This is a story about trying to marry the investment management approaches of jocks and nerds.
Money managers who have sought to combine quantitative (the nerds) and fundamental (the jocks) research in building portfolios since the tech bubble popped in 2000 have run into difficulties because the two approaches sometimes produce conflicting results.
Tony Elavia, director of quantitative strategies at New York Life Investment Management, New York — which in the past year developed its own proprietary risk management and optimization tools — said one problem is the conflicting recommen- dations made by fundamental and quantitative researchers.
"Sometimes, a fundamental manager will recommend a certain weighting to stock, and then that stock will be screened out by the quantitative researchers, and then the fundamental guys ask, ‘What happened?' It's something that's difficult to manage," said Mr. Elavia.
One fundamental large-cap value portfolio manager who did not wish to be identified recalled "shouting matches" between the jocks and the nerds. "We fundamental managers think we're all geniuses in our stock selection, and then quant guys are quick to point out that their models would have weeded out stocks that lost money."
Despite the conflicting recommendations and ego clashes, Merrill Lynch Investment Managers, Putnam Investments, ING Investment Management and New York Life are finding ways to combine the two techniques.
Quantitative portfolio managers weed out stocks from a portfolio via multifactor optimization models — mathematical models that factor in various pricing characteristics of individual securities. The pricing characteristics include earnings momentum, capitalization size, and price to trailing or forward-looking earnings.
Fundamental researchers also consider pricing information, but focus more on tearing apart the balance sheet of the underlying company and using that information as a basis for deciding whether the security should be included in the portfolio, and what weighting it should receive.
The movement toward combining the two approaches has been evident as institutional investors in the past few years have poured more assets in to enhanced index portfolios — tightly risk-controlled portfolios that are meant to notch 200 to 300 basis points over a given index. Enhanced index portfolio managers often use a blend of the two research methods to weed out overvalued securities from a given benchmark as a way to produce excess returns.
Assets of the top 200 defined benefit plans in enhanced index securities in the 12 months ended Sept. 30 increased 27% to $213.6 billion, from $168.5 billion a year earlier. That follows gains of 27% for the year ended Sept. 30, 2003, and 28% for the same period ended 2002 (Pensions & Investments, Jan. 24).
Jeffrey MacLean, president and chief operating officer of consulting firm Wurts & Associates, Los Angeles, said investors have been demanding a greater level of risk management from money managers. "After the tech bubble burst, some money management firms that were overweighted to tech stocks have added aspects of quantitative management into their portfolios, to ensure that they do not select overvalued stocks. Since the bubble burst, plan sponsors have also become more sophisticated and are requiring greater levels of reporting from their money managers."
Kevin Divney, chief investment officer of midcap growth at Boston-based Putnam and lead portfolio manager for its well-known Vista fund, said the firm's portfolio managers consider both fundamental and quantitative characteristics of a stock to determine whether it makes the cut and what its weighting will be. Mr. Divney said he began developing a system that combines fundamental and quantitative management in 2000 to shore up performance in the firm's growth equity portfolios. Previously, the firm kept its quantitative and fundamental research teams as separate units.
Mr. Divney said Putnam resolves conflicts between the jocks and the nerds by strictly sticking to a set stock selection system. "The way to resolve that is that we use the same criteria to value stocks across all of our portfolios. Our criteria look at valuation, quality, growth, and then quantitative and fundamental subfactors of each and consider both. For example, we own XM Satellite Radio, and because it's in a brand new industry you can't really forecast its growth. But because of our system, we don't have a huge position in it." XM Satellite Radio Holdings Inc. is digital radio service that broadcasts signals via satellite.
"The best way to avoid conflict between the quant and fundamental researchers is to align incentives — that's really the best tool you can use." Mr. Divney said. The incentives: performance-based monetary rewards the common desire to achieve client objectives, he said.
Brian Fullerton, head of global risk management at Merrill Lynch Investment Managers, Plainsboro, N.J., said one of the things driving a movement toward quantitative management is a stronger demand from investors for performance attribution. "Clients want to understand what your biases are in your portfolio," he said. "They are very focused on attribution and on what the drivers were in your portfolio and where the returns came from."
"We feel we've overcome the potential difficulties (of having nerds and jocks working together) because of the system we've set up," said Mr. Fullerton. "We have a central quant team that provides inputs. We call the quant team the risk analysis team, but we don't have many disconnects between our portfolio managers and quantitative researchers because the individual portfolio managers own the risk in their portfolios. If there are any disconnects, it becomes their call. What the quant team does is narrow down the universe of securities for the portfolio managers."