Is gold an effective hedge and/or safe haven instrument against stocks, rupee–dollar rate and crude: an empirical investigation from India

Abstract

The current study tries to empirically investigate the hedging and safe haven characteristics of gold against stock markets, crude and foreign exchange movement for India for the period 1 April 2008–31 March 2019 by taking daily closing prices obtained from www.investing.com (converted to returns). The methodology includes using both traditional OLS and quantile regression approaches by setting up both linear and nonlinear relationships amongst the assets. A separate section in the paper has been devoted which covers the Sub-Prime Crisis of 2008. The results of the study showed that gold acted as a hedge asset with respect to Nifty Stock Index and Exchange Rate but not against Crude when the OLS linear and nonlinear models were employed. Further both linear model and nonlinear OLS regression models ruled out the safe haven characteristics of gold with respect to any of the three assets. However, for quantile regression, hedge characteristics of gold were established at only few quantiles with respect to the same two assets, i.e. Nifty and foreign exchange. With respect to safe haven characteristics of gold under OLS, for both linear as well as nonlinear models, gold did not seem to act as a safe haven, while for quantile models, gold acted as a safe haven only at lower tail(s) for Exchange Rate in the linear model and against Nifty Stock Index for nonlinear model. This shows that quantile regression was able to capture the hedging and safe haven characteristics of gold in a superior manner by doing a comprehensive analysis which was based on eleven different quantiles. Such an analysis is much superior to the traditional OLS which gives only an overall broad assessment of the gold’s hedging capacity. Further, the behaviour of gold return as an asset during the sub-period crisis of 2008 was not substantially different from its behaviour during other periods; however, when the yardstick of measurement was changed to return volatility of gold, the behaviour was found to be different from other periods. The asymmetric impact of return on gold with respect to all other assets was also proved in our study.