Key Takeaways
- • Amundi Asset Management researchers (Elisa Baku, Roberta Fortes, Karine Hervé, Edmond Lezmi, Hassan Malongo, Thierry Roncalli, Jiali Xu) argue in a June 2019 paper that carry, value and momentum risk factors can explain a significant share of time-series and cross-sectional currency returns.
- • The authors propose a Fama-French-Carhart style risk factor model for currencies, built on carry (uncovered interest rate parity), value (equilibrium real exchange rate models) and momentum (technical analysis and over/underreaction patterns).
- • Market-based factors are favored over monthly or quarterly economic data because they are observed daily and are synchronous across regions, which makes APT-style modeling of currency returns more tractable.
- • The paper highlights three practical applications for forex portfolio management: basket hedging, overlay management and the construction of alpha strategies.
Summary
A June 2019 research paper from Amundi Asset Management in Paris examines whether factor investing, which has reshaped equity and fixed income portfolio construction, can be applied to currency markets. The authors, Elisa Baku, Roberta Fortes (PhD candidate, University of Paris 1), Karine Hervé, Edmond Lezmi, Hassan Malongo, Thierry Roncalli and Jiali Xu, frame the question against the established use of risk factors in equities, where size, value, low beta, momentum and quality are now standard, and in multi-asset alternative risk premia strategies.
The authors argue that currencies are a special case. Unlike stocks and bonds, currencies are not treated as a traditional risk premium, institutions do not set a strategic currency allocation, and a foreign exchange rate is not measured in units in the way equities or indices are. As a result, the paper notes, the notion of market gains and losses is less straightforward, since one investor's loss corresponds to another's gain.
The paper contends that the leading economic models of exchange rates, which are often presented as competing, are in fact complementary. Its central claim is that these models can serve as the basis of a Fama-French-Carhart factor model for currencies. The authors choose three market-based factors: carry, derived from uncovered interest rate parity; value, derived from equilibrium models of the real exchange rate; and momentum, reflecting technical analysis, trading behavior and over- or underreaction. The authors say these factors can explain a significant part of both time-series and cross-sectional currency returns.
The authors justify using market factors over traditional economic variables such as monetary policy, inflation, trade and capital flows. Economic data are published monthly or quarterly and not synchronously across countries, which complicates an arbitrage pricing theory (APT) framework. Market factors are observed daily and, under the assumption of market efficiency, prices already reflect economic information.
The applications the authors cite are basket hedging, overlay management and alpha strategy construction. The paper's stated aim is to improve understanding of forex portfolio management rather than to set out a specific trading product.
The authors argue that currencies are a special case. Unlike stocks and bonds, currencies are not treated as a traditional risk premium, institutions do not set a strategic currency allocation, and a foreign exchange rate is not measured in units in the way equities or indices are. As a result, the paper notes, the notion of market gains and losses is less straightforward, since one investor's loss corresponds to another's gain.
The paper contends that the leading economic models of exchange rates, which are often presented as competing, are in fact complementary. Its central claim is that these models can serve as the basis of a Fama-French-Carhart factor model for currencies. The authors choose three market-based factors: carry, derived from uncovered interest rate parity; value, derived from equilibrium models of the real exchange rate; and momentum, reflecting technical analysis, trading behavior and over- or underreaction. The authors say these factors can explain a significant part of both time-series and cross-sectional currency returns.
The authors justify using market factors over traditional economic variables such as monetary policy, inflation, trade and capital flows. Economic data are published monthly or quarterly and not synchronously across countries, which complicates an arbitrage pricing theory (APT) framework. Market factors are observed daily and, under the assumption of market efficiency, prices already reflect economic information.
The applications the authors cite are basket hedging, overlay management and alpha strategy construction. The paper's stated aim is to improve understanding of forex portfolio management rather than to set out a specific trading product.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.