As I have discussed a couple of times, in value investing screeners can be a helpful tool if you use them right. In theory one has two possibilities: One can use screeners in order to create a more or less automatic investing strategy or one could use a screener for idea generation.
Screeners for automatic investment strategies
In order to come up with a screen for an automatic investing strategy, you have to do a lot of number crunching and backtesting. Mostly you end up with something which combines a few single attributes (P/E, P/B, past perfomeance) which have performed best in the past. Usually there is a periodic requirement (for instance every year on June 30th) to redo the analysis and adjust the portfolio.
Many people, especially with a mathematical background, like this kind of investing, because it doesn’t really require to know anything about the companies. To be fair, it requires a lot of discipline to hold through this aproach especially if the startegy doesn#t work for a subsequent number of years
In a very interesting article Joel Greenblatt is comparing the performance of “fully automatic” accounts versus accounts where peple manually followed the Magic Formula. Not surprisingly, the auotmatic system outperformed by a wide margin. The reasons for this underperformance seems to be relatively clearly market and/or strategy timing. People buy more after good performence and sell after bad performance.
Without having proof for that, I would nevertheless assume the follwoing: In my opinion many of the strategies work in the long term, but only few people are actually “mentally equipped” to follow them through.
For me personally, it wouldb e really hard to invest in companies I don’t really like so i guess I would not hold through the magic Formula for instance.
Screens for idea generation
Another type of screener would be a screener, which, based on certain pre defined attributes, tries to identify interesting companies to be analysed further. Those screens are not back tested but rather rely on subjective assumtions what could make a stock intersting.
The “Magix Sixes Screen” I often use for instance is a good possibility to find potential “fallen angels”. The only stock out of this screen where I really invested, Autstrada, didn’t work out, so why bother further with those screens ?
As I discussed several times, I have certain ideas what risk characteristics my portfolio should have. For instance I prefer below market volatility because this helps me avoid any market timing actions even in the worst times. As I have only a limited amount of time per day to work on analysis (maybe 1-2 hours) and I want to have at least 20-25 different investments, one has to think about how to distribute the capacity best.
My “special situation” investments are usually rather “high maintenance”, so I prefer for the rest of the portfolio more “low maintenance” stocks. Low maintenance for me means the following primary characteristics:
1. company has a stable unexciting business with respective results over a long period in the past
2. company is cheap compared to “intrinsic value” to limit downside
3. company has shown historically that it adds value above cost of capital
So in the end, I am looking for companies which have a boring (non-volatile) business model and a sexy (cheap) valueation..
I think additionally it makes sense to define what I am not necessarily looking for in the first place
– deep value turnaround situations (too risky)
– net nets (usually no ongoing value creation)
– “moat” stocks (too crowded)
– growth stocks (too risky)
In the next post I will follow up how I actually calculate the “Boss” Score.