Correlation
Correlation is a statistical measure ranging from -1 to +1 that quantifies the degree to which two assets' returns move together, with +1 indicating perfect co-movement, -1 indicating perfect opposite movement, and 0 indicating no linear relationship.
Correlation was one of the two key inputs — alongside individual asset variances — that determined portfolio risk in Modern Portfolio Theory. The mathematical symbol was ρ (rho), and it was calculated as the covariance of two assets' returns divided by the product of their standard deviations. The resulting dimensionless number between -1 and +1 was much more interpretable than covariance because it was scaled to a standardised range.
In portfolio construction, the goal was to combine assets with lower (and ideally negative) correlations to harvest the maximum diversification benefit. If two assets had a correlation of +1.0, holding both produced no reduction in portfolio volatility — they moved in lockstep. A correlation of 0 meant the assets moved independently, and combining them reduced portfolio variance proportionally to the square root relationship. A correlation of -1.0 was the theoretical ideal, allowing complete cancellation of risk through the right weighting — a situation approximated by traditional long-short hedge fund strategies.
In Indian markets, important correlation observations included: Nifty and Sensex were almost perfectly correlated (above 0.98) given overlapping constituents; large-cap IT indices and the broader Nifty had moderate positive correlation (around 0.6 to 0.7), meaning IT stocks provided modest but not substantial diversification within domestic equity; gold and Nifty exhibited near-zero to slightly negative correlation over long periods, making gold a valuable diversifier; and Indian equity versus US equity (S&P 500) showed moderate correlation that increased during global risk-off events, reducing the diversification benefit of international equity precisely when it was most needed.
Correlations were not stable over time, which created a significant challenge for portfolio managers relying on historical data. During market crises, correlations between risky assets typically spiked — equities in different sectors, which normally showed modest correlations, began moving together as investors reduced risk indiscriminately. This correlation breakdown undermined the diversification assumptions of static MPT-based portfolios and motivated the use of dynamic correlation models, conditional correlation estimates, and stress-testing frameworks in institutional portfolio management.
SEBI's investment adviser regulations required that registered advisers consider correlation and diversification principles when recommending portfolios to clients. The growing adoption of quantitative tools among Indian wealth managers and robo-advisory platforms incorporated correlation matrices to ensure that model portfolios were genuinely diversified rather than superficially so — holding multiple funds with similar underlying portfolios was a common source of false diversification that correlation analysis could readily identify.