries some combination of commission, bid-ask spread, platform fee, foreign-exchange conversion charge and, in fund structures, potential entry or exit charges. Small, frequent rebalances can quietly cost more than the drift they correct. Cost and fee disclosure for collective investment schemes sits within international securities regulation standards, and the disclosure documents exist precisely so you can compare these before you tradeSource^source. It is also worth confirming that the intermediary executing your trades is licensed in the jurisdiction you deal in — in the UAE this is verifiable through the relevant regulatorSource^source.

Practical implication: set your bands wide enough that acting is worth it. A 1-point drift is noise. Correcting it is expensive noise.

Tax, wherever you are tax-resident

Selling an appreciated holding can create a taxable event depending on your tax residence and the type of account. Some people invest through wrappers where this does not arise; others do not. The rule of thumb that survives all jurisdictions is simply this: find out how disposals are treated for you before you build a policy that requires frequent selling, and prefer contribution-based rebalancing where selling has a tax cost. This is a matter for a qualified adviser in your jurisdiction, not something to infer from a general article.

The behavioural cost

Threshold rebalancing requires you to look. Looking at a portfolio frequently is associated with more anxiety and more unplanned decisions, not fewer. If checking monthly means you will also fiddle monthly, an annual calendar rule you actually follow beats a sophisticated band rule you use as an excuse to trade.

Choose the policy that survives contact with your own personality.

When not to rebalance

Some situations genuinely call for leaving it alone.

• **The drift is small.** Inside your bands, do nothing. A portfolio at 62 per cent against a 60 per cent target is the same portfolio.
• **The portfolio is tiny relative to contributions.** If you are adding 3,000 a month to a 15,000 pot, contributions will fix everything. Trading is pointless.
• **The costs exceed the benefit.** If correcting a 4-point drift costs a meaningful fraction of the amount being moved, the correction is not worth it.
• **You are about to change the target anyway.** If your circumstances have shifted, decide the new allocation first, then trade once, rather than rebalancing to a target you are about to abandon.
• **The asset has changed character, not just price.** If a holding fell because the underlying thing is impaired — a fund closing, a strategy changing, a company in distress — rebalancing into it is not discipline, it is averaging into a decision you have not re-examined. Rebalancing rules apply to asset classes and diversified holdings, not to single securities you happen to be losing money on.

That last point deserves emphasis. Applying a rebalancing mindset to individual shares is how people build concentrated positions in failing companies while telling themselves they are following a rule.

A rebalancing policy you can write on one page

Write these seven lines, date them, and keep them where you will find them when you are stressed.

1. **Target mix.** The percentage for each sleeve, with one sentence on why each sleeve exists.
2. **Bands.** Absolute points for large sleeves, relative percentages for small ones. State the numbers.
3. **Order of operations.** Contributions and income first. Sell only if still outside the band afterwards.
4. **Review date.** One fixed date a year for a full check, including whether the target still fits your life.
5. **Exclusions.** Anything the policy deliberately does not touch, such as a locked pension, an employer share plan in a vesting window, or an illiquid holding.
6. **What is not a trigger.** Write the sentence explicitly. Market news, forecasts, a friend's opinion, and a bad month are not triggers.
7. **Trigger log.** Every time you act, one line — the date, which band broke, what you did, and one sentence on why.

The trigger log is the part that makes this real. After two or three years you can read back and see whether you actually followed your own policy or quietly overrode it whenever it was uncomfortable. Almost everyone discovers they overrode it at least once, and the log tells you exactly what that cost.

Rebalancing will not make you a better investor. It will stop your portfolio from silently becoming a different portfolio than the one you chose — which, over a long enough period, is most of what discipline actually means.

Source^source: International Organization of Securities Commissions.

Source^source: Securities and Commodities Authority, United Arab Emirates.

Source^source: OECD, Financial Education and Financial Literacy.