Rebalancing Without Wrecking Your Taxes

Two fears sit on top of each other in this lesson, and most people feel both at once. The first: my winners have run so hard that I am quietly carrying more risk than I ever signed up for. The second: if I sell to fix that, the taxman takes a cut every single time I tidy up. Both fears are real. Both have a clean, boring answer — and by the end of this lesson you will be able to fix the risk and keep almost all of the tax.

Start with the thing nobody warns you about. You set an allocation once — say 60% equity, 25% debt, 10% gold, 5% cash — and then you do nothing wrong for three years. You just leave it alone, like everyone tells you to. And yet the mix changes anyway. This slow, unasked-for change is called portfolio drift: because different assets grow at different speeds, the winner swells and the laggards shrink as a share of the whole, so your actual mix wanders away from the one you chose. You didn't make a decision — but the drift acts exactly like one.

Look at where equity landed: 70% of ₹50,00,000, or ₹35,00,000, against a target of 60%, which on this book is ₹30,00,000. That is ₹5,00,000 too much sitting in shares. The other three sleeves are all under target by the same ₹5,00,000 between them. Nothing was mismanaged — this is just what a winning asset does to a static mix. The question the rest of the lesson answers is not whether to fix it, but how to fix it without donating money to the tax department to do so.

It is tempting to shrug at drift. Equity went up — isn't more equity a good thing? Only until it isn't. Back in Lessons 5 and 6 you set an allocation to match your risk capacity and risk tolerance — how much loss you can afford, and how much you can stomach. Drift silently pushes you past that line. The Iyers chose a mix that would fall about 21–27% in a bad year, which they decided they could live through. At 70% equity, a crash hits harder than the plan allowed for.

Put a number on it. Those 10 extra percentage points of equity are ₹5,00,000 of extra shares. In a year the market falls roughly 50% — which markets do, about once a decade — that slice alone loses ₹2,50,000 more than their plan budgeted for. That is real money, and it arrives at the worst possible moment, when everything else is falling too. Rebalancing is not about squeezing out more return. It is about putting the risk back where you deliberately set it, before the market decides to test it for you.

So you rebalance to restore the risk you chose. But how often? Rebalance too rarely and you drift into danger; rebalance too often and you rack up tax and costs for no benefit. There are two clean ways to decide when to act, and one sensible way to combine them.

• Calendar rebalancing — check on a fixed schedule, say once a year every April, or on your birthday. Simple and habit-forming; but you might trade when barely anything has drifted, or miss a big mid-year surge.
• Threshold (band) rebalancing — act only when a class drifts past its tolerance band, a ±X% cushion around each target (a common starting point is ±5 points). You act only when it actually matters, so you trade far less; but you have to glance at it now and then to notice a breach.
• Best of both — check on a calendar, act only on a band. Look once or twice a year and rebalance only what has actually broken its band. You get the discipline of a schedule and the thrift of a threshold.

The Iyers' equity is at 70% against a 60% target — a 10-point drift, well past a ±5 band. That is a breach, and a breach is a signal, not a feeling: it tells them to act now regardless of the calendar. Had equity only crept to 63%, they would leave it alone; 63% is inside the 55–65% no-action zone, and trimming it would just manufacture tax and costs for a rounding error.

Here is where the second fear bites. To rebalance the obvious way, you sell some of the winner. But selling a fund or a share is a taxable event: any gain you realise on the way out is a capital gain, and capital gains are taxed. Rebalance clumsily and you hand over a slice of your money every time you tidy up — which is exactly why so many people never rebalance at all, and drift into danger instead.

The rate depends on what you sell and how long you have held it. These are the FY 2025-26 rules for the sell leg — you met them in full in Lesson 41 and the income-tax track; here is the slice that matters for rebalancing:

Three things jump out, and they drive everything that follows. First, long-term equity is taxed far more gently than short-term — 12.5% versus 20% — so a recent winner is the most expensive thing to sell. Second, only long-term equity gets the ₹1.25 lakh allowance: the first ₹1,25,000 of long-term equity gains each financial year is completely tax-free. Third, a small securities-transaction tax (STT, about 0.1%) is charged on the sale on top. Gold and other funds have their own holding-period rules — see Lesson 41 — but the equity rules above are the ones a typical rebalance turns on.

The fix is an order of operations. Instead of reaching for the sell button first, you fund the rebalance from the cheapest sources first and leave the taxable sale for last — often for never. This ordered ladder is the tax-aware rebalancing waterfall, and it has five rungs:

1. New contributions → the laggards. Point your next few SIP instalments at the underweight sleeves instead of buying more of the winner. This is rebalancing with new money — no sale, so no tax event at all, and no exit load or STT either.
2. Dividends & coupons → the laggards. Steer the year's dividends and debt coupons into the underweight sleeves rather than reinvesting them back into equity. Again, cash you already have — nothing realised, nothing taxed.
3. Spend the ₹1.25 lakh LTCG allowance. Now sell — but only the oldest long-term equity units, and only up to the allowance. The first ₹1,25,000 of long-term equity gains each year is free, so this trim costs 0% tax.
4. Set off a harvested loss. If a holding is under water and you were trimming it anyway, realise the loss and net it against the gain, so the taxable gain shrinks or vanishes (the standalone playbook is Lesson 43).
5. Sell taxable units — last. Only now, if the rungs above didn't finish the job, do you sell units that actually trigger tax — and even then, long-term before short-term.

That equation is the quiet key to the whole lesson. The ₹1.25 lakh allowance is not ₹1.25 lakh of sales you can make — it is ₹1.25 lakh of gains you can realise for free. Because only about half of the Iyers' equity value is profit, they can sell ₹2,50,000 of it and still keep the realised gain down to exactly ₹1,25,000, the full free allowance. Think of the allowance as an annual rebalancing budget the tax code hands you: spend it, and it refreshes next April.

Now watch the waterfall do its work on the Iyers' actual screen. They need to move ₹5,00,000 out of the overweight equity sleeve and into debt, gold and cash. Here is their holdings screen, with the current mix set against the target and the tax-aware fix laid out alongside — the two rows the lesson leans on, the allowance trim and the sell-last row, are tinted.

Read the fix ladder top to bottom. Rung 1: they steer ₹2,00,000 of upcoming SIP money to the laggards instead of buying more equity — no tax. Rung 2: they sweep ₹50,000 of the year's dividends and coupons the same way — no tax. Those two rungs alone move ₹2,50,000, half the job, without selling a thing. Rung 3: they sell ₹2,50,000 of their oldest equity units, which realises exactly ₹1,25,000 of long-term gain — the whole free allowance — so that trim is taxed at 0%. Rung 4 is empty for them (they have no loss to harvest this year — that is Suresh's move). Rung 5 is never reached: the ₹2,50,000 of new money plus the ₹2,50,000 allowance-sized sale already moved the full ₹5,00,000.

Same risk fixed, two very different bills. The Iyers save the entire ₹15,625 — not by dodging anything, but by funding the fix from new money and the free allowance instead of a blunt sale. On a modest book that is a nice-to-have. On a large one, as Suresh is about to show, the same discipline is worth lakhs.

When you do have to sell, which units you sell decides the tax as much as how much you sell. Suresh (55, a Kochi CA on the 30% slab, with about ₹1.8 crore including a large taxable equity book — one crore is ₹1,00,00,000) has let his equity run badly overweight and needs to raise ₹20,00,000 by trimming it. Roughly ₹8,00,000 of gain gets realised whichever units he picks. But he holds two kinds of lots of the same fund, and the choice between them is worth a fortune.

The recent units — bought inside the last 12 months during the run-up — are short-term: taxed at 20%, with no allowance. On ₹8,00,000 of gain that is ₹1,60,000. The old units — held more than a year — are long-term: taxed at 12.5%, and the ₹1.25 lakh allowance comes off first, so (₹8,00,000 − ₹1,25,000) × 12.5% = ₹84,375. Simply choosing the old lot over the recent one — the gentler rate plus the allowance — saves ₹75,625 before he does anything else. Hence the rule of thumb: sell long-term before short-term, and within long-term, the oldest and lowest-gain lots first.

Then Suresh reaches for rung 4. He holds a sector fund that is under water — a bet he wanted out of the plan anyway. Selling it now realises a ₹3,00,000 long-term loss, which sets off against his long-term rebalancing gain (a long-term loss can only offset a long-term gain — that is the Lesson 43 rule). His taxable gain becomes ₹8,00,000 − ₹3,00,000 − ₹1,25,000 = ₹3,75,000, taxed at 12.5% for ₹46,875. Against the naive ₹1,60,000, the lot choice and the harvest together save him ₹1,13,125.

There is one place you can shift risk with no tax event at all: inside a tax-advantaged wrapper. Money moving between options within EPF, PPF or NPS is not a sale in the taxable market, so it triggers no capital gain. NPS is the clearest example — its active choice lets you dial the split between equity (E), corporate bonds (C) and government securities (G), and switching that mix is completely tax-free. If part of your risk-shifting can happen there, do it there first.

For the Iyers this dovetails neatly with how their portfolio is built. Their EPF and PPF already fill most of their debt sleeve (you saw this at the Lesson 40 capstone), and both are EEE — tax-free all the way through. So 'top up debt' can partly mean 'keep contributing to EPF/PPF/VPF', which is new money into the debt sleeve with no sale and no tax. Choosing which asset lives in which wrapper so that rebalancing stays cheap is a whole topic of its own — asset location — the subject of Lesson 44 · Asset Location — the Right Asset in the Right Wrapper.

One warning, so you don't over-learn this lesson. All the cleverness above is about paying less tax to fix your risk — never about not fixing your risk to avoid tax. If your equity has drifted to a genuinely dangerous level and the waterfall can't get you all the way back for free, sell anyway and pay the tax. This is the old rule: don't let the tax tail wag the risk dog. A ₹15,625 or even a ₹1,60,000 tax bill is an annoyance; carrying a crash-sized risk you never chose can cost far more.

The Iyers are the happy case — the waterfall fixed them for ₹0. But if it hadn't, and they were still stuck at, say, 68% equity with a big embedded gain, the answer is still to trim it. The tax is the price of putting the risk back where it belongs, and that is a price worth paying. Optimise the tax; never let it veto the risk decision.

A good wealth manager runs exactly this waterfall for you — and it is worth seeing the move decoded, both to appreciate a good one and to catch a bad one. The skill isn't the trade; it's the sequence.

The tell at the bottom is the one to remember. A manager who 'rebalances' every month or quarter — generating short-term gains at 20%, STT, and trail commissions each time — is churning your book, not managing it. A good one rebalances rarely, on a band, and sequences the tax so you keep more. Turnover is a cost they should be sparing you, not a service they bill you for. If you want the deeper 'is my adviser worth the fee' test, that is Lesson 54.

Because drift is real and 'let me rebalance you' sounds so helpful, it is a favourite cover for churning your account for commission. Here is what the churn looks like when it is dressed up as care — and exactly how to check and report it, without any blame.

The one rule under all four tells: a genuine rebalance is rare, small, and tax-aware, and you never surrender control to do it. Anything that trades your book constantly, or sells you a 'tax-free' loophole (usually bond-washing or dividend-stripping, which the tax department can disallow), is generating fees, not fixing risk. Verify any adviser on SEBI Check, prefer a fee-only Registered Investment Adviser who earns nothing from your trades, and report churning on SEBI SCORES — or the cybercrime portal / 1930 if money was taken.

If you are reading this with a sinking feeling — because your mix has drifted for years untouched, or because you once 'fixed' it with one big sell and a tax bill landed — put the blame down. Drift is invisible, the app never showed you a waterfall, and nobody hands you a band at sign-up. What is still open to you is more than you fear.

The headline: drift is not damage, just a mix that wandered, so one banded check and one pass of the waterfall puts you back on plan — you have lost nothing by being late. If you did pay tax on a naive sell, that was a lawful tax on a real gain, not a penalty, and the risk still got fixed. And you needn't do it all at once: the ₹1.25 lakh allowance refreshes every financial year, so you can trim a little each year at no tax, and a market fall will do half the job for you for free.

• How often should I rebalance? Once a year is plenty for most people — or whenever a class breaks its ±5-point band, whichever comes first. Check on a calendar, act only on a band. More often than that usually just manufactures tax and costs.
• Won't rebalancing just trigger tax? Only the taxable sale does — and that is the last rung, often never reached. New money, dividends, the ₹1.25 lakh allowance, a harvested loss, and switches inside EPF/PPF/NPS all move your mix with little or no tax.
• Do I even have to sell to rebalance? No. If you're still adding money, just point new contributions and dividends at the underweight sleeves. For a growing portfolio, new money alone can keep you near target for years without a single sale.
• Recent units or old units first? Old (long-term) units, every time. They're taxed at 12.5% and get the ₹1.25 lakh allowance; recent (short-term) units are taxed at 20% with no allowance. Within long-term, sell the oldest, lowest-gain lots first.
• Should I rebalance in a crash? Yes — and it's the cheapest rebalance there is. A crash pushes equity below target, so you buy it cheap with new money (no tax), and you can harvest losses to offset future gains. The hard part is emotional, not fiscal.
• What's a sensible band? ±5 percentage points on your main sleeves is a common starting point; some use a ±20% relative band (e.g. a 10% gold target drifting past 8% or 12%). Tighter bands mean more trades; wider bands mean more drift. There's no single right answer — pick one and stick to it.
• Is the ₹1.25 lakh allowance per fund or per year? Per person, per financial year, across all your long-term equity gains combined — not per fund. Spend it deliberately; if you don't use it, it doesn't carry forward.
• My winner is up 200% — won't any sale be taxed to the hilt? The gain is only the profit portion of what you sell, and the first ₹1.25 lakh of it is free each year. Sell in slices across financial years, pair with a harvested loss, and lean on new money, and even a big winner can be trimmed gently.
• Can I rebalance for free forever? Only while new money and the allowance can absorb the drift. Past that, a modest taxable sale is fine — remember the risk dog: fixing a dangerous drift is worth a small tax bill.

Put it together on your own numbers. Enter your equity, debt, gold and cash values, your equity target and band, and the tax-aware levers — the new money and dividends you can steer, how much of your equity value is gain, and how much ₹1.25 lakh allowance you have left. The planner flags the drift, lays out the waterfall, and shows the tax you save against a naive full-sell. It comes pre-filled with the Iyers, so it reproduces their ₹0-instead-of-₹15,625 result exactly — clear it and try your own.

Watch how the tax saved moves when you change the levers. Give yourself more new money and the sale — and the tax — shrinks toward zero. Raise the embedded-gain figure and the same trim realises more gain, so it eats the allowance faster. Drop the equity drift inside your band and the planner tells you to do nothing at all, which is often the right, and cheapest, answer.

• Portfolio drift — the slow change in your actual mix as winners outgrow laggards, leaving you riskier (or safer) than you chose, without deciding to.
• Rebalancing band (tolerance band) — the ±X% cushion around each target inside which you leave the mix alone; you act only when a class drifts past it.
• Calendar rebalancing — checking and rebalancing on a fixed schedule, e.g. once a year, come what may.
• Threshold (band) rebalancing — rebalancing only when a class breaches its band, whenever that happens, rather than on a set date.
• The tax-aware rebalancing waterfall — the order that funds a rebalance cheapest-first: new money → dividends & coupons → the ₹1.25 lakh LTCG allowance → a harvested loss → a taxable sale last.
• Rebalancing with new money — steering fresh contributions to the underweight sleeves so the mix corrects without any sale, and so without any tax.
• The ₹1.25 lakh LTCG allowance — the first ₹1,25,000 of long-term listed-equity gains each financial year, which is tax-free; here, an annual budget for trimming winners at 0% tax.
• Embedded gain — the profit portion of a holding's value; the allowance is spent on this gain, not on the amount you sell.
• Don't let the tax tail wag the risk dog — the rule that a dangerous drift must be fixed even if it triggers some tax; risk control comes before tax optimisation.