Buy What You Know, Study What You Don't

The best way to form a 'correct yet contrarian view' of a stock is to pick a stock to research where you already have some background knowledge on the company, its products or the industry

Author's Avatar
Geoff Gannon
Sep 25, 2017
Article's Main Image

The concept of buy what you know is a Peter Lynch idea. Most investors I talk to think its too simplistic. Like any idea, perhaps it is too simplistic if taken to the extreme. But Ive found talking to other investors that they underestimate how much of an advantage you have in starting your research process on the firm footing of knowing something about the business you are about to research and how it works in the real world.

First, lets put the Peter Lynch quote in context. What Lynch was saying is that if a man worked his whole life in the oil and gas industry why would his first thought be to analyze a pharmaceutical company? His first thought not necessarily his last, but certainly his first should be to look at oil and gas companies. If the man had made a fortune in the oil and gas industry and was looking to buy an entire business, hed never think to buy a whole drug company. So why would he think to buy a piece of a drug company just because this was a stock purchase instead of the purchase of a private business in its entirety?

At least start yourself off on the firm ground of background knowledge you already have. If you know the oil and gas industry, start by looking at oil and gas companies. Otherwise, youre going to get quickly overwhelmed by a deluge of data and trust some expert instead of using your own common sense.

Ill give you two recent examples. I was talking to one person about a stock he suggested. In many ways, its attractive. The fundamentals looked interesting. And with a fair amount of research time it would be possible to decide whether or not this was a good stock to buy. So this isnt a stock that necessarily would end up in the too hard pile. I thought that the two of us researching the stock together and chatting about it over Skype could definitely figure this one out and come to a decision to buy or not.

But there was a problem. The stock we were looking at was a low-cost airline. So I asked him if hed ever owned an airline stock before? No. Had he ever researched an airline before? No. Thats not an insurmountable problem. But it doesnt just double the amount of time youre going to have to spend researching this stock. It probably quadruples or octuples the amount of time you need to research this thing.

I gathered together a bunch of links to

Warren Buffett (Trades, Portfolio) talking about why he invested in airlines recently. I sent a link with Glenn Greenberg (Trades, Portfolio) talking about Ryanair (this was years ago). And I also gathered together a group of four or five peers that I thought he should research together with this stock. There were questions I needed him to focus on like: What happens when an airline has trouble getting new planes and new pilots at the terms they did early on in their development? What happens when a lot of the routes an airline was flying had no competitor and then a competitor starts flying that same route? I just had a pile of questions and resources that I dropped in his lap to look at. None of this would have been necessary if hed had a previous knowledge of airlines. Now, thats fine. This can be his first step into researching airlines. The knowledge he accumulates learning about this stock may serve him well as background knowledge when he studies a totally different airline in two or five or 15 years. Thats the way investing works. You bring a ton of old background knowledge to any new idea. And by the time you make a buy decision, its often going to be based on at least 50% of stuff you already knew before you started researching this stock this time around.

The this time around part is key. I was talking to someone else about two different stocks recently: Howard Hughes (

HHC, Financial) and Green Brick Partners (GRBK, Financial). It was a lot easier analyzing Green Brick with him than analyzing Howard Hughes.

Why? Is Howard Hughes more complicated than Green Brick?

It is. There are probably about seven different projects scattered around the country from Manhattan to Honolulu and Houston to Las Vegas to Chicago that you want to appraise when looking at the company. For Green Brick there is really (for now) only the Dallas and Atlanta homebuilding markets to consider. But the difficulty in analyzing Howard Hughes with this investor was that when I asked if hed ever analyzed a real estate company before he said, No.

There are specific things about the business model of an airline or a real estate company that you need to understand well hopefully, better than the guy selling you shares in this company before you can buy the stock. For airlines, you need to think about things like cost per seat mile, the leases on these planes, what the load factor is in the industry right now, and what deliveries of planes (changes in industry capacity) are coming soon to airlines that operate the same routes as the one youre looking at. For real estate, you need to think a lot about financing and discounting.

There are a lot of wonderful stocks out there that generate enough free cash flow to fund their own growth indefinitely. They dont need to borrow a lot. And you dont need to think much about assets on their books. If they have a lot of assets those assets will tend to be cash, accounts receivable (net of doubtful accounts) and inventory (stated at the lower of cost or market). For a well-functioning business, those are liquid and reliable assets. There are positive and negative aspects to them. Obviously, youd rather have cash than inventory. So you dont want to invest in a company that always puts every dollar of profits it earns right back into growing inventory. But even a business like that say a fast-growing retailer with really low inventory turns is much more cash generative relative to its long-term needs than something like Howard Hughes.

On the other hand, Howard Hughes owns some phenomenal undeveloped assets. They still own large amounts of land in Master Planned Communities that they carefully developed into affluent suburbs. The residential land on their books will eventually have houses put on it. But when?

A stock like Howard Hughes is going to have a lot of debt (though some of it will be nonrecourse to the parent company). Its also going to have a lot of asset sales. But, those sales arent going to be very big if you take cash value of assets sold this year / fair market value of assets on the book. They might only be selling 5% or 10% of what they own each year. So what is the 90% to 95% they still own going to be worth? Take a particularly tough assumption here. Assume they only sell 4% of what they own each year. That means they are going to take 25 years to work through turning all the assets you see on their books into cash. The average asset on their books today will be sold in 2030 (25 years / 2 = 12.5 years from now). Now, Im making this up in the case of Howard Hughes. You can read through what they say about each master planned community and see that theyll actually realize a lot of value in just the next couple of years in a few places and then in like one of the more extreme examples yes, theyll still be selling things in 2030 I believe.

All of this is in investor presentations, the companys filings with the SEC, etc. But how you evaluate this information is not. What is an acre of undeveloped residential land in a Houston suburb worth? What is an acre of undeveloped commercial land in Summerlin, Nevada, worth? What is the South Street Seaport worth? Not only are these assets in different places. Theyre also going to be monetized at different times. So for the South Street Seaport you may be able to just value the asset and thats an accurate idea of what value you should add to your appraisal of the company. Thats a project that is close to completion. But there is some land in some of these communities that isnt going to be developed for like a decade.

What is a dollar in 2027 worth compared to that same dollar in 2017?

There are ways to solve this problem. But it helps if youve had practical experience worrying about things like discount rates and real estate values before. The value of the land is going to be higher in 2027 than it is in 2017. But the value of receiving $1 in 2027 is lower to us (and Howard Hughes) than receiving that same $1 today. So we care about the difference between how much we want to get paid to wait each year and how much the land is going to appreciate in value each year.

Thats the right way to frame the problem of analyzing Howard Hughes. When I lay it out like this, it seems simple and intuitive. But I can tell you from experience that when someone first reads a real estate developers 10-K, they dont immediately cut through all the fog of details and lay things out like I just did. They often rely a lot on posts at Value Investors Club, blogs, etc., or presentations by investors who propose this idea as a long at some conference or who propose this idea as a short at some conference. The more background knowledge you bring to analyzing a stock, the more comfortable you will be framing the question in a way that makes logical sense to you instead of the way its conventionally done.

If youre going to bring a different and better approach to valuing a stock you need to have the knowledge to know where conventional wisdom is wrong.

For me, this is what happened with banks. For a long time, I did not feel qualified to analyze banks. One of the biggest reasons why I didnt feel qualified to analyze banks is that I read what analysts, bloggers, the financial media, etc., said about how you value a bank stock and it didnt make much sense to me. For example, value investors would say you should value a bank on price-book (P/B). This didnt make sense to me because a bank earns money lending out its deposits not lending out its equity. I had some experience investing in insurance stocks by this point and I knew that valuing insurers on their book value was a poor idea in the long run.

Buying an insurer at a low P/B at the right time in the cycle (a bad time) and selling at a higher P/B at the right time in the cycle (a good time) made sense as a trade. But the really big money in investing in insurance stocks is not made from buying them at below book value. The really big money in insurance stocks is made from buying insurers with better business models even when they trade at a higher P/B ratio than others.

Insurers should be valued based on only two things: premiums and float. An insurer with a combined ratio less than 100 can make an underwriting profit each year. So for such insurers, every $1 of premiums adds some number of cents in profit. So price/premiums makes sense there. And then you also have float. Even if an insurer writes at a combined ratio over 100, they can invest the float. If an insurer can make say 5% a year on their float then you have to take Float * 0.05 = 5% of float in underlying profit (on investments) and add that to your estimate of normal earnings.

I owned a small insurance stock once before (bought at a deep discount to book value) which I believed should not trade below book value. I believed that because it tended to make about an 8% annual underwriting margin. It also had some float. If you add the return on float to the direct return on underwriting you get a number that says you should pay book value or more. But what made much more sense to me is you got an appraisal of the company that was based not on book value but on premiums.

I also analyzed a company called Car-Mart (

CRMT, Financial). They sell used cars. But really what they do is make very risky car loans. I valued them based on price-receivables not P/B. Once I had a background valuing stocks like Progressive (PGR, Financial) and Car-Mart, I felt confident enough to go and value banks likeFrost (CFR, Financial), Bank of Hawaii (BOH, Financial)and Wells Fargo (WFC, Financial). I was confident that the approach I had looking at the total economic cost of these banks deposits was a better method than anything based on book value or efficiency ratios or anything like that.

Until you know a subject well enough to think you have your own way of valuing a stock, you are going to have problems coming up with a correct yet contrarian view. Those are the most valuable views.

You can also act with more conviction. The person I was talking to about Green Brick had found that stock on his own and actually owned it in the past. Its true that David Einhorn and Daniel Loeb also own the stock. But thats not what attracted him to it in the first place. With Howard Hughes, this person had first learned about the company by watching

Bill Ackman (Trades, Portfolio)s presentation on the company. Ackman is a large shareholder in Howard Hughes. You are never going to be as confident investing a lot of money for a long time in an idea that a big shareholder of the company brings to your attention. Youre going to be way more confident investing in a company that is local to the area (as Green Brick is for this investor I was talking to) and which youve already owned in the past.

In fact, some of the best investments you can make are often revisits of stocks youve owned or researched before. Right now, Im looking at Omnicom (

OMC, Financial) around $75 a share. I bought it for the first time at under $28 a share (like eight years ago now) and I wrote a research report on it at a higher price than it is at now (about a year ago). Because I feel like I know the advertising industry in general well and Omnicom in particular well, Id be willing to bet big on that stock and hold it long enough for the idea to work out. I also have a more precise idea of what price Id like to pay for the stock because Ive come up with a specific value for Omnicom twice before in my lifetime.

Ive also been following Howden Joinery(

LSE:HWDN, Financial) and Cheesecake Factory (CAKE, Financial) for years. Compare this to a stock like AutoZone (AZO) which I only researched for the first time this year. It would be harder for me to feel confident enough that I see AutoZone more clearly that I have a better frame for analyzing the stock than the guy selling his shares to me. So Id be more likely to make a smaller bet on AutoZone (put a smaller percent of my portfolio into it) and Id be more likely to sell the stock sooner if something unexpected happened. Thats not a recipe for long-term success in the stock market.

Unless an idea has the potential to get you to 1) put a lot of your portfolio into it and 2) hold it for a long time, its not worth your time. The value of any idea is Position Size * Years Held * Annual Return. We value investors often focus too much of our research time on what we think the annual return in the stock will be and not enough on how big were going to make the position and how long were going to hold it. It doesnt matter much if a stock goes up if: A) Youve already sold it or B) you didnt buy enough of the stock to move the needle in your portfolio.

So start with the stocks you know best. And then expand your potential circle of competence out to new stocks that are related in some way to stocks you already know. If an idea looks great but you know nothing about it educate yourself about the company, its business model, its competitors and the industrys history. The most important thing to keep in mind is that researching a stock in an industry youve looked at before is going to take you about 10 times longer than researching a stock youve already studied in the past.

Ask Geoff a question

Disclosure: Long Frost.

Also check out:
Rating:
5 / 5 (6 votes)
3 Comments
Load More

Please Login to leave a comment

Please Login to leave a comment