Government Securities — G-Secs, T-Bills & SDLs

Say the words 'government bond' out loud and a certain picture appears: a trading desk in a bank, a screen full of yields, men in suits moving crores. It sounds like something only banks and big institutions buy — a market with a locked door, and no obvious handle for someone like you. So the honest first reaction is a small shrug: *that's not for me, and even if it were, I wouldn't know where to start.*

Here is the reassurance, up front, before a single yield. An ordinary individual can own these directly — the exact same government securities the banks buy, bought at the same auction, at the same price, sovereign-backed, with zero fees. Since November 2021 the Reserve Bank of India has run a scheme (RBI Retail Direct) that opens that locked door to any resident with a bank account and a PAN. This lesson is about *what* is behind the door — the safest debt in the country, and why it's the anchor every other return is measured against. The *how-to* — opening the account, placing the order — is the very next lesson, so you never have to guess at the mechanics.

We follow two people you've met before. Harpreet Singh — 53, a garment shopkeeper in Ludhiana, about seven years from retiring — keeps his safe money (₹3,00,000 of savings, alongside an LIC policy and some gold) in bank fixed deposits, renewed every year or two. He's conservative to the bone and has never owned a bond. He's about to discover a way to lock a government-backed rate for a decade. Lakshmi Rao — 64, a retired schoolteacher in Hyderabad, a widow living off a ₹95,00,000 (₹95 lakh — one lakh is ₹1,00,000) corpus that must throw off about ₹50,000 a month — needs certainty above all, plus a safe place to park money she'll need in a few months. Government securities speak directly to both of them.

Start from what a bond already is. In Lesson 32 you learned that a bond is a loan with a receipt: you hand over money (the face value, also called par — the amount printed on the bond, usually ₹100), the borrower pays you a fixed coupon (the interest, usually twice a year), and on the maturity date you get your face value back. You are the lender; the bond is the IOU.

A government security — a G-Sec for short — is simply a bond where the borrower is the government. When the Government of India spends more than it collects in taxes (as almost every government does), it borrows the difference, and one big way it borrows is by issuing these securities. Buy one, and you have lent money to the Government of India. It pays you a coupon along the way and returns your money at maturity. That's the whole idea: a G-Sec is the government's IOU to you.

Why does this borrower matter so much? Because of who it is. A company can go bankrupt and fail to repay a bond — that's credit risk (or default risk), which you met in Lesson 5. The Government of India, borrowing in its own currency, is a different animal: it controls the rupee and the tax system, so it can always meet a rupee promise. Its default risk on a rupee G-Sec is treated as essentially zero — this is sovereign risk, and for the domestic government it's about as close to 'no risk of not being paid back' as finance offers. That single fact — a borrower that (in rupee terms) cannot default — is what makes everything else in this lesson tick.

Check yourself before moving on: when you buy a G-Sec, who owes whom, and what is the one risk that has (in rupee terms) essentially vanished compared with a company's bond? (You are the lender and the government owes you; the risk that essentially vanishes is credit/default risk — the government can always repay in its own currency.)

'Government security' is really a family name, not a single product. Three members matter to you, and they differ mostly in two things: how long they last, and how they pay you. Meet the family first, then we'll take each one properly.

• Dated G-Sec — the long one. A dated security simply means a bond with a fixed maturity date years out — commonly 5 to 40 years. It's issued by the central government, pays a fixed coupon twice a year, and returns your face value on the dated day. This is Harpreet's long rate-lock.
• Treasury Bill (T-Bill) — the short one. A bill that matures in under a year — 91, 182, or 364 days. It pays no coupon at all; instead you buy it below ₹100 and it grows to ₹100 at maturity. That's Lakshmi's place to park near-term money.
• State Development Loan (SDL) — the state one. A dated bond just like a central G-Sec, but issued by a state government (Punjab, Telangana, Maharashtra…) rather than the centre. It pays a coupon twice a year and usually yields a touch more than a central G-Sec.

Notice the through-line: all three are the government borrowing from you, so all three carry that near-zero sovereign default risk. What changes across the family is the *shape* — a long fixed-coupon loan (dated G-Sec), a short grow-to-par bill (T-Bill), or a state's version with a small extra yield (SDL). Pick the member that matches your need — a decade of certain income, a few months of safe parking, or a slightly higher state-backed rate — and you're choosing within one very safe family, not stepping outside it.

Check: your neighbour says he 'bought a government bond that matures in three months and pays no interest.' Which family member is that, and how does he make any money? (A Treasury Bill — a T-Bill — which pays no coupon; he earns by buying it below ₹100 and getting the full ₹100 back at maturity.)

Take the long member first, because it's the one Harpreet has been quietly needing for years. A dated G-Sec is a central-government bond with a maturity date far out — say a 10-year bond. It carries a fixed coupon, paid every six months, and on the maturity date you get your face value back. Because the coupon is fixed the day you buy, you know — to the rupee — every payment you'll receive for the whole life of the bond. That certainty is the entire appeal.

Put Harpreet's money on it. Suppose he moves ₹3,00,000 of his savings into a 10-year G-Sec at a yield of about 6.75% (roughly where the 10-year traded in mid-2026). His interest for the year is 6.75% of ₹3,00,000 = ₹20,250 — arriving as ₹10,125 every six months, like clockwork, from the Government of India. He knows that number for all ten years; no bank can revise it, no market mood can cut it. At the end of year ten, his ₹3,00,000 comes back in full. Over the decade he collects ₹20,250 × 10 = ₹2,02,500 in coupons and gets his ₹3,00,000 back — ₹5,02,500 in total, every rupee of it a fixed, sovereign promise made on day one.

There's one honest caveat, and it's the mirror image of the certainty. Locking 6.75% for ten years is wonderful if rates fall — Harpreet keeps earning 6.75% while new bonds pay less. But if rates *rise*, he's still stuck at 6.75% while newer bonds pay more (and if he tried to sell, the seesaw would mark his bond down). A fixed rate-lock cuts both ways: it protects you from falling rates and denies you rising ones. That's not a flaw, it's the deal — certainty in exchange for giving up the chance to re-price. Whether to lock long or stay short is a rate-cycle judgement, which is Lesson 9's territory.

Check: Harpreet buys a ₹3,00,000 ten-year G-Sec at 6.75% and holds it to maturity. Two years in, market yields jump and his bond's screen price drops. Has he lost money? (No — if he holds to maturity he still collects ₹10,125 every six months and gets ₹3,00,000 back on the dated day; the screen price only matters if he sells early.)

Now the short member — and it works in a way that surprises people the first time, so slow down here. A Treasury Bill (T-Bill) matures in under a year: the three standard tenors are 91 days, 182 days, and 364 days (roughly three months, six months, a year). And here's the twist: a T-Bill pays no coupon at all. It is a zero-coupon security. So how do you make money on a bond that never pays interest?

You buy it for less than ₹100 and it grows to exactly ₹100 at maturity. That's the whole mechanic — it's issued at a discount, redeemed at par. The gap between the discounted price you pay and the ₹100 you get back *is* your return; there's no separate interest cheque because the interest is baked into that gap. 'Zero-coupon' and 'issued at a discount' are two ways of saying the same thing.

Walk the numbers, because the return is easy to under-read. Say a 91-day T-Bill can be bought at ₹98.50 (per ₹100 of face value). You pay ₹98.50 now; in 91 days you receive ₹100. Your gain is ₹1.50 on an outlay of ₹98.50 — that's 1.50 ÷ 98.50 = 1.52% in 91 days. It looks tiny, but it's earned in a quarter of a year. Stretch it to a full year and it's 1.52% × (365 ÷ 91) ≈ 6.11% annualised. So a bill that 'pays no interest' is really yielding about 6.11% a year — you just collect it as a lump at the end instead of as coupons along the way.

Check: a 91-day T-Bill is offered at ₹98.80 instead of ₹98.50. Without a calculator, is its annualised yield higher or lower than 6.11% — and why? (Lower — a higher price means a smaller ₹1.20 discount, so less return for the same 91 days; price up, yield down.)

The discount mechanic is neat, but the reason it matters is *what a T-Bill is good for* — and that's Lakshmi's problem exactly. She's 64, living off her corpus, and every few months a lump comes due: an insurance premium, a property-tax bill, a grandchild's school fee, a cushion she wants ready in three months but not spent today. Where does that money sit meanwhile? A savings account pays about 2.7% — it loses to inflation (Lesson 1). A short FD works but locks her in with a penalty for breaking early.

A 91-day T-Bill is built for this. Say Lakshmi has ₹3,00,000 she'll need in about three months. She buys ₹3,00,000 of face value at ₹98.50 — paying ₹2,95,500 today — and in 91 days the government pays her the full ₹3,00,000. She's earned ₹4,500 (that same 6.11% annualised) on money that was doing nothing, and it matures exactly when she needs it, from the safest borrower in the country. No default risk, no lock-in penalty, no market timing. When the next lump is three or six months out, she rolls into a fresh 91- or 182-day bill.

Check: why is a 91-day T-Bill a better home than a savings account for money Lakshmi needs in three months — name two reasons? (It earns far more than a savings account's ~2.7% — about 6.11% — and it's sovereign-safe and matures exactly on time, with no early-exit penalty.)

The third member is the one most retail investors have never heard of. States borrow too — to build roads, run schemes, cover their own gaps — and when a state government issues a dated bond, it's called a State Development Loan (SDL). Mechanically it's the twin of a central dated G-Sec: a fixed coupon twice a year, a maturity date (often around ten years), your face value back at the end. The RBI runs these auctions for the states just as it does for the centre.

One difference matters: an SDL usually pays a little more than a central G-Sec of the same maturity. That extra is called a spread — the gap in yield between two bonds. Why the pickup? A state's credit is seen as a notch below the centre's: the central government prints the rupee, a state doesn't, so a state is fractionally less bulletproof (though still extraordinarily safe, and effectively guaranteed within the system). The market prices that sliver of extra risk as a slightly higher yield. Historically the 10-year SDL spread over the central G-Sec has run about 0.25% to 0.50% (25 to 50 basis points — a basis point is one-hundredth of a percent); it widens when states borrow heavily (recently it has been nearer 0.80–0.90%).

So on the same ₹3,00,000, a 10-year SDL at ~7.15% would pay Harpreet about ₹21,450 a year versus the central G-Sec's ₹20,250 — roughly ₹1,200 more a year for taking a state's credit instead of the centre's. Is that worth it? For most conservative savers the answer is 'a small, sensible yes' — the extra risk is tiny and the extra yield is real. It's the gentlest possible step out on the risk ladder: still government, still sovereign-grade, just a state instead of the centre.

Check: an SDL yields 0.40% more than a central G-Sec of the same maturity. In one line, what are you being paid that extra 0.40% for? (For taking a state government's credit instead of the central government's — a fractionally-less-bulletproof borrower, so the market pays you a small spread.)

Now the idea that makes government securities matter far beyond the people who buy them. Because a G-Sec (in rupee terms) essentially cannot default, its yield is treated as the risk-free rate — the return you can earn with, as close as finance gets, no risk of not being paid back. You met the phrase in Lesson 1; here's where it becomes a real, quoted number: about 6.75% for the 10-year, set in the market every day. It's the floor, the baseline, the zero point.

And once you have a true zero point, you can measure everything else against it. Every other yield in this whole course is really quoted as 'the G-Sec plus (or minus) a spread' — the extra you demand for taking on some risk the government doesn't carry. A corporate bond that can default? It must pay *more* than the G-Sec, and the gap is the price of its credit risk. A bank FD insured only to ₹5 lakh? It tends to sit right around — even a touch below — the sovereign yield. That's why a G-Sec is called the risk-free benchmark: it's the ruler every other return is held up against.

Read the ladder from the anchor out. A savings account at ~2.7% is more than four percentage points *below* the risk-free rate — safe, but leaking to inflation (Lesson 1, made concrete). A comparable bank FD at ~6.4% sits just *below* the G-Sec — and it's only insured to ₹5 lakh, while the G-Sec has no such ceiling. The SDL at ~7.15% is +0.40% for a state's credit. A top-rated (AAA) PSU/corporate bond at ~7.5% is +0.75% — the market's price for the chance, however small, that a company defaults. A weaker (A-rated) corporate bond at ~9.0% is +2.25% — a much fatter spread for much more credit risk. And the Nifty's earnings yield at ~4.5% sits *below* the G-Sec, which surprises people until you see why: you buy shares for growth and future capital gains, not for today's income, so investors accept a lower current yield in exchange for the upside. Every one of those numbers is 'G-Sec ± a spread.'

Check: a friend is offered a corporate bond yielding 9.0% and calls it 'way better than a boring 6.75% government bond.' What is the 2.25% difference actually compensating him for? (The company's credit/default risk — the government can't default in rupees, the company can, so the extra 2.25% is the price of that risk, not a free lunch.)

One more idea, and it falls out naturally from everything above. A T-Bill matures in 91 days; a dated G-Sec in 10 or 30 years. Each maturity has its own yield — and if you plot them, short to long, you get the yield curve: the government's yield at every length of time, drawn as a line. It's one of the most-watched pictures in all of finance, and now you can read it.

Usually the curve slopes up: longer bonds yield more than shorter ones, because when you lock your money away for longer you take more risk (rates could move, inflation could surprise) and you demand a little more to do it. In our illustrative mid-2026 snapshot the 91-day T-Bill sits near 6.11%, the 1-year around 6.30%, the 5-year ~6.55%, the 10-year 6.75%, and the 30-year ~6.95% — a gentle upward climb of under a percentage point from three months to thirty years. That's the same 6.11% and 6.75% you've already met, now shown as two points on one line.

The *shape* carries a message. A steep upward curve often says the market expects growth or higher rates ahead. A flat curve says short and long are priced alike — uncertainty. An inverted curve, where short yields sit *above* long ones, is rarer and often read as a warning that rates (and maybe the economy) are expected to cool. You don't need to forecast from it — that's well beyond this lesson — but you should know that when the news says 'the yield curve steepened' or 'flattened,' it's describing this exact line, built from the very securities you just learned.

Check: on a normal upward-sloping curve, does a 30-year G-Sec yield more or less than a 91-day T-Bill, and what's the basic reason? (More — locking money away for longer carries more risk over time, so the market pays you a higher yield for the longer commitment.)

So who actually *issues* these, and where do they come from? The Reserve Bank of India (RBI) is the government's banker and debt manager: it conducts the auctions through which new G-Secs, T-Bills and SDLs are sold, on behalf of the central and state governments. An auction is exactly what it sounds like — buyers bid for a new batch of bonds, and the bids set the price (and therefore the yield). This is the primary market: brand-new securities, sold straight from the issuer.

Once a bond exists, it can change hands again before maturity — one investor selling to another. That's the secondary market, where prices move day to day with yields (the Lesson 32 seesaw in action). The two markets matter to you differently: you *buy new* in the primary auction, and you *can sell early* (or buy someone else's bond) in the secondary market if you ever need to exit before maturity. For retail investors the RBI's platform links both — new issues by auction, and a secondary window called NDS-OM to trade existing ones.

Here's the part that used to keep individuals out — and no longer does. In the auction there are two ways to bid. Big players (banks, primary dealers, insurers) bid competitively: each one quotes the *exact yield* it's willing to accept, and risks being left out if it bids wrong. That takes a trading desk and market expertise — genuinely not a game for a shopkeeper in Ludhiana. So the RBI carved out a second lane: non-competitive bidding. A retail investor simply says *how much* they want — say ₹50,000 or ₹2,00,000 of the bond — without having to quote any yield at all, and is allotted the security at the weighted-average rate that emerges from the competitive bids. You get the same price the professionals collectively set, without having to out-guess them.

Check: Harpreet wants ₹3,00,000 of a new 10-year G-Sec but has no idea what yield to 'bid.' Does that shut him out — and why not? (No — he uses non-competitive bidding: he just states the ₹3,00,000 amount and is allotted the bond at the auction's weighted-average rate, without quoting any yield himself.)

For Harpreet, this is the real decision. His safe money has always lived in bank fixed deposits. A G-Sec and an FD are cousins — both are 'lend money, earn a fixed rate, get it back' — so which is better? The honest answer is 'it depends,' and here's the honest comparison, not a sales pitch for either.

Start with yield, and resist the temptation to oversell it. On ₹3,00,000, the 10-year G-Sec at ~6.75% pays ₹20,250 a year; a comparable bank FD at ~6.4% pays ₹19,200 — the G-Sec is ahead by only about ₹1,050 a year. That's real but modest. The bigger differences are structural: (1) the rate-lock. The G-Sec fixes 6.75% for the *full ten years*; most bank FDs run five years or less, so an FD saver keeps renewing and re-prices at whatever rates prevail — if rates fall, their income falls with them. The G-Sec removes that reinvestment guesswork for a decade. (2) The safety ceiling. A bank FD is insured only to ₹5,00,000 per bank (DICGC, Lesson 1); above that you're trusting the bank. A G-Sec is the sovereign itself — no ceiling — which matters little for ₹3,00,000 but a great deal for Lakshmi's ₹95 lakh. (3) Liquidity. Break an FD early and you usually pay a penalty; a G-Sec you can *sell* in the secondary market at the going price (which may be higher or lower — the seesaw).

So which should Harpreet pick? There's no single right answer — that's a suitability judgement (Lesson 6), and at a real decision a fee-only SEBI-registered adviser earns their keep. But the shape of it is clear: if he wants a long, certain, sovereign rate-lock and values not having to renew or worry about the ₹5 lakh ceiling, the G-Sec fits beautifully. If he wants a shorter tenor, the familiarity of his branch, or the flexibility to break in an emergency, an FD is perfectly sensible. They're not enemies — most conservative savers end up holding some of each.

Check: name the two advantages a 10-year G-Sec has over a bank FD that have nothing to do with the headline interest rate. (It locks the rate for the full ten years — no reinvestment guesswork — and it's sovereign-safe with no ₹5 lakh insurance ceiling; a third is that it can be sold in the secondary market rather than broken with a penalty.)

Wealthy families and their advisers use government securities constantly — and often make it sound like a proprietary skill. It isn't. Here's the move they make, the plain logic under it, and the fact that you can now do the very same thing yourself, for free.

The move decoded: park near-term cash in T-Bills, build a row of dated G-Secs for the certain-income leg, and reach for SDLs where a small extra yield is worth a sliver more risk — all sovereign-grade, all bought at the RBI auction. The logic is nothing exotic: match each maturity to when you need the money, and anchor the safe part of the portfolio to the borrower that can't default. And the punchline: every one of those securities is available to *you* directly on RBI Retail Direct, at zero fees, at the same auction price. So the 'is my manager worth the fee?' test is sharp here — if all they're doing is buying you G-Secs and T-Bills and charging a percentage for the privilege, they're charging you to access something that is free and (after Lesson 34) genuinely simple. A good adviser earns their fee on judgement — which maturities, how much, when — not on gatekeeping a public auction.

Check: a relationship manager offers to put your money into 'exclusive government-securities strategies' for a 1% annual fee. What's the DIY substitute, and what's the fair question to ask? (Buy the same G-Secs/T-Bills/SDLs yourself on RBI Retail Direct for zero fees — and ask whether the manager's judgement is worth 1% a year versus doing it directly.)

The word 'government' is a magnet for fraud, precisely because it means 'safe' to everyone. So the con is predictable: a scheme, a WhatsApp forward, or a slick app promises a 'government-guaranteed' or 'sovereign' bond paying 12%, 15%, even 18% — draping the sturdiest word in finance over a return no government security actually offers. You now have the one fact that dismantles it instantly.

The real 10-year G-Sec yields about 6.75%. That is what the Government of India genuinely pays to borrow. So a '12% government-guaranteed bond' is claiming to pay nearly double the actual sovereign rate — with, supposedly, zero risk. That is impossible, and the risk-free benchmark from earlier is exactly why: nobody can offer far more than the risk-free rate *without far more risk*. A genuine 12% return exists somewhere (in equities, in junk credit) but never with a government guarantee attached. Anyone welding 'government' to a double-digit 'guarantee' is lying about one or the other — usually both. The number itself is the tell.

Check: an app advertises a '100% government-backed bond, 14% assured.' Using only the risk-free benchmark, why can you reject it on sight? (Because the real government (risk-free) rate is ~6.75%; nothing can pay 14% with a government guarantee and no risk — a return that far above the sovereign rate always means real risk, so the 'guarantee' is a lie.)

Maybe this lesson lands with a small pang. Perhaps, like Harpreet, you've kept your safe money in bank FDs and a savings account for years — renewing, re-pricing, occasionally watching the rate drop — never once considering a G-Sec, because government bonds seemed like something for banks, not for you. If that's you, read the next paragraph before anything else.

Set the blame down. You weren't lazy or naive — the door was genuinely shut to individuals for most of India's history, and only opened to retail in November 2021. Nobody at your bank branch was ever going to point you to a product the bank doesn't sell. Keeping safe money safe in an FD was a *reasonable* choice, not a mistake; you protected your capital, which is the first job. What's changed is only that a better-fitting tool for part of that money now exists and is reachable. You haven't lost years — you've arrived exactly when the access did. The concrete next step is small and specific: the very next lesson walks you, screen by screen, through opening the RBI Retail Direct account and placing a first order. Nothing to undo, nothing to regret — just a door you can now choose to walk through.

Can an ordinary individual really buy G-Secs — or is it only for banks? Yes, genuinely. Since November 2021 any resident individual with a bank account and a PAN can buy G-Secs, T-Bills and SDLs directly from the RBI through the free RBI Retail Direct scheme — same auction, same price the banks pay, zero fees. The how-to is Lesson 34.

What exactly is a T-Bill, in one line? A short government security (91, 182 or 364 days) that pays no interest — you buy it below ₹100 and get the full ₹100 back at maturity, and that discount is your return. Perfect for parking money you'll need within a year.

G-Sec or fixed deposit — which should I pick? They're close on yield (a 10-year G-Sec ~6.75% vs a comparable FD ~6.4%). The G-Sec's edge is a longer rate-lock, no ₹5 lakh insurance ceiling (it's sovereign), and tradability; the FD's edge is familiarity and easy (if penalised) early exit. Many conservative savers hold some of each; on tax they're identical (both taxed at slab).

Are SDLs as safe as central G-Secs? Almost — a State Development Loan is a state government's bond, seen as a fraction less bulletproof than the centre (which prints the rupee), so it pays a small spread more (historically ~0.25–0.50%). For most conservative investors that extra yield is a sensible trade for a sliver more risk.

Why is a G-Sec called 'risk-free' when its price moves around? 'Risk-free' means the government won't *default* — you'll be paid your coupons and principal for certain (in rupees). The *price* still moves with yields if you sell early, but if you hold to maturity that movement doesn't touch you. Risk-free refers to credit risk, not to day-to-day price wobble.

What return do government bonds actually pay right now? In mid-2026, roughly: a 91-day T-Bill ~6.11% annualised, the 10-year G-Sec ~6.75%, a 10-year SDL ~7.15%. These move at every auction, so always check the current figure (RBI / CCIL / RBI Retail Direct) before you buy.

How is the interest taxed? G-Sec, SDL and T-Bill returns are taxed as interest, at your income-tax slab — the same as a bank FD. There's no special exemption here (the tax-advantaged government-adjacent bonds — SGBs, 54EC, tax-free PSU — are Lesson 35). Full treatment: Lesson 36 and the india:income-tax track.

Do I have to 'bid' in an auction — I wouldn't know what yield to quote? No. You use non-competitive bidding: you state only how much you want and receive the auction's weighted-average rate. Quoting a yield is for banks and primary dealers (competitive bidding), not for you.

Can I get my money out before maturity? Yes — you can sell a G-Sec in the secondary market (via NDS-OM on the RBI platform) at the going price, which may be above or below what you paid (the price–yield seesaw). Held to maturity, you simply get your face value back. Liquidity is decent for central G-Secs, thinner for some SDLs.

Is a 'government-guaranteed 12% bond' real? No. The real sovereign rate is ~6.75%; a 'guaranteed' 12% government bond doesn't exist and is a classic scam using the word 'government' as bait. Buy only via RBI Retail Direct or a SEBI-registered broker; report suspicious offers to SEBI SCORES or cybercrime 1930.

Put it together on a live tool. Pick an instrument — a dated G-Sec, a T-Bill, or an SDL — enter an amount and the terms, and watch the return appear, along with its spread over the risk-free 10-year G-Sec. It starts on Harpreet's ₹3,00,000 G-Sec at 6.75% (reproducing the ₹20,250 a year you computed); one button loads Lakshmi's 91-day T-Bill parking at ₹98.50.

Play with the two levers that teach the most. Switch to the T-Bill and push the price down from ₹98.50 toward ₹98.20 — watch the annualised yield climb as the discount widens (the seesaw, laid bare). Then switch to the SDL and watch its spread over the G-Sec turn positive: the small state-level pickup, in rupees, on your own amount. If you can predict which way each number moves before you touch it, you've understood the family.

• Government security (G-Sec) — a bond issued by the government to borrow money; you are the lender, it pays you a coupon and returns your face value at maturity. In rupee terms it essentially cannot default.
• Dated security — a government bond with a fixed, far-out maturity date (commonly 5–40 years), paying a fixed coupon twice a year; the long member of the family.
• Treasury Bill (T-Bill) — a short government security (91, 182 or 364 days) that pays no coupon; issued at a discount to ₹100 and redeemed at ₹100, so the discount is your return.
• Zero-coupon / issued at a discount — a security that pays no periodic interest; you buy it below face value and it grows to face value at maturity, and that gap is the interest. (T-Bills work this way; dated G-Secs and SDLs do not.)
• State Development Loan (SDL) — a dated bond issued by a state government (via RBI auction), paying a coupon twice a year; usually yields a small spread above a central G-Sec of the same maturity.
• Sovereign risk — the (very low, for a domestic government borrowing in its own currency) risk that a government fails to repay; for Indian rupee G-Secs it's treated as essentially zero.
• Gilt — an old synonym for a government security (from gilt-edged certificates); the RBI retail account is a Retail Direct Gilt (RDG) account.
• Risk-free rate / risk-free benchmark — the G-Sec yield (e.g. ~6.75% for the 10-year), treated as the return earnable with essentially no default risk; the baseline every other yield is measured against.
• Spread — the gap in yield between two bonds; the extra a riskier bond must pay over the risk-free G-Sec (e.g. an SDL at +0.40%, a corporate bond at +0.75% or more). A basis point (bp) is one-hundredth of a percent.
• Auction (competitive vs non-competitive bidding) — the RBI's sale of new securities; big players bid competitively by quoting a yield, while retail bids non-competitively — stating only an amount and receiving the auction's weighted-average rate.
• Yield curve — the line of government yields plotted from short to long maturities; usually slopes up (longer = higher yield). Its shape (steep, flat, inverted) reflects market expectations.
• Primary vs secondary market — the primary market is where new securities are auctioned by the RBI; the secondary market (for retail, NDS-OM) is where existing securities are bought and sold before maturity, at prices that move with yields.