Currency can provide significant diversification and added returns. However, the exposure is left unmanaged at your peril, says James Binny at Gartmore
It is now a well accepted fact of investment that, in order to enhance returns and reduce risk, diversification of investments is paramount. For European investors that diversification is disappearing as the reality of monetary union nears. EMU and the creation of the euro currency will lead to greater synchronisation of economic performance. The economies and markets of the participating countries will become more integrated, resulting falling diversification benefits of cross European border investments. Of course, any diversification benefit introduced by currency variations across the region will completely disappear. For many European investors entering equity markets for the first time, the competing desires to reduce risk through diversification while not venturing too far from home have led to investment across Europe, but frequently not too much further. If investors are to maximise their diversification benefits they will now need to look further afield. With that foreign investment comes the side-effect of currency exposure.
Currency set to become an increasingly important part of portfolio return
Global equity markets have shown spectacular returns over the past few years. There is a huge amount of debate, beyond the scope of this article, as to whether such returns can be maintained. One possible result of the tremendous run up in prices is that they now catch their breath, and stabilise at current levels. If this is the case, then, assuming currency volatility remains, the relative impact of foreign exchange returns will increase.
Over the last decade, the search for diversification has led to increasing overseas allocations by global investors. The trouble is, the more that investors have sought to exploit this attribute of international investment, the more the attribute has disappeared. As the markets around the world have become increasingly interconnected, so the correlation between them has increased. Research at the Groupe HEC School of Management in France1 showed that this is not so much that day-to-day returns are correlated, but that correlation tends to be higher at times of higher volatility, particularly during crashes. For instance, in 1987 all equity markets crashed at the same time and, in 1994 all bond markets crashed at the same time. The research also showed that volatility tends to be contagious between related markets, as was amply demonstrated in South East Asia in 1997. Therefore, just when diversification is most needed, it disappears. As a result, the diversification provided by foreign exchange becomes more important. It should be noted that the foreign exchange diversification will not decrease in the same way that asset market diversification has decreased. This is because foreign exchange is a medium of exchange not an asset - portfolio flows will travel both ways.
For these reasons, the relative impact of returns due to foreign exchange is increasing and is frequently greater than the impact of market or stock selection. In the past it had been possible to hide inadequate currency risk management behind the stellar performance of the underlying equities. However, if the returns generated by equities decrease, then the lack of a sensible currency management strategy will become increasingly apparent.
The diversification provided by the currency markets can be viewed positively, but that should not be the end of the process; currency risk is too important to be ignored. One reason given for inaction is frequently the complexity and number of decisions involved in the issue. Below, we attempt to break those decisions down into a more manageableseries of questions:
Manage currency exposure?
Unhedged or some degree of hedging
Static or active hedge?
Full hedge, static partial hedge, managed hedge
Currency managed separately?
Managed by underlying asset manager or specialist
Target of management?
Method of management
Variety of techniques and models
Fully hedged, unhedged, 50/50
Should foreign exchange exposure be managed?
Some investors assume that a nation's currency and its equity market tend to be negatively correlated. This view appeared validated by the Yen and the Nikkei between 1990 and 1995; the Nikkei index was falling but these losses were partially offset by the strength of the Yen. Many commentators suggested that this relationship was natural, and inaction was a clear way of reducing risk. However, from 1995 the Nikkei has continued to weaken, but so has the Yen. Currency exposure has actually added to rather than offset the loss. Therefore it would be dangerous to rely on currency to provide diversification in such a passive unhedged fashion. The annual returns of the Nikkei and Yen against the Deutschemark can be seen in graph 1. This shows the varying correlations of the two assets; it also shows that the return due to currency is frequently as large or larger than that due to equities.
Correlations between equity and currency returns are not only unstable, but are extremely difficult to predict. When investing overseas it is safest to assume a close to zero correlation.
There is a school of thought that currency is a zero sum game" - the impact of currency is small and tends to cancel itself out over the longer term. However, even if the currency moves do cancel out over time, the time horizon can be punishingly long. In between there could be multi year trends with substantial deviations from "fair value". If you had invested in U.S. securities in 1985 when the dollar was worth over 3.00 Deutschemarks (see graph 2) you would still be waiting for the negative currency move to be cancelled out, with the exchange rate still only at 1.80, despite the dollar's strength over the last two years.
Even if it is assumed that, over the long term, the currency impact will be neutral, the medium term risks of remaining unhedged could be significant. Following a decision to leave currency risk unhedged, consider two simplistic but possible scenarios:
Scenario 1: there are problems with the start of the euro (perhaps caused by political interference with the European Central Bank). As a result the euro depreciates sharply and the unhedged decision looks inspired, at least from a return standpoint, over the short term.
Scenario 2: it becomes increasingly clear that monetary union is proceeding successfully. The story of European restructuring is accepted by increasing numbers of international investors. As they rush to buy European assets, the portfolio flows drive up the euro. In addition, central banks reallocate reserves to include the euro at an allocation to take proper account of the euro and the relative size of the European economy. As a result the currency contribution to return is strongly negative.
A decision to leave the currency exposure unhedged is, therefore, effectively to take a bet on future events, in this case, on the success or otherwise of monetary union. The danger is that often no risk control is applied if the bet goes wrong. Prudent investors should address currency risk. Leaving currencies unhedged is an active decision.
An alternative decision would be to remove all risk of currency fluctuations by fully hedging the currency exposure. However, this would also remove all of the potential diversification benefit of foreign currency exposure. In addition, the hedges will generate cash flows. If the foreign currencies depreciate then losses on the foreign currency exposure will be offset by positive cash flow on the hedges. However, supposing the opposite occurred; then unrealised gains on the foreign currency exposure will be offset by losses on the hedges. These losses will have to befunded, possibly by selling some of the underlying stocks, thereby upsetting the underlying equity managers' strategy. This is not only a cash flow problem, it is also an opportunity cost; if it wasn't for the hedge, considerable gains would have accrued to the investors.
Static or active hedge?
The two extremes of fully hedged and unhedged exposure have drawbacks as detailed above. So perhaps a static hedge somewhere in between would be the answer - for instance 50% hedged/50% unhedged? However, investors will still lose if foreign currencies weaken, and will have to fund some large cash flows if foreign currencies strengthen. This is potentially the worst of both worlds.
In summary, all of the various static strategies have drawbacks. If the international portion of investors' assets is significant, we believe it becomes increasingly important to manage that currency exposure more actively.
Aside from the theoretical discussion, a recent survey by Brian Strange of Currency Performance Analytics2 analysed the actual performance impact of currency overlay programs on a total of 152 individual accounts, representing $40bn of currency exposure. This showed that, on average, currency overlay managers added 1.9% a year to international investment returns relative to a variety of static hedges. In addition, the increased returns were delivered with lower volatility than leaving the currency risk unmanaged. This resulted in an average information ratio (added value divided by volatility) of 0.5.
Our own experience is even better. In the 10 years that Gartmore has managed client currency exposure, we have added on average 2.3% per year excluding fees to U.S. clients' international performance with an information ratio of 0.9. Fees are typically 10 to 15 basis points.
To ignore currency overlay is not only to accept extra risk, it is also to miss out on an extra source of return. Currency overlay adds value.
Should currency be managed separately?
One possibility is for the manager of the underlying assets to manage the foreign exchange risk. Our experience is that fixed income fund managers are capable of managing the foreign exchange positions. The forces driving foreign exchange rates are the same as those driving bond markets. However, the skills required to manage equity portfolios tend to be very different - the currency decision is a distraction. Even if equity managers do place some hedges, they are unlikely to take significant risk relative to the benchmark; to do so would risk underperformance from an area where they feel less confidence. As such, most funds with significant overseas exposure would benefit from hiring a specialist currency overlay manager for at least the equity portion of their international exposure.
Target of currency overlay
Having decided to hire a specialist currency overlay manager, the next decision is which type of manager to hire. Historically the field of currency management has been divided into two opposing camps - forecasters and risk managers. The forecasters seek to predict the direction of exchange rates and adjust the hedges accordingly. Pure risk managers claim that it is not possible to forecast exchange rates as the market is too efficient. However, they claim that it is possible to forecast risk and only the risk should be managed. It should be noted that through the control of risk in the short term, positive returns relative to the benchmark can still be achieved in the medium term - it should not be a question of choosing between return generation or not.
At Gartmore we would claim that such a dichotomy is unduly simplistic. Our investment process starts from a risk control standpoint - the primary role of a currency overlay manager should be risk control. However, it is naive to believe that it is impossible to forecast exchange rate direction, at least to some extent.
Currency markets throw up many hazards and opportunities - our forecasting model enables these hazards to be negotiated and opportunities exploited so that the risk model can continue its work of achieving the long term goal more effectively. Forecasting can add value to the process by adding to the consistency of returns and avoiding unpleasant episodes.
Method of management
The next decision is the selection of managers. It is beyond the scope of this article to examine all the different techniques and models. However, there are certain properties of exchange rates which have changed as the foreign exchange market has evolved.
One of the main properties that managers seek to exploit is trending - if foreign exchange rates have risen recently, many models assume that they will continue to rise. Models that were developed in the 1980s and early 1990s are dependent on long-term currency trends. Such trends existed in the 1980s but are less evident in recent years as demonstrated in graph 2. While it is still true that foreign exchange markets trend more often than not, reflecting, among other things, the differing time horizon of investors, such long-term trends are much less evident than previously.
Trending is still the dominant environment. However, there are other environments which must be taken into account: sometimes the trends reverse; sometimes currencies wander up and down without any clear direction; sometimes currencies jump up or down suddenly and violently. Currency risk managers must be able to adapt to these changing environments. We have developed our adaptive risk model so as to prosper from the trending periods while still recognising that trends do come to an end and that there are periods with no trends at all .
The choice of benchmark depends on the client's perception of risk. If the greatest concern is that absolute losses are minimised, then a fully hedged benchmark is the most appropriate. If, however, cash flows or underperformance of unhedged currencies are of greater concern, then an unhedged benchmark is more appropriate.
In practice the decision is not so clear cut and a 50% unhedged/50% fully hedged benchmark may be the best compromise - a position of "least regret". Over the full currency cycle of five to 10 years a currency manager should expect to deliver similar returns for the range of benchmarks, and we would expect that return to be in excess of either extreme of fully hedged or unhedged.
The analysis by Currency Performance Analytics referred to earlier also showed that managers added more value when given a benchmark that was not at an extreme of fully hedged or unhedged. This is because managers are able to take positions either side of the benchmark. Therefore, in order to allow a manager to add value, a benchmark such as 50% hedged/50% unhedged would be appropriate. Such a benchmark is also a greater test for the manager as there are no periods when a static benchmark position is the best position that can be taken.
In summary: as monetary union becomes reality, in order to improve the risk and return characteristics of investments through diversification, investment outside Europe will become more necessary. With such investments come the associated currency exposures. Such exposures represent a risk which should be managed. However, they also offer the opportunity for further diversification and return generation. There are many issues to be resolved regarding such management. However, this should not be taken as an excuse for inaction and the issue can be broken down into more easily tractable questions.
1 "International market correlation and volatility", Solnik, Boucrelle, Le Fur, Financial Analysts Journal, September/ October 1996
2"Currency overlay managers show consistency", Pensions & Investments, June 15, 1998
James Binny is a senior investment manager at Gartmore Investment Management in London"