This is an interesting idea:
https://dl.dropbox.com/u/28494399/Blog%20Links/Floats_and_Moats.pdf
Maybe this quantitative metric for changing floats/moats can be structured into a screen?
You know what I still enjoy doing? Throw darts at the newspaper. I don’t do it often, but it’s fun to pick 10 companies at random and start digging. I’ve found some good picks that way. Lots of times it’s not the company I start investigating, but it could lead me to a competitor or supplier that I wouldn’t have thought of before. Or it forces me to look at an industry I wouldn’t normally gravitate towards.
Not exactly screening, but good for idea generation.
^ Very true, I started investigating ATU and ended up buying competitor PH…or I started investigating CTXS and bought competitor RHT…
How would you target this? Moats are usually visible in the operating numbers. Float is limited to a handful of business models. I don’t know how you would cross reference the two quantitatively without finding stocks that already have these elements mostly priced in – this is not Warren’s era, which lacked computers. The best hope for this screen would probably be to produce some good GARP candidates that have float and moat with a tail. That’s not a bad screen at all, but I don’t think it would be a pancea of screening. What would be more helpful is to find the company that has a moat where the moat hasn’t shown up in the numbers yet.
Anyway, my experience has been anything that the market thinks has a moat and will “grow forever” trades at insane multiples, and more often than not, these companies neither have the moat they are supposed to have, nor do they grow forever, so those kind of stock picks are quite risky. There are exceptions of course.
I think part of the issue is also the timelyness of the data. With a quant screen we are typically looking at financial results that are historical nature. With a more qualitative analysis we may be able to pull data points that have not been fed into the stock price because they are prospective changes that have not yet been incorporated into the financials. you feel me?
I’d like to know how you created this qualitative screen, I am more interested in the general principles of how you collected/collated the data and then presented it to yourself in a meaningful way. My head hurts just thinking about the type of work that goes into creating something like that, which is why I am in sales. hah.
My experience with screeners has been mixed but in terms of my process I simply use them to whittle away names so I can focus on 5-10 names that I can focus/do further research on. Just two days ago while riding the train back from work I decided that I wanted to create a model portoflio of North American stocks with very low beta’s and high ROE and CF per share and back test it just to satisfy a curiosity I had after reading an academic journal talking about the relative outperformance of low beta stocks. Essentially, thats the type of stuff I use screeners for. I am still dealing with an internal conflict in terms of deciding whether I should manage my own money using a form of enhanced indexing based off of fundamentals that I screen for and just rebalance back to a preset SAA or whether I should follow a full blown klarman style active investing approach. I don’t think I really have the time/expertise for the latter but I do enjoy fucking around with screeners and reading 10-ks which is where the conflict arises.
I really like this idea, how did you find a screen that presents that information? or did you make it yourself?
Sometimes I wish I had the aptitude to code, god damn ADHD.
Totally agree.
I won’t get too detailed because this took me a really long time to develop. But generally, over time I began to recognize certain patterns of events or circumstances that would cause stocks to go up, and I began to formulate a mental list of these. One advantage I have is that I’ve looked at lots and lots of stocks over a fairly long time period, perhaps 500 or more domestically. People say stock prices are random, but they sure seem to correlate strongly to particular actions, even if that correlation doesn’t occur right away (the market sometimes takes a while to come around to what is happening, especially in under followed stocks).
So basically, create a list of events, then figure out how to screen for those events, and then evaluate the output of the screen on a stock-by-stock basis. It was very difficult to create this, and you need both the screen and the expertise to read it – neither one is very valuable in isolation. Whether the results are any good or not will have to be seen over time.
This is the great debate for people trying to manage their own money. Do you invest in 10 names you know really well? Better be right on most of the 10. Or do you find an indexing approach instead where you basically track the market? The issue is return on time. I don’t know the answer, but I do this full-time 6-7 days a week and even I would not be comfortable owning only 10 stocks. I prefer to be much more diversified than that, because anything can and does happen even if your research is solid (which itself is time consuming). You can have the best research in the world, but if a meteor falls out of the sky and lands on your company, all bets are off.
Look at a 10 year chart on MGI. I bet people in 2006 were like, yeah, this stock is a cash cow with growth and it’ll just keep going up over time, right?? But MGI took the float and invested it in high risk securities and blew the company up reaching for an additional 1% yield in their fixed income portfolio (d’oh!). Stock went from a split-adjusted price of nearly $300 at peak aaaaaaallllllllllllll the way down to $8. It was really safe until it wasn’t.
Also depends on what you’re owning.
I think 10 stocks equally weighted is decent if these are stocks that you expect to have a proverbial “forever” holding period for…but if you’re into much more speculative issues…then diversified is a must.
Bromion does the latter whereas I focus on the former, so I prefer to take 10% positions in stocks like MSFT/GOOG/BRK/RHT/PH etc etc…but I sometimes don’t follow my own advice and put 10% each in RSKIA and DJCO… but I don’t invest much money. 
One thing to look for, but it is hard to screen for, is changes in the language in the 10-K and 10-Q (or other country forms). These can be hard to find but important if you can find something. Most companies have the boiler plate pretty much set so anything be added or taken away can give insight into direct concerns/impacts/benefits. I would really watch this for young companies as they will typically just use standard language at first adopted for their own needs, which is probably more prone to being adjusted down the line than a company that has existed for a bit…