Has Schwab Found the Holy Grail for Investors?

Charles Schwab (SCH Quote) claims to have unearthed the holy grail for investors: purely objective investment research.

This is surely great news, assuming, of course, that you believe that there is a holy grail to be found.

The "Schwab Equity Ratings" model, recently announced by the company, was designed by the quantitative analysts who came on board when Schwab acquired Chicago Investment Analytics in November 2000. (CIA co-founder Greg Forsythe now works as Schwab's director of equity research.) Schwab promises coverage of more than 3,000 stocks with more than $75 million in market value. The brokerage will provide a weekly screen for its customers based primarily on fundamental factors, such as earnings and cash flow, but will also take into account analyst upgrades and downgrades.

The stocks will be ranked, with the top 10% earning A's, the bottom 10% warranting F's and the middle 40% representing a solid C. 'A' rated stocks are expected to strongly outperform the market over the next 12 months, while 'F' rated stocks are likely to greatly underperform the market.

The system seems sound in its impartiality, although Schwab wouldn't comment on the specifics or the limitations of its model. In its big marketing push to attract high-net-worth individuals, Schwab is clearly taking advantage of the skepticism surrounding research provided by investment banks.

What's Its Value to Investors?

But Schwab's promise of objectivity raises an important question: Even assuming such a promise is possible, what good is it to investors?

"There are two types of objective research," says Art Micheletti, chief investment strategist at Bailard Biehl & Kaiser, portfolio managers for high-net-worth individuals in Foster City, Calif. "In the first type, the analyst is objective in the sense that he or she does the best job possible, without bias or an ax to grind. They have no incentive other than to make the right call. The second kind uses historical data, makes models and crunches data to get an answer as to the best stock."

Both types of objective research are fallible.

"The problem with the first is that [the analysts] can get the call wrong: They can look at the data and make their judgments and still get it wrong," Micheletti says. "The problem with the second is that the future doesn't look exactly like the past. Models blow up."

Schwab research surely is objective in the first sense: The firm has no investment banking ties that would taint its research. But the idea of a perfect quantitative model has most observers skeptical, to say the least.

"If there really was this magical algorithm, wouldn't someone have figured it out already?" asks Patrick Dorsey, an analyst with Morningstar. "And if someone had figured it out, wouldn't they be running a hedge fund from the Caymans?"

Schwab Equity Ratings seems to be derived from what's called a "factor-loaded" model, in which analysts look at years of stock data to determine what factors (such as earnings surprises, industry attractiveness, relative strength) correlate with positive stock returns. The model then takes these factors into account in determining which stocks are likely to outperform in the future.

"You can always develop a strategy that works in a sample," says Larry Swedroe, author of Rational Investing in Irrational Times. "But even if that strategy is effective going forward -- and most of them are not -- others will copy it, and your advantage disappears. The very act of exploiting an anomaly makes it go away."

That may prove particularly true in Schwab's case, given the scope of its customer base. "Schwab has a big audience," says William Valentine of Valentine Ventures in Bend, Ore. "There's a great chance that people will rush in and take out any advantage almost immediately."

Hope Springs...

But just like the enduring myths of elusive chalices, the idea of a model or screen that can objectively proffer sure-to-outperform stocks holds much allure.

Indeed, Schwab's Equity Ratings is just the latest incarnation of such an ideal. Value Line, the investment research firm widely touted for its unbiased and successful equity-ranking system, has been using its proprietary modeling to offer the best stocks to subscribers of its service. And while academic studies have suggested that the so-called Timeliness system, which uses earnings and price-momentum indicators to pick stocks, has consistently beaten the market historically, implementing the research to achieve real-world benefits has proved exceedingly difficult.

The (VLIFX Quote)Value Line fund is driven by the research firm's ranking system. The large-cap growth fund, however, has trailed the S&P 500 for the past six years. "Academic studies don't usually take trading costs, bid/ask spreads and the like into account," Dorsey explains. "All those real-world costs can really eat into returns, even if the model seems to outperform."

In other words, Schwab's research might be fine for investors looking to tweak their holdings or for those who will incorporate it into an amalgam of research, but don't create a portfolio around "A" stocks.

Objective research has its limits, particularly when it's model-driven. Remember, for instance, that Schwab's "A" stocks are graded on a relative basis; they're simply the best of the bunch. "If there comes a point when very few companies are attractive, wouldn't that be useful to know?" Dorsey points out. "Certainly, there was a time last October when stocks were much more fairly valued in general than in mid-2000 or even 1999."

Morningstar also provides objective investment research, but does not use a model. The firm uses a discounted cash-flow analysis: the use of information from a company's financial statements to project its future cash flows, then the discount of those flows back to the present to arrive at a fair value for the stock. The Morningstar analysts who perform this calculation, though, use their extensive industry knowledge and interaction with the company to supplement the stock analysis.

"If the point of research is to provide valuable and unique insight, then the more you get into systematizing the research through screens and consensus figures, the less valuable it becomes," Valentine says. "It's not that models hold no value, just less."