Choosing Hedged or Unhedged Returns for Global Equity Models
Summary
The document asks whether a global tactical equity allocation model should compare country equity returns in local currency, currency-hedged form, or unhedged form. The model may already include currency positions, so the choice concerns which return series best supports prediction rather than the investor’s final implementation alone. Inputs mentioned include macroeconomic data, yield curves and differentials, valuation, and momentum.
The responses offer two perspectives. One recommends studying research on optimizing currency risk and reward in international equity portfolios. Another argues that either representation can work in principle when currencies are modeled as assets in their own right, and suggests choosing the series that aligns with the model’s intended interpretation of its weights. The thread does not compare predictive performance empirically or prescribe a universally superior series. Its central practical point is to align the return definition with how currency exposure will be represented and interpreted in the model.
Key ideas
- The choice between hedged and unhedged returns affects how country equity performance is represented.
- If currencies are modeled as separate assets, either return convention may be usable in principle.
- Return definitions should align with how the model represents currency exposure and portfolio weights.
- Research on currency risk and reward in international portfolios can inform the choice.
- The document provides no empirical test establishing which series predicts returns better.
Tags
Full text
# Should I use currency hedged or unhedged returns for a global equity allocation model? # Should I use currency hedged or unhedged returns for a global equity allocation model? I am building a global tactical equity allocation model. The model will help determine an optimal allocation amongst a number of major developed and emerging stock markets (represented for my purposes by the MSCI country indices). The model will use inputs such as macroeconomic indicators, local currency yield curves and differentials, valuation (e.g. P/E), momentum, etc. Question: Should the returns series input into the model be currency hedged or unhedged? In practice, we will have the option of making separate currency bets. Is the fundamental series which should be compared across countries the hedged or unhedged series? To clarify, the goal in making this decision is to maximize our predictive power. For example, if I were building an options relative value model, I would certainly compare only delta-hedged returns, even though my actual hedging may differ I practice. in this case, however, adding currency hedging may actually be adding noise rather than reducing it. ## Answer by ZAxisMapping (score 5, accepted) https://quant.stackexchange.com/a/1695 Check out: "Universal hedging: Optimizing currency risk and reward in international equity portfolios," Fischer Black - Financial Analysts Journal, 1989. as well as many of the subsequent research that references this article (via Google Scholar, for instance). Good luck. ## Answer by Brian B (score 0) https://quant.stackexchange.com/a/1669 Since currencies are explicitly part of your asset set, it does not matter in principle which choice you make (currency hedged or unhedged) for the other securities. In order for you model weights to have the most intuitive meaning, you should choose unhedged if you think you will generally be neglecting to hedge the risk, and hedged otherwise.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.