Estimation of time varying risk premia for the Nikkei 225 stock index futures contracts

The purpose of this research is to examine the intertemporal relation between risk and expected equity returns. In particular, I investigate whether the excess holding period return on the Nikkei 225 stock index, defined as the expected return on a stock index minus the risk-free interest rate, is positively related to risk as measured by the conditional volatility of the Nikkei stock index returns.;If the degree of uncertainty in asset returns varies over time, the compensation (or risk premium) required by risk averse investors for holding risky assets must be varying. This is called time varying risk premia (TVRP). To derive the TVRP, a simple index arbitrage in the stock index futures market is employed for the equilibrium value of the intertemporal asset pricing model which was modified to allow for risk aversion. The time varying risk premia are applied to the hedging model. Generalized Autoregressive Conditional Heteroscedasticity in mean (GARCH-M) models of excess returns on the Nikkei 225 stock index are used to estimate the time varying conditional variances and a mean-variance ratio that represents the risk-return trade off. In the ARCH model, the nature of heteroskedastic error terms over time is captured by the conditional variance, which depend upon past available information. Thus, the conditional variance is an useful variable for measuring uncertainty and provides a model of the TVRP.;The empirical results show that there was strong evidence of both ARCH and GARCH effects and the time varying risk premia, as reflected in heteroscedastic error terms over time, using data on spot and futures prices of the Nikkei 225. This suggests that an investor's risk premium changes over time. Another finding is the persistence of shocks to the Nikkei 225 stock index returns volatility. My results also show a negative relationship between excess returns on the Nikkei 225 stock index returns and conditional volatility using the GARCH-M model. The findings of this paper support that an increase in riskiness can either increase or decrease the stock index returns.;The specification tests demonstrate that the expected returns on the Nikkei 225 stock index can be improved by adding its own lagged endogenous variable and the first order of moving average which contain important information.