ELSS, and the lock-in that is doing more work than the deduction

What an ELSS actually is

An equity-linked savings scheme (ELSS) is an ordinary open-ended equity mutual fund with two statutory features attached: the amount you put in qualifies for a deduction from taxable income under the old tax regime, and every unit is locked for three years from the day it was bought.

That is the whole product. Underneath the tax label it is a diversified equity fund holding listed shares, valued once a day at the per-unit worth of everything it owns — its NAV — run by a manager against a benchmark, and charging an annual fee on the assets under management, the expense ratio. Everything that decides the outcome of any equity fund decides this one's too.

The label does a lot of damage, because it invites people to file the decision under tax rather than under investing. An ELSS bought in March to reduce a tax bill is still an equity commitment made in five minutes, and the equity part is the part that determines what the money is worth in three years.

So the useful order is the reverse of the usual one. First, whether the deduction is available to you at all — which depends entirely on which regime you are taxed under. Then what the lock-in does. Then whether the fund underneath is one you would hold without either.

The deduction belongs to one regime only

This is the fact that decides whether the rest of the article is relevant to a given reader, so it comes first.

The deduction sits in section 123 of the Income-tax Act 2025, read with Schedule XV. Almost everything written about ELSS calls it Section 80C, which was its number in the Income-tax Act 1961 — an Act repealed with effect from 1 April 2026. The rates barely moved in the replacement; the section numbers all did. If you are still finalising a return for income earned in the year to 31 March 2026, that year is governed by the 1961 Act and the old numbering is the correct citation for it.

The second point matters more. Under section 202 of the 2025 Act, the new regime is the default. You are taxed under it unless you actively choose otherwise. And the new regime carries no section 123 deduction — no deduction for an ELSS, and none for the health insurance premium (s.126) or education loan interest (s.129) that people usually bundle into the same March exercise.

Read the last row twice. The lock-in does not go away when the deduction does. Someone who moved to the default regime and left a March ELSS instruction running is paying a three-year restriction for a benefit they can no longer claim. It is an expensive thing to get wrong quietly, and nothing on the account statement flags it — the statement has no idea which regime you are in.

Working out which regime applies to you is arithmetic, not judgement: total the deductions you would genuinely claim — not the ones you could theoretically claim — compute the bill under each rate table, and compare the two numbers. That subtraction is worked through in old regime versus new. The regime decides whether an ELSS has a tax case at all, and nothing further down this page can settle it for you.

What the lock-in actually locks

The lock-in is not on the scheme, the folio, or the instruction. It is on each unit, running three years from its own allotment date.

For a lumpsum that distinction is invisible. For a SIP it is the whole thing, because a SIP does not create one holding — it creates a stack of separate purchases, each with its own age. The instalment bought in April 2026 becomes free in April 2029. The instalment bought in March 2029 stays locked until March 2032, whatever the age of the SIP that bought it.

Hence the most common ELSS misunderstanding, which costs people a redemption they planned for: “my ELSS SIP completes three years next month, so I can take the money out.” No. What completes is the first instalment, and on a monthly plan that is one payment out of thirty-six. The other thirty-five are still serving their own clocks, and the last one bought is the last one free.

Run that forward and it turns into something more useful than a warning. Once a monthly SIP has been going for three years, the ladder is full: at any moment from then on, exactly the last three years of instalments are locked and everything older is free. The locked portion stops growing. The commitment is not “my whole ELSS holding” — it converges on a fixed slab equal to three years of contributions, with a corpus that keeps compounding around it.

One more timing detail. The deduction attaches to the financial year in which the money went in; the lock-in runs from allotment. An instalment made in March therefore earns its deduction for a year that is nearly over, and stays locked for the full period starting from that March — which is why late-March investing quietly costs the most in committed time per rupee of deduction.

Why the lock-in is doing work rather than holding you hostage

A lock-in on a bank deposit is a cost with nothing on the other side — the deposit was never going to behave differently for being untouchable. Equity is the one asset where the restriction interacts with what the asset does, and it does so in three distinct ways.

It removes an option at the moment the option is most likely to be used badly. The mechanical part is flat: for three years, redemption is not available. What follows from that is a judgement rather than a certainty — that the exits which do most damage cluster in the first sharp fall after someone invests, and this is exactly the window in which an ELSS unit cannot be sold. It is a reasonable reading, and it is worth holding as a reading rather than a law. The mechanics of that particular exit are in SIPs during market crashes.

It is shorter than the horizon the asset needs anyway. A portfolio that falls by a third needs to rise 50% to get back to level, and gains of that size do not arrive on a schedule — which is why equity is matched to horizons measured in years rather than months. Against that, three years is not the binding constraint on when you can sensibly take the money; your own goal date is. This is the part that makes the lock-in cheap: it restricts a period during which most holders of an equity fund had no business redeeming in the first place. An ELSS bought for money needed just past the lock-in is a horizon mismatch, and the restriction neither caused it nor fixes it. That ordering is set out in goal-based investing.

It changes the shape of the fund's liabilities. An ordinary open-ended equity fund can in principle be asked to return its entire corpus tomorrow, and is managed knowing that. In an ELSS a rolling slab of the corpus is not redeemable by anyone, at any price, for any reason. That is a structurally stickier liability. Whether managers actually run ELSS portfolios differently as a result is an empirical question, and not one this article answers — the mechanism is real, the behavioural consequence is unmeasured here.

Put the two halves of the usual claim apart, because only one of them is solid. ELSS is conventionally described as the shortest commitment among the instruments that earn this deduction; that is a comparison worth checking against the current Schedule XV rather than taken on faith, since the other terms are administered and can be revised. What does not depend on the comparison is the second half. The rest of the basket is debt, an insurance contract, or a retirement account, and their years are pure waiting. Here the years overlap almost entirely with a holding period the asset required regardless. The shape of each instrument's lock-in is laid out next to the rest of the basket.

How the exit is taxed, and the rate that can never apply

An ELSS is a domestic equity fund by construction, which puts it past the threshold — more than 65% of total proceeds in domestic equity shares — that makes a scheme an equity-oriented fund for tax purposes. Redemption therefore takes equity treatment under section 198 of the Income-tax Act 2025 — the provision that was numbered 112A until the 1961 Act was repealed.

Long-term treatment applies once units have been held for 12 months. The long-term rate is 12.5% on gains above ₹1.25 lakh in a financial year, and that threshold is a single annual allowance across all your equity gains — not one per scheme and not one per folio. Short-term gains on equity-oriented units are taxed at 20% under section 196, formerly 111A. The full mechanics are in capital gains tax.

Now put the two rules next to each other, because their interaction is the part nobody states. The lock-in is three years. The long-term holding period is 12 months. The lock-in is the longer of the two, and it applies to every unit without exception.

So no ELSS unit can be redeemed short-term. The section 196 rate is unreachable in this category — not unlikely, not rare, structurally out of reach. Every rupee of gain you realise from an ELSS is long-term by construction. In an ordinary equity fund the short-term rate is a live hazard for anyone who redeems inside the holding period; here the statute has already removed it. The one thing that could qualify this is an event that resets how the holding period is reckoned — a transmission, a merger, a segregated portfolio — and that is an open question rather than a known exception.

Note what does not change with regime. The deduction is regime-dependent; the capital-gains treatment is not. An ELSS held by someone on the default new regime is taxed on exit exactly as one held by someone on the old regime.

And the arithmetic mistake this section exists to prevent: the deduction is not a return. It reduces the tax payable in the year the money went in — a one-off reduction in what the units effectively cost you, which does not recur in years two and three. Adding your marginal rate to the fund's annualised return produces a number that describes nothing at all. If you want the two in one figure, the honest way is to treat the tax saved as a reduced outflow on the date it actually lands, and compute the rate of return that reconciles cashflows arriving on irregular dates — XIRR.

The ceiling is shared, and the lock-in is not capped

Section 123 read with Schedule XV allows a deduction up to ₹1.5 lakh, under the old regime. That figure is a combined ceiling across every qualifying item on the schedule, not an allowance for ELSS.

Here is the specific mistake. A salaried taxpayer treats the ceiling as empty each April, when a large part of it may already be spoken for by claims that happen automatically — the provident fund contribution deducted from salary each month is commonly the largest of them, and the one people forget to count because they never chose it. If the automatic claims already fill the ceiling, an ELSS instalment on top of them buys no deduction whatsoever. Which items draw on the shared ceiling is set out in tax-saving investments.

And this is where the two features come apart. The deduction is capped. The lock-in is not. Every rupee invested above the ceiling gets the full three-year restriction and none of the tax benefit. At that point the like-for-like comparison is not “ELSS versus nothing” — it is an ELSS against an ordinary diversified equity fund with a comparable mandate, identical tax on exit, and no lock-in at all. Which is a comparison with an obvious structure to it, and one that almost no March-deadline conversation ever gets around to having.

So the sequence worth running once a year, before any instalment: which regime am I on, how much of the Schedule XV ceiling is already consumed, and is this rupee inside it or outside it.

What the lock-in costs you

Nothing in finance is one-sided, and an option removed is an option removed in both directions. The lock-in that stops a panicked exit stops a considered one just as completely.

• You cannot leave a scheme you no longer want. The manager changes, the mandate drifts, the expense ratio goes up — and you wait. In an ordinary fund those are reasons to switch; here they are reasons to wait out a clock.
• You cannot rebalance the locked portion. If equity runs hard and your allocation drifts away from what you set, the locked slab cannot be trimmed. That breaks the mechanism described in rebalancing, which depends on being able to sell what has grown.
• It is not available in an emergency. Not at a discount, not with an exit load, not at all. Which is an argument for the emergency fund sitting somewhere else entirely, and for never letting a March tax exercise reach into money that has a job inside the lock-in.
• Selection errors compound for the whole period. A poor choice in an ordinary fund costs you until you notice. A poor choice here costs you until the statute lets go.

That last point flips the usual framing of ELSS as the low-effort tax option. The irreversibility raises the stakes on picking the scheme, rather than lowering them.

It is still an equity fund, and that is what to examine

Once the regime question and the ceiling question are settled, what remains is a fund-selection problem with everything the tax framing distracted from still in it.

The category rules do not pin an ELSS to a market-capitalisation band the way a large cap or mid cap mandate is pinned, so two schemes wearing the same three letters can hold very different portfolios — one concentrated in large companies, another running a meaningful mid and small cap weight. Comparing their returns without comparing their holdings compares two different risks.

The rest is the ordinary checklist, and it applies with more force here because the decision is not reversible for the length of the lock-in:

• Cost — charged annually on the whole corpus, including through the years you are not allowed to leave.
• Direct or regular plan — identical portfolios, different expense ratios, and the gap compounds across a holding period the statute has made long.
• Every available three-year and five-year window rather than the single one ending today — rolling returns — because a trailing number depends heavily on where its window happened to end.
• The worst peak-to-trough fall the scheme has been through, and how long it took to recover — its maximum drawdown. Here it is not a question of whether you would have held on, since you would have had no choice, but of whether the corpus was where you needed it when the lock finally opened.

And the allocation question sits above all of it: how much equity belongs in the portfolio at all is decided by asset allocation, not by how much deduction is left on the ceiling.

Testing it on the real series

Two things about an ELSS are worth simulating rather than assuming: what a monthly contribution would actually have produced across the real NAV path, and how deep the fall was during the years you would not have been able to exit.

FNOTrader's Mutual Funds app runs any contribution schedule against a scheme's full NAV history — around 34 million AMFI NAV rows, refreshed nightly at 22:30 IST — and reports XIRR, invested against value, maximum drawdown, and the rolling-return distribution for the period. The drawdown figure carries extra weight in this category, since it describes a stretch you could not have shortened.

Past performance describes what happened over the period measured and is not an indication of future returns. FNOTrader is not a SEBI-registered investment adviser and does not recommend schemes.

Common questions

Continue reading

More in Tax Planning · App: Mutual Funds · Definitions: glossary · Free tools: calculators · All: every article