Growth vs value is probably the longest-lasting dichotomy within the investment industry. Although money managers of each stripe will forcefully argue their cases, most retail investors opt for a middle-of-the-road blend. But for Joseph Piotroski, who teaches at the University of Chicago, there is no choice: value stocks are the preferred quarry for investors willing to do a bit of digging.

To determine if a stock is cheap, the most common measure thrown around is book value — the difference between assets and liabilities. If you buy a portfolio of stocks with a low price/book value ratio and hold them for at least a year or so, on average you’ll beat the market, suggests Piotroski’s research.

To prove his value thesis, Piotroski looked at a 20-year period in the U.S. market, concentrating on the bottom 20% of the price/book universe at the end of each year.
He discovered that even though the one-year performance of these stocks beat the market by almost six percentage points annually, the bulk of the gains actually came from fewer than half of the companies.

Piotroski, an accountant by profession, then decided to separate the winners from the losers by ranking these stocks on a numerical scale he developed. The best value stocks ought to be those with solid, preferably improving financials, while the worst ought to be at the opposite end of this spectrum, he reasoned.

Piotroski concentrated on readily available balance sheet values, assigning up to four points for profits, followed by a point each for positive earnings, positive cash flow, year-over-year earnings growth and cash flow that exceeded earnings. Companies earned one point if the past year’s ratio of long-term debt to total assets declined; a second if its current ratio improved; and a third if the company stayed away from issuing any common stock.

The net result: companies near the top of Piotroski’s rankings beat the market by 13 percentage points over one year, whereas companies near the bottom of the range trailed the market by almost as much. The pattern continued for two and, sometimes, three years, Piotroski reports.

Before you jump on board, however, keep in mind that many analysts believe that modern-day balance sheets fail to capture the true value of intangible assets such as human resources and patents, skewing the data accordingly. Still, this hasn’t stopped investors from embracing this or similar methodologies, screening for a miscellany of stats through any number of software programs and Web sites.

Many of these “quant” screens are available through a little-known Web site (www.validea.com). Rather than try to develop his own screen, owner John Reese, a computer programmer who became interested in the stock market in the 1990s, has adopted the investment methodologies of several lesser-known analysts such as Piotroski, as well as more familiar names such as James O’Shaughnessy, Peter Lynch and Ken Fisher, duplicating their investment approaches using various computerized formulas. Not that they have anything to say about it.

Reese’s program analyses a large universe of stocks to see if they meet the criteria that the various analysts would, according to their own writings and public statements, be expected to set. Reese then seizes the top 10 picks of each model and compiles them in virtual “guru portfolios.”

Overall, Reese’s returns look impressive, particularly the portfolio using Piotroski’s strategy. But with only 15 months of data, it’s hard to judge whether the strategy will produce consistently good returns in different market conditions.

A similar Web-based guru screener is offered by the vast Web site of the American Association of Individual Investors (www.aaii.com). It too features Piotroski’s work, along with several other even more obscure analysts. Both are pay-per-view sites. IE