Chapter 6 · Derivatives on NSE: Futures and Options

6.1 Futures: Pricing, Margins, Leverage and Hedging

A futures contract fixes today the price of a trade that happens later. Price it by cost of carry, follow the basis to zero at expiry, settle gains and losses every day through margins, measure what leverage really does, and hedge a portfolio with index futures using its beta.

📍 The situation

Kavya manages her family's ₹50 lakh portfolio of Indian large-cap shares. Its beta against the Nifty 50 is about 1.2 (§3.2). An important policy announcement is due within the month and she would like to protect the portfolio for those few weeks without selling dozens of shares, paying taxes and costs, and buying them all back. Her broker suggests "selling a few lots of Nifty futures". How many? What does it cost? What happens to her account if the market rises instead?

Her cousin Arjun has a very different plan. He has read that "one lot of Nifty futures needs only about ₹1.3 lakh of margin" and wants to buy a lot to profit from a rise. He has not asked what happens on a day the index falls 2%, why the futures price on his screen is about 100 points above the index, or why his broker keeps talking about "MTM". This section answers both of them. The mathematics is light — compounding (§1.1), a telescoping sum, one regression coefficient — but the consequences of getting it wrong are not.

⚠️ About the numbers on this page

Index levels, prices, interest rates and dividend yields here are hypothetical. Lot sizes, margin percentages, expiry days and contract rules are set by NSE and its clearing corporation under SEBI regulation and change from time to time; this page uses an illustrative lot size of $Q=50$ units for an index contract and an illustrative margin of 12% of contract value. Before trading, check the current contract specifications and margin files on the NSE website and the latest SEBI circulars. Transaction costs and taxes are ignored unless stated. Names are hypothetical and nothing here is investment advice.

Forwards, Futures and Their Payoff

A farmer who agrees in March to sell wheat in June at a price fixed today, and a jeweller who fixes today the price of gold she will buy next month, are both using a forward contract. Futures are the exchange-traded, standardised version of the same idea.

📘 Definition 6.1 — Forward and futures contracts

A forward contract is an agreement made at time $0$ to buy or sell an asset (the underlying) at a future date $T$ (the expiry or delivery date) for a price $F_0$ fixed today (the forward price). Nothing is paid at time $0$. The buyer holds a long position; the seller a short position.

A futures contract is a forward contract that is standardised (underlying, quantity, expiry date) and traded on an exchange, with a clearing corporation as the counterparty to every trade. Gains and losses are settled every day through margin accounts (Definition 6.5). On NSE, index futures (on the Nifty 50 and other indices) are cash-settled: at expiry only the difference between the final settlement price and the contract price changes hands. Stock futures are generally physically settled (shares are delivered) — check the current rules.

📐 Theorem 6.1 (Formula) — Payoff of a forward or futures position

If $S_T$ is the price of the underlying at expiry, one unit of a position entered at price $F_0$ pays at expiry

$$\text{long: } S_T-F_0,\qquad \text{short: } F_0-S_T.$$

For a contract of $Q$ units (the lot size) multiply by $Q$; for $N$ contracts by $NQ$. The two payoffs add to zero: futures are a zero-sum game between buyers and sellers, before costs — and a negative-sum game after them.

Why. At expiry the long side buys for $F_0$ something worth $S_T$ (or, under cash settlement, receives $S_T-F_0$ in cash, which is the same thing); the short side is on the other side of the same trade. The payoff is a straight line in $S_T$ with slope $+1$ (long) or $-1$ (short) — compare the bent lines of options in §6.2. ∎

📘 Definition 6.2 — Contract specification: lot size, contract value, expiry

The exchange fixes for each contract the lot size $Q$ (units of the underlying per contract; for an index contract, rupees per index point), the tick size (smallest price step), the expiry date and the settlement method. The contract value (notional) of one lot at futures price $F$ is $$\text{contract value}=F\times Q.$$

On NSE, several monthly expiries trade at the same time (near, next and far month); the expiry weekday and the number of contracts are set by the exchange and have changed over time. SEBI also sets a minimum contract value, and NSE revises lot sizes periodically so that contracts stay above it. The lot of 50 used in this chapter keeps the arithmetic simple; real index lots are larger, because a contract must be worth at least that minimum. These are exactly the details you must check in the current NSE contract specifications.

✏️ Example 1 — What Arjun is really buying
Arjun buys 1 lot of index futures at 22,100 (lot size $Q=50$, illustrative). Find the contract value, and his profit or loss if the final settlement price at expiry is (a) 22,600, (b) 21,700. What does the seller make in each case?
Contract value $22{,}100\times50=$ ₹11,05,000. Arjun controls an index exposure of about ₹11 lakh — whatever margin he actually deposits.
(a) $50\times(22{,}600-22{,}100)=$ ₹25,000 profit (Theorem 6.1).
(b) $50\times(21{,}700-22{,}100)=-$₹20,000, a loss.
Seller Exactly the opposite: −₹25,000 in (a), +₹20,000 in (b). Every rupee one side gains, the other side loses; brokerage, taxes and charges are paid by both.

Cost of Carry: What Should a Futures Contract Cost?

Why does a one-month index future trade about 100 points above the index? Not because "the market expects a rise". It is because there are two ways to own the index at expiry: buy it today and hold it (paying interest on the money, receiving dividends), or buy a future today and pay later. If these two routes had different costs, an arbitrageur would use the cheap one and sell the expensive one.

📘 Definition 6.3 — Cost of carry and cash-and-carry arbitrage

The cost of carry is the net cost of holding the underlying until expiry: interest on the money tied up (at the risk-free rate $r$), minus any income the asset pays (dividends at yield $q$ for an index), plus storage costs for commodities.

A cash-and-carry arbitrage buys the underlying with borrowed money and sells it forward through a short futures position; a reverse cash-and-carry sells (or sells short) the underlying, invests the proceeds and buys futures. For an index, "buying the underlying" means buying the whole basket in index weights or an index ETF.

📐 Theorem 6.2 (Formula) — Cost-of-carry (fair) futures price

With continuous compounding (§1.1), risk-free rate $r$, time to expiry $T$ in years, and no arbitrage, transaction costs or restrictions on borrowing and short selling:

  1. an underlying paying no income: $$F_0=S_0\,e^{rT};$$
  2. an index with continuous dividend yield $q$: $$F_0=S_0\,e^{(r-q)T};$$
  3. a stock paying known dividends with present value $D$ before expiry: $$F_0=(S_0-D)\,e^{rT}.$$

For a few weeks the simple-interest version $F_0\approx S_0\,[1+(r-q)T]$ (used in many market handbooks) differs only in the second decimal place. If $r\gt q$ the future trades above spot; if dividends are large ($q\gt r$, or a big dividend before expiry) it can trade below.

Derivation (no arbitrage). Take case 2. At time $0$, borrow $S_0e^{-qT}$ and buy $e^{-qT}$ units of the index; reinvest the dividends as they arrive, so the holding grows at rate $q$ to exactly one unit at $T$. Also sell one future at $F_0$ (it costs nothing). At $T$: the index unit is worth $S_T$, the short future pays $F_0-S_T$, and the loan costs $S_0e^{-qT}e^{rT}$. Total: $$S_T+(F_0-S_T)-S_0e^{(r-q)T}=F_0-S_0e^{(r-q)T},$$ a sure amount from a strategy that cost nothing. If it were positive this cash-and-carry would be free money; if negative, the reverse trade (sell the index, invest, buy the future) would be. So $F_0=S_0e^{(r-q)T}$. Case 1 is $q=0$; in case 3 borrow only $S_0-D$, because the dividends repay the part $D$ of the loan. ∎

✏️ Example 2 — The fair value of a one-month index future
The index is at 22,000. Take an illustrative risk-free rate $r=7\%$ and dividend yield $q=1.2\%$ (both continuously compounded) and 30 days to expiry. Find the fair futures price and the premium over spot. Compare with the simple-interest version.
Set-up $T=30/365\approx0.08219$, $(r-q)T=0.058\times0.08219\approx0.004767$.
Fair price $F_0=22{,}000\,e^{0.004767}\approx22{,}105.1$ — about 105 points (0.48%) above spot.
Simple interest $22{,}000\times(1+0.004767)\approx22{,}104.9$. The two agree to a fraction of a point.
Meaning The 105 points are about one month of interest minus one month of dividends on ₹22,000 per index unit. They say nothing about where the index "should" go.
✏️ Example 3 — A stock future with a dividend, and an arbitrage
Shares of Alpha Bank Ltd (hypothetical) trade at ₹1,500. A dividend of ₹20 per share goes ex in 30 days; the stock future expires in 61 days; $r=7\%$. (a) Find the fair futures price. (b) The future is quoted at ₹1,520. Describe the arbitrage and its profit per share and per lot of 500 shares (illustrative), before costs.
(a) $D=20e^{-0.07\times30/365}\approx19.885$; $F_0=(1{,}500-19.885)\,e^{0.07\times61/365}\approx$ ₹1,497.53 — below spot, because the dividend (≈ 1.3% of the price) exceeds two months of interest (≈ 1.2%).
(b) Trade The quote is rich, so do cash-and-carry: borrow ₹1,500 at 7%, buy one share, sell one future at ₹1,520.
Cash flows Day 30: receive ₹20 and invest it to expiry: $20e^{0.07\times31/365}\approx20.12$. Expiry: deliver the share for ₹1,520 (or, equivalently, the share and the short future together are worth ₹1,520). Repay $1{,}500e^{0.07\times61/365}\approx1{,}517.65$. Profit $\approx1{,}520+20.12-1{,}517.65\approx$ ₹22.47 per share.
Per lot $500\times22.47\approx$ ₹11,234, locked in whatever the share price does — before brokerage, taxes, the bid–ask spread, margin on the future and the true borrowing rate, which is what keeps real gaps small.
📘 Definition 6.4 — Basis, premium and discount

The basis at time $t$ is $b_t=F_t-S_t$, the futures price minus the spot price. (Some texts use $S_t-F_t$; always state the sign convention.) A future trading above spot is at a premium (the market is in contango); below spot, at a discount (backwardation). The implied cost of carry is the annual rate $c$ with $F_t=S_te^{c(T-t)}$, that is $c=\dfrac{\log(F_t/S_t)}{T-t}$.

📐 Theorem 6.3 — The basis converges to zero; calendar spreads

Under the cost-of-carry price of Theorem 6.2, at time $t\le T$ $$b_t=S_t\left(e^{(r-q)(T-t)}-1\right)\approx S_t\,(r-q)(T-t),$$ so the basis shrinks roughly in proportion to the time left and is exactly $0$ at expiry: $F_T=S_T$ (convergence). For two expiries $T_1\lt T_2$ on the same underlying, $$F^{(2)}=F^{(1)}\,e^{(r-q)(T_2-T_1)},$$ and the difference $F^{(2)}-F^{(1)}$ is the calendar spread.

Why. Apply Theorem 6.2 at time $t$ with $T-t$ years left; at $t=T$ the exponent is $0$. Convergence also follows directly: on the expiry day a future to buy the index "right now" must be worth the index itself, otherwise buying the cheaper one and selling the dearer one is riskless. Dividing the two cost-of-carry prices gives the calendar relation. In real markets the basis wobbles around this line from day to day (supply and demand, costs, short-selling limits) — that is basis risk — but it is pulled to zero by expiry. ∎

✏️ Example 4 — Reading the basis and the calendar spread
(a) With the index at 22,000 and the 30-day future at 22,140, what cost of carry is the market implying? (b) Using $r-q=5.8\%$, what should the basis be 20 days later if the index is then at 22,300? (c) With the near-month (30-day) future fair at 22,105, find the fair next-month future (28 days later) and the calendar spread.
(a) $c=\dfrac{\log(22{,}140/22{,}000)}{30/365}\approx7.72\%$ a year, about 1.9 percentage points above our fair $r-q=5.8\%$: the future is slightly rich (by about 35 points), which may be costs and demand, or an opportunity for a cash-and-carry trader with cheap funding.
(b) 10 days left: $F=22{,}300\,e^{0.058\times10/365}\approx22{,}335.5$, basis ≈ 35.5 points — a third of the 105 points of the 30-day contract, as Theorem 6.3 predicts, even though the index has moved.
(c) $F^{(2)}=22{,}105\,e^{0.058\times28/365}\approx22{,}203.6$, spread ≈ 98.6 points. A hedger who rolls a short position from the near to the next month sells the next month about 99 points higher — the carry for another four weeks.
Interactive 1 — One Expiry Cycle: Carry, Convergence and Daily Mark-to-Market

A simulated index path (blue) and a futures price (red) that trades at its fair value (Theorem 6.2) plus a small simulated wobble, over one expiry cycle. The lower panel shows either the basis $F-S$ (green) shrinking to zero around its fair value (dashed), or the margin account of a trader who is long one lot (illustrative lot 50): green is the balance, the amber dashed line the margin required (rate you choose × contract value); triangles mark days the trader had to add money. Play walks through the trading days; New path draws another simulated month.

7.0%1.2%
fair F₀ on day 0—
day / index—
futures / basis—
today's MTM—
cumulative P&L—
money added (margin calls)—

Margins, Mark-to-Market and Leverage

A forward contract has a weakness: by expiry the loser may have disappeared. Futures solve this by making both sides deposit a margin and by settling gains and losses every single day, so that no one is ever allowed to owe much.

📘 Definition 6.5 — Initial margin, mark-to-market and margin call

To open a futures position each side deposits an initial margin with its broker (who passes it to the clearing corporation): a fraction $m$ of the contract value, set by a risk model to cover a large one- or two-day move. In India this consists of a risk-based margin plus additional (exposure) margins, revised daily and higher in volatile markets.

At the end of each trading day the position is marked to market (MTM): with daily settlement prices $F_0,F_1,\dots$, the long side receives $$Q\,(F_t-F_{t-1})$$ in cash on day $t$ (pays it if negative), and the short side the opposite.

If losses take the account below the margin required, the broker issues a margin call: money must be added, usually by the next morning, or the broker may close the position. (Markets elsewhere often use a lower "maintenance" level; in this course we use the simple rule "top up whenever the balance falls below $m\times F_t\times Q$".)

📐 Theorem 6.4 (Formula) — MTM adds up to the payoff; the leverage factor

For a long position of one lot opened at $F_0$ and held to expiry, the daily MTM flows add up to the payoff of Theorem 6.1: $$\sum_{t=1}^{n}Q\,(F_t-F_{t-1})=Q\,(F_n-F_0)=Q\,(S_T-F_0).$$

If the trader's own money in the position is the initial margin $mF_0Q$, a futures price change of $x\%$ changes that money by $$\frac{Q\,\Delta F}{mF_0Q}=\frac{x\%}{m},$$ so the leverage factor is $1/m$ (8.3 for $m=12\%$), and an adverse move of $m$ (12%) wipes out the whole initial margin. More generally, with capital $C$ supporting $N$ lots, leverage $=NF_0Q/C$.

Why. The sum telescopes: $$(F_1-F_0)+(F_2-F_1)+\cdots+(F_n-F_{n-1})$$ $$=F_n-F_0,$$ and by convergence $F_n=S_T$ on the expiry day. The leverage statement is the same ratio divided by the money put in. So margins and MTM do not change what you make or lose in total — they change when you pay it: daily, in cash. ∎

✏️ Example 5 — Five days of MTM in Arjun's account
Arjun buys 1 lot at 22,100 ($Q=50$, margin 12% of contract value) and deposits ₹1,50,000. The daily settlement prices are 22,180, 22,010, 21,830, 21,590 and 21,860. Track his daily MTM, his balance, the margin required and any margin calls. What is his total profit or loss?
Day 0 Initial margin $0.12\times22{,}100\times50=$ ₹1,32,600; he deposits ₹1,50,000, a cushion of ₹17,400. Each day the requirement is $0.12\times50\times F_t=6F_t$.
Ledger
daysettleMTM (₹)balance (₹)required (₹)added (₹)
122,180+4,0001,54,0001,33,080—
222,010−8,5001,45,5001,32,060—
321,830−9,0001,36,5001,30,980—
421,590−12,0001,24,500 → 1,29,5401,29,5405,040
521,860+13,5001,43,0401,31,160—
Total MTM sum $=4{,}000-8{,}500-9{,}000-12{,}000+13{,}500=-$₹12,000 $=50\times(21{,}860-22{,}100)$, as Theorem 6.4 says. Check: he put in $1{,}50{,}000+5{,}040=$ ₹1,55,040 and has ₹1,43,040.
Lesson A fall of only 2.3% from his entry price (to 21,590) forced Arjun to find more cash within a day. Had he been unable to, his broker could have closed the position at the low — and he would have missed the next day's recovery.
✏️ Example 6 — What leverage does to a 2% day
(a) With a 12% margin, what does a 2% fall in the futures price do to the margin on one lot bought at 22,100? (b) Arjun now has ₹5,00,000 and buys 3 lots at 22,100. What is his effective leverage? What does a 5% fall do to his capital, and what fall wipes it out? (c) What do two successive 3% falls do?
(a) Loss $=0.02\times22{,}100\times50=$ ₹22,100 on a margin of ₹1,32,600: −16.7%, i.e. $2\%/0.12$. The leverage factor is $1/0.12\approx8.3$.
(b) Exposure $3\times22{,}100\times50=$ ₹33,15,000 on ₹5,00,000 of capital: leverage ≈ 6.6. A 5% fall loses ₹1,65,750 — a third of his capital. The capital is gone after a fall of $5{,}00{,}000/33{,}15{,}000\approx15.1\%$, and long before that the margin calls start (the required margin alone is ₹3,97,800).
(c) The futures price is multiplied by $0.97^2=0.9409$: loss $=33{,}15{,}000\times0.0591\approx$ ₹1,95,917, about 39% of capital in two days. An investor who had bought ₹5 lakh of an index fund instead would be down 5.9%.
Moral Leverage multiplies the percentage moves of the market by the leverage factor — in both directions. Nothing about it improves the odds; it only enlarges the bet.
⚠️ Leverage and the individual trader — read this before trading F&O
  • Losses are magnified exactly as much as gains. At a leverage of 8, an ordinary 3% day is a 25% day for your money. A few bad days in a row can take more than you planned to risk — and in a short futures or short options position, more than the margin you deposited.
  • MTM is paid in cash, every day. You can be right about where the market ends up and still be forced out on the way, because you could not meet a margin call. Gaps (a large overnight move) can jump straight past any stop-loss.
  • It is a zero-sum game before costs and a losing one after them. Every rupee gained by one side is lost by another; brokerage, statutory charges and taxes, and the bid–ask spread are paid by both. On the other side of a retail trade are often professional firms with better information, faster systems and lower costs. Frequent trading and short-dated contracts multiply the costs.
  • Look at the evidence, not the advertisements. SEBI has published studies of the profits and losses of individual F&O traders and has tightened rules for retail participation; read the latest study and circulars on the SEBI website and the risk disclosure document your broker must give you. Treat tips, "sure-shot" calls and screenshots of profits with deep suspicion (§8.2).
  • Hedging and speculating are different activities. Kavya's short futures position reduces the risk she already has; Arjun's long position creates a leveraged risk. The same contract can do either.
  • If you trade at all: never with borrowed money or money you need; decide the maximum loss in rupees before you enter and size the position from it (§7.2); know where the margin money for a bad week would come from.

Hedging a Portfolio with Index Futures

Kavya's problem is the reverse of Arjun's. She already carries the market's risk; a short index future gains when the market falls and so offsets part of her loss. How much to sell depends on how strongly her portfolio moves with the index — its beta, estimated by regressing the portfolio's returns on the index's returns (§3.2).

📘 Definition 6.6 — Hedge, hedge ratio and basis risk

A hedge is a position taken to reduce an existing risk. A short hedge (selling futures) protects someone who owns the asset or a similar one; a long hedge (buying futures) protects someone who will buy later. The hedge ratio is the futures exposure per rupee of exposure being hedged. Basis risk is the risk left over because the hedged asset and the futures price do not move exactly together — a different portfolio from the index, a basis that is not exactly at its fair value when the hedge is lifted, or a hedge horizon that differs from the expiry.

📐 Theorem 6.5 (Formula) — Number of index futures contracts for a beta hedge

A portfolio of value $V$ with beta $\beta$ against the index is hedged (its market risk removed) by selling $$N^*=\frac{\beta\,V}{F_0\,Q}$$ index futures contracts, where $F_0Q$ is the contract value. To change the portfolio's beta from $\beta$ to a target $\beta^*$, trade $$N=\frac{(\beta^*-\beta)\,V}{F_0\,Q}$$ contracts ($N\lt0$: sell; $N\gt0$: buy). In practice $N$ is rounded to a whole number of lots.

Derivation. Over the hedge period write the index return as $R_m$. The market model of §3.2 says the portfolio's return is $R_P\approx\alpha+\beta R_m+\varepsilon$, with $\varepsilon$ the stock-specific part. The portfolio's rupee gain is about $V\beta R_m$ (plus $V\alpha+V\varepsilon$). A short position in $N$ futures gains $NQ(F_0-F_1)\approx -NQF_0R_m$ if the futures price moves in proportion to the index. The total sensitivity to $R_m$ is $V\beta-NQF_0$, which is zero for $N=N^*$. With $N$ contracts the new sensitivity is $V\beta+NQF_0=V\beta^*$, which gives the second formula. ∎ The hedge removes the $\beta R_m$ term only: the stock-specific term $\varepsilon$ stays, and so does any error in the estimate of $\beta$.

🧭 Setting up an index-futures hedge, step by step
  1. Estimate the portfolio's beta against the index from past returns (§3.2) and note its $R^2$: the hedge can remove at most about that fraction of the variance.
  2. Choose the contract: the expiry should cover the hedge period (otherwise plan the roll).
  3. Compute $N^*=\beta V/(F_0Q)$ with the current futures price and lot size, and round.
  4. Arrange the initial margin and a cash reserve for MTM: if the market rises, the short futures lose money every day, while the gain on the shares is not cash.
  5. Lift the hedge by buying the futures back (or let them expire), and judge the result as "portfolio + futures", never one leg alone.
✏️ Example 7 — Kavya's hedge
Kavya's portfolio is worth ₹50,00,000 with $\beta=1.2$. The index is at 22,000 and the 30-day future at its fair price 22,105 (Example 2), lot size 50. (a) How many lots should she sell? (b) Suppose that at expiry the index is down 6% and her portfolio has moved exactly $\beta$ times as much. Find her loss on the shares, her gain on the futures and the net result.
(a) $N^*=\dfrac{1.2\times50{,}00{,}000}{22{,}105\times50}=\dfrac{60{,}00{,}000}{11{,}05{,}250}\approx5.43$: sell 5 lots.
(b) Shares $1.2\times(-6\%)=-7.2\%$ of ₹50 lakh $=-$₹3,60,000.
(b) Futures The index settles at $22{,}000\times0.94=20{,}680$, and by convergence so does the future. Gain $=5\times50\times(22{,}105-20{,}680)=$ ₹3,56,250.
Net $-3{,}60{,}000+3{,}56{,}250=-$₹3,750: a 7.2% fall became a 0.08% dip. The small gap comes from rounding 5.43 down to 5 lots, partly offset by the 105 points of carry earned by selling the future above spot.
✏️ Example 8 — Turning beta down, or up
With the same numbers ($V=$ ₹50 lakh, $\beta=1.2$, $F_0=22{,}105$, $Q=50$): (a) Kavya wants to cut her beta to 0.5 without selling shares; (b) her brother, with an identical portfolio, wants beta 1.5 for a month. What should each do?
(a) $N=\dfrac{(0.5-1.2)\times50{,}00{,}000}{11{,}05{,}250}\approx-3.17$: sell 3 lots (achieved beta $1.2-3\times11{,}05{,}250/50{,}00{,}000\approx0.54$).
(b) $N=\dfrac{0.3\times50{,}00{,}000}{11{,}05{,}250}\approx+1.36$: buy 1 lot (beta ≈ 1.42) — and he now carries leverage, with every warning of the ⚠️ box above.
Note Lot sizes make fine-tuning coarse for small portfolios: one lot here moves the beta by about 0.22. For a ₹5 crore portfolio the same lot would move it by only 0.022.
Interactive 2 — Beta Hedging: What a Short Index Future Removes, and What It Cannot

400 simulated one-month scenarios. Each dot is a scenario: the index return across, the portfolio's profit or loss up. Red: the portfolio alone; blue: the portfolio plus the short futures (index future at fair value, lot 50, illustrative). The portfolio moves $\beta$ times the index plus a stock-specific surprise. Slide the hedge from 0% to 150% of $N^*$ and watch the blue cloud rotate flat — and then tilt the other way. Play sweeps the hedge.

1.20100% of N*
N* (exact)—
lots sold / margin (12%)—
σ of P&L, unhedged—
σ of P&L, hedged—
variance removed—
worst scenario: alone / hedged—
📈 Market case — Kavya hedges for one month, end to end

Data (hypothetical): portfolio ₹50,00,000, $\beta=1.2$; index 22,000; illustrative $r=7\%$, $q=1.2\%$; 30-day future at fair value 22,105; lot 50; margin 12%. For clarity assume the portfolio's price moves exactly $\beta$ times the index's (no stock-specific surprise, dividends ignored).

  1. Size. $N^*\approx5.43$ (Example 7). She compares 5 lots (slightly under-hedged) and 6 lots (slightly over-hedged).
  2. Cash needed. Initial margin $=0.12\times11{,}05{,}250\times5=$ ₹6,63,150 for 5 lots (₹7,95,780 for 6). Brokers may accept pledged securities for part of this under exchange rules — check — but MTM losses must be paid in cash.
  3. Outcomes at expiry (rupees; portfolio P&L $=V\beta(S_T/S_0-1)$, futures P&L $=NQ(F_0-S_T)$):
index at expiryportfolio alone5 lots: futures5 lots: net6 lots: net
20,240 (−8%)−4,80,000+4,66,250−13,750+79,500
22,000 (0%)0+26,250+26,250+31,500
23,760 (+8%)+4,80,000−4,13,750+66,250−16,500
  1. What the hedge did. A range of ±₹4.8 lakh shrank to about −₹0.14 lakh to +₹0.66 lakh. With the exact (fractional) $N^*$ the result would be the same in every scenario: about ₹28,600, which is $\beta V\,(F_0/S_0-1)$ — the carry. A fully hedged portfolio earns roughly the risk-free rate (the carry on the futures plus the dividends the shares pay), which is exactly what no-arbitrage says it must.
  2. The cash-flow catch. In the +8% month the 5 short lots lose ₹4,13,750, paid in cash day by day, while the ₹4.8 lakh gain on the shares is only on paper. Without a cash reserve Kavya might have to sell shares — the very thing she wanted to avoid.
  3. What the table hides. Real portfolios do not move exactly $\beta$ times the index: the stock-specific part (Interactive 2) is untouched by the hedge; beta estimated from past data may be wrong for this month; and if she lifts the hedge before expiry the basis may not be at fair value. Brokerage and taxes on the futures add a small cost.

The mathematics, not a recommendation. A hedge gives up the upside to remove the downside; whether that is worth doing depends on Kavya's horizon, taxes, liquidity and nerves. An index put (§6.2) is the alternative that keeps the upside — at the price of a premium.

⚠️ Common mistakes and myths
  1. "The futures price is the market's forecast." It is spot plus carry (Theorem 6.2). A premium of 105 points says interest exceeds dividends, not that the index will rise 105 points.
  2. "I only need ₹1.3 lakh, so I am only risking ₹1.3 lakh." You are exposed to the full contract value (₹11 lakh in Example 1). Losses beyond the margin must still be paid.
  3. "MTM losses are only paper losses." They are debited in cash every day; failing a margin call can get the position closed at the worst moment (Example 5).
  4. "Leverage improves returns." It multiplies returns, good and bad, by the same factor $1/m$; it does not change the odds, and it raises the chance of being wiped out by a run of bad days.
  5. Hedging with $V/(F_0Q)$ contracts. That ignores beta: a ₹50 lakh portfolio with beta 1.2 needs about 5.4 lots, not 4.5.
  6. Judging a hedge by the futures leg alone. "I lost ₹4 lakh on my hedge" in a rising market is the hedge doing its job; look at portfolio + futures.
  7. "A hedged portfolio cannot lose." Stock-specific risk, beta error, basis and rounding remain (Interactive 2), and the hedged portfolio's expected return falls to about the risk-free rate.
  8. Forgetting dividends. A large dividend before expiry pushes a stock future below spot (Example 3); that is carry, not bearishness.
  9. Treating a cash-and-carry gap as free money. Costs, taxes, margins and the true borrowing rate eat most visible gaps; what is left is a small, capital-hungry business.

Exercises

Unless stated otherwise: index lot size $Q=50$ (illustrative), margin 12% of contract value (illustrative), continuously compounded rates, 365-day years, no costs or taxes. Round rupees to the nearest rupee and index levels to 0.1 point.

I
Payoffs and contract value.
  • 1 A trader buys 2 lots of index futures at 22,300; final settlement is 22,640. Find the profit.
  • 2 Another trader sold 1 lot at 22,300. Find her result at the same settlement.
  • 3 A stock future is at ₹845.50 with an illustrative lot of 1,250 shares. Find the contract value.
  • 4 Explain why the total profit of all buyers and all sellers of a futures contract is zero before costs. What is it after costs?
  • 5 A trader buys 3 lots at 22,450 and closes them before expiry at 22,380. Find the profit or loss.
  • 6 In one or two sentences each, explain cash settlement and physical settlement, and say which applies to NSE index futures.
1 $2\times50\times(22{,}640-22{,}300)=$ ₹34,000.
2 $50\times(22{,}300-22{,}640)=-$₹17,000.
3 $845.50\times1{,}250=$ ₹10,56,875.
4 Each contract has one long and one short side and their payoffs are $Q(S_T-F_0)$ and $Q(F_0-S_T)$, which add to zero (Theorem 6.1); summing over contracts gives zero. After costs the total is minus the total brokerage, charges and taxes paid: negative.
5 $3\times50\times(22{,}380-22{,}450)=-$₹10,500. (Closing before expiry: the MTM flows add up to $Q(F_{\text{exit}}-F_{\text{entry}})$.)
6 Cash settlement pays only the difference between the settlement price and the contract price; physical settlement delivers the asset against payment of the contract price. NSE index futures are cash-settled; stock futures are generally physically settled (check current rules).
II
Cost of carry, basis and calendar spreads.
  • 1 Index 21,500, $r=6.5\%$, $q=1.4\%$, 45 days to expiry. Find the fair futures price and the basis.
  • 2 Recompute part 1 with the simple-interest formula $S_0[1+(r-q)T]$.
  • 3 A stock at ₹2,400 pays no dividend before expiry; $r=7\%$, 60 days. Find the fair futures price.
  • 4 A stock at ₹640 pays a ₹12 dividend in 20 days; the future expires in 50 days; $r=7\%$. Find the fair futures price and compare it with the no-dividend price.
  • 5 In part 1 the future is quoted at 21,650. Which arbitrage is indicated, and roughly how much does it lock in per lot before costs?
  • 6 Index 22,000, 18-day future at 22,060. Find the implied cost of carry (continuous) and the simple annualised premium $\frac{F-S}{S}\cdot\frac{365}{18}$.
  • 7 The near-month future (30 days) is at 22,105 and $r-q=5.8\%$. Find the fair next-month future (28 days later) and the calendar spread.
  • 8 With 5 days to expiry and the index at 21,900, find the fair basis ($r-q=5.8\%$). What will the basis be at the close of the expiry day?
1 $T=45/365$: $F_0=21{,}500\,e^{0.051\times45/365}\approx21{,}635.6$, basis ≈ 135.6 points.
2 $21{,}500\times(1+0.051\times45/365)\approx21{,}635.2$, 0.4 points lower.
3 $2{,}400\,e^{0.07\times60/365}\approx$ ₹2,427.78.
4 $D=12e^{-0.07\times20/365}\approx11.95$; $F_0=(640-11.95)\,e^{0.07\times50/365}\approx$ ₹634.10, against $640e^{0.07\times50/365}\approx$ ₹646.17 without the dividend — the dividend turns a premium into a discount.
5 The quote is about 14.4 points above fair: cash-and-carry (buy the index basket or an ETF with borrowed money, sell the future). Locked gain ≈ $50\times14.4\approx$ ₹719 per lot at expiry, before costs — which in practice would likely exceed it.
6 $c=\log(22{,}060/22{,}000)/(18/365)\approx5.52\%$; simple: $(60/22{,}000)\times365/18\approx5.53\%$.
7 $22{,}105\,e^{0.058\times28/365}\approx22{,}203.6$; spread ≈ 98.6 points.
8 $21{,}900\,(e^{0.058\times5/365}-1)\approx17.4$ points; at expiry the basis is $0$ (convergence, Theorem 6.3).
III
Margins, MTM and leverage. A trader buys 2 lots at 22,000 and deposits ₹3,00,000. Settlement prices on the next three days are 21,850, 21,700 and 21,950.
  • 1 Find the initial margin required.
  • 2 Find the MTM on each of the three days.
  • 3 Track the balance against the margin required. Is there a margin call?
  • 4 After day 1 (balance ₹2,85,000), below which day-2 settlement price would a margin call have been triggered?
  • 5 What are the leverage factor and the percentage change in the initial margin if the futures price rises 1.5%?
  • 6 With ₹4,00,000 of capital the trader buys 3 lots at 22,000. Find the effective leverage and the index fall (in points and %) that loses half of the capital.
  • 7 "I'm sure the index will recover by expiry, so I can ignore this week's MTM losses." Explain what is wrong with this reasoning.
1 $0.12\times22{,}000\times100=$ ₹2,64,000 (2 lots = 100 units).
2 $100\times(-150)=-$₹15,000; $100\times(-150)=-$₹15,000; $100\times250=+$₹25,000.
3 Balances ₹2,85,000; ₹2,70,000; ₹2,95,000 against requirements $12F_t$: ₹2,62,200; ₹2,60,400; ₹2,63,400. No margin call. Total MTM $=-$₹5,000 $=100\times(21{,}950-22{,}000)$.
4 Balance after day 2 is $2{,}85{,}000+100(F_2-21{,}850)$; requirement $12F_2$. A call occurs if $88F_2\lt19{,}00{,}000$, i.e. $F_2\lt21{,}590.9$ — a fall of under 1.9% from entry.
5 Leverage $1/0.12\approx8.33$; a 1.5% rise changes the margin by $+12.5\%$.
6 Exposure $3\times50\times22{,}000=$ ₹33,00,000: leverage 8.25. Half the capital, ₹2,00,000, is lost after $2{,}00{,}000/150\approx1{,}333$ points, about 6.1%.
7 MTM losses are paid in cash daily. If the account falls below the required margin and the trader cannot add money, the broker can close the position; then the later recovery does not help. Being right "eventually" is not enough with leverage.
IV
Hedging with beta. A portfolio is worth ₹80,00,000 with $\beta=0.9$. The index is at 22,100 and the future at 22,200.
  • 1 How many lots should be sold for a full hedge?
  • 2 At expiry the index is down 5% and the portfolio has fallen exactly $\beta$ times as much. Find the net result with the number of lots from part 1.
  • 3 How many lots reduce the beta to 0.3?
  • 4 How many lots, bought or sold, raise the beta to 1.2?
  • 5 Explain why a fully hedged equity portfolio should earn about the risk-free rate.
  • 6 The portfolio's monthly volatility is 6% and its $R^2$ against the index is 0.6. Estimate the monthly volatility left after a perfect beta hedge, and the fraction of variance removed.
1 $N^*=\dfrac{0.9\times80{,}00{,}000}{22{,}200\times50}$ $=\dfrac{72{,}00{,}000}{11{,}10{,}000}\approx6.49$: sell 6 lots.
2 Index $22{,}100\times0.95=20{,}995$; portfolio $-4.5\%=-$₹3,60,000; futures $6\times50\times(22{,}200-20{,}995)=+$₹3,61,500; net $+$₹1,500.
3 $(0.3-0.9)\times80{,}00{,}000/11{,}10{,}000\approx-4.32$: sell 4 lots.
4 $(1.2-0.9)\times80{,}00{,}000/11{,}10{,}000\approx+2.16$: buy 2 lots.
5 Shares plus a short future at the cost-of-carry price form (approximately) a riskless position; anything riskless must earn the risk-free rate, or the cash-and-carry argument of Theorem 6.2 would give an arbitrage. The hedger earns the dividends plus the futures premium $\approx(r-q)T$ — together about $rT$.
6 The hedge removes the market part, a fraction $R^2=60\%$ of the variance; the remaining volatility is $6\%\times\sqrt{1-0.6}\approx3.8\%$ a month. The stock-specific risk remains.
🎯 Practice corner NISM-styleCase studyInterview

E1. NISM-style A share trades at ₹1,000 and pays no dividend before expiry. With an interest rate of 8% a year (simple interest), the fair price of a three-month future is closest to (A) ₹1,000 (B) ₹1,080 (C) ₹1,020 (D) ₹1,002.

Compute $1{,}000\times(1+0.08\times\tfrac14)=1{,}020$ (continuously compounded: 1,020.20).
Distractors (A) ignores carry; (B) uses a full year; (D) uses one month at about 2.4% a year.
Answer (C).

E2. NISM-style As a futures contract approaches expiry, its basis (futures minus spot) (A) grows steadily (B) approaches zero (C) becomes equal to the dividend (D) is always negative.

Reason $b_t\approx S_t(r-q)(T-t)\to0$; at expiry the future and the spot are the same thing (Theorem 6.3).
Answer (B).

E3. NISM-style A fund manager holds a ₹1 crore portfolio with beta 1.5. Index futures trade at 20,000 with a lot size of 50. To hedge fully she should (A) sell 10 lots (B) sell 15 lots (C) buy 15 lots (D) sell 100 lots.

Compute $N^*=1.5\times1{,}00{,}00{,}000/(20{,}000\times50)=15$, and a long portfolio is hedged by selling.
Distractors (A) ignores beta; (C) doubles the risk; (D) forgets the lot size.
Answer (B).

E4. Case study Rahul sells 2 lots of index futures at 22,400 and deposits ₹2,80,000 (margin 12%, lot 50). Settlement prices over four days: 22,550, 22,760, 22,690, 22,300. (a) Find each day's MTM. (b) When does he face a margin call, and for how much (top up to the requirement)? (c) Find his total profit and the return on the cash he put in. (d) What would his result have been if he could not meet the day-2 call and was closed out at 22,760?

(a) A short gains when the price falls: $-100\times150=-$₹15,000; $-$₹21,000; $+$₹7,000; $+$₹39,000.
(b) Day 1: balance ₹2,65,000 vs required $12\times22{,}550=$ ₹2,70,600: call ₹5,600. Day 2: $2{,}70{,}600-21{,}000=$ ₹2,49,600 vs ₹2,73,120: call ₹23,520. Days 3–4: ₹2,80,120 and ₹3,19,120, above the requirements (₹2,72,280 and ₹2,67,600).
(c) Profit $=100\times(22{,}400-22{,}300)=$ ₹10,000; cash put in $2{,}80{,}000+5{,}600+23{,}520=$ ₹3,09,120, final balance ₹3,19,120: about 3.2%.
(d) A loss of $100\times(22{,}760-22{,}400)=$ ₹36,000 — the same view, a very different outcome, decided by liquidity, not by the market's final level.

E5. Case study A fund manager runs a ₹5 crore portfolio with $\beta=0.85$. The index is at 22,000; illustrative $r=6.8\%$, $q=1.3\%$; 35 days to expiry; lot 50; margin 12%. (a) Find the fair futures price. (b) How many lots hedge the portfolio? (c) What initial margin is needed? (d) If the index falls 7% by expiry and the portfolio moves exactly $\beta$ times as much, find the net result.

(a) $F_0=22{,}000\,e^{0.055\times35/365}\approx22{,}116.3$.
(b) $N^*=\dfrac{0.85\times5{,}00{,}00{,}000}{22{,}116.3\times50}\approx38.4$: sell 38 lots.
(c) $0.12\times22{,}116.3\times50\times38\approx$ ₹50.4 lakh — about 10% of the portfolio, plus a reserve for MTM.
(d) Index 20,460. Portfolio $0.85\times(-7\%)=-5.95\%=-$₹29,75,000. Futures $38\times50\times(22{,}116.3-20{,}460)\approx+$₹31,47,000. Net ≈ $+$₹1.72 lakh: the carry on the hedge, slightly reduced because 38 lots is a little below 38.4.

E6. Interview "The one-month index future is at a discount to spot even though the dividend yield is below the interest rate. Free money?" What trade would exploit it in theory, and why might the discount persist in practice?

Theory With $r\gt q$, Theorem 6.2 says $F_0\gt S_0$, so the future is cheap: reverse cash-and-carry — sell the index basket short, invest the proceeds at $r$, buy the future. At expiry buy back the basket at $S_T$ through the future's payoff; the gain is $S_0e^{(r-q)T}-F_0\gt0$.
Practice Short-selling 50 stocks is costly and restricted (borrowing shares, margins, settlement rules); an investor who already owns the basket could sell it instead, but faces taxes and costs; the true lending rate may be lower than $r$; dividends may be expected to be larger than the trailing yield; and heavy hedging demand from sellers can push the future down. Arbitrage pulls prices back only when it is cheap and easy to do.
Interview answer "In theory a reverse cash-and-carry, in practice limited by the cost and difficulty of shorting the basket — the no-arbitrage band is wide on the short side." Mentioning the asymmetry is what interviewers look for.
🧑‍🏫 For the teacher

Suggested plan — 2 lectures × 60 min.

  1. Pricing. Start from the 📍 situation: put a live (or the page's hypothetical) index and future quote on the board and ask "why is the future higher?" Collect answers ("people expect a rise") and then derive Theorem 6.2 with the two routes to owning the index at expiry. Examples 2–4; project Interactive 1 with the lower panel on "basis" and change $r$ and $q$.
  2. Margins and hedging. Example 5 as a class ledger on the board (students compute each day's MTM). Switch Interactive 1 to the margin account and to "cash = initial margin" — count the margin calls over a few New paths. Then Theorem 6.5, Example 7, the 📈 Market case and Interactive 2 (sweep the hedge; raise stock-specific risk). Close with the ⚠️ leverage box.

Class activity. Give each pair a different hypothetical portfolio (value, beta, $R^2$). They size a one-month hedge, compute the margin and a cash reserve for a 5% adverse move, and present the hedged and unhedged outcomes for index moves of −8%, 0 and +8%.

Misconceptions to watch for. Futures price = forecast; margin = maximum loss; MTM losses are not real until expiry; leverage improves the odds; hedge with $V/(FQ)$ ignoring beta; a hedge that loses money in a rally "failed".

Suggested homework. I(1, 3, 4), II(1, 4, 5, 7), III(2–4, 6, 7), IV(1, 2, 5, 6); Practice corner E3, E4, E5.

🛠️ Do it yourselfSpreadsheet mini-project · about 60–90 minutes

Project. Build a futures "account simulator": a simulated month of index prices, the fair futures price each day, a margin ledger with margin calls, and a beta hedge of your own (hypothetical) portfolio.

Why it matters. Reading that "MTM is settled daily" is one thing; watching your own ledger demand cash on day 4 of a month that ends in profit is another. The same sheet turns the hedge formula into numbers you can check.

What to do.

  1. Inputs: spot S0 in B1, r B2, q B3, calendar days to expiry B4, lot size B5 (illustrative), margin rate B6, daily volatility B7 (e.g. =0.14/SQRT(252), the 252-day convention).
  2. Rows 10–31 (22 trading days): day k in A; spot in B (B10 =B1, B11 =B10*EXP(B$7*NORM.S.INV(RAND()))); days left in C =B$4*(1-A10/21); fair future D =B10*EXP((B$2-B$3)*C10/365); basis E =D10-B10. Paste the spot column as values once so it stops changing.
  3. Ledger: MTM F11 =B$5*(D11-D10); required margin G =B$6*D10*B$5; balance H (H10 = your deposit), top-up I11 =MAX(0,G11-(H10+F11)), H11 =H10+F11+I11. Check that =SUM(F11:F31) equals =B5*(D31-D10) (Theorem 6.4).
  4. Hedge sheet: portfolio value, beta (estimate it with =SLOPE(fund returns, index returns) on monthly data you download — for example a fund's NAV history from AMFI or its AMC and index history from the NSE website, §3.2), N=beta*V/(F0*Q), and a table of outcomes for index moves from −10% to +10%.
  5. Python version: S=[S0]; for k in range(n): S.append(S[-1]*exp(sig*random.gauss(0,1))), then F=[S[k]*exp((r-q)*(T-t[k])) for k in range(n+1)] and the ledger in a loop.

Hand in: a chart of spot, futures and basis; the margin ledger with every margin call highlighted; the number of margin calls in 20 recalculated months when the deposit equals the initial margin versus 1.5 times it; and your hedge table.

Going further. Add a second expiry and simulate a roll; or regress your portfolio on the index and compare the hedged volatility with $\sigma\sqrt{1-R^2}$.

Instructor note: when the balance sits exactly at the requirement, any down day triggers a call (the balance falls by $Q\,\Delta F$, the requirement only by $mQ\,\Delta F$). In our own simulation (14% annual volatility, 21 trading days, 12% margin) about 9 months in 10 brought at least one call when the deposit equalled the initial margin, about 4 in 10 with a 1.25× deposit, and almost none with 2× — a good discussion of why brokers ask for buffers. Check that students paste the random path as values before comparing runs.

🧪 Self-check quiz

0/5 correct

Q1. With $r\gt q$ and no arbitrage, a one-month index future trades

Q2. A long futures position is held from $F_0$ to expiry. The sum of all its daily MTM flows equals

Q3. With a margin of 10% of contract value, a 1% fall in the futures price changes a long trader's initial margin by

Q4. To hedge a ₹40 lakh portfolio with beta 0.8 using index futures at 20,000 with lot 50, you should

Q5. An equity portfolio fully hedged with index futures at fair value should be expected to earn about