Stock Analysis Process - How Geoff Researches Stocks

by Geoff Gannon


Someone who reads the blog sent me this email:

I'm interested to know when you analyse a company, do you follow a particular order? Do you always start from a screen? Do you look at its overall business and competition landscape before diving into its financials? And when you look at its financials, do you follow a particular order? Do you look at its income statement first and then balance sheet?

I don't have a particular order for finding a stock/company. I do screens and stuff like that. I read blogs. I look at company's competitors. I just go through some foreign stock exchanges from A to Z. Or some states from A to Z. Or some industries from A to Z.

Basically, I'm looking through lists of companies the way I figure Warren Buffett flipped through Moody's Manuals. I'm moving quickly to see if the company is really cheap compared to past earnings and current tangible assets and current sales. But I'm not doing any math, I'm just using websites like GuruFocus or Morningstar or MSN Money or the stock exchange sites, or the company's 5-year or 10-year financial summaries. You can find something like that online for a lot of companies and then you quickly just run your eyes over those numbers. Are they pretty ordinary looking? If so, just move on. If something pops, stop and look at the stock.

If any number catches my eye - like a ton of excess cash, or low p/b, or low p/s, or low EV to past EBIT, FCF, etc. I look at the company description. Usually I'll use Bloomberg for this or the company's own website. What does the company say it does? If it says it invests in real estate, copper mining, is an investment bank, etc. I drop it there. If it says it does something I think I can visualize if I work real hard at it, then I keep going. I look for words like "niche", "specialized", etc. I look for business descriptions that sound non-capital intensive. Do you test or monitor or score or report? That's good. Do you make capital goods? That's bad. Do you make something cheap and repeat purchased, that's good? Do you distribute? Good. Produce? Bad. Do revenues sound recurring? Are you a one of a kind company? A possible "hidden champion"?

This is all from the one paragraph description. Some are pretty inaccurate. But a lot aren't. You get interested in Bunzl real fast when you read about it, because of what the business does. This is the Bloomberg description for Bunzl:

Bunzl plc is a distribution group supplying a range of non-food consumable products for customers to operate their businesses but which they do not actually sell. The Company partners with both suppliers and customers in providing outsourcing solutions and service oriented distribution. Bunzl's main customer markets include grocery, foodservice, cleaning and safety.

Really, those are the only words I saw. They distribute. It's not food. It is consumable (great!). The customers need it, but they don't sell it. What could be better? And then who are the customers? They're "grocery, foodservice, cleaning and safety". I think I can understand those.

Then I probably go the LSE site if I haven't done that already and just run my eyes over operating income, return on capital, and whether revenue is up/down over last 3 years and magnitude of move. I still haven't calculated anything, just used my eyeballs.

Finally, I go to the company website. I want to make sure it's geared to customers instead of investors - unless it's a holding company or decentralized to the point where customers see name different from the one on the HQ. Except for that possibility, I want to make sure this isn't a promotional company. They can report their financials and have reports and stuff, but I'd rather not see a lot of investor oriented stuff on the front page. I want them to use the site to communicate with customers. I don't want it to feel like an IPO. The company should be unknown, unloved, unadvertised. That would be ideal.

Now, I go to the annual report (if the company has one - otherwise I settle for the 10-K). I open it up and peruse it. I read the entire letter from the CEO/Chairman. I glance at graphs/charts/table they include. What factors do they focus on? Do they talk about free cash flow? Good. Combined ratio? Good. Record Sales? Eh. Consecutive reading breaking years? Great.

At this point, I go to the financial statements and check operating cash flow and cap-ex. I'd like free cash flow to be positive. I check for operating income - though I probably saw this on the internet already - and want it to be positive too. I shouldn't see negative numbers anywhere. It should look like a business. Again, this is all eyeballs. I can tell a 15% profit margin from a 1.5% profit margin just by glancing at profit and sales I don't need a calculator. I can tell a 0.4 asset turn from a 4.0 asset turn. That's the kind of eyeballing we're talking about here. Nothing normal is interesting. I want especially high or low numbers in certain areas. I want a way to understand the business. If it has huge margins or really fast turns or something, I can start to understand it. If it has close to zero cap-ex I can start to understand it. If free cash flow is greater than reported earnings, I can start to understand it. I'm thinking how this could be even better than I first imagined. I want to know you can take capital out of the business every year and still grow the business. I want it to be like owning some timberland where maybe your land grows 7 new trees for the 100 you started with, you cut down six and next year you still have 101 trees instead of 100 trees. I'm looking for something like that. I'm not looking for something that grows but can't be harvested.

At this point, I've decided whether the company is really something I'm going to study as a possible investment or not. I've probably thrown out at least 95% of companies by this point. I'd say it's more like 99%. It's a big, big number. Very few stocks make it past this first impression of 10 - 20 minutes tops.

Once I get to this point, I create a folder on my computer to store all the company's annual reports. I download all the annual reports. Then I read them all, taking the data and putting it in a Microsoft Excel workbook I created. I use the last 11 years of data (really 10, but I need some year-end data from the 11th year for some calculations).

It can take me a while to enter all the data. For U.S. stocks, it's all in EDGAR. I just need their most recent 10-Q and their past 10 10-Ks. I have to do this all by hand, because I make adjustments to lines like "cash" to adjust for other marketable securities they have on the balance sheet that might be non-current. And I'm adjusting capital expenditures to add in certain spending on intangibles, pre-publication expense, etc. that a website might not include. Anyway, depending on how it goes this can take 20 minutes or 60 minutes. I do a split-screen in Windows 7, where I've got Excel on the left and the financial report on the right. I've gotten pretty fast with this. It would take a new person a lot longer, because they can't navigate the reports quite as fast as I can because I know where things normally are.

Within 20 minutes to 1 hour, I have all the 10-year data in. So now I have all the data I like to look at. I have the 10-year average free cash flow margin, return on capital, the F-Score, the Z-Score, the coefficient of variation on different stuff, price/multiples etc., I also have a sheet that shows me the DuPont analysis for all 10 years. And I can look at 10-year data for each of the variables I care about. I can basically do a Value Line type sheet with everything from free cash flow margin and return on invested capital to tangible asset turns to cash conversion...everything. I have these workbooks for every stock I've looked at. So, I can compare them on these numbers.

Finally, if I like what I see in terms of the numbers, I will then do 2 more things. I'll do a full historical workup on the company's past financial data. In the U.S., this usually means going back 15-17 years instead of just 10. So, I'll have a Value Line type page that goes back 15-17 years with earnings, sales, margins, turns, ROC, etc. I'll also do any bonus nuggets they throw in like store counts, square footage of retail space, unit volumes, etc. I'll look for any less common ratios that are important here, like railroads and cruise ships and airlines and hotels and movie theaters and store owners and stuff like that have industry specific ratios you want to look at going back 15-17 years and comparing to industry averages, competition, etc. So I spend another hour or whatever on that.

If I'm still interested, I read all the past annual reports from oldest to newest in order. Preferably, in one sitting. I'm not looking at notes to the financials and stuff like that. I'm reading everything other than the financial statements, notes, etc. I'm just looking for commentary on customers, segments, products, etc. I'm looking at capital allocation. I'm looking at management's way of talking about things.

Finally, I print out the latest 10-K, the latest 10-Q, and the latest 14A. I get a highlighter and a pen. I read the 10-K first, then the 14A, and finally the 10-Q. I go through highlighting and marking the margins with notes.

Once I do all that, I'll sometimes do other things like look for interviews with management, investor presentations, any references on the internet from trade publications, local newspapers, etc. Just looking for some additional coverage. When I'm doing this, I have specific questions. I probably want to know more about how/why customers choose between competitors.

Throughout this whole process I always keep a notepad on a clipboard with me wherever I go. I have this notepad, calculator and pen at my side at all times. And I just doodle calculation like say in 5-years the net margin is 4% on today's same dollar volume of sales, slap a 10 P/E on that, what's the compound annual return on the stock. Assume they don't pay any dividends, do buybacks, etc. for 10 years. Then a special dividend pays out all accumulated cash, what would that annual return be over 10 years. Is it better than the stock market overall is likely to do? Then, I can afford to get "stuck" in that stock. I sketch out scenarios real fast like that.

If I ever get to this point with a stock, I'm basically going to buy it. I would have dropped the idea earlier if I was ever going to drop it at all. I sometimes do other things like go through 8-Ks and 13Ds for the last few years. I'll check who owns a stock and what events have happened and stuff like that. But that's really only going to provide color on my investment at this point. Because if I made it this far, there's a much better than 50/50 chance I'll buy the stock.

So, I can't say there's one point where I have a real debate over buy/don't buy. Instead I just have a process that makes me eliminate 99.99% of stocks or whatever before letting me get to the end. I get uncomfortable about the stock at some point before the end of the process and then I just start over with the next one.

Talk to Geoff About Analyzing Stocks