Bonds From Scratch — Coupon, Yield, Duration, the Seesaw

Harpreet Singh, 53, has run his garment shop in Ludhiana for thirty years. He has about ₹3,00,000 saved, an old LIC endowment, roughly 200 grams of gold, and about seven years until he’d like to slow down. Every rupee he has ever set aside went into something he could touch or that a person he trusted promised him. The word “bond” makes his eyes glaze over — “boring, and full of words like coupon and yield and duration that sound designed to trip me up.”

Lakshmi Rao, 64, a retired schoolteacher in Hyderabad, has the opposite worry. She lives off a corpus of about ₹95 lakh (₹95,00,000) and needs roughly ₹50,000 a month, so certainty is everything to her. Someone told her bonds were “safe” — which is why she was startled when a relative’s “safe” bond fund showed a loss on the screen after the news said interest rates had moved. “If it’s safe,” she asked, “how can it be down?”

Both fears are fair, and this lesson answers both. By the end, “coupon, yield and duration” will be three plain ideas, not a trap — and you will know exactly why a safe bond can dip on a screen and yet never cost a careful holder a single rupee. We will build the whole thing from one ordinary ₹1,000 bond, and add nothing you can’t see coming.

Strip away the jargon and a bond is the simplest thing in finance: a loan, with you on the lending side. When Harpreet buys a bond, he is lending his money to whoever issued it — a government or a company, called the issuer — and in return they put two promises in writing: a regular interest payment, and his money back in full on a fixed future date.

That is the whole definition. A bond is a loan you make to a government or company — on paper, and (unlike your FD) one that can be bought and sold. You are the lender; they are the borrower who owes you.

It helps to contrast it with owning shares (Lesson 22, Stocks — What You Actually Own). As a shareholder you own a slice of the company and ride its fortunes up and down. As a bondholder you own no slice at all — you are simply owed a debt. You don’t share the good years, but you also don’t suffer the bad ones the same way: you get your fixed interest whether the company soars or stumbles, and if it collapses, bondholders are paid before shareholders see a rupee. Less upside, far more certainty — which is exactly why bonds are the steady, stabilising half of a portfolio (Lesson 7, Diversification and Asset Allocation).

Harpreet already understands a fixed deposit, and an FD is also lending for interest: he hands the bank a sum, it pays him a fixed rate, and it returns his money at the end. So why the fuss about bonds — what’s actually different? One thing, and it changes everything: a bond can be sold to someone else before it matures.

His FD is a private contract between him and the bank. Nobody else can buy it, so it never has a “market price” — it simply sits there paying its rate until maturity. (He can break it early for a small penalty, but there is no price that moves up and down.) A bond, by contrast, trades. And the moment a thing can be re-sold, it gets a secondary-market price — the price other buyers will pay for it today — which can be more or less than what Harpreet originally paid.

That single fact — a bond has a secondary-market price — is the source of everything else in this lesson: the seesaw, the three yields, and the mystery of the “safe bond that lost value.” An FD’s safety is that it has no price to fall. A bond’s price can wobble — but, as we’ll see, the wobble is only ever real if you choose (or are forced) to sell early. Both are debt; the bond just happens to trade.

Every bond, however grand it sounds, is built from just four parts. Meet them on Harpreet’s ₹1,000 bond, and you can read any bond in the market. The picture below lays them out as the bond’s actual schedule of cash — money out today, coupons in each year, and your principal back at the end.

Face (par) value — ₹1,000. This is the sum returned to Harpreet at the end, and the number the interest is figured on. It is not necessarily what he pays for the bond (that can differ, as we’ll see); it’s the amount printed on the loan.

Coupon — 8%, which on a ₹1,000 face means ₹80 a year. The coupon is the fixed interest the issuer pays, set the day the bond is issued and never changing afterwards. (The name is literal history: bonds once came with paper coupons you clipped off and took to a bank for your interest.) Rates can double, the bond’s price can lurch — Harpreet still collects exactly ₹80 a year. That fixedness is the anchor for everything that follows.

Maturity — 10 years. This is the day the loan ends: the issuer pays back the ₹1,000 face and the bond ceases to exist. A bond can mature in 91 days (a T-bill) or in 40 years; the length is what you choose to match to your needs.

Issuer — the government or company that borrowed from Harpreet. This is the part that decides whether the promise is worth anything: a Government of India bond is about as safe as a rupee promise gets, while a shaky company’s bond is only as good as the company. We’ll weigh that carefully in the credit-risk section.

Read the schedule left to right and the machine is obvious: ₹1,000 goes out today; ₹80 comes back every year for ten years — ₹800 of interest in all; and the ₹1,000 itself returns at maturity. Harpreet turns ₹1,000 into ₹1,800 of cash over the decade, and — this is the point — the ₹80 and the ₹1,000 are promised in writing. (A real government bond splits its coupon into two half-yearly payments of ₹40; the idea is identical.)

Now Lakshmi’s question — how can a safe bond lose value? Here is the machine behind it. Suppose Harpreet buys his ₹1,000 bond when 8% is the going rate, so he pays exactly ₹1,000. A bond priced at its face value like this is said to trade “at par.” Then, a while later, the market pushes rates up: brand-new bonds are being issued at 10%.

Harpreet’s bond still pays only ₹80 a year. If he wants to sell it, who would pay ₹1,000 for an ₹80-a-year bond when a fresh ₹1,000 bond now pays ₹100? Nobody. To find a buyer, his bond’s price has to fall until that fixed ₹80 (plus the ₹1,000 at the end) works out to the new 10% — which happens at about ₹877. So his “safe” bond is suddenly worth ₹877 on the screen: a fall of 12.3%, caused by nothing but a change in rates.

The reverse is just as true, and just as mechanical. If rates fall to 6%, new bonds pay only ₹60, so Harpreet’s ₹80-a-year bond looks generous — buyers bid its price up to about ₹1,147, a rise of 14.7%.

This is the price–yield seesaw: market yields and existing bond prices sit on opposite ends of a plank. Push rates up, prices go down; push rates down, prices go up. You met the seesaw in Lesson 9 (Reading the Rate Cycle); here you can see the rupees behind it, and — crucially — what did and didn’t move. The coupon didn’t change. The ₹1,000 due at maturity didn’t change. Only the resale price today moved.

And it isn’t a textbook curiosity. Over the past year the 10-year government bond yield eased from about 7.7% to about 6.75% as the RBI cut its repo rate to 5.25% (mid-2026). Rates fell — so, exactly as the seesaw predicts, the price of every bond already in the market rose. Anyone holding one was quietly up, for doing nothing at all.

Two quick words for the two sides of that seesaw. When a bond’s price sits below its ₹1,000 face — like Harpreet’s at ₹877 — it is trading at a discount, and it’s called a discount bond. When it sits above face — ₹1,147 — it’s at a premium, a premium bond. At exactly ₹1,000 it’s “at par.” These are nothing more than names for below, at, and above the face value.

They matter because they quietly reveal your real return. Buy at a discount and you get the ₹80 coupons and a bonus on top: the price is pulled up from ₹877 to ₹1,000 by the maturity day — an extra ₹123 in your pocket. Buy at a premium and you still get the ₹80 coupons, but you swallow a slide from ₹1,147 back down to ₹1,000 — losing ₹147 of what you paid. That pull toward face value is the piece the coupon rate completely hides — which is exactly why the coupon alone is a poor guide to what a bond will earn you, and why we need a sharper yardstick.

A bond is always quoted with a “yield,” but there are really three of them, and mixing them up is where beginners quietly overpay. Take Harpreet’s one bond and look at all three at once, across a discount, a par, and a premium price.

The coupon rate — 8% — is simply ₹80 ÷ ₹1,000 face. It’s fixed forever, and it tells you the rupees, not your return, the moment the price isn’t ₹1,000.

The current yield corrects for that: it’s ₹80 ÷ the price you actually pay. Buy the bond at a discount for ₹877 and the current yield is 9.12%; buy it at a premium for ₹1,147 and it’s 6.97%. Better than the coupon — but it still ignores the pull back to ₹1,000.

The yield to maturity (YTM) is the real one. It’s the single annual return you’ll actually earn if you buy at today’s price and hold to maturity, counting every coupon and the pull back to ₹1,000. For the discount bond it works out to 10%; for the premium bond, 6%; at par, 8%.

Notice how the three line up, and then flip. At a discount: coupon 8% < current yield 9.12% < YTM 10%. At a premium: coupon 8% > current yield 6.97% > YTM 6%. Only at par do all three agree at 8%. So a bond “paying 8%” bought for ₹1,147 really earns you 6% — and if you compared it to an FD on its coupon alone, you’d overpay badly.

The seesaw tells you which way a bond’s price moves. Duration tells you how far. Take the same 8% bond and the same 1% rise in rates, and change only the maturity — watch what happens.

A 3-year bond’s price falls about 2.5%. A 10-year, about 6.4%. A 30-year, about 10.3%. Same coupon, same rate move — wildly different pain. The reason is intuitive: the longer the bond, the more years of below-market coupons a buyer is stuck with, so the price has to drop further to make up for it. That sensitivity is the bond’s duration — roughly how many percent the price moves for each 1% change in rates. A duration of about 6.7 (the 10-year here) means a 1% rate move shifts the price about 6.7%.

Two things lengthen duration: a longer maturity, and a lower coupon (a bond that returns your money more slowly). Both mean you’re waiting longer to be made whole, so the price is more sensitive to rates in the meantime.

This is precisely why Lakshmi, who needs certain income and can’t stomach big swings, leans short. A 3-year bond barely twitches when rates move; a 30-year bond is a rollercoaster she has no reason to ride. Duration is the dial: turn it down (shorter, plainer bonds) for a calm ride matched to a near-term need; turn it up only if you can genuinely wait out the swings. Match the bond’s length to your time horizon and duration works for you instead of against you.

Now Lakshmi’s fear gets put fully to bed. Go back to that ₹877 dip. Here’s the thing the red number on the screen never mentions: a bond’s price is pulled back to its ₹1,000 face as maturity approaches — because on the maturity day it pays exactly ₹1,000, no more, no less. The dip isn’t a one-way loss; it’s a temporary sag that heals itself by the finish line.

Lakshmi uses this deliberately. She matches each bond to when she’ll need that slice of money — a bond that matures the year she needs the cash — and she holds it to maturity, which just means keeping it until the end rather than selling early. Along the way, rates rise and fall and her bond’s price wanders; but because she never sells, the wandering is pure noise. She collects her coupons and gets her ₹1,000 back, exactly as promised. The seesaw is someone else’s problem.

The only person the seesaw actually hurts is one forced to sell early into a dip. Sell in year three, when rates have jumped and the price is ₹903, and a paper sag becomes a real loss of ₹97 per ₹1,000 — a 9.7% hit locked in. But notice what that really was: not the bond failing, but a mismatch between the bond’s length and when the money was needed.

One small mechanic you’ll meet the first time you buy a bond in the market, so it doesn’t startle you. Coupons arrive on set dates. If Harpreet buys a bond three months after its last ₹80 coupon was paid, he will collect the full ₹80 at the next coupon date — even though he only held it for the final stretch. That would be unfair to the seller, who held it for the first nine months and gets nothing.

So the market squares it up. The buyer pays the seller the interest that has built up since the last coupon — the accrued interest. On an ₹80-a-year coupon, three months’ worth is about ₹80 × ¼ = ₹20. Harpreet pays the quoted “clean” price plus that ₹20; the all-in figure he actually hands over is the “dirty” price.

He isn’t losing ₹20 — he gets it straight back inside the full ₹80 coupon a few months later. Accrued interest is just the market splitting one coupon fairly between the seller and the buyer. The app shows you the all-in number, so don’t let the words rattle you: this is bookkeeping, not a cost.

Remember the quiet assumption tucked inside YTM — that you reinvest every coupon at the YTM. Here’s why it matters. Harpreet’s bond hands him ₹80 a year. To actually earn the 8% the YTM promised, he has to put each of those ₹80 payments back to work at 8% too. If rates have fallen and he can only reinvest his coupons at 6%, his realised return quietly comes in below the YTM. That gap is reinvestment risk.

Put numbers on it. Ten years of ₹80 coupons, reinvested at 8%, grow to about ₹1,159. Reinvested at only 6%, they grow to about ₹1,054 — roughly ₹105 less on a ₹1,000 bond. Same bond, same coupons; a lower reinvestment rate shaved the outcome without anyone doing anything wrong. The YTM was never a guarantee — it was a promise conditional on where you can re-home the cash the bond keeps sending you.

Reinvestment risk is the mirror image of the seesaw. The seesaw hurts when rates rise (existing prices fall); reinvestment risk hurts when rates fall (coupons re-home at less). You can’t escape both at once — which is one more reason a bond ladder (several bonds maturing in successive years) is so calming: you’re always reinvesting a little at whatever rate prevails, never all of it at the worst possible moment. A “zero-coupon” bond — one that pays no coupon at all, just a single lump at maturity — sidesteps reinvestment risk entirely, because there are no coupons to re-home.

Everything so far quietly assumed the issuer keeps its promise. That assumption is free for a government bond and costs you for everyone else. The chance the borrower can’t pay a coupon or return your ₹1,000 is credit (default) risk — you met it in Lesson 5 (Risk, Truly Understood); here it decides which bonds actually deserve the word “safe.”

A Government of India bond — a G-sec — is the risk-free floor: the government issues the rupee, so a rupee default is close to impossible, and it yields about 6.75%. Every step down the credit ladder — a top-rated AAA company, then AA, then A, then BBB, then “high-yield” (the polite name for junk) below that — pays a higher yield precisely because the chance of not being repaid climbs. Rating agencies (CRISIL, ICRA, CARE) grade this from AAA at the top down to D for default, and the rating is the first thing to check on any bond that isn’t a G-sec.

“Safe-sounding” is not the same as safe. In 2020, about ₹8,415 crore of Yes Bank’s AT1 bonds — bank bonds, held by people who thought them rock-solid — were written down to zero, effectively overnight. (The Bombay High Court set that write-down aside in 2023, and the matter is now before the Supreme Court — a live case.) The point isn’t to scare you off bonds; it’s to fix the rule in place: the extra yield you earn down the ladder is rent for risk, never a free lunch.

For most people building a safe sleeve, that argues for staying high on the ladder — G-secs and top-rated bonds — and leaving the junk to those who are paid to analyse it. Holding to maturity removes price risk, but it does nothing about credit risk: a defaulting issuer won’t be there to pull the price back to ₹1,000. The full credit ladder, and how corporate bonds and debt funds are priced and taxed, is Lesson 36.

One more risk, quieter than the others. Holding to maturity is the plan — but life happens, and sometimes you must sell early. Can you, at a fair price? That’s liquidity risk (again from Lesson 5): how easily you can turn the bond back into cash without accepting a poor price.

Government bonds are highly liquid — there is almost always a buyer, and the price you get is close to fair. Many corporate bonds, especially smaller issues, trade thinly: days can pass with no buyer, and to sell in a hurry you may have to accept a haircut well below the bond’s fair value. That means an illiquid bond can force a worse loss on an early seller than the seesaw alone would suggest.

So liquidity is part of “safe.” A bond you can’t sell without a discount isn’t quite as safe as its coupon implies. For money you might genuinely need at short notice, favour liquid government bonds — or, honestly, keep that slice in an FD or a liquid fund, which is the emergency-fund logic of Lesson 3. Bonds are for money with a known horizon, not for the cash you may need next week.

Add it up and a bond carries several risks — price (the seesaw), credit (the issuer), reinvestment (re-homing coupons), and liquidity (selling when you must). So why hold them at all? Because none of those is the risk that actually wrecks portfolios. That risk is equity’s gut-lurching 30–50% crashes — and bonds are the shock-absorber that steadies the ride.

When shares plunge, high-quality bonds usually hold their value or even rise — central banks often cut rates in a crisis, and the seesaw then works for you. So a bond sleeve is the dry powder that lets you sleep through a crash, rebalance calmly, and avoid selling your shares at the very bottom. That’s the job: not the highest return, but the steadiest hand. It’s the “debt” half of the asset-allocation idea from Lesson 7, now seen up close.

One practical note before you go further. A bond’s coupon is interest, taxed at your income-tax slab — just like FD interest. So a headline yield is a pre-tax number: for a high earner the after-tax yield is meaningfully lower, while Lakshmi, largely below tax and helped by the ₹50,000 senior-citizen interest deduction (80TTB), keeps far more of hers; Harpreet, on the old regime, sits in between. The debt-fund route is taxed differently again since 2023. We don’t compute any tax here — the investing slice is Lesson 36, and the full treatment lives in the income-tax track.

The moment you understand yield and credit, a whole category of fraud becomes obvious — and it targets exactly the saver this lesson is for: someone tired of ~6.5% FDs, told there’s a “safe bond” paying far more. Here’s how to see straight through it, using only what you now know.

The tell isn’t the glossy brochure; it’s the arithmetic. A “guaranteed” 16% sits seven to eleven points above the government’s own ~6.75% borrowing rate — and the credit ladder just showed the only way to earn that much more is to take a real chance of not being repaid. “Guaranteed” and “secured” are words doing the job the numbers can’t. Demand the credit rating, verify the seller on SEBI Check, remember that a genuine listed bond trades on the NSE or BSE with a SEBI-filed prospectus, and never let “closes today” rush you. Report anything suspect to SEBI SCORES (scores.sebi.gov.in), and to 1930 or cybercrime.gov.in for money already sent — it protects the next person.

The professionals’ bond trick is unglamorous and completely free to copy — which is the whole point of decoding it.

The move is duration-matching: buy a bond that matures around when you’ll need the money, and hold it to that date for a known sum on a known day. You can do it yourself — the exact G-sec that matures your target year, bought free on RBI Retail Direct (Lesson 34), or a plain FD of the right tenor. The tell that an adviser isn’t earning their fee: money you need in a year parked in a long-duration bond fund, or a reach for a high-yield “credit” fund to flash a bigger number. Both mismatch duration or credit to your horizon — and charge you for the privilege. The right move is boring, and boring is the compliment.

If a “safe” bond has already rattled you, read this before anything else. This is not the scam radar — nobody defrauded you. These are the ordinary first bruises of learning a new asset, and both are forgivable.

Two common stumbles: panic-selling a bond fund after it dipped on a rate rise, or steering clear of bonds entirely because you heard “you can lose money in them.” The same fix answers both — the seesaw, plus duration-matching, plus holding to maturity. Selling into a dip is the only thing that turns a paper sag into a real loss; and a ladder of good, short-to-medium bonds held to maturity is one of the calmest places your money can sit. Set the blame down; the price seesaw fools people who do this for a living.

• How can a “safe” bond lose money? Held to maturity with a sound issuer, it doesn’t — the price seesaw only bites a forced early seller. A bond fund can show a loss because it’s valued at market prices daily, but it recovers as its bonds roll into newer, higher-yielding ones.
• Coupon or yield — what’s the difference? The coupon is the fixed rupees (₹80 on a ₹1,000, 8% bond). The yield adjusts for the price you paid. YTM — the yield that also counts the pull back to face value — is the one to compare bonds on.
• What is duration, simply? Roughly how many percent a bond’s price moves for a 1% change in rates. Longer maturity and a lower coupon mean more duration, and bigger swings.
• Should I hold to maturity or sell? If the money is for a dated goal, match a bond to that date and hold — you get a known sum on a known day. Sell early only if you must, knowing you accept the day’s price.
• Are corporate bonds safe? Some are (AAA); many aren’t. A company bond pays more than a G-sec precisely because it can default. Stay high on the credit ladder unless you can analyse credit yourself — the detail is Lesson 36.
• Bond or FD — which is better? Different tools. An FD has no price wobble and DICGC cover to ₹5 lakh; a bond is tradable, can be a G-sec (safer than any bank), and can be laddered to exact dates. Many people hold both.
• Do bond prices really move like shares? They move, but far less for short and medium bonds, and for the opposite reason — rates, not company profits. A 3-year G-sec is a world calmer than a stock.
• What happens at maturity? You get the face value (₹1,000) back, plus the final coupon, and the bond ceases to exist. All the interim price drama simply vanishes at that point.
• Can I lose my whole investment in a bond? In a G-sec, effectively no. In a low-rated corporate, yes — a default can take most or all of it (Yes Bank’s AT1 bonds went to zero). Safety is issuer quality first.
• Are bond coupons taxed? Yes — as interest, at your slab, like FD interest. Debt funds are taxed differently since 2023. We don’t compute it here; see Lesson 36 and the income-tax track.

Here’s the whole lesson in one tool. Enter a bond’s face value, coupon, years to maturity, and the market yield buyers now demand, and it prices the bond, shows you all three yields, and reveals how far the price swings if rates move 1% either way. It opens on the lesson’s discount bond — ₹1,000 face, 8% coupon, 10 years, and buyers now wanting 10% — which should read ₹877, a 9.12% current yield, and a 10% YTM.

Then experiment. Drop the market yield below the coupon and watch the price climb above ₹1,000 into premium territory, with the YTM sliding under the coupon. Lengthen the years and watch the ±1% swing grow — that’s duration, right in front of you. Set the yield equal to the coupon and see all three yields collapse to one number at par. When the tool stops surprising you, you can read any bond in the market.

• Bond — a loan you make to a government or company, on paper and tradable; they pay you interest and return your money on a fixed date.
• Face (par) value — the sum returned at maturity, and the base the coupon is figured on (₹1,000 here). Not necessarily what you pay for the bond.
• Coupon (rate) — the fixed interest the issuer pays, set at issue and never changing (8% = ₹80 a year on a ₹1,000 face).
• Maturity — the date the loan ends and the face value is repaid.
• Issuer — who borrowed from you (a government or company); their credit quality decides how safe the promise is.
• Secondary-market price — what other buyers will pay for the bond today; it can be above or below the face value.
• Price–yield seesaw — market yields and existing bond prices move in opposite directions: rates up, price down; rates down, price up.
• Discount bond — one whose price is below face value; its coupon lags current rates, so a buyer also gains a pull-up to face by maturity.
• Premium bond — one whose price is above face value; its coupon beats current rates, but the price slides back to face by maturity.
• Current yield — the coupon divided by the price you actually pay (₹80 ÷ ₹877 = 9.12%).
• Yield to maturity (YTM) — the true annual return if you buy at today’s price and hold to maturity, counting coupons and the pull to face; the yield to compare bonds on.
• Duration — how sensitive a bond’s price is to rates: roughly the percent it moves per 1% rate change. Longer maturity and lower coupon mean more duration.
• Accrued interest (clean vs dirty price) — interest built up since the last coupon that a buyer pays the seller; the clean price plus accrued interest is the dirty (all-in) price.
• Hold-to-maturity — keeping a bond to its maturity date rather than selling early, so the interim price swings never become real.
• Reinvestment risk — the risk that the coupons you receive must be reinvested at a lower rate than the YTM assumed, reducing your realised return.
• Credit (default) risk — the chance the issuer fails to pay a coupon or return your face value; the reason yields rise down the credit ladder.
• Liquidity risk — the risk you can’t sell the bond quickly at a fair price when you need to.