Studies On Futures Hedging

In this thesis, we mainly provide an extension on the futures hedging strategy proposed by Barbi, Romagoli (2014) and investigate whether this extension could result in better hedging performance.Barbi, Romagoli (2014) presented a futures hedging strategy, where we could get the optimal hedge ratio by minimizing the quantile risk measure (QRM) of the hedged portfolio. It dominates the existing hedging strategies with two main advantages:As a generic risk measure, QRM could include Vaule-at-Risk (VaR), Conditional VaR-at-Risk (CVaR) and other risk-averse-related risk measures, which makes the strategy applicable for different risk measures and investors. Moreover, copula is incorporated in the strategy to reflect the dependence between spot and futures. Under this strategy, the quantiles of the hedged portfolio could be represented as the function with respect to this copula. According to the empirical study, Barbi, Romagoli (2014) has proved that the hedging performance of this strategy is in general superior compared to others. However, the model suggests that we use a static copula to fit observations during a long period. In this way, it would be harder to capture the dynamic spot-futures dependence structure and thus leads to lower hedging effectiveness. Motivated by this drawback, we expanded this future hedging strategy by incorporating a regime-switching copula to capture the dynamics of spot-futures dependence.Our studies start with the description of the model and implementation of the hedging strategy in Barbi, Romagoli (2014). We use Monte Carlo simulation to test the sensitivity of the hedging effectiveness to different factors. It is shown that the spot-futures dependence and investors’ risk aversion levels could have great impact on the hedging performance. Monte Carlo simulation is also applied to the original hedging strategy and our extended hedging strategy to compare the hedging performance. The extended hedging strategy based on a regime-switching copula proves to be more accurate to capture the dynamic dependence between spot and futures, but dominates in the hedging effectiveness only when there are significant regime shifts in the spot-futures dependence. Finally, we apply both of the two strategies to our empirical data and compare their hedging performance. The results from FTSE100 stock index and its corresponding futures contract reconfirm our conclusions getting from the simulated data.