Pair ratio density trade
This is the simplest form of pair trading: no regression, just the ratio of one stock's price to another's. The ratio of AXISBANK to ICICIBANK futures is placed on a normal density curve built from its last 500 sessions. When it lands in the far tail, the strategy bets the ratio will drift back towards its average, buying the cheap stock and selling the rich one in equal rupee amounts.How it trades
Results
PRE-BUILT · DEFAULTSWhere it made and lost money
There is no evidence to split here. On default settings the strategy took no trades between 1 Oct 2025 and 30 Sep 2026, so nothing can be said about how it fared against the NIFTY market backdrop, whether by India VIX, session type or trend. The cause is the narrow entry band: the AXISBANK/ICICIBANK ratio had to close in the outer 2.5% of its two-year density curve, but not past the 0.3% stop level, on a day the stocks' returns also correlated at 0.7 or more. That combination did not occur in this year.
The same trades, split by the market conditions around each entry, before costs.About this strategy
- Each evening, divide AXISBANK's near-month futures close by ICICIBANK's and measure the ratio's average and standard deviation over the last 500 sessions.
- Turn today's ratio into a percentile on the normal curve (its cumulative probability).
- Check the filters: daily returns of the two stocks must correlate at 0.7 or more, the window must have no one-day ratio jump above 15%, and neither stock may be in the F&O ban.
- If the ratio sits between the 0.3% and 2.5% percentiles, buy 1 lot of AXISBANK and sell ICICIBANK lots of about the same rupee value at 09:20 next session; at the mirror-image top tail, do the reverse. Both legs sit in one position.
- Close at 09:20 once the ratio is back inside about one standard deviation (target) or beyond about three (stop).
- On monthly expiry both legs close at 15:15 and reopen, re-sized, in the next month at 09:20 the following session.
Why traders use it
A price ratio is easy to chart and easy to explain: if AXISBANK usually trades at a steady fraction of ICICIBANK, an unusual reading suggests one has run ahead of the other. Putting the ratio on a density curve turns that into clear entry, target and stop bands. The method needs no statistics package, which is why it is widely taught to Indian traders as a first step into pair trading with stock futures.
The year in plain words
On default settings from 1 Oct 2025 to 30 Sep 2026 the ratio trade on AXISBANK and ICICIBANK made no trades, so there is no profit or loss before or after costs. The code re-measures the ratio every session and opens a pair only when it sits between the 0.3% and 2.5% tail of its 500-session curve, with return correlation of at least 0.7. In this year the ratio never closed in that band with the filters passing. This means the rule set was untested on this pair over the year, not that it worked or failed.
What the settings do
Risks
- A ratio far from its two-year average may reflect a real change in the businesses, and the trend can continue to the three-deviation stop.
- Whole-lot sizing means the rupee values of the two legs rarely match exactly, leaving some net market exposure.
- The time stop is off by default, so a trade drifting between target and stop can stay open for months, rolling at each expiry.
- Each round trip is four futures orders, plus four more at every monthly roll, so costs add up.
Who it suits
Traders who want a simple, chart-friendly introduction to pair trading with stock futures, have margin for two futures positions at once, and are prepared for long spells with no signal.
Questions traders ask
Is the ratio pair trade profitable on AXISBANK and ICICIBANK?
This test cannot say. With default settings it made zero trades from 1 Oct 2025 to 30 Sep 2026 because the ratio never reached the 2.5% tail of its two-year curve while the filters passed, so there was no profit or loss before or after costs. Past results, or the lack of them, do not guarantee anything about future ones.
How is the ratio method different from regression pair trading?
Regression methods fit a hedge ratio and trade the leftover residual. The ratio method simply divides one price by the other and asks how unusual today's reading is against its own history. It is easier to follow but assumes the two stocks should be held in equal rupee amounts.
What does the density curve tell you?
The ratio's past 500 readings give an average and a standard deviation. Placing today's ratio on a normal curve with those values gives a percentile: 2.5% means only about one reading in forty would be expected this low. The strategy enters in that tail, takes profit near one deviation and stops beyond about three.
How can I make the strategy trade more often?
Raise the entry percentile, lower the return-correlation filter, shorten the lookback, or choose a different second stock. Each widens the set of days that qualify. None of them says whether the extra trades would make money.