This paper considers the strategic currency hedging optimization problem for institutional investors with globally diversified portfolios. We argue the currency hedging decision should be made at the portfolio level on a currency-by-currency basis. It is suboptimal to develop hedging strategies by asset or to impose uniform or asset-specific hedging constraints at the portfolio level. Extant research on currency hedging usually confines the analysis to single asset classes rather than multi-asset portfolios (see, for example, Perold and Schulman (1988), Campbell, Serfaty-de Medeiros and Viceira (2010), Schmittmann (2010)). As a result, the recommended currency hedging strategies based on these analyses are often asset-specific.1 Some studies extend the analysis to the portfolio level but only consider uniform hedge ratios (Eun and Resnick (1988); Liu and Jacobsen (2014)). These approaches are also popular among practitioners globally.2 We quantify the potential efficiency loss of these common approaches relative to the recommended one. In our sample optimizations for Australian and Japanese investors, adopting currency-specific hedge ratios can reduce conditional value at risk (CVaR) by up to 1.5% and 3%, respectively, without reducing the expected return of the portfolio.

Download PDF