What is Correlation?

Indian market context (NSE)

Reference levels: Nifty 50 at 24,300, Reliance Industries at ₹1,300, Bank Nifty futures at 55,000 (lot size 30). Examples below show how Correlation shows up on Indian index, equity, and futures books — update to live quotes in your journal.

Nifty 50 perspective

Correlation on Nifty (24,300): define rupee risk per trade before the 9:15 open; index gaps on global cues can skip planned correlation levels — use exchange-supported stop types and size for gap beyond stop.

Reliance Industries perspective

Correlation for Reliance (₹1,300): stock circuits and 20% band limits can trap positions past your planned exit; keep correlation outside circuit freeze zones where possible.

Bank Nifty futures perspective

Correlation on Bank Nifty (55,000): span margin changes intraday — a valid correlation at entry may be too large after a margin hike; recheck buying power before adding lots.

How to validate

• Minimum sample: 30 closed trades on one strategy tag before trusting Correlation.
• Check for one outlier week inflating Correlation — export largest winners and losers.
• Recompute Correlation after including brokerage, STT, and slippage on F&O tags.
• Compare Correlation on the same date range as profit factor and max drawdown.

How to track in TradeLyser

• Open Strategy Board or analytics → filter by strategy tag and review period.
• Locate the widget or column reporting Correlation (or export trades to compute manually).
• Store snapshot values in weekly review: Correlation, profit factor, drawdown, trade count.
• If Correlation is custom, add a spreadsheet column fed from TradeLyser CSV export.

Best practices

• Publish Correlation per strategy, not only at account level.
• Use the same calculation window (weekly vs monthly) year-round.
• Pair Correlation with sample size in every review slide or note.
• Reconcile Correlation with broker statements before tax filing.

Common pitfalls

• Changing rules after fewer than 20 trades because Correlation moved slightly.
• Mixing intraday and positional tags when computing Correlation.
• Ignoring costs so Correlation looks better than banked P&L.
• Letting one outlier trade dominate the Correlation reading.

How to use this in TradeLyser

Log correlation assumption on pairs tags; review divergence tail events quarterly.