Extensive empirical evidence documents relatively consistent if modest predictability in excess bond returns and excess currency returns.
The theory that active investors’ ability to add value on a risk-adjusted basis is proportional to their forecasting skill motivates the authors to extend empirical forecasting models to new sorts of trades-curve steepness positioning, and cross-country spread trading.
The authors test the performance of increasingly complex trading strategies between 1992 and 2002, from using single indicators to predict specific trades, to pooling indicators into one forecasting model for each trade, and then to diversify across several trades.
The broad composites produce the best risk-adjusted performance. One may gain an edge from the limited ability to forecast returns and magnify this edge through diversification across strategies.