This Issue In Spanish

This Issue In Chinese

Archives

About

Contact Us

Separating Winners from Losers

Investing in firms with book values that are higher than their market values isn't always a bad investment proposition.

Research by Joseph D. Piotroski

High book-to-market (B-to-M) firms tend to be in poor financial health, as reflected by their low stock prices and poor earnings performance. Yet research consistently shows that a portfolio of these "value" firms outperforms both the overall market and portfolios comprised of low B-to-M "glamour" firms.

The reason for this is because a small number of high B-to-M firms are strong enough to raise the portfolio's mean performance, compensating for the many high B-to-M firms that under-perform the market. Wouldn't it be great to have a way to distinguish prospective winners from likely losers? A University of Chicago Graduate School of Business professor thought so, too.

Joseph D. Piotroski, assistant professor of accounting at the University of Chicago Graduate School of Business, suspected that applying simple financial statement analysis techniques to a high B-to-M portfolio could differentiate strong firms from weak firms.

The approach was designed to allow "investors to form a value portfolio that consists of only the firms with the strongest performance prospects," says Piotroski. Such a portfolio, he says, "should therefore outperform a generic high B-to-M portfolio."

His results, reported in the paper "Value Investing: The Use of Historical Financial Statement Information to Separate Winners from Losers," show that he was right. The average return earned by a high B-to-M investor can be increased by at least 7.5 percent annually. And when Piotroski applied his investment strategy across a 20-year time period (1976-1996), he generated a 23 percent annual return.

How to Predict the Future

"The success of the strategy," writes Piotroski, "is based on the ability to predict future firm performance and the market's inability to recognize these predictable patterns." The market's treatment of high B-to-M firms, according to Piotroski, is based on several factors.

First, poor historical performance appears to lead to overly pessimistic expectations about future performance, causing the market to temporarily misprice these firms. Second, these firms suffer from investor neglect: analysts typically do not follow high B-to-M firms and many investors are not interested. Instead, analysts and investors favor low B-to-M "glamour" companies with strong positive momentum. As a result, both analyst forecasts of earnings and stock recommendations are often unavailable. Finally, high B-to-M firms have limited access to informal channels of information distribution and voluntary disclosures about future performance lack credibility due to their poor recent performance.

How then did Piotroski predict future returns for these firms? The answer lies in their financial statements. Because high B-to-M firms are financially distressed, their valuation should focus on near-term economic fundamentals such as leverage, liquidity, cash flow adequacy and profitability trends. For these types of firms, financial statements represent both the most reliable and most accessible source of information about their financial condition.

Other studies have used accounting and market data to predict future earnings and returns. Piotroski believes that, while successful, these studies are limited by their incorporation of "complex methodologies and a vast amount of historical information." By contrast, his strategy is simpler to understand and easy enough for both financial analysts and average investors to implement.

Take Out the Calculator

Rather than examine the relationships between particular financial signals and future returns, Piotroski created portfolios based on a firm's overall financial health, or F_SCORE. He chose nine fundamental signals that attempt to measure various dimensions of each firm's financial condition.

Based on each signal's realization, he assigned a "1" for "good" signals about firm performance or a "0" for a "bad" signal. The sum of those signals, ranging from 0 to 9, is the firm's overall fundamental signal. These nine signals measured three areas of a firm's financial condition: profitability, financial leverage/liquidity, and operating efficiency. The higher the score, the greater the likelihood the firm will earn positive future returns.

When measuring profitability, Piotroski assumed that positive cash flow or positive net income demonstrates the firm's ability to generate funds through operating activities. Similarly, a positive earnings trend suggested an improvement in the firm's ability to produce future positive cash flows, while cash flow in excess of net income demonstrated that earnings are not driven by noncash accruals.

The second set of signals measured changes in the firm's credit risk. "Since most high B-to-M firms are financially constrained," writes Piotroski, "I assume that an increase in leverage, a deterioration of liquidity, or the use of external financing is a bad signal about financial risk."

Leverage was measured as the historical change in the ratio of total long-term debt to average total assets, while liquidity was measured as the change in the ratio of current assets to current liabilities. An improvement in leverage suggested an improvement in long-term solvency, while an improvement in liquidity was a good signal about the firm's ability to service current debt obligations.

Finally, Piotroski assumed that raising equity capital was a negative signal since it highlighted the firm's need to use external financing to support operations and investment activities.

The remaining two signals were designed to measure changes in the efficiency of the firm's operations. The eighth signal measured changes in the firm's gross margin percentage. According to Piotroski, an improvement in margins was a good signal, indicative of a potential improvement in factor costs, a reduction in inventory costs, or a rise in the price of the firm's product.

The ninth signal measured the change in the firm's asset turnover ratio. An improvement in asset turnover signified greater productivity from the asset base through either more efficient operations or increased sales demand.

While Piotroski acknowledges that these are not necessarily the optimal set of performance signals, they are a set of metrics that are commonly used to assess a firm's economic condition.

"They are not 'optimal' in the sense that they do not represent outputs from a mathematical model or equation," he explains. "Instead, they are an intuitive set of performance measures." Moreover, they are easily implemented for analyzing one firm or an entire portfolio of firms. But most importantly, Piotroski's results show that these metrics work.

Putting the Theory to the Test

In conducting his research, Piotroski needed a substantial amount of data. Using COMPUSTAT-Standard & Poor's database of financial information on publicly traded companies-he identified firms with enough data to calculate their B-to-M ratio for each year between 1976 and 1996. After classifying them into quintiles by their B-to-M ratios, he determined that the average firm in the top quintile had a mean B-to-M ratio of 2.444 and a market capitalization of $188.5 million. Consistent with their poor operating history, the average high B-to-M firm experienced declines in net income, gross margin, and liquidity, and an increase in leverage.

For his final sample, Piotroski kept only those firms in the top quintile with sufficient financial statement data to calculate all nine signals. This methodology yielded 14,043 high B-to-M firms. The aggregate F_SCORE signal was calculated for each of these firms. Based on this measure, 396 firms with a score of "0" or "1" were expected to have the worst stock performance, while 1,448 firms with scores of "8" or "9" were projected to have the best subsequent returns.

His calculations show that his intuition was right on target. Although the complete portfolio of high B-to-M firms earned a market-adjusted return of 5.9 percent over the one-year period following portfolio formation, more than half of the firms in that portfolio actually earned negative market-adjusted returns. Splitting the portfolio by these firms' fundamental signals, Piotroski documents that firms with the lowest overall signals had a one-year market-adjusted return of -9.6 percent. By contrast, firms with the highest overall signals show a one-year market-adjusted return of 13.4 percent, outperforming the generic high B-to-M portfolio by 7.5 percent and the low-scoring firms by 23 percent.

The research also shows that a majority of the high B-to-M portfolio's winners are firms with low share turnover and no analyst following. Consistent with those general observations, further analysis shows that Piotroski's investment approach works best for small- and medium-sized firms, thinly traded firms, and firms with no analyst following. Moreover, the success of the strategy does not appear to be driven by purchasing firms with low share prices.

Another noteworthy finding is that low-scoring firms are more than five times as likely to de-list for performance reasons as high-scoring firms. "These results are striking because the observed return and subsequent financial performance patterns are inconsistent with common notion of risks," writes Piotroski. Normally, one would expect riskier firms to earn higher rates of return. The evidence, he explains, seems to indicate instead that the market is slow to react to good news embedded within a high B-to-M firm's financial statements.

Finally, Piotroski finds additional evidence supporting the claim that the market underreacted to the previous financial statement information. Specifically, he documents a positive relationship between these scores and subsequent quarterly earnings announcement reactions. Firms with high scores earned almost 5 percent over the next four earning announcement windows, while firms with low scores earned less than 1 percent during those same windows.

Piotroski plans to continue investigating whether financial statement analysis techniques can be used to better understand the source of firm profitability and quantify a company's risk. Moreover, he will continue to examine how financial markets interpret and price financial performance data. In the meantime, he says, analysts and investors who carefully analyze the information in a firm's financial statements, particularly for thinly followed and neglected firms, should be able to get a jump on the market.

Joseph D. Piotroski is assistant professor of accounting at the University of Chicago Graduate School of Business. His paper, "Value Investing: The Use of Historical Financial Statement Information to Separate Winners from Losers," was awarded Chicago GSB's 2000-2001 Ernest R. Wish Accounting Research Award.