Are small-cap stocks running out of steam? Stephen Dean, co-manager of the Laudus Rosenberg International Small-Cap/Inv (RISIX), doesn't think so. "They've had a tremendous run in the last five years," Dean acknowledged, and as a result they are no longer as undervalued as they once were. "But there's no obvious measure that says small-cap stocks are overvalued relative to large-cap stocks either in the U.S. or outside," he added.
The $395.3-million fund invests in shares of international companies with market capitalizations of $20 million to $2 billion. As of June 30, it returned 25.3% annualized over three years, versus 18.2% for its international small-cap fund peers, and the 20.5% for its benchmark, the S&P/Citigroup EMI World Ex U.S. Index. Over the five-year period, the fund's gain is 11.4% annualized, versus 3.8% for its peers, and 7.4% for the benchmark. Its volatility, as measured by standard deviation, is slightly lower than its peers.
Dean summed up the fund's investment strategy in two words: "Fundamentals matter," he said. "The price you pay for the earnings potential of companies is really the driver of subsequent investment returns." The management team strives to buy stocks that are underpriced by 15%, and that have attractive near-term earnings potential.
To find those stocks, the company puts its Research Center to work, a team of roughly 25 headed by AXA Rosenberg co-founder, Barr Rosenberg, who also co-founded the risk management consultancy Barra. The team has developed two proprietary quantitative computer models that digest some 200 variables, from detailed balance sheet and income statement data, profitability ratios, and off-balance-sheet and footnote items. "You could almost think of it as a break-up analysis," Dean said. "What we're trying to do is maximize the potential return as identified by our models, versus the risk of deviating from the benchmark."
Working from the company's Orinda, Calif. headquarters, the research team combines the two quantitative models to produce a single number that ranks the relative attractiveness of each stock. "Then they go about building a portfolio that looks a lot like the benchmark from a risk profile [view], with similar industry weights, country exposures, and market cap profiles," Dean said. The management team accepts the average valuation for each industry as being fair. "We just want the cheap stock within each industry," he said.
To feed its computer models, the firm purchases data from about 50 suppliers, relying most on three of the largest: Compustat, a unit of Standard & Poor's; Global Vantage; and Worldscope. "We like to get multiple sources so we can find data errors," Dean explained. In addition, a staff of research data analysts also comb through the data to ensure accuracy.
Despite the degree of computerization, Dean stresses the fund's strategy is not akin to program trading. "We're picking individual stocks," Dean said. "We're in the quantitative camp, but rather than focus on a handful of unique stocks, we're trying to apply this discipline across 9,000 stocks and 200 different items."
With 1,039 holdings as of June 30, 2005, the small-cap fund's universe of stocks comprises the approximately 3,600 issues in the S&P/Citigroup EMI World Ex U.S. Index, plus others outside it, for a total of about 9,000 stocks. Dean estimates that 80% of the portfolio is invested in index-listed stocks. "The ones we pick outside the benchmark are typically relatively small stocks," he said.
Share are bought and sold for the fund by traders at the firm's London, Tokyo, and Singapore offices. (The firm also has a branch in Hong Kong, but does not trade on that city's exchange.) Traders can exercise personal judgment by stopping a transaction in response to breaking news, such as a just-announced merger or other information that has not yet made its way into the computer models.
Traders can also select from the daily computer-generated list, based on which issues are most liquid and available at the best price; a trading cost hurdle figured into the transaction. "We've kept our trading costs quite low," Dean said. "No particular name is crucial to us. If it's not available in the marketplace, we can just move down the list." However, traders are not empowered to choose stocks that are not on the daily list.
The fund holds its issues in the local currency of the exchange on which it was purchased, while its returns are translated into dollars. It does not hedge its currency risk. Traders sell the fund's holdings once they become more fairly valued, and as other issues with similar risk profiles become more attractive.
As director of AXA Rosenberg's Global Product Strategy Group, Dean has been co-manager of the team-led fund since its inception in October 1996, while also overseeing the firm's eight other funds. About 40 people in offices around the globe, including Dean's 11-person strategy group, review the model-building and trading processes to "monitor and provide perspective across a number of our strategies," Dean said.
For the past several years at least, the process has yielded strong results. "We're going to get 60% of them right, and 40% of them wrong," Dean estimated. "But if we can do that consistently, the law of averages will work out for us. We spread the bets over a large number of names, so no one name is going to make or break the portfolio."
The fund roughly follows the index's sector weightings, though its managers may vary those weightings slightly according to where the rewards are greatest and the risks lowest. "If a sector is less than 5% of the index or benchmark, we might be out of it completely," Dean said. "But if it's above a 5% weight, we're going to be in that sector." Individual holdings typically represent less than 1% of fund assets. "If the largest stock in the benchmark is only half a percent, we're not going to hold 10% of our portfolio in it. It's just too big a bet," Dean said.
The fund also tries to avoid what Dean called the "value trap," defined as buying stocks of companies that may go bankrupt, or that may stay cheap "for so long that you don't have the patience to stick with them."
As of June 30, the fund's turnover was 59.7%, compared with 101.3% for its peers. However, the average holding period is about a year -- a product of combining a roughly six-month earnings forecast and a two- to three-year projected valuation. "It doesn't go up to 200% or 300%, and it's not a long-term, 15%, Warren Buffet-type of buy-and-hold strategy," Dean said.
Originally named Rosenberg Institutional Equity Management, the firm sold an interest to AXA Investment Managers in 1999, and was renamed AXA Rosenberg. In early 2004, its funds were "adopted" by supermarket Charles Schwab, which took over their business functions such as marketing and distribution. AXA Rosenberg then became subadvisor in charge of managing the fund portfolios, which Schwab rebranded with the Laudus Rosenberg name.
"It's very information intense," Dean said of the fund's process. "If it can be automated, then automate it. Let human beings do the stuff they're good at, which is handling unique and exceptional cases." The firm employs some 250 people worldwide to run its nine funds, not including Schwab's contingent.
"There's a mess out there," Dean said, referring to the vast amount of data that the staff digests. "But if you actually look at it from various perspectives, it all starts to make sense -- if you're patient, and disciplined."
Contact Bob Keane with questions or comments at: firstname.lastname@example.org.