Try These Strategies for Deep Value

I recently gave an overview of a new stock-screening tool called RobotDough. Its innovative, equation-style screening allows investors to build complex screens to find specific types of stocks, then review them for any red flags and keep track of the strategy as time goes on. One great thing about RobotDough is its growing community of shared screens, some of which could stand to make investors a lot of money.

A haven for Grahamites
Many of the shared screens on RobotDough make use of various ideas gleaned from Ben Graham, the father of value investing. Indeed, a few of the top-ranked strategies use Graham's idea of net net working capital as their main criterion. Net net working capital is essentially a way of discounting a company's balance sheet to what might be a liquidation value. Graham would then look for companies with a market cap of two-thirds this value to find deep discounts. It can be calculated with this formula:

NNWC = (Cash and Marketable Securities (Accounts Receivable * 0.75) (Inventories * 0.5) - Total Liabilities)

Trading according to this strategy would have returned 62% over the past year, but some of the better performers on RobotDough added a couple of other, harder to find screening criteria -- the Piotroski Score and the Altman Z-Score.

The secret ingredients
The Piotroski Score, first published in 2000 by University of Chicago professor Joseph Piotroski, assigns one point for each of nine different criteria relating to profitability, balance sheet health, and operating efficiency. Piotroski found that by buying stocks with a high score and shorting those with a low score, an investor would have had 23% annual profits from 1976 to 1996.

The Altman Z-Score was developed by New York University professor Edward Altman and similarly assigns a score based on profitability and balance sheet health. What Altman found was that a very low score could predict the future bankruptcy of a company with a high degree of accuracy.

After adding criteria for Piotroski and Altman scores that weed out the danger of bankruptcy, the returns rose to 107% over the past year.

Not as easy as it sounds
Unfortunately, this screen shows off a feature of RobotDough that may be a mixed blessing. RobotDough can scan stocks on any exchange in the world, so many of the picks this screen makes are over-the-counter microcaps -- not easily traded, and not for the faint of heart. There is another nearly identical screen that only allows stocks traded on major U.S. exchanges, but even these are all thinly traded microcaps.

Other things to try
Still, these are good springboards for other strategies. The screen I proposed in my previous article works off a similar principle to net net, and while it would not have done as well as these two over the past year, the securities are considerably more liquid, and RobotDough's yet-unreleased backtesting feature shows the strategy would have returned 193% over the past three years.

You can also use the principles found in successful screens and invert them to find short candidates. While the successful screens look for stocks with healthy Piotroski and Altman scores, a screen for unhealthy scores can find a strategy that would lose a lot of money. As an example, I developed a screen with the following criteria:

• Only stocks listed on the NYSE, Nasdaq, or AMEX.
• Market cap > $300 million.
• Volume > 250,000.
• Price > 10.
• Altman Z-Score < 1.81.
• Piotroski Score < 3.
• Free cash flow (TTM) < (Market cap * -10%).

This would find liquid stocks, losing massive amounts of cash, with dangerously low scores, and a price high enough to avoid some of the wild recoveries penny stocks can undergo. Over the past three years, the strategy would have lost a total of 26%, with the market up 6.4%. While not necessarily short recommendations, these are the stocks currently in the screen:

Go ahead and try out your own strategies, see how they stack up against others, and refine them for maximum profitability. But be careful: It's a little addictive.