Integrated dynamic models for hedging international portfolio risks

Abstract

We develop scenario-based stochastic programming models for hedging the risks of international portfolios using options. The models provide increasing level of integration in managing market and foreign exchange (FX) risks. We start with a single-stage model with currency options for selective hedging of FX risks, while market risk is addressed through diversification, we add stock options to hedge market risks, and add quantos and currency options to develop an integrated model, using an innovative method to price quantos on the scenario tree underpinning the stochastic program. The models are extended to multi-stage settings. Backtesting on market data over a 14-year period that includes the global financial crisis of 2008, we demonstrate the effectiveness of taking increasingly integrated views of risk management. Simultaneous hedging of market and FX risks using stock and currency options has the best ex post performance. The differences are economically significant, and statistical significance is established through rigorous hypothesis testing. The models are particularly effective during the crisis. Test results show that two-stage models outperform their single-stage counterparts, regardless of the hedging strategy.