Managed Futures Matchmaking: Program Categorization the Sexy Way

 It was last week that we sent out our end of year recap of how managed futures had fared over the past 12 months. Bypassing the idea of doing a program by program breakdown in favor of strategy performance analysis, we found ourselves facing a common question- how, exactly, should these programs be classified?

We eventually settled on a fairly conventional approach to strategy categorization- guided by the elements that drive the performance of a program- and while these categories are certainly helpful when trying to understand the general performance of various strategies, when it comes down to evaluating individual programs and making decisions about your portfolio, it may not be the most elucidating method of strategy division. Let us explain…

The Problem

Anyone who has ever had to conduct due diligence research on a host of managed futures programs will likely tell you the same thing: the traditional labels we apply to these programs are rarely a perfect fit. Part of this might be attributed to the evolution of strategy development for CTAs as a whole. As managers have attempted to enhance their opportunities for profit and better manage their risk exposure, various elements of strategy have melded into a host of hybrid programs that are not so easily sorted into distinct piles.

For instance, the Rosetta program is the poster child for defiance of conventional managed futures categories. Their fundamental trading style intuitively places them in the discretionary trader category, and yet they are incredibly distinct from, say, a Dighton Capital. They also engage in complex spread trading strategies, and specializing in agricultural trading, but their discretionary approach makes them very different from systematic agricultural trading program Global Ag, and even more different from the hybrid spread trading program we see with Emil Van Essen. In this light, effectively categorizing these programs under traditional labels becomes exceedingly difficult.

For the organizational nutcases among us, the intuitive response is to develop some sort of strategy category hierarchy- a road map of reflective questions that ultimately spits out a very specific, final categorization. Even this approach, as appealing as it seems in theory, falls victim to a logical fallacy, insofar as it assumes that these categories are mutually exclusive, which is the only way that a branching categorization system works.

For some, the category conundrum may seem silly. At a certain point, you have to wonder, Why bother with sorting through the details at all? Why not compare the programs based on performance and risk metrics alone? Why not simply group the programs by their correlation levels?

You’re talking to a group of huge statistic junkies. We aren’t going to tell you that performance, risk and correlation aren’t important considerations when constructing a managed futures portfolio, because believe in their significance. However, just as these figures, in and of themselves, cannot paint a complete picture of a managed futures program, the numbers can fall short of helping investors really understand what, exactly, they’re investing in. Numbers may construct a vivid picture for the John Nashes and Matrix hackers of the world, but for most, they fall woefully short of establishing a clear explanation of behaviors and philosophies.

Perhaps more importantly, in a world of wild variance in track record length, assets under management and manager experience amongst the various managed futures programs, these data points cannot paint the panorama view of the managed futures landscape. Categorization provides an equalizer that is not dependent upon anything other than the defined mechanisms of trading associated with a given program. As such, an understanding of the strategic categories underpinning these programs can provide the context that connects the dots in a way that nothing else can.

Proper categorization can  provide valuable perspective for investors and portfolio advisors alike, but the industry’s attempts to batch up the entire universe of managed futures programs into neatly defined categories just doesn’t seem as effective they could be. After countless hours white-boarding out the issues, crunching the numbers, and writing and rewriting this piece, we finally realized what the solution was… in the most absurd of analogies.

A Relational Perspective on Program Categorization

Those involved in the managed futures space will tell you that advisers and CTAs don’t really sell managed futures. The asset class doesn’t fit into supermarket-style aisles of options (though, really- are supermarket aisles ever perfectly arranged, either?). Applying that kind of retail taxonomy to a process that only vaguely resembles the business models historically associated with selling is a mismatch of sorts.

No, advisers and CTAs don’t sell to clients- they court them.  We are the matchmakers, and our clients and managers are the potential lovebirds.

Stay with us, now.

Every investor is unique- they come with their own combination of prior investing experience, risk capital, risk appetite and tolerance levels, and understanding of portfolio construction. Sure- similar levels may result in similar investors, but no two clients are going to be exactly alike. In the same vein, managed futures programs are all unique as well. Sure- they may share common characteristics that make them easier to comprehend, but at the end of the day, the reason new programs are developed every year is because there will always be slight differences in how these characteristics are arranged. When you look at things from this angle, the matching of investors with managed futures programs requires a more nuanced understanding of personality composition than it does pitching a sale.

Now, admittedly, matters of the heart aren’t exactly black and white. Opposites can attract. Cultural divides can be bridged. Age gaps can be irrelevant… unless you’re Demi and Ashton. While investing certainly has its grey areas, we wouldn’t go so far as to say that this matchmaking analogy is perfect. We all know someone who just has a knack for setting up matches, often guided by intuition. Here, it’s not enough to have a “hunch” that an investor will hit it off with an individual program- you have to “psychoanalyze,” if you will, the compatibility of the two. This point is furthered when you think about managed futures categorization in the context of personalities. We’re not aware of anyone who requires a Myers-Briggs examination of their dates prior to dinner (and would be creeped out if we did), but by seeking to categorize managed futures programs, we essentially do just that… and it’s anything but creepy here.

For context, the Myers-Briggs assessment is a psychometric personality test that evaluates an individual’s association within four characteristic pools based on established dichotomies related to attitudinal, functional and lifestyle determinations. Based on the realm of possible combinations, the test outlines 16 different personality types. There may be variations within each four-characteristic combination, but the idea is to provide a general overview of a specific grouping of qualities.

In our mind, it’s high time that managed futures developed a Myers-Briggs test of their own.

If you look at a managed futures program’s strategy as its personality- or grouping of characteristics- the methodology of categorization- and investor analysis- shifts dramatically. It’s no longer just about including a spread trader in your portfolio or wanting some ag exposure; it’s about finding a program that most closely matches an investor’s risk tolerance, trading preferences and diversification on a totally different level. We won’t go as far as to we’re trying to be the Myers-Briggs of our space… but then again, maybe we will.

Introducing- the Attain Capital Program Category Assessment– or PCA.

While the differences between managed futures programs may be widespread and difficult to smash into a preconceived hierarchy of categories, the ways in which these programs can differ are more easily outlined. Based on our thousands of pages and hours of due diligence research, the major areas of distinction that we note are as follows:

Let’s break it down:

  • What They Trade- What kind of contracts is the program trading? Are they generally constraining themselves to outright futures trading, or are they participating in options trading? Does their program trade both kinds of contracts? If none of the above, the program likely does not qualify as a managed futures program. This element is an important consideration, as different contract structure participation can carry unique risk profiles.
  • Where They Trade-What markets does the program participate in? Do they, like a Cervino Capital Gold Covered Call Program, only trade one market? Are they focusing on a given sector, as stock index traders such as Paskewitz do? Have they built their program around a set of select sectors, like Rosetta’s trading of grains, softs and meats? Maybe, like Clarke Capital Worldwide, the program participates in a wide array of sectors. This characteristic is an important one for investors to consider. On one hand, focus on a single market or sector, particularly when traders rely on fundamental or discretionary signals, can play a significant role in the ability of a program to succeed, whereas diversification of market exposure may help other programs hedge against one trend wiping out the positions of an entire program.
  • How They Trade- If we know what contracts are being traded in which markets, the next obvious question is what the guiding philosophy behind the trades might be. The most common classification here is that of trend following or momentum trading, where a manager will attempt to latch on to the momentum of an emerging trade in order to capitalize on the price movements therein; these are the more traditional managed futures programs. Another strategy might be mean reversion. Call it contrarian or counter-trend if you will; this strategy bets that the trend will reverse. The medium between these two would be a relative value strategy, where positions are offset in such a way that they culminate in neutral market exposure- a strategy that some might include spread trading under. A program that doesn’t quite fit within these strategies may be using a hybrid of these perspectives, such as a program embracing a variety of trading models. This distinction can be indicative of the amount of risk taken on by a manager, and the volatility of a program in general.
  • Why They Trade-What makes a program place a trade? Is it a systematic signal, such as a breakout from a moving average? Is it a fundamental signal, such as a report on industrial demand for copper? Is it a blend of the two, where the manager executes their discretion by interpreting a variety of signals in order to gauge the prudence of a given trade? Maybe the program is a hybrid of these ideas, with a discretionary overlay when  it comes to a specific sector, and pure systematic reliance in another. The why provides an element of predictability- not necessarily in performance, but in process- that may appeal to investors of varying risk appetites.
  • When They Trade- Speaking of timing… this may be the more complex elements of categorizing managed futures programs, as the interpretation of the brightlines dividing their meaning may change depending on who you talk to. On our side of things, we want to know if a program looks to get in and out of trade within a one day period, a 2 to 7 day period, or longer. It’s entirely possible that a program will have a collection of models that trade on different timeframe levels. This element of evaluation can be incredibly important to understanding how a program works, as different timeframe guidelines for trades may perform differently under specific market circumstances.

That’s great… now how do I use it?

As we played around with the various combinations of characteristics outlined here, we found the descriptions they provided us with wound up being pretty accurate. This was a bonus… until we started to do a “strategy breakdown” following these lines in the sand; at that point, we were functionally were back to doing program by program reviews. That’s not necessarily a bad thing, but for functional purposes, it made the relational approach to categorization a little unreasonable for instances where you’re trying to look at things from 20,000 feet.  In fact, using this methodology, there are 764 different category profiles in the managed futures space- enough to make Myers and Briggs insanely jealous and drive the average investor crazy.

The relational approach to program categorization may not be useful to those doing big picture industry analysis, but its specificity can be of great use to the individual investor when it comes to understanding the differences between programs available and figuring out which are best suited for inclusion in their portfolio. In essence, the relational approach serves as a framework for evaluation of a program’s makeup; a 3D view of its composition and ideology. It basically provides a compatibility assessment for an investing relationship, forcing the investor to look behind the impressive physique of a program’s performance and the age of the track record to what lies within.

Now, just as we would never advise making a major relationship decision based on the outcome of the Myers-Briggs assessment alone (again, creepy- and are we really supposed to believe there’s only 16 kinds of people out there?), the relational approach outlined here is not the end-all, be-all for managed futures program selection. It’s just one piece of the puzzle, and is ultimately most helpful when used in conjunction with evaluation of past performance, various risk metrics, non-strategy related program attributes (assets under management, track record length, etc.), and the qualitative due diligence that dedicated IBs like Attain can provide. We understand how these pieces work together, and, further, that the road to allocation is a long and complicated one. That’s why we’ve invested so heavily into creating tools that make that journey a little more comfortable for the wary investor. In our experience, when you see the full picture- well, that’s when the magic begins. That kind of self-awareness and general understanding of the field in which you’re operating, in our minds, is a match made in heaven (sorry, couldn’t help ourselves).

Managed Futures 2011 Performance: Strategy Breakdown

As 2011 draws to a close and everyone begins to reflect over the happenings of the past year, those with money ‘at work’ usually become particularly pensive. Could I have done better? Could I have done worse? Does my portfolio need adjusting? The self-examination process this year is perhaps even more strenuous than usual. With managed futures set to round out the year with losses overall, those invested in CTAs may have more questions than usual as they look at what worked (if anything), and what didn’t amongst the managed futures programs they follow.

These are loaded questions- especially in the chaos that has been 2011- so we took some extra time to make sure the perspective here was what we needed. Plus, there will be no newsletter the next two weeks with our offices closed Mon. the 26th and Mon. the 2nd. We typically end the year by breaking down how individual programs had performed throughout the year, but usually end up repeating ourselves quite a bit through that process (i.e. this program, like the rest of the trend followers, did xyz because of the same market environment). To avoid this repetition, we instead looked this year to break down managed futures performance by strategy type.

The difficulty in this kind of analysis is establishing what categories are most appropriate to use. There are a wide variety of factors that can differentiate one program from another. That being said, in our analysis of a wide universe of managed futures programs, we found a series of elements that created unique enough distinctions to warrant specific categories.

To learn about the categories we decided upon, and how managed futures programs within those categories performed during 2011, click here.

Did ‘Risk On’ just sink Managed Futures again?

While the buy-and-hold crowd join the robots in pumping prices up- and then down, and then up and so on – are rejoicing today after stock market gains of 3% and more – managed futures managers may be saying “Not again…”

While not as fully short as they were at the beginning of October, most systematic multi market managed futures programs are still on the short side of the “risk on” assets (stocks, grains, metals, foreign currencies) and long side of the “risk off” assets (USD and bonds), meaning they are essentially betting on a continued ‘risk off’ environment. This has pushed many programs to small gains for December, as evidenced by the Newedge CTA Index up about 1% for the month.

A good portion of those gains will likely be given back after today, unfortunately. This is the same sort of pattern we saw at the beginning of October, when the sharp rally caused losses, and the same stance (although greatly reduced) they had at the end of November.  Both of those ended in less than ideal results (although the November move was more an open trade loss than realized losses), and now today is threatening more of the same.

Being In the holiday spirit, we’re looking for a silver lining here – thinking that maybe another failed move lower means we’re just one step closer to the real move happening. As we’ve explained before, traditional managed futures programs with a trend following type profile are designed to participate in many false breakouts, so that they can be certain of being in the real breakout when it does happen.

While each move may be independent of the ones preceding and following it, they aren’t as independent as a completely random event like the flipping of a coin. In that instance, despite the probability of a coin coming up heads or tails being 50% over many, many flips – the fact that the last flip was heads has no bearing on whether the next flip will be heads. We’re not so sure there isn’t some interdependence in market moves, however. Unlike the coin, the fortunes, hopes, and fears of market participants are interrelated and dependent on what has transpired before, and what they believe that experience means for the future. Those interdependences are infinitely complex, and managed futures managers aren’t trying to find the Da Vinci Code to unlock them – but it isn’t too far of a stretch to think that another trend reversal may mean we’re another step closer to the day/week/month when the trend extends, instead of reversing… is it?

Disclaimer: Past market movements are not necessarily predictive of future moves.

Conflict of Interest Behind Conflicting Testimony?

As testimony on Capitol Hill has continued, MF Global’s collapse has become (if possible) even more dramatic. Corzine’s strategy of pushing plausible deniability in his testimony fell apart with the testimony of the CME’s Terry Duffy, who argues that the law was broken and Corzine knew about it.

At this point, the question has become one of motive- not what Corzine’s motivations might have been as he dipped into those accounts, but whether the resounding indictment is a red herring ploy by the exchange to deflect criticism. One of the articles being passed around today questions Duffy’s testimony with a timid skepticism, though the headline- CME’s Duffy vs MF Global’s Corzine: A question of trust– pushes the idea with all the gusto of a Shakespearean rivalry:

The drama over the meltdown of the brokerage firm MF Global pivots around a clash between two veteran traders who rose from relatively humble roots to the very top of the futures-trading business.

One is Jon Corzine, the firm’s former CEO who just testified in Congress about the mystery surrounding some $1 billion in customer money that vanished from MF Global before it failed. The other is Terrence Duffy, the chairman of CME Group Inc, the huge Chicago exchange where MF Global did most of its trading.

At stake is not only Corzine’s reputation – and whether his career on Wall Street and in politics comes to an ignominious ending – but investors’ trust in Duffy, the CME and the U.S. futures industry, which is largely self-regulated…

A desire to shore up the CME’s image helps explain the forceful and at times personal tone of Duffy’s testimony against Corzine, said federal officials familiar with the matter.

“It doesn’t surprise me they are being so aggressive, they don’t have a choice,” one official said of CME’s handling of the matter. “They have a lot of people who lost money.”

Dennis Hastert, the former Republican speaker of the House who has known Duffy for years and who now sits on CME’s board, says of Duffy: “His reputation, his business, everything he’s ever worked for is on the line.”

Noting that MF Global was one of the largest traders on the exchange, Hastert says “the whole business works on trust, and when somebody breaks that trust, it jeopardizes the system.” Duffy “was not amused by the situation at all,” Hastert added.

In some ways, the logic here makes sense. Duffy and the CME find themselves in a perilous position. If Corzine was doing something fishy, to what extent are they liable for not catching it? It had crossed our minds that the CME would be feeling the heat in all of this, especially with the industry anger over their initial reaction to the MF Global collapse. But would that heat be enough to compel them to such tactics?

We’re betting no.

For starters, testifying in front of Congress is not for the faint of heart. Not only is your testimony broadcast to the world, but a misstep carries major legal repercussions, providing a disincentive to try to lie. This doesn’t mean that people don’t perjure themselves. If Duffy is right, that means Corzine already has. The difference here is what’s at stake. At this point, Corzine’s entire life has imploded; he has nothing left to lose. Duffy, on the other hand, would have a long way to fall if disgraced, and would leave himself and the CME open to potential civil legal action if the testimony had been fabricated.

Second, the CME has historically been very careful about what they’ll say and when they’ll say it. They carefully curate and guard their reputation, as one would expect when dealing with such a massive financial entity. This sort of disciplined message development is what drew the ire of so many in the days following MF Global’s filing for bankruptcy. Those calling the CME for comment or explanation found themselves quickly turned away; no one was about to say anything until they could do so with certainty. Ignore Duffy’s risk in this mix- the CME’s board and communications department would not have stood behind Duffy’s testifying if they didn’t have that certainty.

In our minds, this perspective makes more sense than the vindictive testimony angle- particularly when you consider the numbers. One might expect that November would have seen a drop in trading volume as MF Global, one of  the largest FCMs in the business, froze up billions in client trading funds, or as a result of concerned investors who hadn’t been directly impacted by MF Global pulling out of the markets. We certainly expected  the CME’s volumes to be down, and have been monitoring volume on a weekly basis looking for a trend to this effect. Turns out, that has not been the case.

You’ll notice that November saw some drops in equity index trading, energies and forex. You’ll also notice that, despite these drops, volume year to date is up across the board. In other words, the CME isn’t seeing a massive drop in business that’s compelling them to throw Corzine under the bus. This doesn’t mean that the CME doesn’t have more long-term concerns, nor does it mean that confidence in the system hasn’t been shaken. We’re just not seeing an immediate compulsion for recklessness, which, given the risk-oriented nature of our industry, definitely has us leaning towards believing Duffy over Corzine.

The Risk in What’s Being Risked

One of the managers we work with, Dean Hoffman of Hoffman Asset Management, has an interesting piece up on his blog today. As he points out, there are thousands of CTAs for an investor to choose from when they start investing in managed futures, and dozens of statistics you can reference as you sort through your options. Some of these are metrics you’ll find us frequently referencing on the blog, such as max drawdowns, length of drawdown and risk ratios like Sharpe, Sortino and Sterling. However, as Hoffman points out, these statistics, no matter how excellent, are going to fall short of what many think they can do:

What investors want (or should want) is excellent risk adjusted performance, but in my opinion, the standard performance measures only succeed at hindsight reporting. Those same measures perform miserably when trying to predict future risk adjusted performance.

In other words, metrics of evaluation for CTAs relate to past performance, which (say it with us, now) is not necessarily indicative of future results. In fact, Hoffman provides some pretty interesting charts which pretty effectively confirm this idea. What was most interesting to us, however, was his reliance on a metric which, in our opinion, is not often enough considered in the evaluation of a CTA: their margin-to-equity ratio.

The margin-to-equity ratio indicates what percentage of a CTA managed account is posted as margin, on average. Essentially, it tells us how much money they have tied up in margin at any given point in time relative to the nominal investment amount. For example, if you have a $1,000,000 nominal investment in a CTA, and the margin requirement on that account is $100,000 – the margin to equity on that account is 10% (100k/1mm).  Note that it is on the nominal amount, so if you have $200K traded as $1 million through the use of notional funds, and the same $100k in margin , the ‘official’ margin to equity is still 10%, even though it would be 50% on a cash basis.

Hoffman concludes:

…margin-to-equity ratios can be an excellent way to predict future drawdowns. Empirical data show us that higher margin usage leads to higher average and maximum drawdowns. Also, unlike returns and drawdowns, margin-to-equity ratios are fairly easy to predict.

… lower margin-to-equity ratios, correlate to superior risk adjusted performance.

In summary, we believe that to help prevent serious drawdowns and get superior risk adjusted performance that one is better off with managers who have low margin-to-equity ratios than with managers who have high margin-to-equity ratios.

While the conclusions drawn here are certainly worthy of contemplation, the word that jumps out at us is “predict.” It appeals to the most basic inclination of an investor- the desire to effectively see into the future and determine returns before allocating funds. Unfortunately, despite what the data in this piece may suggest, we don’t agree with Mr. Hoffman’s crystal ball theory.

The data analyzed here in regards to margin-to-equity comes from the Barclay Hedge database. The listed margin-to-equity ratios found for a program there are not calculated levels, but are submitted to the database by the managers themselves. Given that, while they give a rough idea of the average margin-to-equity ratio for each program, they aren’t necessarily reflective of actual margin-to-equity ratios. And unfortunately, CTAs rarely keep this statistic up to date in the various performance reporting databases. While they will update their monthly performance regularly, other information in the database can become a little stale, especially since updating that information is not a requirement for continued listing in the database.

Investors can always ask a manager about past margin-to-equity levels (and indeed, if a manager can’t provide a detailed report on their margin usage – that’s a red flag from a due diligence standpoint), but even here, the ratio can change substantially depending on the market environment. In fact, in our experience, this ratio, for most managers, ends up being fairly fluid, fluctuating between 10-30%, as managers engaged in profitable trades will often have higher margin allocations than they would at another point in time. The moving target here means that, unless you’re monitoring the margin-to-equity ratio for all of these programs in real time, the relationships you find between their ratios and performance are tenuous at best.

We’re not saying margin-to-equity ratios aren’t an important consideration; we use it regularly in our analysis of CTAs. It’s just that the metric should be more specific – with the low, average, and maximum margin to equity levels reported – not just a single number that we’re left to interpret (though it’s likely the average level). Much like the Sharpe ratio and max drawdown and other metrics, the margin-to-equity ratio is not a magic bullet when it comes to CTA evaluation. It’s only in the context of a myriad of other statistics that it has value, and, much to the chagrin of investors everywhere, it still can’t predict the future.