Bucking the trends
Martin Steward examines the big claims that are made for managed futures' non-correlation with traditional assets, other hedge funds and even each other
When global equities fell 40% and managed futures returned 20% during 2008, institutional investors sat up and took notice, and the numbers gave the marketing machine a tonne of evidence for its well-worn claim that the strategy is not correlated with equities, bonds or other hedge funds.
Equity and credit are both "long the economy", says Abraham Trading Co president Salem Abraham. Short sellers are short the economy; and so are US Treasuries (if you don't mind the low absolute rate of return). "Only managed futures offers a high absolute rate of return that is agnostic about the economy. That's exciting for a diversified portfolio."
Christopher Keenan, senior managing director with Welton Investment Corporation, agrees. "The only non-correlated strategies are managed futures and global macro, which suggests that other hedge funds are mainly driven by equity beta." Welton's optimisations suggest that a 60/40 balanced portfolio will give you 10% a year with 10% volatility and a 32% maximum drawdown. A 39% equity, 26% bonds and 35% managed futures portfolio would deliver the same return with 8% volatility - but only half the maximum drawdown; while a 22/15/63 portfolio would squeeze 12% from your 10% volatility budget while reducing maximum drawdown to just 9%.
Some fear that this has all been oversold. "Perhaps there is zero correlation between managed futures and the financial markets over longer periods," says Winton Capital Management managing director David Harding. "That doesn't mean that that isn't made up of periods of strong positive and negative correlation. Managed futures will not necessarily perform well when the economy is performing badly."
Managed futures programmes are trend-following, not trend-setting: the traditional market starts the trends, so that necessarily entails some (positive or negative) correlation. More pertinently, there is no rule that managed futures will always be short equities when markets tank. By autumn 2008 we were a year into a bear market, so it is no wonder managed futures was generally short. But if a crash interrupts a bull market, they'd be long.
"Big slow surprises are good for some managed futures strategies," says Harding. "But the ‘insurance' argument is another that's in danger of being oversold: a very sudden catastrophe - like a terrorist attack - would not necessarily be insured against with managed futures."
If a bear market takes hold gradually, medium-term trend followers will suffer for a few days or weeks until they detect the new trend and go short. But if they are long in the face of a sudden crash, a stop loss may cause them to miss the sharp rebound once the market realises the world has not ended - effectively losing twice. This stuttering pattern of trend reversals characterised 2009, when managed futures indices lost around 5%. Risk markets like grains and energy bounced when governments announced their rescue measures in March, but then spent the rest of the year range-trading at their new levels - oil doubled in three months and has stuck at $70-80/bbl ever since. Similarly, bonds sold off in the spring and then range-traded. In FX, the dollar maintained a decent bear trend through the year - but returns were given back with December's sharp reversal.
Medium-term trend followers may have missed much of the bounce because they take time to catch-up with markets, and because those markets were shifting from a super-high volatility regime into a normalising one. When markets are volatile programmes tend to dial down their leverage, and vice versa, so when markets switch quickly from high-vol to low-vol, programmes find themselves without enough risk to exploit whatever trends exist. Positions at Man Group's $22bn AHL were "significantly downsized" during 2008, according to investment manager Harry Skaliotis. From a typical gross leverage-to-NAV of five to six times, the flagship Diversified programme shrank to 1.5 times during 2008 - with which it returned 33% for the year.
"It was probably Q3 of 2009 before programmes got back to their more average levels of risk calibration," says Jean-Christophe Wicker from the portfolio committee for the fund of managed futures funds run by IAM. Even in markets like metals and equities, where bull trends did stretch into Q3 and Q4, medium-term trend followers struggled to extract much juice.
What about short-term traders: could they not exploit both the first big bounce and the ranging markets later in the year? Programmes with five to seven-day time horizons fared best in 2009, according to Aurélie Vincent, head of CTA and global macro for Lyxor's managed accounts platform. There are more of these around nowadays because the style outperformed medium-term trading in the low-vol markets of 2003-07. Shorter-term programmes will often also include mean-reversion strategies. "Before 2006 we were long-term trend followers, but since then we've added short-term and mean-reversion strategies to our programme," says Abraham. "That's helped differentiate us: we were down 87bps in January 2010, whereas the managed futures universe has lost about 5% in those choppy markets."
This brings us to the claim that managed futures is highly internally diversified. Vincent says there is "a lot of dispersion between medium-term and short-term strategies". Malcolm Leigh, an investment consultant with Mercer, reports "considerable dispersion even between managers trading similar markets with similar styles".
But there are cautionary voices - sometimes voices one might expect to make the differentiation argument. When Skaliotis describes the evolution of AHL, he emphasises efforts to diversify the programme - doubling its markets, adding CDS indices, developing carry-trade strategies for FX. "If you stand still, you lose diversification as markets become more globalised and developed," he says. The AHL programme, developed by Michael Adam, David Harding and Martin Lueck, was bought by Man in the early 1990s; Adam and Lueck left to found Aspect Capital and Harding left to set up Winton - so you might expect Aspect and Winton to insist that their programmes are not correlated with AHL through family resemblance.
John Wareham, Aspect's chief commercial officer, suggests that medium-term trend followers were probably more homogenous 10 years ago, before their "shared DNA" had a chance to evolve differently - a process "exaggerated between those shops that have invested heavily in research and those that have not". However, he adds: "It remains a fact that medium-term trend followers returns are highly correlated at a high level." Harding at Winton is even more blunt: "Most managed futures programmes are strongly correlated with each other - it's as simple as that."
If this is true is it necessarily bad? What does it say about the most striking and attractive aspect of managed futures' distribution of returns: the big positive skew? A correlations-based explanation might be that when markets trend, programmes correlate as they all latch onto the same phenomena; and that when prices are range-bound, dispersion increases as they hunt fruitlessly for trends across hundreds of different markets. That would suggest that the strategy's right-tail risk could easily switch to a left-tail risk: if the 5% down year is explained by a few programmes losing 90% and dragging down the rest, the prospect of a chance correlation on the downside would be horrifying.
In fact, very few programmes make massive losses: in 2009, most of the industry did correlate, around losses of 5-10%, and a big chunk of those losses came from trading costs, rather than price movements, as stop losses cut losing bets. As Harding points out, while monthly returns show a fat tail on the right, daily returns show a fat tail on the left: "But by cutting those losses on a day-to-day basis you end up with significant positive skew in your monthly returns distribution." That positive skew is a robust effect of systematic risk management - "the essence of a self-healing, diversified systematic strategy", as Wareham puts it - and not of diversification.
That positive skew makes a strong case for managed futures, but two of the most common claims for the strategy should be scrutinised closely: there is no reason why managed futures should not be correlated with large losses in equity markets, especially those caused by exogenous events during low-volatility bull markets; and while it pays to diversify among the main differentiators (discretionary-versus-systematic, time horizon, market tilts), it may also pay to limit yourself to a handful of managers.