A second popular strategy for active currency managers is the “trend-following” strategy, which involves using several technical moving averages to provide trading or hedging signals. Active currency managers can use this strategy to either trade around their benchmark in order to add alpha, or alternatively to provide a hedging signal. The academic backing for trend-following strategies is as deep as that for the differential forward strategy, including works by Bilson (1990, 1993), Taylor (1990), LeBaron (1991) and Levich and Thomas (1993), which showed that these strategies can indeed produce consistent excess returns over sustained periods of time. I would suggest however that the seminal breakthrough in this area came in the form of the note by Lequeux and Acar (1998), which gave the strategy more specific properties by suggesting that in order to be representative of the various durations followed by investors an equally weighted portfolio based on three moving averages of 32, 61 and 117 days would be most appropriate. Simply put, the core idea behind this strategy is to go long the currency pair when the price is above a moving average of a given length and to go short the currency pair when it is below. More specifically, if the spot exchange rate is above all three moving averages, hedge the foreign currency exposure 100%. If it is above only two out of the three moving averages, hedge one-third of the position. In all other cases, leave the position unhedged.
Archive for the ‘Strategies’ Category
Trend-Following Strategy
July 3rd, 2009Differential Forward Strategy
July 2nd, 2009The core idea behind this is that of “forward rate bias”, or the reality that forward rates are poor predictors of future spot exchange rates, in contrast to the theories of covered interest rate parity and unbiased forward parity. We have looked at some of the academic backing for this admission earlier in this blog, notably by Fama (1984), Kritzman (1993) and finally Bansal and Dahlquist (2000), who suggested that the negative correlation presented by Fama between future exchange rate changes and current interest rate differentials is crucially linked to changes in macroeconomic variables.
As outlined by Acar and Maitra (2000), the differential forward strategy seeks to take advantage of the apparent market inefficiencies as represented by “forward rate bias” by hedging the currency risk only when the interest rate differential is in favour of the hedger. That is to say, only when the forward points are at a discount. Conversely, the currency manager should not hedge currency risk when the forward points are at a premium and consequently the interest rate differential would reflect a cost. More specifically, when the interest rate differential pays the investor to hedge, the currency manager should have a hedge ratio of 100%. Conversely, when the interest rate differential costs the investor to hedge, the hedge ratio should be zero. Thus, if the currency manager is operating under a symmetrical benchmark, the manager would go overweight the hedge by 50% when the interest rate differential is in their favour and underweight by 50% when it represents a cost.
The results of this strategy have proved to be extremely robust and have been tested across some 91 currency pairs.
Of necessity, when the interest rate differential is favourable, the differential forward strategy will have the same returns as a full forward hedge. Equally, when the interest rate differential represents a cost, the differential forward strategy will have the same returns as an unhedged strategy. Thus, overall, the returns of the differential forward strategy will be a function of both fully hedged and fully unhedged strategies. The advantages of such a strategy are the following:
As established, it has consistently added alpha for active currency managers.
Equally, it has also reduced risk relative to benchmarks.
The strategy combines the decisiveness of a full hedge with significant flexibility when used with a symmetrical benchmark.
The expected returns of the differential forward strategy are a function both of the expected returns of the fully hedged and fully unhedged strategies.
Given that the differential forward strategy is based on exploiting the principle of “forward rate bias”, it must follow to an extent that its expected returns are also a function in turn of the extent of that forward rate bias and thus of the interest rate differential relative to the expected future interest rate differential. For any given interest rate differential, the hedging strategy will perform best when the correlation between the hedged and unhedged returns is more negative.