By LESLIE P. NORTON
Over the past year or so, Vanguard chief Jack Bogle has become one of the biggest nags in the mutual-fund universe. He has loudly declaimed, to just about anybody who'll listen, against what he views as the evils afflicting the industry, including excessively large mutual funds, high management fees, mediocre performance and insensitivity to tax consequences. While some fund companies have taken steps in line with Bogle's suggestions-Fidelity, for example, last week closed three of its funds, which it said had grown too big ("Closing the Door at Fidelity") -- the industry as a whole, he argues, is ignoring its investors' best interests.
To some fund watchers, Bogle's criticism hasn't been unexpected. After all, one of fundom's biggest challenges is presented by indexing -- and Vanguard is probably the king of indexing. Less well known, however, is that some of the inspiration for Jack Bogle's musings derives from his son, John Jr., and the younger Bogle's partner, Lang Wheeler. The two, ironically, are active managers who run Numeric Investors, a small money-management firm in Cambridge, Massachusetts.
Look behind some of Bogle pere's more contentious pronouncements over the past year, and you'll find the footprints of Bogle fils and Wheeler. To bolster his argument that largeness damages performance, Bogle senior has cited excessive trading costs. Source? Bogle Jr.'s analysis of thousands of trades made by his own firm. He discovered that the more assets a manager supervises, the longer it takes to buy a stock, and the greater the difference between the target price and the actual price. Bogle Jr. calls this an "opportunity loss." And that leads to worse performance. Last summer, Bogle pere took on the latest rage sweeping the fund industry: tax efficiency. His conclusion: Indexing usually is superior to active management in reducing the tax burden on individual investors. One source: Bogle Jr.'s ongoing search for ways to ease taxes for his own investors. (He hasn't yet succeeded.)
"John has become something of a missionary," chuckles his father. "He says things about the industry that I couldn't say. So I quote him instead."
Bogle Jr. and Wheeler are acutely sensitive to such issues because of their aggressive investing style. Numeric is a quantitative investing shop, boasting 125 computers and nearly 18 miles of cable wire. Its analytical models rely heavily on quick-trigger momentum investing. Turnover exceeds 200% a year. In his analysis of trading costs, for example, Bogle Jr. discovered that the prices of stocks identified by his firm's small-cap growth model rose 2.9% in the first 10 days following the first purchase, while those of stocks in the value model climbed 0.5%. The problem worsened as the portfolio got larger. Bogle and Wheeler would argue that no matter what type of investing you practice, the bigger your portfolio, the bigger the opportunity loss.
Numeric limits the size of its funds, and caps the bite of its fees at 1%, a fairly low level for active managers. As for returns, they've been impressive.
Numeric has $4.2 billion under management. About $3.8 billion of this is in managed accounts, which consistently have outpaced their benchmarks. The rest -- around $410 million -- is in the Numeric N/I Mutual Funds. These funds, most of which were founded in 1996, have done well against the industry's merciless bugaboo, the S&P 500.
Bogle and Wheeler are upfront about their methodology's limitations. Bogle Jr. figures their stock picks are successful just 55% of the time in strategies that use a complex combination of earnings momentum and value models. But skillful trading also helps them to cut losses early. "It's very much like playing blackjack," quips Bogle. "If you play enough, you can do very well." Assuming, of course, that you know how to play your hand.
All the portfolios rely on a combination of quantitative models, which assign scores to each of the 2,400 stocks Numeric follows. One model, called "Estrend," is based on the premise that when a company reports an earnings surprise, analysts underreact, upgrading forecasts slowly because they don't like to stray from the herd. That creates opportunities. Estrend finds stocks for which profit forecasts are rising, and it can identify sell candidates whose estimates are falling. The model adjusts for risks related to the forecasts. For example, in its view, stocks followed by fewer analysts, or for which estimates are widely spaced instead of closely grouped, deserve bigger discounts.
Another of Numeric's models tracks "fair value," based on the premise that volatile stock prices eventually revert to a mean related to fundamentals, including earnings, book value and the quality and growth of profits. The Numeric managers find that momentum works better for some stocks, and fair value for others. Some sectors, such as biotech and gold, are insensitive to both. You won't find Numeric buying stocks in these groups.
Most of the money Numeric runs in managed accounts is in portfolios that can own both long and short positions and whose holdings primarily are determined by the Estrend or fair-value model. With a long/short strategy, Numeric contends it can add value both ways. To ensure exceptional returns, the accounts bear performance incentives for the managers.
A sampling of Numeric's current favorites shows that it isn't afraid to pay up. The models are signaling a buy on Arterial Vascular Engineering, which makes tiny metal scaffolds to prop open blood vessels after a balloon angioplasty procedure. Arterial's version of these scaffolds, called stents, were recently approved by the FDA. The shares have soared -- recently, they were at 40 -- and have become a target of short-sellers. But Numeric has picked up shares, with Bogle citing dramatic growth and estimates for fiscal 1999, ending June, jumping to $1.67 a share from $1 a share three months ago. Numeric expects estimates to rise further. If the shorts are right, of course, Numeric could just as easily pull the plug on the stock.
The models also dislikes a raft of companies that they believe have deteriorating earnings outlooks, including Nike and Monsanto.
If Bogle and Wheeler are obsessed with one thing, it's called alpha. That represents consistent outperformance of a passive benchmark like the S&P 500, what they call "excess return." The term crops up over and over again in their speech. Says Bogle, "If you can generate substantial excess returns, your clients will let you take a very fair fee."
To that end, they're constantly tinkering with the models to ensure the best returns possible. The models can be subject to what Bogle describes as secular decay, a phenomenon that occurs when so many people use a strategy that returns from it markedly diminish. Bogle observes, for example, that more and more folks are using earnings surprises to pick stocks. Corporate managers also are playing to the strategy. That's reduced the power of the model, although Bogle points out that the effect will never completely vanish because analysts will continue to act as a herd.
Numeric is always seeking ways to slash costs. The study on transaction fees, for example, continues. Over the past 2 1/2 years, Numeric has measured 50,000 of its own trades, worth some $50 billion.
The firm also has examined its models for signs of potential trouble. Recently, for example, the short sales triggered by Estrend haven't quite worked out. Why? Because some stocks that have collapsed on worse-than-expected earnings news have bounced back quickly -- sometimes on the same day. So Wheeler is creating a vast database of price reactions to study this behavior and determine whether it should adjust its trading patterns. Similarly, the firm constantly adjusts its models.
Bogle Jr. and Wheeler seem an unlikely couple. Bogle, 38, is invariably sunny. Wheeler, 55, often is gloomy and apocalyptic. Wheeler is a former Eagle Scout who attended Yale and served in Vietnam. Lanky and good-looking, he describes himself as a "dork" who has never held one job too long. Bogle, on the other hand, brims with enthusiasm and has a long background buying stocks. As a teenager, he decorated his bedside lamp with ticker symbols of stocks he owned, then studied finance and economics at Vanderbilt University. However conventional his upbringing, however, Bogle comes from a long line of rabble-rousers. His great-grandfather, Philander Armstrong, was an insurance executive who once penned an industry expose, entitled License to Steal.
At State Street Bank, however, Wheeler and Bogle developed a tight friendship based on a shared interest in using computers to profit from investors' behavior patterns. When Wheeler left to found Numeric in 1989, Bogle followed soon after. Mark Engerman, now 30, arrived in 1994. While trying to pitch Bogle and Wheeler on a software package developed by Barra Inc., Engerman was hired by the two instead, and is now a partner in the business. The firm has now grown to 30 employees, including three more portfolio managers: Arup Datta, Steve Cusimano and Shannon Vanderhooft. But Bogle has provided the articulate public presence that most firms that make their living crunching numbers dream about.
So if the managers are continuously closing funds, how do they grow the business? By starting different portfolios. While the firm began by offering momentum-investing portfolios, it's now migrated into value investing and other strategies. Numeric's offices overlook the Charles River and are a stone's throw from the campus of the Massachusetts Institute of Technology. Naturally, this gives the group access to a vast pool of statisticians and econometricians.
Recently, they hired Robert Furdak from State Street. Furdak specializes in international equities and will start investing in foreign stocks. Numeric nearly has completed a Japanese database and believes that momentum models will start yielding profits in Japan. Also in the works is a European portfolio. Ultimately, though, asset levels will determine what funds they offer. Thus, though market-neutral funds are a new fad in the mutual-fund industry, you won't find one at Numeric. They already run enough money in this style; any more, says Bogle, and they'll lose their edge.
Needless to say, these views aren't popular with other fund companies. Numeric managed mutual funds for the Quantitative Group for almost four years. But in 1996, Wheeler and Bogle left in a spat about fees, questioning high fund expenses and the continuation of the funds' 12b-1 marketing fees at a time when the funds were closed to new investors. That made the folks at Quantitative roll their eyes, because Numeric had known from the start that the funds were sold by brokers, who desired ongoing compensation.
How long will they be able to keep their winning edge? Bogle & Co. are convinced that bigness is the route to mediocre returns. So they intend to stay small. That makes Don Phillips, president of Morningstar Inc., the Chicago-based fund-rating concern, a fan. Says Phillips: "If you ask me to name one fund manager in America who has the best chance of continuing to beat the index, I'd pick John Bogle Jr."
So, what happens now? At any other firm, bigger assets are the natural outgrowth of success, but hugeness isn't in the cards.
Says Wheeler: "It is such a rich garden. Why do you need to own half of it?"
Questions: