A Framework for the Pursuit of Quantitative Alpha

Where do we go from here?

Friday, February 14, 2014 , By E. Paul Rowady, Jr.

Even as the outlook for many of the traditional quant spaces dims, the need for firms to develop quantitative and automated capabilities is as great as ever. But where are the new opportunities to find alpha?This is a question being asked within most quant shops these days -- and, frankly, within most asset management firms as well.

The heady, pre-GFC days for high-turnover strategy profitability are over, and may never return in the same form again. Coordinated central bank intervention will continue to pour cool liquidity on anything resembling a hot spot for volatility spikes. And with the jury still out on whether high-frequency trading is a force for good or evil, regulators in some regions continue to contemplate tolls and taxes on speedy traders. Add the incremental costs for more speed, more computational horsepower and/or more data, and the outlook for some of the more traditional quant spaces seems a little sour.

TABB Group (my employer) has been firmly on the record (since the publication of "Quantitative Research: The World After High Speed Saturation" in June 2011) that strategy development will need to shift for most quant teams, but that quantitative research efforts will continue to grow and permeate more of the workflows for capital markets and asset management use cases.

In short, our recommendation has consistently been that most firms should opt to continue to develop quantitative and automated capabilities, even if the profitability is not as effervescent as in prior periods.

So, where do we go from here? The framework below is a concept that I have used for the past several years to describe the ranking of opportunities for highly automated (aka quantitative) strategies.

My position here is that the goal is to leverage quantitative/automated methods to optimize (not maximize) risk-adjusted returns (or Sharpe Ratio) and theoretical capacity. (Maximizing the Sharpe Ratio has a negative impact on theoretical capacity.)

Generically, this means that standardized products in liquid markets with high levels of automation (or straight-through processing (STP)) are going to be the leading factors for success with quantitative methods. Therefore, U.S. equities and equity index futures naturally became leading targets for quantitative pioneers. Exotic OTC derivatives, emerging markets or any other low-liquidity/low-automation scenario would fall at the opposite end of that spectrum.

Of course, the devil is in the details and, consequently, this framework is subject to significant debate. I think the area for greatest debate is along the "STP axis" in our framework model above. In past versions, I have labeled this axis "TOF," for "turnover frequency." Since automation, STP and TOF are highly correlated, they have periodically become interchangeable.

Drawing upon what I know today, the variables that should certainly be given heightened consideration include market structure complexity (think: exchange pricing models), liquidity fragmentation, technical maturity, evolving regulatory frameworks and perhaps even regional tax regimes. Moreover, maybe the whole idea of automation should be redefined to be more holistic. Perhaps "friction" would be a more holistic concept to introduce here, with technical and regulatory friction, among others, included under the broader umbrella.

The Trouble with Labels

Another major point here is the semantic distinction between asset classes and product classes. I often find that people in our business are too cavalier about the use of labels. Asset class vs. product class is one of the more common ones. Yes, certain asset classes are becoming more automated. But the implications for new trading strategies only happen at the product level -- sometimes, frustratingly, one product at a time.

Consider that quantitative trading in U.S. equities usually means the components of the S&P 500,sometimes means the optionable universe of about 3,000 names and never means every listed equity in the region. Why, then, should we expect any other asset class or region to be different?

Case in point:  today, quant opportunities in credit derivatives might only mean the top few CDX and iTraxx (multi-name) indices. Eventually, with more market structure evolution and regulatory alignment, credit derivatives opportunities for quants might include the emergence of a "credit index arbitrage" strategy with the underlying single-names.

Another example is that while quant opportunities in cash bonds mainly mean the top 200 issuers (at least in the U.S.), quant opportunities in interest rate swaps mainly mean cleared OTC U.S. dollar and euro denominated swaps (and might eventually mean IRS futures at CME and Eris). Moreover, for firms with the right portfolio of skills, the cross-listing of JGB futures in Singapore vs. Japan could certainly be very attractive. And the list goes on …

Parting Thoughts

It's important to remember -- and emphasize -- that speed is not the only way to go, even though it is developing a rich track record for strategies with the highest Sharpe Ratios. Dragging a well-worn concept from previous research out again, we remind you that faster, smarter and "dirtier" are the three pillars of trading strategy "DNA."

Chances are, with the increasing democratization of speed today, that new strategy development will need to place more emphasis on being smarter and digging where no one else is willing to dig.

E. Paul Rowady, Jr., is a senior analyst and director of the Data and Analytics practice at TABB Group, where he focuses on data management, risk analytics, technical infrastructure and OTC derivatives. He has roughly 25 years of capital markets and proprietary trading experience, with a background in research, risk management and trading strategy development. He also has specific expertise in derivatives, highly automated trading systems and enterprise data management initiatives.

TABB Group is spending the next several weeks looking at the quant landscape and collecting information on the road ahead from our unique community of thought leaders. We will also be launching a short survey in the coming weeks to help quantify some of these signals. We want to know: where do YOU, as a quant, go from here? Please join in the conversation by reaching out to me directly atprowady@tabbroup.com. We will target some unveiling of our findings toward the end of the first quarter 2014.