Reality versus theory
Although currency rates have played important roles for a long time and complex foreign exchange theories have been developed, it’s only in the past decade that pension funds have begun to manage their currency risk with currency overlay programmes. These pioneering funds had found their international equity managers didn’t hedge much, or if they did, they didn’t hedge effectively. What can plan sponsors learn from these pioneering funds? The short answer: reality is different from theory.
Investment theory postulates that an investor should use optimisation techniques to determine how much of the currency exposure should be hedged to attain an optimal risk/ return profile over time. Once the hedge ratio has been determined a passive hedge can be applied to achieve the indicated outcome. However, in reality it’s not that easy.
Since optimisations are input-sensitive, the use of different expected currency returns, correlation coefficients etc can and do result in different hedge ratios. Not one of them is certifiably correct. Even so, Evaluation Associates, a pension consultant in the US, has found that hedge ratios usually fall in a range of 30–70%. Hence, a passive hedge of half of a pension fund’s currency exposure could be expected to produce acceptable results over time. But it doesn’t.
The key words are: “over time”. A pension fund in Euroland that hedged 50% of its dollar exposure in the past four years would have forfeited half of the dollar’s rise, while a dollar-based pension plan would have avoided only half of the currency loss on its Euroland assets caused by the dollar’s rise. In the first half of the 1980s, the dollar rose for five years and then fell in the next three against the Deutschmark. Since an optimisation’s projected outcome may take more than 10 years to pan out, if ever, passive hedging of a portion of a portfolio has been found to be unsatisfactory in the one to three years over which pension fund executives strive for success.
What’s more, an optimisation does not account for the cashflows that are associated with a passive hedge programme. Simply put, in a year when the dollar rises 10%, a 50% hedge will have a 5% loss that must be paid. In a year when the dollar declines 10%, a 50% hedge will result in a 5% gain that offsets only half of the translation loss on the underlying investment. If the dollar continues up another year, which is not unusual, the cumulative loss could become 10%, 15%, 20% or more. Since currency losses are not sufficiently reduced with a passive hedge in the short- to intermediate-term, and cashflows that result from gains and losses on hedges can be large and disruptive, pension plans are adopting active hedging as a preferred alternative to a passive hedge.
In an active currency overlay programme the currency exposures are hedged from time to time to protect against currency declines while allowing an investor to capture currency gains by being unhedged when currencies rise.
Many investors believe that currency returns offset each other over time, so why hedge? First, it’s not true that currencies have no expected long-term returns. They do. Second, pension funds have found that the currency risk is larger than theory predicts. As a result, they have realised that it’s prudent to appoint currency specialists to manage their currency risk.
A look at the monthly price changes in the Deutschmarke versus the dollar from 1969–98 demonstrates that they were close to normally distributed. There were many small price changes and a few large ones (Figure 1). The data suggests that positive returns will cancel negative returns over time. However, they don’t. In 1969–98, the DM rose on average 3.5% a year against the dollar with a standard deviation of 12.8. From a European perspective the dollar depreciated by about 3.5% a year. The yen rose about 5% per year against the dollar over the same period with a standard deviation of 14.4.
Monthly returns are not relevant for a pension fund that measures results in calendar years. When changes in the Deutschmark/dollar are measured over the 29 calendar years since 1969 a graph of them does not produce a bell curve (Figure 2). Contrary to the theory that returns would be normally distributed, they were not. Most of the returns were in the tails of the “bell curve” – not in its middle. In fact, the Deutschmark rose or fell less than 6% against the dollar in only six of the 29 years (20%). In the other 23 years (80%), it rose or fell more than 6%.
Based on the actual data on how currencies have moved (the statistics for the yen and sterling are similar), a pension fund can expect that the euro/ dollar exchange rate will change by 10% or more in two out of three years! And, because the dollar has a history of trending for two to three years, the currency risk is much larger in the medium-term than most pension funds realise. That’s why they are turning to active currency overlay managers to reduce their currency risk.
However, the emphasis on risk reduction is changing. Pension funds with programmes that focus on reducing the risk at the expense of foregoing returns are not always fully satisfied with their overlay managers. A few years ago there was little data on how overlay managers have performed. But last year, Brian Strange of Currency Performance Analytics published the results of an extensive study of 152 currency overlay accounts managed by 11 overlay managers with combined exposures of over $40 billion. Strange found that the average currency manager had added an extra return of 1.9% per year, on average, to the client-selected benchmarks in the 10 years to 1997. He also found that the managers reduced risk by about the same regardless of the decision style they used.
Strange’s study, and performance data collected by pension consultants like Watson Wyatt, Frank Russell and William H Mercer in the past year, have made it easier for pension funds to select overlay managers that can produce a high added value while reducing risk as well. Adding an extra average return of up to 1.9% per year will become more important in the years ahead when it’s likely that equity returns will be lower than in recent years. In fact, currency overlay is emerging as an important tool for increasing returns.
In a recent comparison of overlay managers’ skill at adding value versus those of the average large-cap US equity and US investment-grade fixed income managers, Rob Zink of Bridgewater Associates found that the currency overlay managers added more value. They had added +1.9% a year, on average, while large cap managers had reduced returns with –1.1% on average in the 10 years to 1997. The average fixed income manager had added +0.20%. And, overlay managers had a higher information ratio (added value/standard deviation of the added value). It proves they know how to hedge.
As the currency overlay industry continues to mature and overlay managers’ skills at adding value become more recognised currency overlay will become as accepted as investing across borders has become in the past twenty years. It’s easy to see why. With little increase in active management risk the implementation of a currency overlay programme may boost returns by 1.9% a year.
Ulf Lindahl is chief investment officer at AG Bisset & Co in Connecticut