How to improve stock screens

Stock screens are a very useful tool for investing. We can’t go through every single stock, there are thousands in the US alone, so cutting that list down by some parameters that make the companies more investable saves time and energy.

But it is also something that most retail investors do very poorly, and most follow the same sorts of rules as each other which means bargains are so much harder to come by. Take this common example: a screen for companies selling at a P/E less than 14, with a Return on Equity (ROE) greater than 10%. You will come up with a list of companies which earn good returns and are apparently cheap. But going through this list you will soon find these companies are going through a raft of different problems which make their valuations suitable.

That’s because companies selling so obviously undervalued don’t stay that way for long. If it was so easy for you to find it with a simple screen, you can bet every other retail investor can easily find it, as well as every professional fund manager. So against this collective competition, do you really think you have a chance?

Stay ahead of the crowd

The best way to screen is to find companies that normal screens wont capture and that are more likely to go unnoticed in the market. But we still want to find companies that satisfy our criteria, for me these are threefold:

1. Low debt
2. Cheap relative to future earnings
3. Good returns on investment

I’ll go through how I structure my screens for each of these and hopefully it will demonstrate how to capture those companies that fall through the cracks of a typical screen and are more likely to be mis-priced by the market. Then you can design your own individual screens in the same fashion.

(One important note – the metrics below are what I use for screening only – I will use other metrics when valuing and analyzing a company that reflect their true earning power to shareholders.)

Low debt

This is probably the easiest to screen for, but the easiest place to let a suitable company fall through the cracks. Investors will typically choose one of the following: Interest coverage > 5, Debt/Equity < 50%, Debt < 5 x Net profit…etc.

But each of these have problems. Firstly, if a company has been suffering reduced profits due to write-downs then this will affect both interest coverage and net profits and make each look very depressed. However these are the kinds of situations where temporary problems create good buying opportunities so you really don’t want to screen them out.

Debt/Equity has big problems with businesses that have very strong returns on equity. Take IBM for example, which has Debt/Equity of ~200% yet debt is less than 3x net income, far from excessive. These are also the strong types of company that you can invest in long term, and you definitely don’t want them screened out.

The solution is to screen for debt based on profits rather than equity, and on profits that exclude these possible exceptional items. I use Gross Profit, but if that isn’t available in your screener then EBITDA is a decent substitute. I screen for Total debt < 1 x Gross Profit as I like to invest in companies with very low debt, but this can be adjusted to cover what you are comfortable with.

Cheap relative to future earnings

A lot of investors will use the standard P/E ratio to screen for cheap companies but this is plagued with the problems of temporary write offs affecting earnings as discussed in the last section. There is also the problem that everyone else is doing it! Some may think they are ahead of the curve by screening based on forward P/E but this has 2 problems.

1. Analyst estimates are not always correct and reflect market sentiment which will probably be priced in already
2. Not all stocks have estimated earnings – especially smaller companies

The second is the most crucial in my opinion; the biggest advantage a retail investor has is they can invest in very small companies which aren’t covered by analysts and fund managers. This makes mis-pricing of shares much more likely. If you screen based on forward P/E and a stock doesn’t have forward guidance it will likely be screened out.

Some investors like to screen by P/Free Cash Flow but in my experience this is pretty much useless. Cash flow can be so erratic for most companies that you will likely get many companies you don’t want included because they have a random good year for cash flow, and many companies you do want are excluded because they have a random bad year of cash flow.

So if we can’t use earnings, cash flow or forward earnings, then what can we use? Well it’s back to profits higher up the income statement that wont be affected by write-offs or exceptional items. I use EBITDA and the common ratio EV/EBITDA to search for companies cheap relative to earnings. This isn’t anything new to seasoned investors but it is actually harder to find a screener that can facilitate such a screen. I recommend finding one that does.

Good returns on investment

By now you hopefully know what I’m going to say about the usual ROE screen – earnings can be affected by temporary factors and you will end up screening out good companies if you do it based on ROE. So let’s first decide to use a more stable form of earnings – EBITDA or Gross profit.

But now there is the problem of leverage. If you screen based on EBITDA / Equity then a company with high debt levels will do well. Conversely a company holding a lot of cash will do very poorly and these are great companies to invest in as they may return this to shareholders. So a better measure is one similar to Return on Invested Capital (ROIC), which adds debt to equity and subtracts cash.

I screen for EBITDA / Invested Capital greater than 20% which usually equates to a ROIC of 10% or higher under normal earning conditions. Net profit / EBITDA typically varies from 35-65% so adjust your EBITDA/IC ratio accordingly to what ROE you are targeting.

Hints of undervaluation

Finally I’ll share a couple of other subtle signs you can search for when looking for undervalued stocks.

One of my favorites is insider buying. Luckily you don’t actually need to screen for this, websites like GuruFocus have already done it for you, but it can be useful to include as part of a bigger screen. Just make sure they are actually buying shares and not just exercising options – look at the price paid per share.

I also like to screen for companies with capital expenditure < 0.8 x depreciation charge, as this indicates a company is generating more in free cash flow than suggested by the income statement and given investor’s focus on earnings, can lead to undervaluation.

The last is short interest in shares, those with a very high short interest will have cheaper valuations, but be careful as these companies are usually heavily shorted for a good reason and you need a solid thesis to bet against that.