Overnight gap tilt
Research on stock returns often finds that overnight moves and daytime moves behave differently, with some of the overnight gap reversing during the session. This single-stock version applies that to one F&O stock. Each morning it compares the stock's overnight gap with NIFTY's; if the stock opened at least 0.5% weaker than the market, it buys the stock future at 09:16 and sells it at 15:20.How it trades
Results
PRE-BUILT · DEFAULTSWhere it made and lost money
With six trades on RELIANCE, no condition has enough trades to judge. The two trades on quiet sessions both won, together ₹25,600, while three on choppy days lost ₹10,300 and the single trend-day trade lost ₹6,350. That fits the idea loosely: a stock that gapped down more than the market recovers best when the rest of the day is calm, and keeps sliding when the session swings widely. The four trades taken on NIFTY weekly expiry days all lost. With no stop, each result is simply the stock's move from 09:16 to 15:20. Figures are before costs.
About this strategy
- Each morning, compute the stock future's overnight return: today's open versus yesterday's close.
- Subtract NIFTY futures' overnight return to get the stock's gap relative to the market.
- If the relative gap is -0.5% or worse, buy one lot of the stock's near-month future at 09:16.
- Skip the day if the stock is in the F&O ban, or if it is the session right after a monthly futures expiry.
- Sell at 15:20 with no stop or target. At most one trade a day, long only.
Why traders use it
A stock that opens far weaker than the market may be reacting to overnight selling pressure that eases once regular trading begins, so some of the gap is often recovered. Measuring the gap relative to NIFTY separates stock-specific weakness from a broad market move. The rule is simple, entirely mechanical and holds only during market hours, avoiding overnight risk. A momentum setting tests the opposite idea: buying stocks that gapped up more than the market.
The year in plain words
On RELIANCE futures from 1 Oct 2025 to 30 Sep 2026, with default settings, the relative gap passed -0.5% on only six days. Two trades won and four lost, for ₹8,950 before costs and ₹7,060 after ₹20 per order and bid-ask slippage. One April trade made ₹20,950, more than the year's total, and April overall made ₹16,400; September lost ₹6,350. Six trades in a year cannot show whether the reversal effect holds for this stock. The original idea ranks every F&O stock daily, which a single-stock test cannot reproduce.
What the settings do
Risks
- There is no stop loss on a full stock futures lot for the whole session.
- A stock that gaps down on real news may keep falling rather than rebound.
- One stock gives only a handful of signals a year, so results depend heavily on single trades.
- The original cross-sectional idea, ranking all stocks, is not captured by trading one stock.
Who it suits
Traders interested in overnight-gap effects in Indian stocks who want a simple intraday rule to test across different F&O stocks, and who have futures margin for a full lot without a stop.
Questions traders ask
Do stock gaps get filled in India?
Often partly, but not reliably. Studies find that some overnight moves reverse during the day, especially for stock-specific gaps. This test on RELIANCE had only six signals in a year, too few to say how often its gaps filled.
Is the overnight gap reversal strategy profitable?
On RELIANCE futures from 1 Oct 2025 to 30 Sep 2026 it made ₹8,950 before costs and ₹7,060 after ₹20 per order and bid-ask slippage, over six trades. One trade made more than the total. Past results do not guarantee future ones.
Why compare the stock's gap with NIFTY?
If the whole market gaps down, a stock falling with it says little. Subtracting NIFTY's gap isolates how much weaker the stock opened than the market, which is the signal this rule trades.
Can I run it on other stocks?
Yes. The symbol can be set to any F&O stock. Running it on many stocks gives a rough picture of the cross-sectional effect, though not the original daily ranking.