The global asset allocation, which has been advocated by academic scholars and investment practitioners for the last two decades, produces a puzzling effect on the portfolios of pension plans. The puzzle stems from the fact that international assets offer ex-post diversification benefits to institutional portfolios while introducing an additional source of risk, namely due to fluctuations in the foreign exchange rates. Nevertheless, until the early 1990s, the impact of currency movements on the performance of the total portfolio has often been underestimated or neglected. The possible reasons for this behaviour could be summarised as follows:
q Merits of active management have not been demonstrated effectively, particularly the opportunity to add value by reducing the risk due to short-term volatility.
q Currency movements have not been adverse enough to raise a flag at the total fund level.
q A small number of plans had significant allocations to international assets.
q Large investment management firms did not have enough incentive to document the importance of the volatility arising from currency movements in an attempt to push the overlay business, which is not known for generating serious dollars due to its low profit margins.
The volatility arising from investments in securities denominated in foreign currencies has a negative effect on risk-adjusted performance measures such as the Sharpe ratio. This is mainly due to the increase in the overall dispersion of the returns of the total fund because of the currency exposure, which minimises the impact of the diversification gain produced by mixing domestic securities with international ones. Given the fact that currencies have little or no correlation with other assets in the portfolio, one would argue that leaving the currency exposure unhedged would reduce the risk of the total portfolio due to the diversifying effect of currency positions. Moreover, actuarial smoothing is believed to work in favor of leaving the currency exposure unhedged given a positive correlation between foreign currency return and inflation in the US. In this case, foreign currency could be viewed as a good diversifier in the asset-liability context, provided that the forward rate is an unbiased predictor of the future spot rate.
Two factors have caused pension plans to take an active view toward the currency exposure of the total fund: (a) the empirical evidence of the added volatility arising from international allocations, and (b) inefficiencies and trending nature of the foreign exchange markets. Since institutional investors realised the relevance of currency hedging on the overall fund management process, different approaches have been utilised to address the issue within the guidelines of the investment policy. Some plan sponsors did not like the fact that currencies, in the long-term, do not provide any ex-ante return, but increase the volatility of the investment portfolio. This observation brought along the risk-controlled currency overlay product, which attracted a great deal of interest from volatility-sensitive plan sponsors. More recently, an increasing number of pension plan sponsors are beginning to discover the value-added potential of active currency management, joining the more sophisticated ones that have been instrumental in making a case for the active currency hedging decision by institutional investors.
The success of the currency overlay universe is well documented in Watson Wyatt’s currency overlay manager database. Through December of 2000, overlay managers have posted impressive information ratios. US dollar-based accounts with unhedged mandates, on average, posted information ratios in the range of over 0.5, which is very competitive compared to the ones witnessed in other asset classes (see Figure 1).
Moreover, the average annualised information ratios, which are computed using a time-weighted scheme since the inception of each account, are 0.82 and 0.60 for the unhedged and 50% hedged benchmarks, respectively. Given the strength of the US dollar against EAFE-weighted currencies during the last five year period ending December of 2000, it should not come as a surprise that currency managers, on average, provided higher information ratios in unhedged benchmarks versus the 50% hedged. In addition to the empirical observation regarding the foreign exchange movements, one could also argue using the forward rate premium bias literature that hedging more of the low-interest rate currency exposure would result in the US-dollar based accounts with unhedged EAFE mandates outperforming their 50% hedged counterparts as far as excess returns are concerned.
Our conclusions regarding the account and benchmark returns reported as of the end of 2000 could be summarised as follows:
q There is strong empirical evidence that skillful management of currency exposure can produce attractive alphas and portfolio information ratios.
q Separation of security/country selection and currency exposure management improves the financial efficiency of international equity portfolios.
q The information diversification as it pertains to the sources of manager insight in currency movements is the key element for building efficient manager structures.
The two-step procedure is preferable to no hedging at all as the efficient frontier analysis has shown that the difference between optimal one-step and optimal two-step solutions differ only in the risk taking behavior of the client. The two-step procedure of selecting securities and countries first, and then letting an expert handle the currency exposure from both a risk-controlling and alpha extraction perspective seems to be a more prudent and realistic option for institutional investors due to following reasons:
q the ability of most currency managers to generate excess returns irrespective of the success of active stock and bond managers,
q the appeal of separate performance attribution for the currency component from the governance perspective, and
q the fact that currency return correlations with other asset returns are less stable than stock-bond correlations.
We recommend that the determination of the neutral (strategic) hedging position should be performed via the full-blown optimisation procedure, which includes assets and liabilities of the client as well as specific risk and operational preferences. To this end, one should refrain from capturing any active element in the management of the currency exposure. The overwhelming evidence of the non-specialist’s ability to forecast currency movements makes us uncomfortable to tackle the issue of return predictability of currencies at the policy level.
Behavioural factors that might affect the long-term consistency of the hedging policy, such as the issues of regret minimisation, ought to be considered in the beginning of the process, and no attempts should be made to (tactically) time the best ‘strategic’ hedge ratio.
The long-term characteristics of foreign exchange markets should be addressed in a risk budgeting context, and active risk be allocated to an optimal group of overlay managers by the currency consultant.
For pension plans that have already hired international equity and fixed income managers without any currency hedging guideline in place and have recently decided to add an currency overlay programme face a tough challenge as it relates to the expertise of current international equity and fixed income managers in the currency markets. We feel the separation of duties between stock/country/sector selection and management of currency exposure is very appropriate and intuitively appealing for institutional investors.
Overall, the utilisation of the existing international managers to manage both the underlying foreign securities and the associated currency exposures does not have the potential of improving the financial efficiency as an effectively designed overlay structure, which would be accomplished with two or three managers. Currency overlay managers bring a highly focused level of expertise to the management of currencies as they concentrate their time and energy on the foreign exchange markets that are known to possess unique dynamics compared to equity and fixed income markets.
We believe that the sophistication offered by overlay managers in terms of the design of creative investment processes as well as the efficient use of a full range of currency instruments could not be overemphasised enough. To this end, however, pension plans should be careful that the ex-ante return potential and risk reduction are not more than offset by increased management fees.
As currency managers actively aim at finding uncorrelated alpha sources in the quantitative investment process, they usually end up with ‘hybrid’ structures that are designed by picking a group of fundamental and technical indicators, which provide the best diversification benefit to the currency manager. The traditional and naïve approach of picking one manager from each style and allocating equally between managers could be unrealistic due to the disappearance of managers that bet on one particular style.
More importantly, the ‘hybrid’ manager might engage in timing of styles by allocating more weight to technical indicators that are known to outperform the fundamental approach in trending markets. Going forward, it is only natural to think that more managers will attempt to capture the time-dependence nature of capital market returns and volatilities, and thus, utilise advanced statistical and econometric techniques to ‘seize the moment’.

We also do not want to take the a priori view that the so-called return-enhancing strategies, which include fundamental, technical and hybrid styles, and risk-controlling approaches are complementary in an overlay programme. There is no scientific reason to believe that picking one of the best in each strategy would produce better results than two carefully selected return-enhancing managers with diversifying investment processes. We feel that various styles that are used by overlay managers in the management of currency exposure should not be determined beforehand. To the contrary, every manager should be included in the traditional mean-variance optimisation, which helps us further refine the optimal weights to reflect the issue of management fees. In an environment, where individual styles are continuously augmented by one of the other overlay styles, the lure of style diversification disappears. Moreover, the multi-layered approach to portfolio management is inefficient due to the limited number of overlay managers. That is why the burden is on the currency consultant to make sure the client preferences are properly evaluated, and an effective combination of managers is chosen to provide the highest probability to outperform the selected strategic benchmark net of management fees along with volatility reduction.
In this study, monthly excess returns of nine managers are used to draw the efficient frontier for the 1996-2000 period in 50% hedged $US-based mandates (see Figure 2). Each manager is constrained not to have an allocation higher than 50% in the custom portfolios. The number of managers is limited to three with the exception of custom portfolio #2 (see Figure 3) to build currency overlay structures that are easy to manage and monitor. Using fewer managers results in less efficient portfolios, which illustrates the necessity of diversification among currency overlay managers.
If there were only one quantitative or judgmental methodology dictating the process of active hedge ratio management, the success of the currency overlay programme would be in jeopardy. Moreover, this would also lead to institutional investors second-guessing the initial currency hedging decision, and more importantly, ending up with preference reversals. Annualised information ratios of custom portfolios range from 1.31 to 1.67, which is remarkably high in the management of institutional funds, and a significant improvement over the single-manager alternative. Figure 2 and 4 show the importance of effective combination of overlay managers to improve the efficiency of the portfolio in the risk-return space. One could achieve high information ratios at every level of active risk with the added benefit of minimising the probability of monthly cash outflows that might be detrimental to the success of the overlay programme. The efficiency of custom portfolios not only increases the range of information ratios but also significantly lowers the number of large and small returns.
Overall, custom portfolio #3 and #4 do not have any outliers and provide the best information ratio. They could be viewed as the apparent winners among the six portfolios designed with the help of the mean-variance optimiser. Moreover, the financial efficiency of the programme as depicted by the information ratio decreases as one increases the amount of active risk beyond a certain level, illustrating the notion that there is an optimal level of tracking error in the management of the currency exposure
To summarise, we want to emphasize the importance of capturing opportunities that are at the moment not used by the plan sponsor universe to lower the volatility of the total fund while increasing the probability of achieving higher returns. It is our view that setting up successful currency overlay programmes to actively hedge the currency exposure in international portfolios provides a case, in which the investor lowers the volatility of the portfolio by engaging in active currency hedging, and at the same time, increases the return potential of the program by effectively selecting the optimal combination of overlay managers.
The value added of the currency consultant is to make sure the overlay programme is structured so that no explicit bets are taken in a particular style and risk is controlled by picking the most appropriate group of managers that collectively provide the highest information ratio for the client. To this end, we postulate that the only way to extract consistent and significant information ratios from the currency overlay program is to concentrate on the understanding of diverse investment management processes used by overlay managers and evaluate the correlations between them in different economic periods to derive the best combination.
Lastly, we would like to stress that any investment programme that includes an allocation to international assets would not be complete until there is an institutional view toward currency hedging in the investment policy statement.
Kurtay Ogunc is with Watson Wyatt & Co in Washington and Brian Hersey is with Watson Wyatt Investment Consulting in Atlanta

Topics