Capital-Gains Exemptions & the Tax-Free-Bond Play (54EC/54F)

The sale went through. The money — more than you have ever held at once — is sitting in your account, and instead of relief you feel a cold little knot: *how much of this does the tax office take, and did I just do something expensive without knowing it?* You half-remember someone saying “put it in bonds” or “buy another flat,” but you don't know what any of it means, and worst of all there's a feeling you can't shake — that a clock started the day you sold, and it's ticking somewhere you can't see.

Sit with that fear, because it's the right one — and this whole lesson exists to disarm it. Here is the calm centre of it: on a big capital gain, paying the tax is not your only option. The law lets you *reinvest* the gain in specific ways, and if you do it right, and in time, the tax on that gain drops to zero. It isn't a loophole and it isn't a trick — it's written into the Income-Tax Act, and it's free to claim. What it *does* ask of you is a decision: which route, and whether the exemption is worth what it costs. By the end you'll be able to make that decision with numbers instead of dread.

Two people carry the lesson. Tanvi Kapoor — 28, a marketing manager in Gurugram earning ₹12,00,000 a year (₹12 lakh — that is, ₹12,00,000) — has just received ₹50,00,000 (₹50 lakh) from selling a flat she inherited. She has never invested a rupee, she's still grieving, and the windfall feels less like luck than like a test she didn't study for. Suresh Menon — 55, a chartered accountant in Kochi, in the 30% tax slab — is selling a plot with a gain far larger than Tanvi's, and he's about to discover that these exemptions have a ceiling. Their two situations teach the whole toolkit.

Check: name your own version of the fear in one line — *is it that I'll overpay tax, that I'll lock the money up wrongly, or that I've already missed a deadline?* Each of those has an answer in here.

Before we rescue a single rupee, you need the number you're rescuing it *from* — the tax you'd pay by doing nothing. Everything else is measured against this baseline, so let's set it plainly.

First, an important distinction most people get wrong. The ₹50,00,000 in Tanvi's account is not all “gain.” Part of it is simply her money coming back — the value the flat already had when it passed to her (its cost). Only the *increase* over that cost is the capital gain, and only the gain is taxed. Say the flat's stepped-up cost works out to ₹10,00,000 (how that figure is built — from the previous owner's cost or a 2001 fair value — is the income-tax track's job). Then her long-term capital gain (LTCG) — the taxable profit on an asset held long enough to count as long-term — is:

Now the tax. Since 23 July 2024, long-term gains on land or a building are taxed at a flat 12.5% (with no indexation). So doing nothing costs Tanvi:

One more honest note on the baseline. Tanvi acquired this through inheritance, and the flat is old, so she *may* also have the option to compute the gain the older way — 20% with indexation — and pay whichever is lower. That election, and the CII maths behind it, is squarely the income-tax track's. For our decision it changes only the size of the baseline, not the logic: whatever the exact bill, the question is the same — *pay it, or reinvest the gain to make it disappear?*

Check: can you separate, in Tanvi's ₹50,00,000, the part that's *gain* (₹40,00,000) from the part that's *cost coming back* (₹10,00,000) — and say why only the first is taxed? If yes, you're ready to shelter it.

The first escape route is the cleanest, and you've already met the tool. Section 54EC says: take the *gain* from selling land or a building, and within a set window put it into special government-backed capital-gains bonds — and that gain becomes exempt. No tax on it, at all. These are the 54EC bonds you met as an instrument in Lesson 35; here you're using them for their real purpose — sheltering a gain.

The rules are specific, and each one matters, so let's name them on Tanvi's numbers. To shelter her whole ₹40,00,000 gain, she buys ₹40,00,000 of 54EC bonds. Four features define the deal:

• Who issues them — and only them. Just four government-backed issuers: REC (Rural Electrification Corp), PFC (Power Finance Corp), IRFC (Indian Railway Finance Corp), and HUDCO (added in April 2025). You buy at ₹10,000 face value per bond, directly, on the issuer's own site (or via RBI Retail Direct / your broker). *NHAI stopped issuing 54EC bonds back in 2022* — a detail worth carrying into the Scam Watch later.
• The ₹50 lakh cap. You can put at most ₹50,00,000 into 54EC bonds for a given gain (counted across the year of sale and the next). Tanvi's ₹40,00,000 fits comfortably under it — the cap won't bite for her (it will for Suresh).
• The 5-year lock. The money is locked for 5 years, and the bonds are non-transferable — you can't sell, pledge, or exit early. This is the real price of the exemption.
• The 5.25% coupon — and it's taxable. The bonds pay 5.25% a year, paid out annually. That's a modest rate, and unlike the gain you sheltered, the *coupon* is fully taxable at your slab (no TDS for residents — you report it yourself). So you're locking money at a low, taxed yield in exchange for wiping out the tax on the gain.

Put together, Tanvi's 54EC move is: ₹40,00,000 into bonds → her ₹40,00,000 gain is exempt → she saves the ₹5,00,000 of tax. The ₹10,00,000 that was only her cost coming back stays fully liquid — she never had to lock that. Here's the application she'd actually fill, so the four features stop being abstract and become fields on a form:

Read down that form and the exemption stops being a mystery: it's a plain bond purchase, in your own name, at face value, with four numbers that decide everything — the cap, the coupon, the lock, and the clock — the one that bites. Notice too what the form is *not*: there's no agent, no fee, no “scheme.” The tinted rows are exactly what a good adviser would point at — and exactly what a bad one hopes you won't read.

Check: for a property gain, can you state 54EC's four numbers from memory — the ₹50 lakh cap, the 5-year lock, the 5.25% (taxable) coupon, and the 6-month window? Those four *are* the decision.

The second route sends the money somewhere very different — into a home. Section 54F says: take a long-term gain from selling *almost anything* (land, gold, shares — any long-term asset that *isn't* already a residential house), reinvest the whole net sale consideration into one residential house in India, and the gain is exempt. Where 54EC hands your gain to a bond, 54F turns it into a place to live.

The phrase that does the work is net sale consideration — a term worth defining precisely, because 54F hinges on it. It's the full sale price *minus* the costs of selling (brokerage, legal charges). It is not the gain; it's the whole amount you received. And that's the catch: for the *full* exemption you must reinvest the entire net consideration, not merely the profit.

So if Tanvi went the 54F way, she'd have to put the whole ₹50,00,000 into one house to exempt her ₹40,00,000 gain and save the ₹5,00,000. Reinvest less than the whole, and the exemption shrinks *proportionately*:

Work it on a partial house. Suppose Tanvi buys a ₹25,00,000 flat — half her sale amount. Then only half her gain is exempt: ₹40,00,000 × (25 ÷ 50) = ₹20,00,000 exempt, leaving ₹20,00,000 taxable, which at 12.5% is ₹2,50,000 of tax. Half the reinvestment, half the shelter. To zero the tax, the whole ₹50,00,000 has to go in.

54F also carries conditions, because the government is exempting your gain to house you, not to fund a portfolio of flats:

• The one-house rule. On the date you sell, you must not already own more than one residential house (other than the new one). 54F is aimed at people putting a windfall into a home, not serial property investors.
• Hold it. Broadly, you must hold the new house for 3 years and not rush out to buy or build *another* one within the window, or the exemption is clawed back.
• A ₹10 crore ceiling. Since 2023, the amount of consideration counted for the exemption is capped at ₹10 crore — a limit that only touches very large sales (not Tanvi's, not Suresh's).

Check: can you say why a ₹25,00,000 house exempts only ₹20,00,000 of Tanvi's ₹40,00,000 gain — and what she'd have to invest to exempt all of it? If the proportion clicks, you've got 54F.

Every exemption so far comes with a deadline, and the deadline is the whole game — the single most common way people *lose* an exemption they fully qualified for is simply running out of time. So let's make the clock visible. It starts the day you sell (the date of transfer), and it has three separate hands.

The three hands move at wildly different speeds, and that gap is the whole danger. The §54EC hand is the short, unforgiving one — 6 months from the sale to buy the bonds, after which the door simply shuts and the gain is taxed. The §54F hands are far kinder — years, not months (2 years to buy a house, 3 years to build) — and one of them even reaches backwards: a house bought up to 1 year *before* the sale can still count, the only window that opens before you've even sold. The third hand, in gold, is the CGAS checkpoint at the filing due date — the safety valve we'll meet next.

Check: for a property sale, which is the shortest deadline, and what happens the day after it passes? (Six months, for 54EC — after which that route is gone for this gain.)

Here's a very real problem. The 54F house windows run for *years*, but your tax return is due much sooner — 31 July 2026 for someone like Tanvi. What if the deadline to *file* arrives and you still haven't found the right house? Must you rush into a flat you don't want, or forfeit the exemption? Neither. The law's answer is the Capital-Gains Account Scheme (CGAS) — a holding pen.

It works like this: if you haven't reinvested by the return due date, you deposit the unutilised amount into a special Capital-Gains Account at a public-sector bank *before* you file. That deposit is treated as if you'd reinvested — so the exemption stays alive — and you then have the full 2-to-3-year window to actually buy or build. The account comes in two flavours: Type A (savings-like, you can withdraw as you spend on the house) and Type B (a term deposit, up to 3 years). If you ultimately *don't* use the money for the house within the window, the sheltered gain becomes taxable then — CGAS defers the decision, it doesn't cancel it.

For Tanvi, that means she could deposit her ₹50,00,000 in a CGAS account by 31 July 2026, keep the 54F option fully open, and take her time (up to 2 years to buy, 3 to build) to decide whether a home is really what she wants — without the tax clock forcing a panic purchase.

Check: if you want the 54F house route but can't buy before you must file, what one action preserves the exemption — and which route does that action *not* help? (Deposit in CGAS by the due date; it does nothing for 54EC.)

You now have all four moves — pay, 54EC, 54F, CGAS. The way to actually *choose* is to lay them against the same gain and see what each does to two things at once: the tax, and your money's freedom. Here is Tanvi's ₹40,00,000 gain run through all four.

The table doesn't decide for you — it reframes the decision. Every route carries a price; they're just charged in different currencies. The two routes that wipe the tax to nothing charge you in *access* to your money — five years locked in a bond you can't sell, or your whole windfall committed to a single house you can't quickly exit. The route that keeps your money free charges you in *tax*. And the holding-pen route charges you only in *time*, postponing the choice without forfeiting it. So the sharper question isn't “how do I pay the least tax?” — it's “which of these prices can I least afford right now?” The next two sections turn that instinct into arithmetic: first the lock-in-versus-pay maths on Tanvi's gain, then the ceiling that reshapes the answer for a bigger one like Suresh's.

Check: in one sentence each, what does each route *cost* Tanvi — the tax, the lock, the house, the delay? If you can name all four prices, you can advise the decision.

Now the question that actually matters, and the one most advisers skip: is the exemption even worth taking? It sounds obvious — of course you'd rather pay ₹0 than ₹5,00,000. But zero tax isn't free. 54EC's ₹5,00,000 saving comes chained to a *five-year lock at 5.25%* — a low, taxed yield on ₹40,00,000. To judge it honestly, you have to fold the saving and the lock into a single number and compare it to what the money could do *instead*.

Here's the honest way to do that. Think of 54EC as an investment and ask what it really *yields*. It pays the 5.25% coupon — plus, by saving ₹5,00,000 of tax on ₹40,00,000 locked for five years, it hands you an extra return of about 2.38% a year (that's the ₹5,00,000 saving spread over the five-year lock). Add them up and 54EC behaves like a ~7.6% a year, low-risk, five-year instrument — once you count the tax you didn't pay.

That reframing is the whole insight. Suddenly 54EC isn't “free money” *or* “a bad low-yield lock” — it's a ~7.6% five-year bet, and whether you take it depends entirely on your alternative. Line them up: a safe FD pays about 6.5% (taxable), so 54EC's ~7.6% beats it — if your realistic alternative is playing safe, lock in. But a diversified equity portfolio is assumed around 11% (an optimistic-but-defensible long-run figure, never a promise), which comfortably beats ~7.6% — so if you'll genuinely invest the freed money in equity *and hold*, pay the ₹5,00,000 and stay liquid. The break-even sits around 7.5%: earn more than that and paying wins; less, and the lock wins.

Check: what return would the freed money have to beat for paying to win over 54EC — and is that above or below what a safe FD pays? (About 7.5%; above a ~6.5% FD, so the answer turns on whether you'll really take equity risk.)

Tanvi's gain fit neatly under the 54EC cap. Suresh's won't — and that's exactly why he's here. Suresh is selling a plot with a long-term gain of ₹90,00,000 (₹90 lakh). Do nothing and the tax is 12.5% × ₹90,00,000 = ₹11,25,000. Naturally he reaches for 54EC — and hits the ceiling.

The cap does something Tanvi never had to face: it leaves a chunk of gain *stranded*. 54EC can shelter at most ₹50,00,000 of gain, saving 12.5% × ₹50,00,000 = ₹6,25,000. But the other ₹40,00,000 of his gain is over the ceiling — 54EC simply can't reach it — so it's still taxed at 12.5% = ₹5,00,000. Maxing 54EC takes Suresh from an ₹11,25,000 bill down to ₹5,00,000, not to zero. Above ₹50 lakh of gain, the bond route alone can never get you all the way; for the residual he must either pay, or absorb it with a 54F house (the two can be combined, but *how* is the income-tax track's department).

Check: on a ₹90,00,000 property gain, how much can 54EC exempt, how much is left taxable, and could Suresh put his *equity* gains into the same bonds? (₹50 lakh sheltered, ₹40 lakh still taxed; and no — 54EC is land/building only.)

When a large property gain lands on a wealthy client, a wealth manager reaches — almost reflexively — for exactly these tools. It's a genuine, legal, valuable move. But it's also one you can run yourself, and the difference between a manager worth their fee and one who isn't shows up in *how* they use it.

The card lays it bare: the “product” at the centre of the move is a government bond bought at face value, or a house you choose, or a bank deposit — none of which needs a manager's allocation skill to access. You can buy 54EC bonds yourself, direct from the issuer, within the 6 months; you can open a CGAS account at any public-sector bank; you claim the exemption on your own return. What you might reasonably *pay* for is judgement — the timing, the paperwork, and a second opinion on a large, once-off call. What you should *not* pay for is access to a bond the government sells to anyone.

And the tell is the lock-in-vs-pay maths from the last section. A manager who reaches for 54EC without first asking what return you'd earn on the freed money — or who steers you to one specific builder's flat for 54F — is optimising *this year's tax bill*, not *your wealth*. A good one runs the ~7.6%-vs-your-alternative comparison *with* you, and sometimes concludes you should simply pay and stay invested. That willingness to talk you *out* of the shiny tax saving is the sign the fee is buying you something real.

Check: what's the single question that separates a manager optimising your wealth from one optimising your tax bill? (“What will you earn on the money if you *don't* lock it up?”)

Sell a property and, within days, the pitches arrive — because a fresh windfall is visible and the seller is flush and anxious. Almost all of them exploit the same gap: you know there's *some* way to save the tax, but not the details. Name the tells and they lose their grip, because every one is just a push to skip a step you can do yourself for free.

The through-line across all four tells is a single lie: that the exemption is a *product* someone sells you, rather than a *claim you make* on your own return. Real 54EC bonds come from four government issuers at ₹10,000 face value, bought directly — nobody “guarantees” you a bigger exemption than the ₹50 lakh the law sets, and nobody needs to route your money through their account to “arrange” it. The 54F “buy *this* flat and pay zero tax” pitch is a builder using a real exemption to offload unsold inventory — the shelter is genuine, but the house has to fit *your* life, not their sales target. And any urgency to wire money *now* (“the window shuts today” — sometimes even pitching NHAI bonds that stopped existing in 2022) is manufactured to stop you checking.

Check: which three verbs mark a fake exemption pitch — and where do you actually buy real 54EC bonds? (“Sells / routes / guarantees”; directly from REC / PFC / IRFC / HUDCO or via RBI Retail Direct.)

Maybe you're reading this a beat too late. Perhaps you sold last year, panicked, and paid the whole tax bill because “just pay it” was the only advice in the room. Or you'd never heard of a 6-month clock and it ran out before you knew 54EC existed. Before anything else: set the blame down. These rules aren't handed to you at the registry, the deadlines are short and buried, and grief makes them invisible. Feeling foolish is the situation working on you — it is not a verdict on you.

The point of the card is that *what's still open is usually more than you fear.* If you're still inside 6 months of the sale, 54EC is live — buy the bonds now. If your return isn't filed yet, a CGAS deposit before the due date preserves the 54F house route. If you've already filed and paid but it's before the revised-return deadline (31 December 2026 for this year), and you *did* make a qualifying investment in time, a revised return may put the exemption on record — ask your CA. And if this gain's window has truly closed, then you paid a *lawful tax, not a penalty* — the money settled a real due — so carry the 6-month clock into your next sale, and if a fake agent took money on the side, report it for the next person.

Check: name the three still-open doors after a gain — the live 54EC window, the CGAS deposit before filing, and the revised return — and the one question that opens each. (Am I within 6 months? Have I filed yet? Did I actually invest in time?)

Time to put it all in one place. The planner below takes a gain — yours or Tanvi's — and does what you'd do by hand: it computes the tax under each route, the tax saved versus just paying, what the route *costs* you (the 5-year lock at 5.25%, or your whole sale amount in a house), and the deadline clock. Switch the asset type and watch 54EC switch off for anything that isn't land or a building; change the route and watch the trade-off flip.

Play with the two things that decide everything. First, the asset type: flip it to “other long-term asset” and 54EC greys out — a live reminder that bonds shelter *property* gains only, while 54F still works for gold or shares. Second, the route against your alternative return: on 54EC, raise the “return on the freed money” above ~7.6% and the verdict swings from “lock in” to “paying likely wins” — the lock-in-vs-pay maths, live. Try Tanvi's ₹40,00,000 as land (54EC saves ₹5,00,000, ₹40,00,000 locked), then as a ₹25,00,000 54F house (tax ₹2,50,000, because only ₹20,00,000 of the gain is exempt). Same tool, every scenario in the lesson.

Check: put in a gain above ₹50,00,000 as land/building and choose 54EC — does the tool leave a residual taxed (the cap biting, just like Suresh)? If you can make the ceiling appear, you understand it.

The questions people actually ask the week after a big sale, answered straight.

Do I have to buy a house to save the tax? No — that's only 54F. For a land or building gain you can instead put up to ₹50,00,000 into 54EC bonds and keep your money out of property entirely. Buying a house is one route, not the only one — and never a good reason on its own to buy a house you don't want.

What exactly is the 54EC lock-in — and can I get out early? Five years, and no. The bonds are non-transferable and non-marketable: you can't sell, pledge, or redeem them early. You receive the 5.25% coupon each year (taxable at your slab) and your capital back after 5 years. That illiquidity *is* the price of the exemption — factor it in before you lock.

Can I split a gain across 54EC and 54F? For a large property gain, broadly yes — you might shelter ₹50,00,000 in bonds and absorb the rest with a house — which is how someone like Suresh gets past the ₹50 lakh ceiling. But the interaction (what counts against what) is fiddly and is the income-tax track's job; here, just know the ceiling exists and that a house can cover what the bonds can't.

What if I miss the 6-month 54EC window? Then that route is gone for this gain, and — crucially — a CGAS deposit can't bring it back (CGAS is only for the 54/54F house routes). Your remaining options are a 54F house if you're within its longer windows, or simply paying. This is why, if 54EC is even possible, you act on the 6-month clock first.

Is the ₹1.25 lakh long-term exemption not available on my property gain? Correct — it isn't. That ₹1,25,000 free slice is equity-only (Section 112A, for listed shares and equity funds). A property gain sits under Section 112, which has no such exemption, so the whole gain faces the 12.5%.

Is CGAS a way to just avoid the tax by parking money? No. CGAS is a *holding pen*, not a shelter. It keeps the 54/54F exemption alive while you find a house — but if you don't actually buy or build within the 2-to-3-year window, the sheltered gain becomes taxable then. It buys you time to decide, not a permanent escape.

The 54EC coupon is only 5.25% — isn't that a terrible return? On its own, yes, it's low, and it's taxable. But that's the wrong comparison. Fold in the tax you *saved* and 54EC behaves like a ~7.6% five-year, low-risk instrument — which beats a safe FD. It only looks “terrible” next to equity you'd actually hold for years. Judge it against your *realistic* alternative, not against the best case.

Does 54EC or 54F work for my mutual-fund or share gains? 54EC — no, it's land/building only. 54F — technically yes (it covers any long-term asset that isn't a residential house, so shares or gold can qualify) *if* you reinvest the whole net consideration into a house. But sinking a share portfolio into a single flat is rarely what you want; equity gains are usually better managed by the ₹1.25 lakh annual harvest (Lesson 43).

Someone offered to “get me” a bigger exemption for a fee — legit? No. The ₹50 lakh cap and the exemptions are fixed in law; no one can enlarge them, and no fee is needed to claim them. Real 54EC bonds are bought directly from REC / PFC / IRFC / HUDCO at face value. A “guaranteed bigger exemption,” a markup, or a request to route your money is a scam — verify and report (see Scam Watch).

I have no idea what to do with the windfall while I decide — where does it sit safely? That's the right instinct — don't rush it into anything. Parking a windfall safely (liquid funds, T-Bills, sweep FDs) while you learn is its own skill, and it's covered in the earlier lessons on safe parking and in Lesson 62 · The Windfall. The exemption clock and the deployment decision are two separate problems — solve the clock first, then take your time on the deployment.