Every futures contract has a quiet twin in the spot market. NIFTY trades in the cash segment; NIFTY futures trade a few points away in the F&O segment. A stock sits at one price; its future sits at another. That little distance between the two prices is called the 

, and a whole family of strategies exists for one reason only: to capture it cleanly, without betting on whether the market goes up or down.

This is the trade that mutual funds quietly run with thousands of crores. It's the trade your "arbitrage fund" actually does under the hood. And it's one of the cleanest ways to understand what "carry" really means. Let's build it from the ground up.

The basis is simply the difference between the futures price and the spot price of the same underlying:

If NIFTY spot is 24,000 and the near-month future is 24,120, the basis is 

 to spot. This is the normal state for index and most stock futures, and it has a name: 

 spot — say the future is 23,950 when spot is 24,000. The basis is 

. In Indian equities this usually shows up around dividends or when there's heavy demand to short a stock.

The single most important fact about the basis: 

 On the last day, the futures contract settles to the spot price. There's no choice about it — that's how cash settlement works. So whatever the gap is today, it is guaranteed to shrink to nothing by expiry.

That guaranteed convergence is what makes the trade an 

2. Why does the gap exist at all? The cost-of-carry idea

 — the cost of holding the underlying until the futures expiry.

Think about it from the seller's side. Suppose you want to deliver NIFTY at expiry. You could buy the basket of stocks 

 and hold it. But buying today means tying up money. That money has a cost — it could have earned interest in a fixed deposit or saved you interest on a loan. To compensate for that financing cost, the future should sell for 

So for an equity index, the fair value of a future is roughly:

Fair Futures = Spot × (1 + r × T) − Dividends

For a broad index most of the time financing wins, so futures sit at a premium. For a single high-dividend stock approaching its ex-date, dividends can win, and the future flips to a discount.

The implied annualised return baked into a futures premium is what NSE displays as the 

, and it's the number a basis trader watches all day:

Annualised basis = (Basis ÷ Spot) × (365 ÷ Days to expiry) × 100

Annualised basis = (120 / 24000) × (365 / 30) × 100 ≈ 6.08%


So the market is paying roughly 6% annualised to be long NIFTY through futures instead of cash. That number is the whole game. If you can fund a cash position for 

3. The cash-and-carry trade, step by step

 — trading above its fair value, i.e. the annualised basis is higher than your cost of funding.

 the position to expiry and let convergence do the work.

Here's why it's market-neutral. Whatever happens to the price:

Add the two legs together and the S_T cancels out completely:

Locked-in profit = (S_T − S) + (F − S_T) = F − S = the basis

You captured the basis. The direction of the market never mattered. The only "cost" against that basis is the financing you paid to hold the spot. So:

annualised basis beats your funding rate.

Say you run a basis book and your effective cost of funds is 

Annualised basis = (25 / 2000) × (365 / 25) × 100 = 

That's far above your 6.5% funding cost — a juicy spread. You:

At expiry, both converge. Gross profit = 25 × 250 = 

Financing cost for 25 days = 5,00,000 × 6.5% × (25/365) ≈ 

 on ₹5,00,000 deployed for 25 days. Annualise that and you're comfortably ahead of a fixed deposit, with essentially no market-direction risk. 

(Hold the celebration — transaction costs and a few real-world catches eat into this. We'll get there in Section 8.)

4. The reverse cash-and-carry (and why it's hard in India)

 — trading at a discount that's too deep to be justified by dividends — you flip the trade:

At expiry, convergence again hands you the gap, and you earn interest on the cash from the short sale on top.

you can't easily short in the cash segment and hold it.

 Intraday shorting in cash is fine, but you can't carry a naked short overnight in the equity cash market. To short and 

 mechanism to borrow the stock — and SLB in India is thin, lending fees can be high, and many stocks simply aren't available to borrow.

This asymmetry matters. It means the "no-arbitrage" band is not symmetric in Indian equities. Futures can stay 

 than fair value for longer than they can stay 

, because the force that would correct a discount (reverse cash-and-carry) is partly blocked. Stock futures running at a persistent discount often reflect exactly this — strong demand to short the future, with no easy way to arbitrage it away.

 work smoothly is in instruments you can short freely — index futures via the cash basket held by institutions, commodities with a borrow/lending market, and currencies.

Here's the part most retail investors never realise. When you buy an 

, you are buying a professionally run cash-and-carry book.

Because the fund is fully hedged, its returns behave like a low-risk, debt-like instrument — but here's the kicker: since it holds ≥65% in equity and equity derivatives, it's 

, not a debt fund. For an Indian investor in a high tax bracket, that tax treatment is often the entire reason to choose an arbitrage fund over a liquid fund with similar gross yield.

Their returns rise and fall with the basis. When markets are euphoric and futures premiums are fat (high annualised basis), arbitrage funds do well. When the market is dull and the basis compresses, their returns thin out. If you've ever wondered why arbitrage-fund returns wobble between, say, 4% and 8% across the year — you're watching the cost-of-carry breathe.

So even if you never put on a single futures trade yourself, understanding the basis tells you 

 what's driving a chunk of the "safe" money in Indian portfolios.

6. Calendar spreads and rollovers: basis between two futures

 — the near month and the far month. The gap between them is the calendar spread, and it carries the same cost-of-carry logic over the time 

This is what's really happening during the noisy "rollover" week you hear about on business TV. Traders carrying near-month positions roll them to the next month by buying the far month and selling the near month (or vice versa). The price at which they roll — the 

 — is just the calendar basis, annualised. A high rollover cost means longs are paying up to stay long: a contango signal. A negative roll means the opposite.

For a basis trader, calendar spreads are attractive because both legs are futures, so margins are lower (the exchange recognises the hedge) and you avoid the cash-segment funding and shorting headaches entirely.

7. Beyond equities: commodities and currencies

The cash-and-carry skeleton is identical across asset classes; only the carry components change.

Commodities (MCX — gold, silver, crude).

 (the benefit of physically holding the commodity, which pulls it lower). Gold is the textbook cash-and-carry commodity because it's cheap to store and easy to finance, so its futures track 

 very tightly. Crude is messier because storage is expensive and convenience yield swings hard — which is how crude futures occasionally go into deep backwardation or, in extreme stress, even negative.

 The currency version of cash-and-carry is 

. The forward/futures premium of USDINR over spot reflects the 

 between the rupee and the dollar. Because Indian rates are structurally higher than US rates, USDINR futures almost always trade at a 

 to spot — that premium is the rate gap, expressed as carry. When the observed premium drifts away from the rate differential, a covered-interest-parity arbitrage opens up.

Same idea every time: figure out the true cost of carry, compare it to the price the market is quoting, and trade the gap.

8. The catches — where the "free money" leaks out

. The leaks are real and they're what separate a profitable book from a losing one.

 Every leg has brokerage, exchange fees, GST, stamp duty, and — the big one in India — 

. You're trading two legs, possibly unwinding both, and on a thin basis these costs can swallow most of the edge. Always model the 

 cost before deciding a spread is profitable.

 are not synchronised. The futures short is marked to market daily; if the market rallies, you must post variation margin 

, while the offsetting gain on your spot sits unrealised until you sell. A position that is perfectly hedged on paper can still create a real cash crunch if you haven't kept buffer. Underfunded basis books blow up not on the spread but on margin calls.

 Your fair value assumed a certain dividend. If a company declares a surprise special dividend (or cuts an expected one), the future re-prices and your locked-in basis shifts. For single-stock cash-and-carry, unexpected corporate actions are the classic way the trade goes wrong.

 The two legs must go on as close to simultaneously as possible. If you buy the spot and the future moves before you can sell it, your basis has already changed. In fast markets this slippage is where the theoretical edge quietly disappears.

. If your cost of funds rises mid-trade (or you were borrowing at a floating rate), the spread you locked in can compress against you.

 If you intend to roll rather than hold to expiry, you're exposed to whatever the calendar basis is on roll day. And in illiquid single stocks, simply getting out of both legs at fair prices can be a problem.

None of these kill the strategy — but they explain why real-world basis returns are 

measured in single-digit annualised percentages

, not the eye-catching 18% from our clean example.

The discretionary version is "watch a screen, spot a fat basis, put it on." The systematic version — the one worth building if you live in Python — is a 

: continuously compute the annualised basis for a universe of stocks and the index, compare it against your true cost of funds, and rank the opportunities net of round-trip costs.

The core calculation is tiny. Here's a clean, self-contained version you can drop into a research notebook:

import numpy as np
import pandas as pd

def annualised_basis(spot, future, days_to_expiry):
    """Annualised cost-of-carry implied by a spot/future pair (in %)."""
    basis = future - spot
    return (basis / spot) * (365 / days_to_expiry) * 100

def net_edge(spot, future, days_to_expiry, funding_rate, roundtrip_cost_pct):
    """
    Net annualised edge of a cash-and-carry trade, after funding and costs.
    funding_rate, roundtrip_cost_pct are in % (annualised / of notional).
    A positive number means the trade is worth doing.
    """
    gross = annualised_basis(spot, future, days_to_expiry)
    # spread costs are a one-off on notional -> annualise them too
    cost_annualised = roundtrip_cost_pct * (365 / days_to_expiry)
    return gross - funding_rate - cost_annualised

# --- toy universe -------------------------------------------------
book = pd.DataFrame({
    "symbol":       ["NIFTY", "STOCK_X", "STOCK_Y", "STOCK_Z"],
    "spot":         [24000.0, 2000.0,    540.0,     1180.0],
    "future":       [24120.0, 2025.0,    538.0,     1192.0],
    "days":         [30,      25,        25,        25],
})

FUNDING = 6.5          # your true cost of funds, annualised %
ROUNDTRIP = 0.15       # all-in round-trip cost as % of notional (STT, brokerage, slippage)

book["ann_basis_pct"] = annualised_basis(book.spot, book.future, book.days).round(2)
book["net_edge_pct"]  = net_edge(book.spot, book.future, book.days,
                                 FUNDING, ROUNDTRIP).round(2)

# only show trades that actually clear the hurdle
opportunities = book[book["net_edge_pct"] > 0].sort_values("net_edge_pct", ascending=False)
print(opportunities)


The whole philosophy of basis trading is in that 

gross basis, minus funding, minus round-trip costs.

 Most "obvious" opportunities die on the second and third terms. STOCK_Y in the toy data shows a 

 basis — the future is at a discount — which a cash-and-carry can't capture (you'd need the reverse trade, and in India that runs into the SLB wall from Section 4).

From here a production version adds: live data feeds for spot and futures, lot-size and capital constraints, dividend-adjusted fair values, a margin/MTM buffer model, and a borrow-availability check before ever flagging a reverse trade. But the signal at the heart of it never changes.

The basis looks like free money until you account for the carry. Once you do, you realise the whole trade is just getting paid — fairly, and in cash — for being the patient one who holds the position while everyone else takes a view.

This article is for educational purposes only and is not investment advice. Derivatives carry risk; size positions according to your own risk tolerance and consult a registered advisor before trading.