Introducing a new five-factor model of the global, top-down determinants of foreign-exchange returns.

In most strategic and tactical asset allocation processes, foreign-exchange exposures—especially emerging markets foreign exchange exposures—are implicit and go unconsidered by investors. At Neuberger Berman foreign exchange has always been considered an explicit alpha source that plays an important role in an effective asset allocation process. Nonetheless, because we assess each currency using bottom-up country-by-country indicators, we have long recognized that our allocation process may be missing the top-down factors behind the performance of individual currencies. Moreover, the higher correlations among emerging markets currencies since the financial crisis of 2008–09 indicate the growing importance of these top-down factors. In this paper we propose an intuitive top-down, five-factor model of foreign exchange returns. We show that this model can be used to separate systematic from genuinely idiosyncratic return drivers in foreign exchange markets, and also to build a simple, systematic long-short strategy that would have substantially outperformed the average EM or DM currency market return over the past 10–15 years.

Executive Summary

  • In most strategic and tactical asset allocation processes, foreign-exchange exposures—especially emerging markets foreign exchange exposures—are implicit and go unconsidered by investors.
  • Many investors that do consider foreign exchange exposures explicitly do so bottom-up and country-by-country despite the fact that higher correlations among emerging markets currencies since the financial crisis of 2008–09 indicate the growing importance of topdown factors.
  • In this paper we propose a top-down, five-factor model of foreign exchange returns, for which the following factors were selected to create a parsimonious model that could nonetheless explain a large systematic part of exchange rate movements in an economically intuitive way:
    • The market factor
    • A carry factor in emerging market currencies
    • A carry factor in developed market currencies
    • A factor based on dependence on crude oil imports among emerging economies
    • A factor based on the spread between emerging and developed currencies
  • We show that this model can be used to separate systematic from genuinely idiosyncratic return drivers in foreign exchange markets, and also to build a simple, systematic long-short strategy that would have substantially outperformed the average emerging or developed currency market return over the past 10–15 years.

The five-factor model would have identified undervalued and overvalued EM currencies over recent years

Performance of 20 EM currencies versus the top and bottom three, as ranked by the Z-scores of the idiosyncratic terms generated by the global five-factor model, 2007 – 2018

Source: Neuberger Berman. The five-factor model is estimated once a month, and Z-scores were calculated using a window of 60 (i.e. five years’) observations at the start of the backtest, extended with the latest available observations thereafter.
Past performance is no guarantee of future results.
Note: When the long-short strategy was implemented using the Z-score threshold of one, there were periods when three currencies did not exceed that threshold at the top and/or bottom of the rankings; in these instances, the portfolio was constructed with however many currencies did cross that threshold; when no currencies crossed that threshold, the existing positions were maintained to limit trading costs.