Executive Summary
In the ongoing search for more diversification - and especially, low correlations as a potential for stabilizing returns in a difficult stock environment - advisors have increasingly shifted in recent years towards "alternative" investments. From real estate and REITs to gold and other commodities to more, a recent FPA survey on Alternative Investments found that 91% of advisors are using some form of alternative investments. Sadly, though, the focus on finding investments that have a low correlation - according to FPA's survey, the number one criteria for choosing an alternative investment - has grown to such an obsession, that we're willing to name anything that has a low correlation as "a new asset class." But the reality is that while some alternatives really are investments that truly have their own investment characteristics unique from stocks and bonds as asset classes, others alternatives - like managed futures - simply represent an active manager buying and selling existing asset classes. Which means it's about time for us to start distinguishing between a real alternative asset classes (e.g., commodities or real estate), and the real value of managed futures.
The inspiration for today's blog post comes from a conversation I had last week with some planners about the use of managed futures in a portfolio as a new "alternative asset class." The planners were considering whether the returns of managed futures had a low enough correlation to be viewed as a different asset class, and were struggling due to the fact that some managed futures funds are more correlated to markets than others.
In considering the question, it may help to remember what a managed futures fund is in the first place. The managed futures industry is compromised of commodities trading advisors (CTAs), who invest in futures contracts, generally either using a proprietary trading system/methodology they have developed or simply by investing with discretion. Depending on the manager style and approach, they may go long or short futures contracts, and may use future contracts in any number of market segments, from metals and grains to equity indexes to currencies to bonds. As futures contracts, the prices will fluctuate up and down with the price of the underlying investment asset price (gold, wheat, S&P 500, the Euro, the 10-year Treasury, etc.), although because only a percentage of the futures contract's value is required to trade it, futures trading implicitly involves leverage. Some managed futures fund managers may try to follow momentum trends; others utilize more market neutral strategies.
So basically, price changes for a managed futures fund are simply a result of price changes in existing asset classes (bonds, equities, commodities, currencies, etc.), which simply happen to be purchased using a futures contract instead of the underlying investment. The overall returns of a managed futures fund are a result of the implemented active trading strategies of the fund manager, and the effectiveness of the manager's trading system or discretionary decisions.
Which means in essence, a managed futures fund is nothing more than an active manager running a go-anywhere long/short strategy implemented using derivatives. If we looking at a fund manager who directly went long the S&P 500, short the Euro, short the 5-year Treasury, long the 10-year Treasury, long gold, and short wheat and cocoa at the same time (without using derivatives), we'd view it as an incredibly eclectic mix of investments for a go-anywhere long/short manager, and hope that he/she was making some good investment decisions. The fact that the manager happens to do the exact same thing, but using futures contracts, does not change the underlying nature of the strategy and investment!
As many point out, many managed futures funds have managed to produce returns that have low correlations to traditional stocks and bonds. To which I say: OF COURSE they have! Many of them are investing in half a dozen or more different underlying asset classes, they're going net long some of those asset classes and net short on others, or they're holding market neutral positions betting on prices/spreads to widen or narrow. OF COURSE they're going to have a low correlation to a single long-only asset class benchmark! But if I really hold the portfolio mentioned earlier - long the S&P 500, short the Euro, short the 5-year Treasury, long the 10-year Treasury, long gold, and short wheat and cocoa at the same time - and generate returns that have a low correlation to stocks and bonds, it's not because my portfolio is invested in a new asset class. It's because I have some a broad mix of long and short positions of a whole bunch of existing asset classes!
Of course, all of this is not to imply that managed futures are a "bad" investment. There are some strong performers out there. The most popular index for tracking managed futures funds is the Barclay CTA Index (no relationship to the Barclays global financial services firm), which shows some remarkably stable returns - only 5 down years in the past 30 (including 2011 year-to-date), and a positive 14.09% return in the difficult 2008 calendar year. Many of the managed futures options available today in mutual fund form have short track records (e.g., Altegris Managed Futures Strategy Fund {MFTAX} and Natixis ASG Managed Futures Strategy Fund {AMFAX}), but some are simply fund-of-funds using outside managers who may have more of a historical track record (e.g., the aforementioned Altegris fund, or the Equinox MutualHedge Frontier Legends Fund {MHFAX}). On the other hand, many of the funds can be expensive, and there are definitely some self-reporting bias problems in the tracking indices, as a Wall Street Journal article pointed out recently.
Notwithstanding some of the costs and performance concerns, there is still a great deal of allure to managed futures funds and the opportunity to earn positive returns in what has been and may still be a difficult ongoing investment environment for equities. But the fact remains that giving your money to a go-anywhere long/short active trader does not itself constitute a new alternative asset class. It simply means you're expecting the implementation of the active manager's strategy to produce favorable (and low-correlation) returns; whether that expectation is reasonable or not depends entirely upon the manager's skill, effectiveness of his/her models, and ability to execute. Which means you'd better know a whole lot about those important factors before handing over your client money. And if you're a passive index investor and believe markets are efficient, you should probably stay away altogether; if you don't believe predictable, tradable pricing events occur in the markets, there's no reason to give your money to a derivatives-based go-anywhere long/short active money manager whose sole job is to capitalize on those anomalies and opportunities!
But again, the bottom line is not that managed futures funds are bad. They may in fact be a decent (or dare I say, good?) investment for some clients - at least if the underlying active manager is actually good! - but nonetheless, generating low correlation returns just because the fund is a go-anywhere long/short active trader using derivatives does not make it a new asset class! It just makes them an "interesting" active manager!
So what do you think? Do you use managed futures funds in your client accounts? Do you view managed futures as a new asset class? Should it be? Is low correlation the only criteria for determining whether something is a new asset class? What other criteria should be involved?
Michael,
I certainly appreciate your thoughts on this. I do not have nor have I employed specific managed futures exposures in my client portfolio models, nor even in my custom accounts. Every time I have evaluated them, even evaluated specific simple hedged investment strategies over different market cycles, getting more than simply non-correlated results, meaning reasonable(non-negative) performance too, seems to have boiled down to the active security selection side of things. Now I have long used asset allocation managers as a core holding in my models. Attempting to blend the efforts of different manager styles, (technical, fundamental, limited discretion, complete discretion) for whom I could be confident of their objective but would have no control over their allocation or execution. When reviewing their most recently disclosed holdings and deriving from that a best-fit index, approximately 17% of that portfolio was classified as “Alternative”. These managers I use all operate in the open-end fund space. I like this approach, although this quarter it helped not at all for me. I have always boiled down assets classes very fundamentally, debt and equity, that’s it. You either own it or you lent what you own. Even cash is a “note”, and leverage is only that in whatever form it ultimately comes down to a contract regarding debt. Even an LP investment I categorize as debt, regardless of the “strategy” as it is secured in the end only by the GP’s ability to redeem the LP units.
Ben
This is exactly right. I have used Managed Futures for many years and in our firm we are careful about how we describe them or any alternative investment. We use the term alternative strategy because that is what it is, a different approach.
To call something a new asset class, or sub asset class, it first needs to be an actual asset, not merely a different way of using existing assets and/or engaging in speculation.
In my way of categorizing things an asset has to have cash flows, or represent something which has cash flows. A speculative vehicle like gold isn’t an asset, as it has no cash flows which can be valued. That doesn’t mean it is unwise to own some gold, or other investments that throw off no cash flows, but they are not assets.
Managed Futures are a great way to diversify clients accounts and to supply high return options when traditional assets are over priced. As I often tell clients when assets are overpriced, I have no idea when, but overpriced assets eventually get cheaper. While we wait I would rather be in a strategy that can make money over time over one which likely cannot. Using just that reasoning one certainly would have been better off over the last five, ten and fifteen years emphasizing Managed Futures over US equities.
Michael,
I agree with most of your article. I use managed futures in client portfolio, and have for years.
While definitely not a “new” asset class (some of the funds/strategies have been around since the 1970s), we view them as a separate asset class from stocks, interest rates, currencies, and commodities (the underlying future markets that are traded by CTAs). The way I define an asset class is a risk/return profile that is significantly different than any existing asset classes or combinations of existing asset classes. For most clients, they are not able to produce the risk/return profile of managed futures on their own. Does that mean that they could be a separate asset class to some people and an extension of various asset classes for others (institutions)? Maybe.
The risk/return profile should also come from separately identifiable risk factors. I tend to favor trend-following systematic managers in this space, and they are the managers often responsible for the non-correlation in periods of equity market drawdowns. I tend to lump global macro CTAs more in the hedge fund space. Despite all of the trend-followers’ models and mathematical talk, I believe managed futures simply provide leveraged exposure to the momentum risk factor (and to a much lesser extent in some currency and interest rate markets, political policies may cause autocorrelation that can benefit trend-followers). While academics like Carhart and firms like DFA recognize a momentum risk factor, the ability to profitably trade it is hampered by transaction costs. The futures arena provides a way to effectively bridge that gap.
Bottom line, there should be a separation of beta, “alternative” beta, alpha. I think financial research will eventually show more and more risk factors or alternative beta – and less and less alpha. I think of it as early physics evolving from knowledge of the atom (CAPM) to understanding the electrons, neutrons, and protons (3-factor model) to where we are today (momentum, the risk factors involved in convertible arb for example, etc).