The Quest for the Holy Stock Screener

January 11, 2011

Many of you may have no doubt noticed that my approach to investing is largely quantitative, concerned with free cash flow yields, interest coverage ratios, the gap between depreciation and capital expenditure, net working capital position, and other financial vital statistics. This should not be surprising; I take the position that corporations are machines that turn cash into (hopefully) more cash, and as a result these basic company-specific figures mean more to me than the penumbras and emanations of economic and sector analyses that characterize the top-down approach of non-value investors.

The emphasis on a company’s facts and figures also serves to give us a good sense of what to look for in a company.  We can identify a potential investment, or at least eliminate an unsuitable one on the basis of the numbers. Fortunately, there are a number of sources that track the financial details of a company in obsessive detail and allow searching. It is, of course, indispensible to follow the advice of Ben Graham in his vital Security Analysis, and go through the accounting in searching detail in order to determine whether the reported numbers are representative of reality or have to be adjusted for nonrecurring events, noncash writeoffs, or nonexistent interest expenses due to convertible debt. But we can at least narrow down our list of candidates if we look for the right numbers, and for that we need the right screener.

The Yahoo! Finance Java-based stock screener (not the basic HTML one) is a good choice, allowing screening for a large family of variables, including dividend yields, return on equity, and the ever-popular P/E ratio. However, the P/E ratio does not tell the whole story, owing to gaps between depreciation and capital expenditures, nonrecurring events, the fallout of an economic crisis, or just simple application of accounting rules. Linn Energy, for example, had a P/E ratio of 1.39 owing to its oil price hedges, because it had to report the income from its hedges but not the offsetting losses in its future production that the hedges were hedging. However, Yahoo’s screener does not have an entry for free cash flow yields; the closest it gets is a screen for entity value to free cash flow ratios, and that measure is often distorted by an unusual cash flow structure. Also, it’s not immediately clear how they compute the entity value or free cash flow, because I have found some missing entries. However, the screener does have the advantage of letting you define your own cutoff points for the screen. If you wanted to know which companies had, say, a price/sales ratio of exactly between .93 and .95, Yahoo! will help.

For net-net companies, Ben Graham’s other favorites, an ordinary screener will not do, because the price/book ratio alone is not specific enough, and combining it with other measures such as the current assets to current liabilities ratio or total debt/equity is likely to produce an endless list of false positives and negatives. Fortunately, someone has assembled a screen that uses data pulled from Yahoo! to identify these companies directly. I have found some issues with Yahoo!’s own data being wrong, which occasionally affects the screen as well, but as I never invest without reading at least a 10-K and the latest 10-Q if applicable, these errors are always caught in time. The author of the screen has also added certain other calculations such as the Piotroski score, which is an estimate of what I would call “fundamental momentum.” I find this a useful screen for identifying potential opportunities on the balance sheet, and it definitely has the advantage of being free.

I have recently discovered a third screen at that actually will screen for price/free cash flow, a thing I have been searching for for some time. However, it does not allow you to set your own limits, and with theirs you have the choice of looking for a multiple of five or ten. Perhaps 5x catches ltoo few and 10x catches too many, but there are additions to the screen that can be added, and I think it is best to cast a wide net, as the vital work of actually finding suitable investments from the screen results is the lengthier (and more fun) part.

However, I have not found a screen that can do what I really want it to do, and that is to center on historical data. I like to look for companies that not only have a high cash flow yield at present, which could be a fluke, but which also have been able to sustain reasonably high earnings in the past, which gives me greater confidence. In fact, many value investors are guided by what they call the Graham-Dodd P/E ratios, which average earnings over ten years. as I tend to be skeptical about growth, and the usefulness of projecting it, I like the concept. However, it is also the case that a company now is almost always not the same as it was ten years ago, so perhaps taking an exponential average, or even calculating an average profit margin over ten years and applying it to current sales might be in order, as suggested by Damodaran in his useful toolbox, The Dark Side of Valuation, as a method for dealing with cyclical companies.

I have also started to see the logic of companies that are aggressively paying down debt. After the financial crisis, many people began to see that liquidity is a thing that can dry up, and beyond that, at any given time there are always companies in the market that need debt reduction. Furthermore, many economists have noted that consumers and corporations have been reducing their debt levels faster than the federal government has been expanding its (and, if you’re a Keynesian, that is the very reason why the government has to keep expanding its debts). But whatever one’s feelings about economics, there are sound reasons why companies should deleverage in uncertain times, and it is wise to get on the right side of the trade. After all, cutting debts certainly improves future earnings power, and as companies with high debts often produce lower profits and nearly always trade at lower multiples, the tax advantages of debt, and the multiple expansion that comes from a distressed debt situation resolving makes the return from paying down high debts a very attractive use of free cash flow. To that end, I want a screener that can track debt repayments and overall debt/equity levels over a period of years, so that I can dig into the reports and know if the company is following a wise strategy.

At any rate, I haven’t found any screener like that online, at least for free. I suppose a screener with all the features I have in mind would be useful enough, if it exists, that its designers can charge for it. So, I shall have to make do with what I can, and keep an eye open for new tools.

Leave a Reply