PAIR TRADING · Z-SCORE · 60-DAY

Pair-trading scanner — where the ratio is stretched.

The stat-arb view of 18 classic Indian pairs — HDFC Bank / ICICI, TCS / Infosys, SBI / PNB, Reliance / ONGC, Sun Pharma / Dr Reddy. Rolling 60-day z-score on the log-ratio tells you when a pair is statistically overextended; rolling correlation tells you whether the mean-reversion thesis is still valid. Both together in one view.

The math in one page

For each pair (A, B): compute daily price ratio r_t = A_t / B_t, take the log ln(r_t), then measure today's ln(r)vs its 60-day mean in units of standard deviation:

z_t = (ln(r_t) − mean_60(ln(r))) / stdev_60(ln(r))

+2 or higher = A is overextended vs B — historically the ratio reverts. Trade: sell A, buy B, hold until z → 0.

−2 or lower = B is overextended vs A — inverse trade.

Correlation < 0.4= pair is regime-broken — the mean-reversion thesis doesn't apply. We flag this as "break". Do not trade z extremes when correlation has collapsed.

Implementation notes for Indian traders

  • Use F&O, not cash. Shorting cash equity is SLB-dependent and expensive. Most of these names have liquid stock futures — sell-buy pair on futures avoids the borrow cost.
  • Size by beta-neutral notional. If β_A = 1.2 and β_B = 0.9, match notional by that ratio (or use our portfolio hedge calculator). Equal-₹ sizing leaves directional exposure.
  • Exit at z = 0 OR at a stop. Statistical arbitrage has tail risk: pairs can stay dislocated for months during regime changes. Typical stop: z = ±3.5 from entry direction.
  • Avoid results week. Pair ratios dislocate sharply when one leg reports results alone. Check the event-positioning calendar before entering.

Related stat-arb tools

ASK AXEL

Which extended pair has the cleanest setup today?

Axel looks at today's high-z pairs, cross-references correlation + results calendar + 200-DMA direction of each leg, and ranks the candidates by cleanliness of the setup. Research framing only.

Ask Axel free

Research tool · not investment advice.

Axel Markets is an information + analytics product. We are not a SEBI-registered Research Analyst (RA) or Investment Adviser (IA). Nothing on this page is a buy, sell, or hold recommendation. Past performance is not indicative of future returns. Verify all data against the authoritative source (NSE, BSE, AMFI, SEBI, company RHP / factsheet) before acting. Ratios and z-scores computed from Yahoo Finance adjusted daily closes over the last ~6 months. Correlation uses log-returns within the same 60-day window. Statistical-arbitrage strategies have tail risk including prolonged dislocations and unlimited loss on the short leg if a leg rallies unexpectedly. Borrow/SLB availability in Indian cash market is limited; futures pair trading requires margin on both legs. Educational research only, not a trade recommendation.