Plans, in general, have widely differing characteristics and objectives depending on plan maturity, workforce/retiree ratio, risk tolerance, legal restrictions, etc. There is no single currency policy that will be optimal for all. Each needs to develop a currency policy that suits its particular characteristics and objectives. Creating a currency policy requires that a number of decisions be made, and this article explores those decisions and their impact.
The level at which international assets start to impact overall plan volatility is commonly taken to be 10%. At this level, plans need to develop a currency investment policy to manage currency consistently with other asset classes.
Investors often spend considerable time developing an investment policy, or strategic asset allocation, for multiple asset class portfolios. For portfolios that include international investments, the policy should include guidelines and a benchmark for the management of the currency exposure. Currency investment policy should address:
• Overall plan objectives
• Benchmark selection - Proportion of currency exposure to be passively hedged.
• Passive versus active policy
• Whether or not to use separate currency management and what style to choose
• Selection of currency guidelines for active positions relative to benchmark
• Hedging vehicles to be permitted (forward contracts, futures or options)
• Proxy hedging currencies
RISK AND RETURN CHARACTERISTICS
Currency is typically an exposure derived from an investment in some underlying asset class such as fixed income or equity, and provides no implicit or explicit return such as earnings growth, dividends, capital appreciation or coupon payments. For this reason, many investors assume currency exposure has no long-term real expected return associated with it. Currency exposure, however, can add substantially to interim volatility of international investment portfolios. For example, the annualized return volatility of sterling, yen, and DM for US$-based investors for the 1988-97 period was 11.4%, 11.4%, and 11.2%, respectively. This compares with international equities (using the EAFE index as an example) which had a volatility of 17.1% unhedged and 15.2% on a hedged basis. On average, over the last 10 years, the magnitude of the currency return has been more than half the local return from EAFE equities. This record suggests how powerful an effect currency can have on a performance benchmark, or on an indexed portfolio, as well as on individual investments.
The absence of an expected return for currency does not mean that investors can assume that currency returns will wash out over the long term. Historical analysis has shown that even over 10-year periods there was only a 22% frequency of currency returns being within the range of +/-1% per annum. There are insufficient historical data to draw any conclusions over very long time horizons, but since most plans review strategic objectives approximately every 5 years, currency exposure clearly implies significant risk over horizons longer than most strategic time frames.
Several theories describe currency movements incorporating variables such as relative purchasing power of currencies, real and nominal interest rate differentials, and trade and financial flows. While none of the equilibrium theories of exchange rates explain relative currency movements in the short term, studies of Purchasing Power Parity (PPP) theory suggest that over longer time frames (3 years and more), currencies revert back to theoretical levels.
While currencies tend to follow their PPPrates in the long term, empirical studies of exchange rates have shown that currency returns exhibit non-random trends and reversals. These studies have implications for currency investment policy. An active strategy or investment policy that uses both fundamental and technical approaches has the best opportunity of consistently adding value.
Several different currency benchmark alternatives are available for investors. These benchmarks include unhedged, fully hedged and partially hedged. The choice of an appropriate currency benchmark depends upon the structure of a plan's overall portfolio, liabilities, risk tolerance and cash flow position. It will also depend on assumptions about the co-movement of currencies with other asset classes and the trending characteristics of currency returns. Some plans incorporate non-zero currency exposure return expectations into the process, although this amounts to an active strategic decision. Most investors assume no expected return from currency exposure as the starting point for strategic analysis of currency. The benchmark and guidelines will typically outline the strategic hedge ratio, instruments to be used, investment ranges, permitted counterparties, and proxy hedging guidelines. For actively managed underlying portfolios, it should be specified whether the currency benchmark arises from the underlying asset benchmark or from the actively managed underlying portfolio.
A set of currency benchmark guidelines for an EAFE-benchmarked active equity fund might include:
PASSIVE VERSUS ACTIVE POLICY
While State Street Global Advisors strongly believes in passive management for stocks and bonds, when it comes to currency we see the case for active management as being more compelling. First, because there is a low probability that currency returns will average out to zero, there is a significant risk that a passive (hedged or unhedged) policy will result in either large losses or missed gains. In other words, passive currency strategies are risky. Second, currency markets exhibit what are considered the three essential characteristics for active management to succeed. These are: (1) observable inefficiencies that can be captured using fundamental and technical approaches; (2) a basis for expecting those inefficiencies to persist, (i.e. the permanent presence of nonprofit maximizers such as central banks, corporate treasurers, and international investors with passive currency policies); and (3) much lower transaction costs than either stocks or bonds. These factors imply that active currency managers have a larger opportunity than stock or bond managers to beat a passive approach.
STYLES OF ACTIVE MANAGEMENT
As a plan is formulating its currency policy, it must decide whether to hire a currency overlay manager for active tactical currency management or to leave currency exposures fixed at the strategic benchmark. Most plan sponsors are not primarily concerned with the volatility of currency returns - their objective is that active management should help them to avoid losses from adverse currency moves. This means that currency overlay managers interpret the expression "managing risk" to mean "avoiding losses." Therefore the objective of most programs is to reduce exposure to foreign currency selectively - the plan wants to keep the upside risk when foreign currencies are rising and avoid the downside risk when they are falling. This means that the performance of the perfect overlay looks like a zero cost option with 100% participation in gains from upward currency moves and 0% participation (or 100% protection) in losses from downward moves (see below). Despite the different philosophical styles of overlay managers, all, in their own way, are seeking to achieve as closely as possible the pay-off of this zero cost option.
There are two generic styles of active currency management. The first group contains the "active" managers, and is generally used to describe managers which use fundamental and/or technical techniques in their strategies. These are essentially proactive approaches which use models or traditional judgmental analysis to anticipate the direction of the market. The second group uses option replication approaches. The vast majority of managers in this group use strategies that derive from "dynamic hedging." Dynamic hedging strategies are more popular than option trading strategies because of the lower costs and greater liquidity associated with using forward contracts. These approaches are more reactive than anticipatory, as positions tend to be taken in response to moves in the market with no attempt to forecast future movements. Currently the market for currency overlay managers is evenly split between the two groups.
METHODS FOR DETERMINING BENCHMARK
Acurrency benchmark should be selected at the same time as determining the strategic positions for the total portfolio. The process should also examine the trade-off between the costs of hedging (transaction costs, cash-flow volatility, and expected return give-up, if any) and the benefits (volatility reduction). If a mean-variance framework is used to assess portfolio return and risk, then estimates of volatility, correlation and expected return are required for each asset class. One way to incorporate the potential for hedging foreign exchange exposures is to develop expected return, correlation and volatility estimates for domestic asset classes, unhedged international asset classes and hedged international asset classes. We advocate using the same relative expected returns for the hedged and unhedged international asset classes (i.e., no expected return from currency exposure relative to hedging) except for the explicit costs associated with currency hedging (i.e., transaction costs, management fees and cash-flow volatility costs). Mean variance optimization reveals that if a U.S investor has less than 5% in international exposure, hedging any portion of the foreign currency exposure is not desirable. It typi cally increases expected portfolio volatility and reduces portfolio expected return. This is because the diversification that unhedged international exposure provides (i.e., low correlation with domestic asset classes) outweighs the volatility reduction within the international component that currency hedging provides. However, if the international component of a portfolio exceeds 10% of assets, optimizers generally indicate that the return/risk ratio improves when some fraction of foreign currency exposure is hedged.
For plans with between 10% and 20% invested in international assets, the optimal portfolios derived from this process typically have strategic hedge ratios of about 50%. The reduction in risk from hedging 50% of the international exposure is minimal, around 10 basis points. Given the sensitivity of the process to the initial assumptions and inputs, one cannot justify the selection of a hedge ratio purely on the basis of optimization alone. However, the "optimal" solution of a 50%-hedged benchmark is supported by three other factors:
First, the 50%-hedged benchmark reduces the pain from choosing the wrong benchmark at the wrong time. This is the situation where a plan selects an unhedged benchmark just before a period of sustained foreign currency weakness and the returns on hedged international assets massively outperform their unhedged equivalent - or when the exact opposite situation proves to be the case. While the plan may have a long-term horizon, it is still hard to accept 3 or 4 years of bad returns which resulted from choosing the worst performing benchmark. In order to avoid this "regret" syndrome, the plan can minimize the possibility of selecting a poorly performing benchmark by opting for the 50%-hedged benchmark. This will never be the best performing benchmark, but offers the comfort of never being the worst.
Second, over a full currency cycle (i.e. when currencies rise and fall such that the unhedged and fully hedged benchmarks have identical performance), the 50%-hedged benchmark actually outperforms both alternatives by a small margin (+/-0.10% per annum). This is due to an effect similar to that of a passive re-balancing strategy where the asset which has risen is sold and the asset which has fallen is bought.
Third, for plans with an active currency program, the 50%-hedged approach allows them to evaluate the currency manager's skill in the usual 3-5 year period. For active programs with a 0% or 100% hedged benchmark, it is only over periods of a perfectly complete currency cycle that one can calculate value added in an unbiased way.
HEDGING TOOLS AND TECHNIQUES
Currency hedging is most commonly carried out with forward contracts, which are agreements with a counterparty to buy or sell one currency against another at a pre-specified exchange rate and time. Forward contracts provide investors with a cost-effective and flexible tool to eliminate spot rate risk. The agreed upon exchange rate (the forward rate) is a function of the existing spot rate and short-term interest rates in the hedged and base currency. The forward price of a currency can be either at a premium or discount to the current spot rate, depending on whether short-term interest rates of the hedged currency are higher (a discount) or lower (a premium) than base currency short-term interest rates. This also has implications for performance measurement of currencies. Since a currency manager always buys or sells a currency at its forward rate, the value of such a decision should be measured by comparing the future spot rate to this forward rate. This measure is called the "currency surprise." In other words, the spot rate to spot rate return on a currency has two components: the fixed element due to interest rate differentials and the element due to the currency surprise.
The "costs" of hedging include the interest rate differential between the hedged currency and the U.S. dollar (or whatever the base currency of the investor), forward contract bid-ask spreads, counterparty credit risk, and the cash-flow volatility that currency hedging implies. Cash-flow volatility occurs when losses on forward contracts need to be funded or gains invested as positions are rolled at the settlement date of the contract. The currency accounting gains and losses of the contracts and the underlying exposures offset one another (less the forward discount or premium), but contract gains need to be invested while contract losses need to be funded from either an outside source or sale of underlying assets. This could mean extra transaction costs from the underlying asset sales and purchases. These costs can be avoided quite easily. At the start of the hedging program, a plan may establish a cash pool used to fund and receive overlay profits and losses. This cash pool can be fully equitized by purchasing equity or bond futures contracts. Extra futures contracts can then be bought or sold as the size of the cash pool varies.
Currency hedging can also be carried out with options. An investor who buys a currency put option eliminates downside currency risk, yet maintains upside potential if a currency strengthens relative to the base currency. In addition to the costs associated with forward contract hedging, option hedging strategies require the upfront payment of the so-called option premium. The option premium is essentially the price of insuring against a decline in a currency below a certain point. An investor can produce an option payoff with a dynamic hedging strategy, which uses forward contracts and options pricing theory to produce the desired put option payoff. The cost of a dynamic hedging strategy can only be estimated since the cost is related to the transaction costs of the dynamic hedge, which in turn depend on the path that currencies follow during the life of the hedging strategy.
Since most developed market international equity portfolios invest in and are measured against EAFE or similar benchmarks, there is often exposure to the 11 different currencies in the index. Hedging this exposure by executing individual trades in all 11 currencies can be more expensive, particularly in some minor currencies. Many currency managers therefore take a basket or replication approach to hedging the minor currencies. The proxy basket is a combination of a number of the major, more liquid currencies (e.g., Euro, pound, yen, Swiss franc and Australian dollar) designed to replicate the wider basket of underlying exposures. This approach must weigh the advantages of lower costs and greater liquidity by using the major currencies with the tracking error risk associated with not selling all the individual currencies. Cross-hedge losses can occur when an underlying basket of currency exposures is hedged naively with one currency. However, more sophisticated approaches keep this tracking error minimal, and generally perform better than individual hedging.
There can be a small degree of duration risk if the manager chooses to use forward contracts which have a significantly longer duration than those used in calculating the benchmark. Most of the major index providers (e.g. Salomon, JPMorgan and Morgan Stanley) calculate hedged returns based on one-month forward contracts. Most managers use one to three month contracts which virtually eliminate duration risk. In addition, forward contract and OTC (over-the-counter) option agreements involve credit risk associated with the counterparty of the transaction. This risk is managed by the plan sponsor and the currency manager agreeing on an acceptable list of counterparty banks. Credit lines are set up and monitored between the plan and the counterparty bank, since both are principals in any transaction, with the currency manager forming the link between them and acting as agent only.
A FINAL WORD
A fully hedged currency benchmark is usually only justified when nearly all portfolio assets are in international investments, where the plan attaches a negative expected return to currency, or where the plan has a very low tolerance to risk. In addition, investors who elect to view the risk of their international investments in isolation may find a fully hedged strategy most appropriate. In any case, an investor's risk aversion should be incorporated into the analysis to determine the level of currency hedging that provides the appropriate level of overall portfolio risk reduction.
HOW AN INTERNATIONAL INVESTMENT IS HEDGED WITH FORWARD CONTRACTS
AUS$-based investor with a 5 billion yen equity investment wishes to fully hedge the currency risk for the next year. The current spot rate is 96.50 yen/$, and one-year interest rates are 5.63% in the US and 0.70% in Japan. The one-year forward rate is then [(l + 0.70%)/(l + 5.63%)] x 96.50 yen/$ = 92.00 yen/$. Since interest rates in Japan are significantly lower than in the US, the forward price of yen is at a significant premium to the current spot price (92.00 vs. 96.50). To hedge the investment, the investor enters into a forward contract to sell 5 billion yen at 92.00 yen/$ one year from today.
Although no funds change hands until the expiration of the contract, the contract accrues gains or losses with changes in spot rates and short-term interest rates. The underlying currency exposure accrues currency gains or losses with changes in the spot rate. If the yen/$ exchange rate ends the year lower than the forward rate (92.00 yen/$), the investor will realize a loss on the contract. If the yen/$ exchange rate ends the year higher than the forward rate (92.00 yen/$), the investor will realize a gain on the contract. Currency losses (gains) accrue to the underlying currency position if the yen/$ spot rate is higher (lower) than the initial spot rate of 96.50 yen/$.
The table below illustrates the currency gains and losses for three different scenarios for the yen/$ exchange rate at year-end. Note that the total currency gain/loss is the same in each of the scenarios. Currency hedging has uncertainty or "currency surprise" associated with currency exposure. It is also interesting to note that this table shows the value of each of the decisions to hedge.
In the first example the decision to hedge was clearly worse than leaving the investment unhedged as the potential currency gain was reduced from 13.5% to 5.0%.
In the third example, the decision hedge is clearly better than leaving the investment unhedged as an 8.1 % loss is reversed into a 5.0% gain. In the second example, hedging might seem to be unnecessary since the spot rate is unchanged, but hedging eliminates the -5.0% currency surprise and leaves the investor with 5.0% more wealth than the unhedged strategy. Hedging the foreign exchange exposure for the one year horizon eliminates currency volatility at a predetermined cost. In this example the cost is negative (5.0% of the portfolio value) because the yen is selling at a significant forward premium. For currencies where interest rates are higher than the base currency the cost is positive because the currency forward rate is at a discount. Hedging always eliminates the currency surprise.