Imagine being able to run a fairly simple stock screen that will outperform the market by 7.5%. To put this in perspective, if you expect the market to return 7% annually for the next 30 years, then a 7.5% outperformance is the difference between the $10,000 you invest today growing to $76,000 or growing to $580,000. And if you have more than $10,000 to invest? Well, then you have 30 years to think about what color you want your Ferrari to be.
Separating wheat from chaff
Before delving into the workings of the winning screen mentioned above, we need some context. Many stock screens focus on identifying out-of-favor companies that have the potential to bounce back, such as the Dogs of the Dow and Pigs of the Dow strategy. These methods can work nicely in situations where the market becomes unrealistically pessimistic about the prospects for a company or an industry.
The challenge, however, is that sometimes companies are unpopular for good reasons, and never do bounce back. Maybe they face new competition in the market, or they are displaced technologically. Whatever the case, it's important that these companies do not ruin the results of the entire portfolio.
The Dow strategies rely on the fact that the companies in the Dow Jones Industrial Average, like Intel
Enter Joseph Piotroski, currently at the University of Chicago. Academically, it's well-known that stocks with a low price-to-book value ratio tend to outperform the overall market. But Piotroski observed that the superior performance of low price-to-book stocks was due to a small number of exceptionally performing firms, while the majority of such firms actually underperform. He thought about it a bit and realized that if he could just find some way of eliminating the underperforming firms, then the remaining portfolio could potentially realize Buffettesque returns. After a bit of tinkering and tweaking, Piotroski eventually came up with a set of criteria that beat the performance of an index by 7.5%. And business at Ferrari dealerships has never been better.
Spinning lead into gold
By this point, you might be curious as to what Piotroski's criteria are. He first ranks all the firms by price-to-book value, and then eliminates all but the lowest 20% (the most bargain priced). He then filters the remaining firms by nine profitability, liquidity, and operating efficiency criteria to focus on the firms most likely to outperform in the future. Each stock is assigned a "pass" or "fail" for each of the nine criteria, and all stocks that receive eight or nine passes are considered a "buy."
The profitability criteria focus on the idea that if a company is making money and improving its performance, then it is likely to be able to sustain itself on its internal cash flows. With this in mind, his first four criteria are:
1. Return on assets greater than zero.
2. Cash flow from operations greater than zero.
3. Return on assets in the most recent year higher than the year before.
4. Cash flow from operations greater than income.
The liquidity criteria represent the belief that companies that can manage their current liabilities and that are decreasing their leverage have less need for external cash, have more flexibility, and are less likely to go bankrupt. The criteria are:
5. The ratio of long-term debt to assets has decreased in the last year.
6. The current ratio -- current assets divided by current liabilities -- has increased in the past year.
7. The firm did not issue shares in the past year.
The remaining two operating efficiency criteria are indications that the firm's operational performance is improving. It will help to eliminate firms being squeezed by competition or that are slowly going out of business.
8. Gross margins have improved in the past year.
9. Asset turnover -- revenue divided by total assets -- has improved in the past year.
What's the screen picking now?
Right now, there are 64 companies that meet eight or nine criteria. The companies have market caps ranging from $12 million to $75 billion, and are in a variety of industries, from tech and communications to mining and retail. There's even a licorice manufacturer. Some of the better-known names include cable operator Comcast
The problems
For the lazy investor, Piotroski's screen might seem like the ideal tool. All you have to do is spend a few hours each year purchasing the picks and then look on smugly as professional fund managers struggle to achieve a fraction of your returns. That seems pretty sweet to me. But there are still challenges with the screen.
First, there are 64 stocks in the screen this year. This is typical, which means that you'd generally have to make about 128 trades a year -- 64 sells and 64 buys -- in order to implement the strategy. With that many trades, commissions would really add up unless you have a huge portfolio. The other downside with yearly trading is tax inefficiency. If you hold Coca-Cola
One other problem with screens is potential over-fitting. This happens when a researcher takes a set of data and tries a multitude of different screens to predict future winners. Eventually, they find a good screen. But the screen never works on any other set of data, either because it was optimized for that particular set of data or because it just happened to find random correlations. For example, if a researcher in the first quarter of 2000 tested a stock screen exclusively using data from the 1990s, then a "buy technology stocks only" screen might look great. That researcher would now be selling pencils on a street corner somewhere.
Over-fitting may be less of an issue when the logic behind the screen seems reasonable, as it does for Piotroski's screen. However, the data Piotroski used to test his screen, a 21-year time period starting in 1976, may not be representative of the long-term performance of the market either. And when stock screens suddenly "stop working" it's not always clear whether the problem was an over-fitted screen or that enough traders are using the screen to negate any potential advantages.
The other major problem with screens, which fans of mechanical investing may see as a strength, is that they do not fully leverage investors' abilities to make reasoned decisions. In effect, investors who rely exclusively on screens are saying that they believe their ability to reason about equities is inferior to their ability to instruct a machine to reason about equities. This strikes me as odd, though I'll admit that the argument "but it works" is persuasive.
Overall, I believe that investors will outperform if they supplement screens with well-reasoned analysis. Investors who successfully identify stocks with strong competitive advantages trading below intrinsic value, like those that Philip Durell analyzes every month in his Inside Value newsletter, will have superior long-term performance. In fact, if you're looking for stocks trading below intrinsic value, you should seriously consider taking advantage of a no-risk, free trial subscription to Inside Value.
Conclusions
For large portfolios, where transaction and tax costs don't outweigh potential benefits, Piotroski's screen may be worthwhile. For smaller portfolios, it may make more sense to not buy the entire screen, but rather use the screen to find companies worthy of further fundamental research.
For more on this topic, check out:
- Dogs of the Dow
- Pigs of the Dow
- The Low Price-to-Book Value Ratio
- Trembling With Greed
- What is Mechanical Investing?
For those who would like to continue this discussion elsewhere, visit our dedicated discussion board.
While writing this article, Richard Gibbons was surprised and disappointed to learn that that tobacco companies are the primary purchasers of licorice. Before now, he thought he was. He owns none of the companies mentioned in this article. The Motley Fool is investors writing for investors.
