Quantile connectedness between Indian sectoral equities, gold, crude oil and the rupee: Time-varying spillovers and portfolio implications.

Abstract

We study return connectedness among the Nifty 50 and seven sectoral indices, gold, Brent crude oil and the US dollar–Indian rupee exchange rate for 2 January 2013 to 25 June 2026 (inclusive), a sample period which covers demonetisation, COVID-19 pandemic, Russia–Ukraine war and tariff shock in year 2025. Using the framework of the time-varying parameter VAR (TVP-VAR) and its quantile extension, we show three main results. As shown, average total connectedness over the sample period is 53.2% and very much event driven: it peaks at 77.9% in response to COVID-19. In particular, equity sectors—led by automobiles, banking and energy—act as net shock transmitters while the rupee is the system's primary shock absorber and more mild net recipients are gold and crude oil. Second, connectedness of state variables is considerably stronger in the tails of the return distribution: total connectedness reaches 86.2% at the fifth percentile and 85.8% at the ninety-fifth percentile, nearly 35 pp above their medians (about {0}). Therefore mean-based spillover estimates significantly underestimate systemic linkages in the market states that matter most for risk management. Third, according to empirical estimates using the dynamic conditional correlation (DCC-GARCH) model, gold behaves as a diversifier rather than a hedge for portfolios—variance reductions of 46–71%, depending upon particular equity–gold cases—and short rupee (long US dollar) positions are the most effective direct short hedges against Indian equity exposure. The findings are robust to alternative forecasting horizons and forgetting factors, as well as the exclusion of the aggregate market index, and carry direct implications for investors, portfolio managers, and policymakers in emerging markets..