Currency overlay managers are frequently measured by their excess return – the difference between an overlay and a benchmark return.
A positive excess return for an equity manager indicates success in increasing the value of a portfolio relative to a passive investment. Since currency managers are often measured against partially hedged benchmarks, a positive excess return does not always indicate that the manager has increased the value of a portfolio. The overlay may actually have reduced the value of a portfolio.
And excess returns do not provide information on the cashflows an overlay generates. They are critical and must be evaluated by investors selecting strategies, benchmarks, and overlay managers. Unfortunately, though, the studies of overlay returns that consultants have published do not include cashflow data.
This article presents a method to measure currency overlay returns against cashflows for a more complete picture of a currency overlay manager’s performance.
Components of overlay returns
Equity and overlay returns cannot be interpreted in the same way because of the fundamental difference between the two management styles. Equity managers buy securities that might appreciate more than a benchmark in a bull market and decline less in a bear market.
Thus, equity managers can add value in bull and bear markets. However, currency managers add real value only periodically, because currency overlay is a process of market timing with three critical decisions: when to place a hedge, when to remove a hedge, and how much to hedge.
For full protection a 100% hedge must be placed when a currency begins to decline and it needs to be removed when a currency begins to rise. When a currency is rising, and it is unhedged, a currency manager cannot add value against a 0% hedged benchmark. That is why excess returns for overlay managers are periodic – large when currencies decline and modest or negative when they rise.
A hedge entails selling a currency forward with the aim of buying it back at a lower price. A hedge will then have a gain, which is received in cash, that offsets a currency’s decline.
However, if the currency appreciates, the hedge will have a loss that must be paid. The two outcomes are illustrated below.
(a) Currency drops 10%; benchmark 0% hedged
Overlay 0% (valuation loss–10% plus hedge gain of +10%)
Excess return +10% (overlay return 0% less benchmark return –10%)
Cash flow cash received equal to 10% of underlying investment
(b) Currency declines 10%; benchmark 0% hedged
Overlay 0% (valuation gain +10% plus hedge gain –10%)
Overlay –10% (overlay return 0% less benchmark return +10%)
Overlay cash paid equal to 10% of underlying investment
(Note that the overlay returns were 0% in both of the examples, whereas the excess returns and cash flows were different.)
The risk of a 100% hedge
Many investors believe currencies have no long-term return even though the Deutschmark and its euro successor rose at an annualised rate of 2.1% against the dollar between 1950 and 2000, the yen rose at a rate of 3.3% and sterling declined at a rate of –0.9%.
When it is assumed currencies have no expected return, 0% and 100% passively hedged strategies are seen as equally acceptable, but they are not. A 0% hedge does not have cashflows while a 100% hedge can have large cashflows.
When hedging losses are modest, cash flows are not important. However, when they are large it is clear that a passive hedge has not removed the currency risk from a portfolio.
An example based on a typical, euro-based portfolio invested in America, England, and Japan, in 1998-2000, illustrates the risk of a 100% hedge. As currencies rose
16%, the 100% hedge incurred a
loss of 6% because of interest-rate differentials.
However, relative to 0% hedged the loss was 22% and it caused a payment of €22 million for a portfolio of €100 million. The cumulative returns for the 0% and 100% hedged strategies are shown in Graph 1 along with the return for a 50% hedge and that of an active overlay.
The 0% and 100% hedged strategies are both risky since currencies have long histories of changing annually, and randomly, by 10% or more and have sometimes changed by over 50% over three to five years.
The polar nature of the 0% and 100% hedged strategies (=benchmarks) causes investors to regret having selected the “wrong” strategy when the other is superior.
Numerous academic studies have found that there is no optimal hedge ratio. That is one reason why the 50% hedged benchmark is used. It halves the impact currency returns have on a portfolio while it halves the cashflows of a 100% hedge. However, an active currency overlay is preferable to passive strategies since it can add return and can have more favorable cashflows.
Measuring overlay performance
In the example, the 0% hedged benchmark rose 16% in 1998-2000 while 100% hedged lost 6%. The 50% hedged return was 4.7%. It was 11.3 percentage points lower than the 0% hedged return, but 10.7 percentage points above the 100% hedged return. Its cash flow was negative at 11.3%.
For the same portfolio, an active overlay return was 14.3%. It was only 1.7 percentage points lower than 0% hedged, while it was 9.6 percentage points ahead of 50% and 20.3 percentage points ahead of 100%. The differences demonstrate that an overlay’s excess return depends very significantly on the benchmark it is measured against.
If the overlay’s excess return is measured against the 50% hedged benchmark, it appears as if the manager’s skill increased the total value of the portfolio. However, measured against 0%, the excess return is negative and suggests the manager had little skill, even though the overlay captured most of the available return; was superior to the 50% hedged return; and had a cashflow profile superior to those of 50% and 100% hedged.
The example illustrates that a comparison of excess returns cannot determine the “real” performance of an overlay. The comparison demonstrates that a manager has outperformed a benchmark, while an examination of cash flows will reveal if a portfolio’s actual value has been increased compared with not having an overlay.
Estimating cash flows
Since overlay managers hedge to different value dates, it is not helpful to collect actual cash-flow data. However, for an overlay with no leverage or cross-hedging permitted, the difference in returns between 0% hedged and the overlay represents a fair estimate of an overlay’s cash flow.
In the example spanning 1998-2000, when currencies rose, the cumulative returns were: 0%; +16%, overlay; +14.3%, 50%; +4.7%, and 100%; –6%. When monthly returns are used to calculated monthly averages, they are +0.44, +0.38%, +0.13%, and –0.17% respectively. When the average returns for the overlay, and 50% and 100% hedged are subtracted from the 0% hedged return, estimates of the average monthly cashflows are obtained: –0.06% for the overlay, –0.31% for 50% hedged, and –0.61% for 100% hedged. The return and cashflow figures can be plotted as shown in Graph 2 with return
on the vertical axis and cashflow on the horizontal axis.
A 0% hedged strategy has no cashflow and will always be plotted on the vertical axis. The alignment of the plots for the other strategies reveals, in this case, that the overlay had a return and cashflow profile that was superior to the 50% hedge strategy, regardless of which benchmark was used to measure the excess returns.
Moving the example to 2001-03, when currencies declined, the cumulative returns were: 0%; –17.9%, 50%; –9.8%, 100%; –0.2%, and for the overlay –2.1% as shown in Graph 3. The 100% hedged benchmark had the best return since currencies fell. When 50% hedged is compared against 0% hedged, its excess return was +8.1% while the overlay, if measured against 100% hedged, had an excess return of –1.9%, suggesting the 50% hedged strategy was superior if only excess returns were compared with no regard to the benchmarks that were used to calculate them.
However, using monthly averages of the returns and subtracting them from the 0% hedged return and then plotting them (Graph 4), the plots (and figures) reveal that the overlay had a superior return and cashflow profile relative to 50% hedged.
Graphs 2 and 4 also demonstrate how excess returns depend, not only on the benchmarks, but on whether currencies are rising or falling over the long term.
To complete the analysis, the monthly returns and cashflows in the six-years 1998-2003 are examined. Since currencies rose and fell, the cumulative returns for 0% and 100% hedged were, coincidentally, almost the same as shown in Graph 5. The return/cashflow graph for the full period (Graph 6) demonstrates that the active overlay was vastly superior to the 0%, 50%, and 100% hedged strategies. The overlay produced a real value that averaged 0.2% per month (overlay return less 0% hedged return). However, keep in mind that the actual cash flows varied significantly from year to year.
Overlay managers who manage against 50% hedged benchmarks, or other hedge ratios, should be able to provide 0% hedged returns since to be able to hedge effectively, an overlay manager must know on a monthly basis what the underlying exposures are . Overlay returns can therefore be more or less standardised and compared on a return versus cashflow basis provided the underlying currencies are similar and the periods compared are the same. More information is available at www.AG Bisset.com.
Ulf J Lindahl is CIO at AG Bisset & Co in Rowayton, Connecticut