The Piotroski screen has been one of AAII’s top-performing screens for a number of years. While the S&P 500 was down 1.0% year-to-date as of February 28, 2010, AAII’s Piotroski screen gained an unbelievable 96.4%. The positive gain can be attributed to two companies: January’s single passing company, J. W. Mays Inc. (MAYS); and February’s single passing company, Highway Holdings Limited (HIHO). As of this writing, no stocks meet the Piotroski criteria.

The Piotroski screen was also the only AAII screen to post a positive gain in 2008 (+32.6%). Each of our benchmark indexes posted a loss that year as well. Since inception, the screen has returned 3,995.9%, the largest cumulative gain in the group.

While these returns are impressive, consider these statistics: the average number of monthly holdings is four; and since the beginning of 1998, the screen has had 20 months where no companies met the screen criteria. Taking a closer look at 2008, there were five months when no companies passed (August 31, 2008, through the end of 2008). As the S&P 500 fell close to 30% during those five months, this screen was not invested in the market at all.

On paper, Piotroski’s selection strategy looks like a winner, but it is hardly an investable strategy in the real world. For this article, we loosened the criteria used in AAII’s fundamental stock screening and research program, Stock Investor Pro, to see if we could create a more investable strategy using Piotroski’s value investing philosophy.

AAII’s Piotroski in Brief

Joseph Piotroski is an accounting professor who examined if it was possible to improve upon a low price-to-book investing strategy by requiring additional measures of financial strength. (He defined low price-to-book as multiples that were in the lowest 20% of all stocks.) Piotroski observed that most companies trade with low price-to-book ratios because they are financially distressed. He developed a nine-point scale to measure improving financial strength.

In his study published in 2000, Piotroski compared the performance of “winners” (a score of eight or nine) and “losers” (a score of zero or one). He found that the winners outperformed the losers over the next year. In addition, when he included a ranking for price-to-book ratio, he found that within the “winners” portfolio (those stocks with a score of eight or nine), stocks in the lowest price-to-book-ratio deciles performed better than those with higher price-to-book ratios.

The nine-point scale rewarded one point for each criterion in three basic categories: profitability, capital structure, and operating efficiency.

Profitability

A company can earn up to four points for profitability. One point is awarded for meeting each of the following conditions (based on the most recent full fiscal year): positive return on assets, year-over-year improvement in return on assets, positive cash flow from operations, and cash flow from operations that exceeds net income. While these criteria may not seem too strict, Piotroski found that 40% of low price-to-book-value stocks reported losses in the prior two fiscal years.

Capital Structure

Three points are awarded for capital structure and the ability to meet future debt obligations. The first requirement is a year-over-year improving (or lower) debt-to-asset ratio. Piotroski measures liquidity by requiring a year-over-year improving (or increasing) current ratio.

Finally, Piotroski avoids companies that issued stock during the last fiscal year. Because the stocks meeting the low price-to-book ratio criterion tend to have depressed prices, he believes this is the wrong time to issue new stock and may signal a firm’s inability to generate sufficient funds internally.

Operating Efficiency

Year-over-year increases in gross margin and asset turnover rate account for the last two points awarded in the Piotroski scale.

Relaxing the Criteria

The screen originally created for and currently used in AAII’s fundamental stock screening program, Stock Investor Pro, requires a perfect score of nine to pass the screen.

See “What It Takes”  for a summary of the screen criteria.

We tested the Piotroski screen three ways using full-fiscal-year data: requiring a score of at least seven, requiring a score of at least eight, and requiring a perfect score of nine (this is AAII’s original Piotroski screen). For example, a score of seven means a company must meet at least seven of the nine criteria.

A potential risk with the screen is that as the fiscal year progresses and a company moves into its third and fourth quarters, financial situations change. The original screen, which uses only full-fiscal-year data, could be looking at information up to nine months old. To remedy this, we used trailing 12-month data (the most recent four quarters of data) and tested for scores of at least seven, at least eight, and a perfect nine.

This resulted in six unique screens, five of them new.

The Results

Performance and Risk

Figure 1 shows the cumulative performance results graphically compared to the S&P 500 index along with the annual performance of all six screens and various indexes for comparison. Each of the six Piotroski screens have handily beat all of our benchmark indexes on a cumulative basis.

Figure 1. Performance of the Piotroski Screens

The original Piotroski screen still beats each of the variations. However, as discussed above, this is not a realistically investable strategy. The best-performing screen of the “investable” variations created for this article uses trailing 12-month data and a score of at least eight (cumulative gain of 1,570.3%).

The worst-performing variation used a perfect score of nine and screened variables using trailing 12-month data. Like the original Piotroski screen, it had many months with no passing companies. On average, only six stocks passed each month—not much better than the average of four for the original screen. The score-of-nine variation also lost 28.3% in 2009 when all of the other screens posted gains. This points to an obvious flaw in a strategy invested in very few stocks: If one or two holdings go sour, it can wreak havoc on the entire portfolio.

From a pure performance standpoint, it seems you will give up large returns by loosening the criteria slightly, but you must also consider risk. In the case of the Piotroski screen, risk is more than just standard deviation; it is in the low number of passing companies each month as well as the monthly turnover.

Not surprisingly, the average number of passing companies rises and the monthly turnover rate falls as we loosen the criteria from a required score of nine to eight and seven.

With a score of at least eight, the average number of holdings each month jumps to 26 using fiscal-year data and to 33 using trailing 12-month data. Allowing a score of seven provides you with an overwhelming average number of monthly passing companies (83 using fiscal-year data and 102 using trailing 12-month data).

Screen Characteristics

Table 1 lists the characteristics of the companies passing each screen variation compared to all exchange-listed stocks as of February 28, 2010. Note the absence of data for both of the nine-score screens as they had no passing companies at the end of February.

The screen that used a score of eight or more and trailing 12-month data has the highest median price-earnings ratio (higher even than the median for exchange-listed stocks) and the highest median 52-week relative strength. Is it possible that this adjustment has captured some elements of momentum better than the other variations? It is hard to say for sure, but this is an interesting development, to say the least.

The median size of the stocks in all the variations is much smaller than that of exchange-listed stocks, which is not surprising given that all of the stocks are in the bottom 20% based on book value. Smaller-cap stocks tend to have lower price-to-book values, in general. Another interesting note is that the best-performing variation (score of eight or greater using trailing 12-month data) does not have the companies with the highest growth rates in sales and earnings per share. In fact, it has the lowest rate of sales growth of all the variations.

Passing Companies

Table 2 shows the companies passing screens that looked for scores of greater than seven and eight based on full-fiscal-year data. Table 3 lists the companies passing with similar scores but using trailing 12-month data instead. The tables are ranked by score and then by lowest price-to-book ratio.

Table 2 contains 18 passing companies (six of which have a score of eight). Table 3 contains 42 passing companies (eight of which have a score of greater than eight).

Innotrac Corp. (INOC) is one of three companies that appears on both lists. The company has a price-to-book value ratio of 0.35, a full-fiscal-year score of eight and a trailing 12-month score of seven. Innotrac provides order processing, order fulfillment and call center services. This is a small company with a market capitalization of $17 million. The company’s outstanding shares increased from 12.30 million to 12.32 million during the last year. This represents a 16% increase in shares outstanding over the year and violates one of Piotroski’s criteria.

Using full-fiscal-year data, Innotrac meets all other criteria on the nine-point scale. However, when we screened for this article, Innotrac had not yet released its financial results for the fiscal-2009 year ending in December, making its fiscal-year data very stale. When using trailing 12-month data, Innotrac violates another of Piotroski’s nine points: an improving asset turnover ratio. The asset turnover ratio fell from 1.85 to 1.80 over the screening period. Over the last year, INOC shares have outperformed the S&P 500 by 50%.

A little digging into the company’s current condition reveals a notice from NASDAQ that INOC is in violation of the exchange’s minimum market value rules. The company has until July 6, 2010, to show a market value of publicly held shares exceeding $5 million for 10 consecutive business days. Currently, about 19% of Innotrac’s common stock is included in the publicly held share category as defined by NASDAQ. This is another good example of the importance of thorough research before investing real money into any security.

Conclusion

While it is easy to be taken in by huge performance numbers, it is important to always look beyond the surface to discover the drivers behind the numbers. AAII’s Piotroski screen offers good performance, but also highlights very few stocks and is a high-risk strategy. As a result, this strategy does not translate to the real world of investing. By tweaking the criteria a bit, we found that though the performance was weakened significantly, more stocks were identified and the strategy’s risk was lowered.

It is important to remember that the passing companies for each screen do not represent a list of recommended stocks. As with all types of investing, it is important to perform due diligence to verify the stock’s financial strength and earning potential. It is also essential to decide if the stocks match your investing style and risk tolerance before committing your investment dollars.