Before I start this evening, a quick story:

One day in March, I got an e-mail from an older gentleman, and he said, “Hi, my name is XXXXXX.? Please call me.”? The name sounded familiar but I couldn’t place it.? So I called him, and he asked, “I own a decent amount of Berkshire Hathaway.? Can you explain to me why the value of the put options Buffett sold have risen from 3Q11 to 4Q11?”? After talking to him for a little bit, he said, “So you really don’t know.”? I replied, “Yes, I don’t really know, but give me a couple of hours, and I can give you a decent answer.”? So I told him I would give him an answer via e-mail, and follow with hardcopy.? As it was, we were both wrong, the value of the put options had fallen.

But then it hit me why the name was familiar — he was one of the Superinvestors of Graham-and-Doddsville.? You never know where you might go, or who you might meet just because you write some obscure blog.? Oh, yeah, you might just get to go to the US Treasury, and meet you-know-who. 😉

And now a question from a reader:

I?ve learned a lot from your blog. Thanks for the effort you put in to maintain it. (Today?s post was great? was just listening to my Intelligent Investor CD while driving last night and was thinking about how Graham talks about the need to be businesslike when investing).

I had a question about your investing process, specifically how you track stocks that look interesting to you.

For instance, I use value line and I flip through the pages on a daily basis, jotting down interesting tickers in a notebook. I then go to a few scans and watchlists, again jotting down tickers I?d like to research further. After that, I take my list of tickers and do some brief research to determine if they should go on my watchlist. This is where I need to change a few things. I use Morningstar to track my watchlists, which is great because it keeps track of all of the financial data automatically, but the problem is that as my watchlist grows, I find certain cheap stocks on the list but I can?t recall why I put them there. There is no spot for notes on Morningstar?s watchlist, so I thought about using a google spreadsheet, but I really don?t want two different things to track.

Walter Schloss was famous for not owning a computer and simply flipping through value line. There must be a simpler way to track stocks that look interesting!

Just wondering if you could share any thoughts on how you track potential stocks and how you refine your watchlist if it starts to get too large. Thanks David.

I don’t have a permanent watchlist.? I generate ideas three ways:

1) I read a lot of articles.? When an idea sounds clever, I write it on a small piece of paper and set it in a pile to age.

2) I study industries more than companies, and look for strong companies in weak industries, and levered companies in industries that are likely to remain strong for awhile.? I use Value Line here, and screen for companies that might yield a 15%/year return, while being above average in terms of balance sheet strength.

3) I read through the 13Ds of a group of ~75 investors that I respect, and look at the companies that are large holdings relative to market cap that are still being acquired.

Then I take all of my ideas once per quarter, shortly after the 13Ds are filed, and compare them against existing portfolio holdings against a variety of valuation metrics, sentiment variables, and other factors.

I then rank all of the companies as a group, looking at where the middle company in my existing portfolio is.? Companies I don’t own that are above it are candidates for purchase.? Companies I own that are below the middle portfolio company are candidates for sale.

I then sit down and analyze, and pair off stocks to sell versus stocks to buy, unless there is some reason to increase or decrease the number of stocks in the portfolio, which is usually around 35.? Presently it is 34.? I? buy/sell 2-3 stocks per quarter.? That keeps things fresh, but allows me to hold companies for longer periods of time.? (My turnover is 25% of the open end mutual fund industry.)

After I make my series of buys and sells, I throw all my work away, and start accumulating data for the next quarter.

My sense is that it pays more to think of your portfolio versus candidates than to track candidates.

The main idea here is that we should always be improving the character/quality of the portfolio.? Trade things that are worse for things that are better.? Human beings can do that trade.? What is hard is for people to assemble the best stocks.? Getting a bunch of “pretty good” stocks can be done, and it does not mean they will outperform all the time.? Getting a portfolio of all of the best stocks is impossible.

So think economically — try to improve your portfolio regularly, regular incremental improvements yield a large economic improvement over a market cycle.? At least, that has been my experience. 🙂