Why Are FX Returns Down?

Because FX participants and their clients have changed

MARHedge
August 28, 2006
By Desmond MacRae

Average annual returns for funds or programs that trade foreign exchange have trended downward since 1998 to just about their lowest levels in the past 15 years. These declines are prompting investors to ask what is happening, and why?

There are several reasons. And as the chart below shows, it doesn’t matter whether the methodologies these programs use are discretionary or systematic, their combined average annual returns represented by the Parker FX Index are still, on average, down.

Notional funding

William Lipschutz, who has been trading currencies since 1982, and who was profiled in The New Market Wizards (Wiley, 1992), suggests that one reason for the decline may be that the increase of notional funding has taken risk-free returns out of the picture because FX trading programs are no longer fully funded as they were several years ago.

Lipschutz, who is now director of trading for $186 million Hathersage Capital Management in South Norwalk, Connecticut, also acknowledges that several other factors are at play, and these are set forth in this article.

Start with his first suggestion. For the last 15 and a half years, discretionary traders’ returns have been less volatile than programs that are strictly systematic. Average annual compounded returns on both trading styles for this period were a satisfying 604.6% and 355.5%, respectively, because these FX returns are widely said to have low correlations to bond and equity returns.

“The risk-free return component of FX trading programs is an immaterial part of their total returns today,” says Lipschutz. “Funds as trading vehicles for absolute return foreign exchange trading programs don’t make much sense given that most of the money invested in them now comes from institutions.”

Institutions can open FX trading accounts with lines of credit extended against little or no collateral.

“In some cases,” he says, “large institutions can obtain bank guarantees that limit potential losses to less than a stated portion of the collateral they put up for FX trading accounts to which they assign trading discretion to FX trading managers they hire,” says Lipschutz.

Some banks can now monitor the value of collateral in trading accounts and communicate with trading advisors continuously.

This means that interest on collateral now contributes very little to total returns.

Precise data about funding of FX trading accounts aren’t available, but the table and the chart show four sets of returns, two for purely systematic programs, and two for purely discretionary programs.

The assumptions are based on Lipschutz’s suggestion that notional funding, as it is called in the futures world, and interbank FX trading accounts funded with credit lines, increased from 10% in 2000 to 90% in 2004. He believes this kind of funding will continue.

The data here include only those FX programs on The Barclay Group database. Systematic programs that use a discretionary override, or discretionary programs that depend in part on systems, have been eliminated. There is also a Parker index for each of these trading styles combined.

In the chart, the yellow line shows total annual returns for discretionary programs. The light blue line shows returns for discretionary programs from which interest based on annual Fed Funds averages—a proxy for risk-free returns—have been subtracted. The dark blue line shows total annual returns for systematic funds, and the dark pink line shows returns for systematic programs with risk-free returns also deducted.

The chart shows clearly the decline of risk free returns as a component of the funding of FX trading programs. It also shows that notional and credit funding are not big factors in the decline of FX trading program returns since 1998. Note also that the very low levels of US interest rates from 2001 through 2005 aren’t much of a factor either.

New clientele

The bulk of assets in absolute return FX trading programs today come from institutions. Several years ago, most assets were allocated by rich individual investors who cared little about volatility of returns as long as their managers made money.

However, virtually all institutional investors report to committees or boards that measure their portfolios’ performances with benchmarks. They shun volatility.

“Our clients have changed,” says Peter Panholzer, who began trading currency futures for clients in 1979. He is now chief executive of $53 million Dynex Corp in Geneva.

Luc Van Hof, chief executive of Analytic Investment Management in Rijmenam, Belgium, agrees. Founded in 1990, AIM has slightly more than $200 million under management on a cash funded basis.

Van Hof offers his services exclusively to institutions. “Trust me (the approach that satisfied rich individuals in the past) has become show me,” he says.

“Institutions want to see daily performance numbers, which we have provided for our clients for the past four and a half years. Daily returns show a much clearer picture of a program’s volatility than monthly returns.

He says institutions require transparency. “Those that look only at monthly returns might be in shock if they saw they had lost 1% or 2% on a given day, even though their returns at the end of each month during a given year were positive.”

The “institutionalization” of currency trading clients is part of the explanation for the decline in FX trading returns. Another may be that as FX trading managers get more assets under management, they become more conservative traders.

“I think this has been documented several times,” says Van Hof. “Most managers tend to deleverage before a reporting performance date comes up.”

All of this makes sense. “Absolute return FX management is more institutional than ever before, and the expectations of variance of returns and absolute level of returns have changed,” says Lipschutz. “Asset managers are ultimately driven by the expectations and the demands of their client bases.”

Panholzer points out that this is true for any sector in the absolute return arenas of asset management, not just foreign exchange.

Maturing markets

For much of his career, Panholzer has been a systems trader, but he has also run FX trading programs that rely on interpretation of market sentiment. He says that the currency “baby markets” of the 1970s were marked by participants having similar outlooks and low liquidity.

“Baby markets” of the 1970s tended to trend well because there were few participants who had similar training styles and outlooks. It wasn’t until 1980 that the Group of Thirty first called for quantifying foreign exchange trading volume. And it would be another 15 years before FX markets saw major participants of significant size that were not commercial or central banks.

“When markets become mature, their inefficiencies that previously provided profit opportunities are ‘arbitraged away’ by more and larger (and different types) of participants,” says Panholzer. “This increases randomness and lowers return opportunities.”

The 2004 Triennial Central Bank Survey of Foreign Exchange and Derivatives Market Activity from the Bank for International Settlements reported a “surge in traditional foreign exchange markets. Average daily turnover was $1.9 trillion in April 2004, a rise of 57% at the then-current exchange rates and 36% at constant exchange rates.

“This increase more than reversed the fall in global trading volumes between 1998 and 2001,” the BIS report said.

In 2004, there was a 33% total turnover between banks and financial customers, up from 28% in 2001. Volume in interbank activity increased, but its share fell to 53% in 2004 from 59% in 2001 and from 64% in the mid-1990s. Trading between banks and nonfinancial customers made up the remaining 14%.

Has the increase in trading volume between banks and financial customers from 2001 to 2004 attracted too many traders for the FX markets to support?

None of those interviewed for this article think that FX markets are overwhelmed by the influx of trading funds. They all believe that despite daily cyclical exceptions, and occasional sudden “hot money” flows that can affect smaller currencies (Thai baht in July 1997; Icelandic kroner in March 2006), FX markets in the major currencies have plenty of liquidity.

Inexperienced traders

The increase in FX trading programs shown in the table (13 trading programs in 1995 versus 111 trading programs 10 years later) suggests that there are a lot of newcomers to FX trading. Can part of the decline in FX profits be ascribed to these newcomers?

“One reason newcomers are not making money is either that they fail to stick to their trading system or plan as they grow, or the system wasn’t designed correctly in the first place, or both,” says Barbara Rockefeller.

Rockefeller publishes a daily morning report on FX markets that posts systematic signals on all major non-US dollar currencies, together with observations on fundamental factors that can influence these markets.

However, it’s hard to blame all of the decline in 2005 and 2006 on inexperienced traders when returns from experienced traders were also down. Lipschutz, Panholzer and Van Hof all say that 2006 isn’t over yet, and that this year could turn out well.

“Inexperience may play a part when trading size,” Van Hof suggests. It’s harder to trade $50 million or more because spreads can widen. Trading size requires experience that, he suggests, newer FX traders may not yet have. Their performances may suffer if their assets grow too quickly.

He’s probably right, but if this is a contributing factor, it’s hard to quantify.

Execution is also important, especially for frequent traders. “A quarter to a third of our performance can be attributed to the quality of our executions,” says Van Hof. His holding periods range from three to 12 hours, so he trades $300 million to $400 million daily. A difference of even one pip per trade can add up. Some newcomers may not yet fully understand this.

Time frames

Using traditional time frames such as days to analyze price data may be another factor contributing to lower FX returns. “FX markets have a compressed time frame compared to other markets,” says Bill Reid, who founded Algorithmic Trading Advisors LLC last year after leaving IBM where he’d specialized in artificial intelligence.

The Fairview, Texas firm’s Algorithmic’s annualized return for 2005 was the outlier in the systematic trading universe cited in the table. Algorithmic has managed between $10 million and $53 million in its 13-month history.

“Things change in FX in hours that take days or months in other markets,” says Reid. “Techniques to handle this compressed time [frame] are lacking in most funds.”

At the end of 2005, there were $6.1 billion under management in 73 systematic programs in the strict universe represented in the table, but only $2.7 billion under management in 38 discretionary programs. If Reid is correct, and if most systematic programs have not adapted, this too may also a contributing factor. But again, it’s hard to quantify.

Algorithmic Trading Advisors uses chaos theory as a basis for price data analysis. Reid hasn’t created rules that are applied to price data to make decisions. Instead, his system learns adaptively, first generalizing on data, then scanning these generalizations to “see” how they are organized before making trades.

Systems like these can read handwritten ideograms used in Chinese, Japanese and Korean writing systems, which rule-based systems can’t do because it is too difficult to write rules for all the variations that occur in individuals’ logographic handwriting.

Better measurements

Comparing returns from different asset classes is difficult. “It is wrong to call a manager who has a higher alpha a better manager,” says Panholzer. “Alpha is a percentage return.” People need volatility-adjusted returns as well as risk-adjusted returns to measure the value of a manager accurately.

“FX trading isn’t the overleveraged business today’s funding practices might lead some people to assume,” says Hathersage’s Lipschutz. “Today’s databases don’t tell you exactly what the funding is, so how do you compare FX traders to a long/short or global macro traders in terms of absolute return?”

A good question to which there are now few if any answers.

The future

Trading foreign exchange is an arcane craft that requires an almost monastic discipline, which may be why so many alternative asset fund managers and their investors appear to prefer systematic approaches. It is not an easy skill to master no matter how one chooses to do it.

The craft’s increasing difficulty, notional or credit funding, dominance of institutional clients, maturing markets, size, scaling and liquidity challenges, and shorter time frames driven by the distribution of sudden news have made trading foreign exchange more complicated. That foreign exchange returns are down should be no surprise.

“It’s interesting to note that FX returns seem to suffer an almost cyclical performance malaise every four or five years,” says Panholzer. A lot of academic work needs to be done to quantify the reasons why.

“The little facts that are part of our fabric of FX thinking are very important, including the apparent cyclical behavior of returns,” he says.

As allocations to foreign exchange trading programs are expected to continue to grow, asset allocators might want to keep this in mind, as well as the need for experienced FX asset managers.

 
Desmond MacRae is a New York-based writer, specializing in banking, finance and investments.