Most short-selling material written for retail investors is American. It assumes you can borrow a stock, hold it for weeks, and cover when the thesis plays out. In India that describes a trade a retail cash account cannot place, and a backtest built on that assumption is modelling something you are not permitted to do.
Here is the actual rulebook, and what a short strategy has to look like once you respect it.
The rule that shapes everything
A short position in the cash segment must be squared off before the session closes. You cannot sell a share you do not own and carry that position overnight in an ordinary delivery account, because settlement requires you to deliver a share you do not have. Fail to square off and you land in the exchange’s auction settlement, where the shares are bought on your behalf at a penal price.
Institutional short selling is permitted, and so is retail short selling — but overnight, both require the same thing: you must actually borrow the stock first.
Stock Lending and Borrowing, and why it is thin
SEBI runs a formal mechanism for this: the Securities Lending and Borrowing framework, set out in SEBI’s Securities Lending Scheme and operated through the exchanges’ clearing corporations as Approved Intermediaries. A lender who owns shares and does not intend to sell them can lend them out for a fee; a borrower posts margin, pays the fee, and gets shares to deliver against a short.
It works. It is also small, and small in a way that matters for strategy design:
- Availability concentrates in the largest names. The stocks with meaningful lending pools are index heavyweights. A screen that finds its short candidates among midcaps and smallcaps will frequently find no borrow at all.
- The fee is variable and reflexive. Borrow cost rises with demand, which means it rises exactly when a name is crowded — precisely the moment a short screen wants to be in it.
- Recall risk is real. A lender can recall shares, forcing you to cover at a time not of your choosing.
In our own data, borrow-fee observations exist for only a fraction of the securities with SLB records. That sparsity is not a collection gap; it reflects a market where most stocks simply are not lent.
What this does to a backtest
The trap is specific. A backtest that shorts a stock for three weeks and reports a profit has assumed:
- borrow was available for that name, every day, for three weeks
- at a cost of zero
- without recall
For a Nifty 50 name, assumption 1 is often fine and 2 is wrong by a modest amount. For the low-momentum smallcap a bottom-decile screen actually selects, all three are wrong, and the reported profit is on a trade nobody could have placed.
Our engine’s capability layer drops holdings it could not have placed rather than pretending otherwise, so a short book that shrinks in the results is information. But it does not yet model borrow cost or check per-day SLB availability, which means overnight short strategies here are still modelled more favourably than reality. That is a stated limitation, and it is the reason this particular template is intraday.
The intraday short: the version that is actually available
Square the position off before the close and the borrow problem disappears. There is nothing to deliver, so there is nothing to borrow. Your broker’s intraday product (MIS, or equivalent) permits it, with margin, and squares you off automatically if you do not.
That constraint turns out to point at a strategy that has an independent justification.
Short-horizon reversal runs in both directions. The same liquidity-provision mechanism that makes oversold large caps bounce makes overextended ones fade: a stock that has run two standard deviations above its own recent average without news has a buyer paying for immediacy, and someone taking the other side collects that premium.
The upside version is less exploited than the downside version, for a straightforward structural reason — far more capital is permitted to buy dips than to short strength. Constraints on short selling are one of the standard explanations for why the low-volatility anomaly persists, and the same logic applies here: an effect that requires shorting to arbitrage away is one that stays un-arbitraged.
Design consequences
Because it is intraday and short, several things are forced rather than chosen:
Large caps only. Intraday shorting needs tight spreads and the ability to exit in size at any moment. In a thin name, an adverse move with no liquidity on the bid is how a small loss becomes a large one.
Costs are the deciding term. At roughly 250 round trips a year, an intraday strategy’s profitability is a function of the cost model. Intraday STT is charged at 0.025% on the sell side only, against 0.1% both legs for delivery — the single structural advantage of intraday trading in India — but slippage at the open is the dominant cost and is not on any rate card. Run the strategy, then run it again with slippage doubled; if the second run is unprofitable, the first was measuring your assumption.
The loss profile is asymmetric. A long position can lose 100%. A short position’s loss is unbounded — a stock can double. Intraday square-off caps the exposure to one session, which is the main defence, but it is not a stop.
No overnight gap risk, in either direction. A book that is flat at the close cannot gap against you. For a short book, where the gap risk is the unbounded side, that matters more than it does for a long book.
Try it
The Intraday Fade template shorts the ten Nifty 100 names most stretched above their own 40-day average at the open, and covers before that day’s close.
Read the result sceptically. Shorting into strength has a modest win rate and is structurally wrong-footed in a melt-up. What the template is really for is showing that a short strategy can be expressed honestly — sized, timed and constrained to what an ordinary Indian account is permitted to place — rather than as a backtest of a trade the rulebook does not allow.
For the research version that ignores the constraint deliberately in order to isolate the pure factor spread, see the market-neutral template.
Further reading
- Market-neutral investing: isolating the factor from the market is where this constraint does the most damage, because the short leg is the entire point of the design
- Backtesting an intraday strategy honestly for the cost arithmetic on the one short structure a retail account can actually run
Glossary: securities lending and borrowing, square-off, securities transaction tax, slippage.