dormouse: Quiet creation of systematic alpha

June 16, 2016 09:00 AM

The first thing you’ll notice when encountering Martin Coward’s new managed futures fund is the unique spelling of its name: dormouse, with a lower-case “d.” Unlike most alternative asset managers, whose names invariably invoke mythological figures, predatory animals or park fixtures like squares, stones and bridges, dormouse is named after one of the most unassuming and inauspicious creatures on the planet: The mouse. 

“Our name — and the way we spell it — really sums up the ethos of our company,” says Coward, well-known through the success of his original quantitative hedge fund, IKOS Asset Management, and whose subsequent split from his partner/spouse Elena Ambrosiadou in 2009 became tabloid fodder for years. “We emphasize scientific rigor. For us, that means we’re primarily concerned with a consistent application of analysis and programming throughout the investment process, and less on developing the typical hedge fund brand. In fact, we want to be the veritable antithesis of your typical hedge fund.”

“Finance is just an application of science, not the science itself,” he continues. “We’re taught all these theories about financial markets, but they don’t actually work, at least not consistently. What does work is the science, so that’s where we focus.”

Dormouse was founded in 2011 by Coward and three associates who previously worked together at IKOS: COO George Dowdye and portfolio managers Hans-Joachim Drescher and Eric Westphal. Both Drescher and Westphal have Ph.D.s in physics, while Coward himself has a doctorate in mathematics from Cambridge. Together, the four have built a robust quantitative systematic electronic statistical arbitrage fund whose performance since inception has been very impressive (see “The mouse that roars,” below). Indeed, Coward’s dormouse was up 30.27% in 2015, compared to a -1.96% loss in the HFRX Macro/CTA Index, and is ahead 6.78% for the year through May 1 —  well ahead of the benchmark’s 0.41% return.

“Our overall philosophy is to equally allocate risk across all alpha sources,” explained Dowdye. “Risk allocation is much more even across the portfolio than with most managers, so we never favor one over another. Accordingly, there is a relatively equal likelihood of each risk happening.”

At its core, dormouse is a systematic multi-strategy fund that uses relative value statistical arbitrage to trade futures. It does not rely solely on any particular strategy such as trend following, mean reversion or momentum strategies common among its peer group. Instead, the fund looks at shorter time periods, other models and the relationships between futures markets and underlying economies to identify over- or underpriced liquid securities across stock indexes, interest rates, currencies and commodities, as well as exploiting the correlations between them. It sounds simple, but it isn’t.

“Our portfolio risk fluctuates closely around the average annualized risk level,” explains Coward. “Construction is done through an aggregation of alpha according to strategy weights, which are then scaled to the target risk of the portfolio using an adaptive co-variance matrix. We look at the whole universe of futures contracts and securities, instead of just one segment or sector, and analyze everything, particularly the relationships between them.” 

“We seek uncommon models,” Dowdye says. “We have found ways to reduce outright factor risk, and methods to neutralize the kind of volatility experienced when markets alternate between risk-on and risk-off phases. That’s one of the real advantages of what we do.”

“We’ve structured dormouse such that our net market exposure is much less subject to that directionality,” Coward says. “These shifts back and forth from risk-on to risk-off — they can be deadly if you haven’t addressed the factor risk. Our system removes, or at least lessens, risk factors common in other funds, such as equity and bond market risk, trend exposure, etc., to create not only alpha, but better risk-return performance.”

Interestingly, when asked whether those direction shifts ever enter into dormouse’s investment opportunity set, Coward is quick to point out that the whole system actually is aimed at exploiting herd behavior to some degree. “Ours is a standalone alpha approach,” he notes. “We focus on performance and providing an independent source of return in an environment with negative interest rates [and] negative carry.” 

And while not totally absent, subjectivity is equally on the sidelines. “We have gut instincts like everyone else,” Coward says, explaining that being a data-driven quantitative shop also means one can constantly adapt and refine models before putting real money at risk. “To some extent, the analysis and the allocation of risk to the model set could be considered subjective, since both are done by humans. But it really is much easier to let the system do the hard part. It’s much simpler.” 

It’s also much more scalable. Malta-based dormouse opened to outside capital in 2014, and now that it has five years of admittedly extraordinary results, is rolling the strategy into a Cayman Islands-domiciled hedge fund. “The hedge fund space has become a bit of a benchmarking business, so we knew that we needed to show true alpha over a significant period before launching the fund,” Coward explains. “From the word go, we built a strategy that we would want to buy ourselves. With a five-year track record and a Sharpe ratio of 1, we now want to bring dormouse offshore, raise some assets around what we’re doing, build out our research team and further refine/develop our strategies.”

Part of the company’s expansion is the establishment of a technical research center in Miami that stands on a separate leg from the Malta-based unit and has been nicknamed Nest. “Nests are places from which one grows, so we thought the name was appropriate,” explains Dowdye. “We intend to grow from nine associates now to around 30, most of whom will be hard-core scientists involved in technical research in Nest.”

The technical center reflects the company’s whole approach. Eventually, the center will be staffed with scientists, mathematicians and programmers, with little emphasis on finance per se and a significant emphasis on the results-driven, peer-reviewed scientific approach to the world. “Hedge fund managers don’t typically operate in a peer-reviewed environment,” Coward says. “But it’s what we live. It is much easier to strip egos when each member of a group is subject to peer review all the time. In science, merit wins, and it is always anchored on data that is both provable and repeatable. No one gets their nose out of joint if a team member points out a flaw in a data set or suggests a tweak to an assumption – it’s a necessary part of the process. We sharpen our pencils and try again. Everyone knows the approach ultimately results in a stronger end product.” 

The company believes capacity in the strategy vehicle can be as much as around $2 billion unless conditions change (which the team considers unlikely), after which dormouse will close to new capital. “We want to be small enough to know each other and our value, and big enough to be credible yet still run the strategy effectively,” Dowdye adds. “When you run tens of billions, you just can’t deploy effectively into niche, unique strategies like ours. We may develop new models in [the] future, but we don’t want size to deteriorate the risk/reward aspects of any program. We’d be cutting off our own nose to spite our face. Being a big business is not that interesting to us; being an effective one is.” 

The new fund will carry two classes, one heavily oriented to large institutional investors and one for the smaller appetites typical among high-net-worth investors and family offices. In each case, dormouse has structured fees to stand out from the crowd. The institutional class, which has limited capacity, carries minimums of $10 million but will charge 1% in management fees and a 10% incentive fee, whereas the much larger “retail” class will have $1 million minimums and charge fees of 1% and 20%, respectively. 

“We’ve listened to all the commentary in the hedge fund industry about low returns and high fees,” says Dowdye, who is heading up the marketing effort. “We decided to see what people thought of high returns with low fees.”

“Many investors try to understand your models; some investors overlook emerging managers because AuM is too small,” he continued. “They are more inclined to invest with large multibillion dollar managers with a lot of employees rather than ‘niche boutiques’ like us. They buy the shiny penny. Instead, they should focus on whether a manager has a sound basis for their models, a good research and investment process and whether the rationale for the model should result in independent repeatable alpha — and sometimes it is best found with the not-so-shiny penny. True alpha starts with individuals and an organization that thinks and does things differently. We believe we have created a process and organization that produces alpha consistently, so that is what should matter. Everything else is really a distraction.”

The dormouse team knows attracting assets to a new fund, even one with an extraordinary five-year track record and well-known founder at the helm, is an uphill battle. The start of 2016 was the worst since 2012 for your average hedge fund, according to industry data provider Preqin. Investors pulled the most from the sector since the second quarter of 2009 and hedge fund closures outnumbered startups in 2015 — 979 to 968 — for the first time since the financial crisis. 

As for assets, another trend in place since the financial crisis is the overwhelming majority of new institutional investment into hedge funds going toward the largest, most established money managers with the strongest brands. “Allocators are human, after all,” Dowdye says. “Allocators take a risk when betting on a small, emerging manager, or an outlier fund with a unique approach. No one has ever been fired for putting money with Bridgewater.” 

Dormouse is also exploring a more liquid alternative strategy. Nicknamed “hedgehog” and still in the incubation phase, the strategy will take a balanced risk approach in exchange-traded funds across different asset classes: High yield bonds, investment grade paper, equities and commodities. “Hedgehog will be our version of a risk-parity program,” Coward says, pointing out that much more testing needs to be done before hedgehog hits the market. Nonetheless, hedgehog would most likely come in a 40-Act mutual fund wrapper or a UCITS structure with low fees and comparatively high liquidity, giving individual investors access to what the company is doing. 

“Hedgehog is a similar idea to our futures-oriented vehicle but takes a longer-term view and works within more traditional equity and bond markets,” Coward says. “We expect fairly low correlation to our short-term strategy, and think overlap between the two investor bases will be relatively low.”

Ultimately, only time will tell if dormouse has actually invented a better mousetrap. The history of quantitative hedge funds is replete with formerly high-flying systems that worked flawlessly — until the day they didn’t. But for moment, the firm is living up to its motto of “nibbling away unnoticed.” 

“There are plenty of people out there delivering what seems to be just leveraged beta,” Coward says. “There has been a tendency on the futures side to take risk off the table over the past five years. It’s understandable, but not where we want to go. We focus entirely on performance so we can be an independent, consistent source of return for our investors.”

About the Author

Steven Lord is Managing Editor of and the founder of Modern Money Group. He has studied bitcoin since shortly after its inception, and has become a leading expert on the digital currency ecosystem. Lord has led several companies in the financial technology, investment media, brokerage and asset management sectors in the United States and Europe.