Quant managers seeking new alpha sources

Active quantitative managers are trying to stand out from the crowd.

Active quant managers suffered a crisis in confidence when most of their strategies plunged in value simultaneously in August 2007 because they were chasing the same alpha sources.

Now, these managers are seeking out new sources of alpha — hoping to stay one step ahead of their competition.

“The first thing you have to do is to get out of the standard well-known sources and bring in factors that are not as well known,” said Robert Litterman, New York-based managing director and head of quantitative resources at Goldman Sachs Asset Management, which had $86.6 billion in quantitative assets under management as of Sept. 30.

The “quant quake” occurred in August 2007 before the financial crisis hit the broader market in 2008, said Daniel Celeghin, partner at Casey Quirk & Associates LLC, Darien, Conn., in an interview. “A lot of signals were being used by leading managers to buy the same stocks. The end result was a series of very crowded trades, and things turned out very badly” for quants.

“What we've seen since then is a lot of soul-searching,” Mr. Celeghin said. “Since '07, I don't know of any quant manager who hasn't taken a fresh look at the (investment) model.”

This is leading quant managers to scour for new factors to add to their investment processes.

“Basically, there are four things that we're doing to try to adapt to the new environment,” Mr. Litterman said at the Quant Invest 2009 conference in Paris earlier this month. “We're focusing on proprietary factors; we're focusing on allocating the risk across factors opportunistically; we're changing overall allocation of risk; and we're looking at event-driven opportunities.”

Managers will have to find new factors, new inefficiencies and generally new sources of alpha — all the while keeping an eye on the sort of extreme risks that trampled on performance in 2007 when the credit crisis hit, said E. Paul Rowady Jr., Chicago-based senior analyst at the TABB Group, a financial research and advisory firm.

Data for total assets in quant strategies are difficult to collect because many investment processes rely on both quant and fundamental drivers, consultants said. However the TABB Group estimated that quantitatively managed assets in U.S. equity — including passive and active strategies — totaled $3.47 trillion in 2009, or 34% of total U.S. equity assets. That's a 30% increase from 2008 and 9% more than 2007.

Managers are considering adding opportunistic investing drivers that include event-driven strategies, catastrophe insurance, commodities; credit, emerging markets and volatility trading. Timing factors also holds potential, according to several managers.

“It's difficult to time factors, so the question is whether it would be worthwhile to try and time factors,” Mr. Litterman said. “Modulating our risk is something else we think will be important going forward. ... What we're thinking about doing is allocating to signals that are less susceptible to financial disruptions, reducing the overall risk” of the portfolio.

“Adaptation is important,” said Richard Lacaille, global chief investment officer at State Street Global Advisors based in London. Quantitative managers will “need to mine new areas and new sectors.”

For example, one factor that might become a possible signal for quant models is human capital — generally defined as the set of skills and experience of employees — and its role in the success of a particular company. SSgA is increasing emphasis on some signals, including those related to timing and top-down analysis, Mr. Lacaille said in an interview. “We are doing more research in that area,” he added. SSgA had $34.6 billion in assets under management in active quant strategies as of Sept. 30.

Quant managers at the newly merged BlackRock Inc., which finalized its acquisition of Barclays Global Investors on Dec. 1, are looking at changes to investment models on four different levels since 2007: risk modeling; return-generating signals; data sets; and the nature of the factors “in terms of generic vs. proprietary,” said Michael O'Brien, managing director and head of the institutional business for Europe, the Middle East and Africa based in London.

“Today's quant models are much more granular and reflect the characteristics of a particular sector,” Mr. O'Brien said. These models aim for excess returns “much in the same way that a fundamental analyst would.”

At Mellon Capital Management Corp., the global quant specialist manager that's part of BNY Mellon Asset Management, officials are more focused on downside protection.

“We've dedicated a lot of research to modify our strategies so they may pay off when the (economic) environment doesn't look so good,” said Michael Ho, chief investment officer at San Francisco-based MCM. “When all assets (in the broader global markets) are falling, we would like to at least break even.”

According to consultants, developments in quant modeling bring a set of different challenges as well: These lesser-known factors likely will be more costly to implement, have less liquidity and limit transparency at a time when investors demand more of it. “Given what has happened, managers should adhere to the side of more transparency, not less,” Mr. Celeghin said.

Furthermore, rapid model development and increased competition in quant management will likely result in accelerated “alpha decay,” Mr. Litterman said, referring to shrinking excess returns as more investors access the same alpha source.

“Secrets don't last forever, so we have to continually look at the factors in our models,” Mr. Litterman said. “That's the key to being a good quant going forward.”

Some consultants question the ability of quant managers to build significantly different models compared to their competitors, given that they all use publicly available data sets to systematically chase alpha.

“There may still be too much money chasing similar signals, particularly in (market) conditions that are far from ideal for quantitative strategies,” said Peter Keutgens, senior investment consultant and a member of the quantitative and systems group at Watson Wyatt Worldwide, Reigate, England. “That still remains a concern.”

However, quant managers who have a process in place to invest in research and continuously challenge the assumptions in their models will have a better chance of succeeding in the future, said Michael Rosen, principal and chief investment officer at Angeles Investment Advisors LLC, Santa Monica, Calif.

It's too soon to tell how much of a long-term effect the changes in quant models might have on performance and overall risk/return profiles of the investment process, consultants and manager said. Some clients are monitoring the progress, however, looking for “strategies that will deliver true diversification,” said Pieter-Jelle van der Sluis, senior portfolio manager within APG's global tactical asset allocation team. APG manages about €205 billion ($302 billion) in assets, most of which belong to the €180 billion Stichting Pensioenfonds ABP, Heerlen, Netherlands.

Within the GTAA division, officials are considering niche strategies in liquid markets such as high-frequency trading and global macro strategies with an emerging markets focus. Such strategies would likely be outsourced to external managers who “can clearly add value,” Mr. van der Sluis said in an interview. APG's quantitative GTAA strategy is included within a broader absolute-return strategies portfolio comprising about 5% of total assets, according to information provided by the company.

APG's internal quant team is in the process of adding factor considerations such as certain behavioral biases. “A typical quant manager will stay with a trend until it reverses,” he said. “Within our systematic approach, we're looking at what we can learn from behavioral finance.”