Using Macroeconomic Forecasts to Improve Currency Mean-Reversion Signals

Summary

The document describes a mean-reversion approach that begins with yield-curve trading strategies and extends the idea to a multi-pair strategy involving major currency pairs. To refine its trading signals, the strategy incorporates machine-learning forecasts of relevant macroeconomic variables. The forecast information is used to optimize the weights assigned to the signals, adding macroeconomic inputs alongside technical ones.

The reported evaluation shows an improvement in annual percentage return over the evaluation period after incorporating the forecasts. The document gives no specific return figures, benchmark comparison, forecast variables, training design, or details about costs and risk. It therefore illustrates a way to combine macroeconomic forecasts with a mean-reversion strategy, but provides too little information to assess the size, robustness, or out-of-sample reliability of the reported improvement.

Key ideas

• The strategy applies yield-curve mean-reversion ideas to a set of major currency pairs.
• Machine-learning forecasts of relevant macroeconomic variables are incorporated into the trading signals.
• The forecast inputs are used to optimize signal weights.
• The document reports improved annual percentage return but omits details needed to assess robustness.

Tags

# Using Macroeconomic Forecasts to Improve Mean Reverting Trading Strategies


# Using Macroeconomic Forecasts to Improve Mean Reverting Trading Strategies









A large class of trading strategies focus on opportunities offered by the yield curve. In particular, a set of yield curve trading strategies are based on the view that the yield curve mean-reverts. Based on these strategies' positive performance, a multiple pairs trading strategy on major currency pairs was implemented. To improve the algorithm's performance, machine learning forecasts of a series of pertinent macroeconomic variables were factored in, by optimizing the weights of the trading signals. This resulted in a clear improvement in the APR over the evaluation period, demonstrating that macroeconomic indicators, not only technical indicators, should be considered in trading strategies.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.