Warren Buffett (Trades, Portfolio) was once asked about the similarities between investigative reporting and investing. He said:
"Well, it is interesting that you mention reporting because Bob Woodward I think back in '73 or '74 when I first got interested in The Post, we had lunch at the Madison and he was saying what he might do with his money and I said: 'Bob why don’t you assign yourself a story, get up an hour early every morning and work on a story you’ve assigned yourself? Now a sensible story to assign yourself would be what is the Washington Post Company worth? Now if [Ben] Bradlee gave you that story to work on what would you do for the next week or two? You go around and talk to people at television stations and you would try to figure out what are the key variables in valuing a television station and you would look at the four (TV stations) that the Post has and apply those standards.'"
The lack of a “reporter’s instinct” is one of the things that has surprised me the most when talking to other investors. Most investors rely a lot on information aimed at investors. So investor presentations, earnings calls, stories in Barron’s, maybe the annual report, the 10-K, the 10-Q, etc. But they don’t just take that off as their launching point and combine it with other easily available – and totally public – to answer questions that are important to valuing the company. They passively wait for information to come to them instead of actively answering the question. It’s like they assume that just because no one has tracked down the information and put it in print yet – they can’t either.
Sometimes, I’m not even talking about going beyond the SEC filings. I just mean using them in a different way. For example, I was talking to an investor who had put together a bunch of information on a specific stock. He had read the 10-K. And he’d put it together with write-ups on blog, Value Investors Club, etc. Because of the way this company he was considering investing in had been formed – the key variable was really just one agreement. What were the terms of that agreement? How long did the agreement run? Questions like that were key. Because the only real economic value this company had was being blessed with this favorable long-term agreement. Anyway, he said that he’d read through all this stuff and nowhere did the company say how long the agreement ran.
“Seventeen years,” I said.
“Where’d they report that?”
“They’ve never reported that,” I said. “But if you carefully read what they say about how they’re amortizing the agreement and then you look at the last two years and compare the balance sheet from one year to the next and compare that to what you see reported in the income statement during the intervening year, well, we can prove mathematically that it’s impossible for there to be more than 20 years or less than 15 years left on the agreement – and, if I had to guess, I’d definitely guess 17 years.”
Now, I understand some people aren’t as confident with a guess like that as they would be with having the company tell you the length of the agreement. But it’s the best you can do. And by triangulating a few facts, you can prove that either the company has made an error in describing how they amortize the deal (this shows up as a footnote in the 10-K), or they made an error on either the balance sheet or income statement, or you made a mistake interpreting those facts. If none of those mistakes were made, we do know roughly how long the deal runs.
It’s not unusual for there to be publicly available information released by a company that analysts and reporters overlook – not just individual investors.
Here’s another example.
I was researching a stock called DreamWorks Animation. It’s now owned by Universal. But it was a standalone company that had been – by then – split off from the live action DreamWorks studio. As a result, this company was only releasing one or two movies a year. Everyone knew which one or two movies those were. They knew the release dates. And the company did earnings calls each quarter. They had an output deal with HBO where HBO got to run DreamWorks movies once they left theaters but before other TV channels could get the rights to the movies.
It was often reported that the terms of that deal were unknown. The terms of that deal weren’t unknown. My newsletter co-writer, Quan Hoang, and I both knew them. And they weren’t even “undisclosed.” DreamWorks was publicly disclosing the terms of the deal to anyone who read their 10-Qs and 10-Ks including the footnotes on when accountants booked certain items. If you could precisely date when revenue from HBO would appear, you could see exactly how much HBO was paying for each movie.
True, the deals' terms weren’t published anywhere. But they were precisely knowable. DreamWorks had a note in its financial statements that said when it recognized revenue from HBO. It also discussed things like receivables due from different studios. If you knew what date a certain movie transitioned into the HBO window and you read the statements, you could tell exactly how much HBO was paying per movie. The same thing was true by the way when DreamWorks switched from an output deal with HBO to one with Netflix (NFLX, Financial). This was then muddied when DreamWorks started releasing TV shows made specifically for Netflix. But until that time, anyone who wanted to know how much HBO or Netflix was paying DreamWorks for each film could figure it out exactly using nothing but DreamWorks’ own 10-Ks and 10-Qs.
This is what I mean by acting like an investigative reporter. A lot of investors think that because a company doesn’t tell you something, you can’t find it out. Sometimes you can find it out exactly (as in the case of how much DreamWorks was paid per movie for first-run premium TV rights). Other times, you can guess.
Usually, your guess can be very educated. And you can educate yourself in about a day’s time.
I’ll give another example. I was reading through the 10-K of a company that was – at the time, this was maybe six years ago – a net-net. So I was just quickly checking through the quality of assets like cash, receivables and inventory. I noticed slight anomalies as I scanned the footnotes. Time after time, the company was being unusually aggressive in its depreciation. So if a category of assets could be depreciated between three and 10 years, you’d see they were doing it in three. They also owned things you’d expect them to lease. And then they wouldn’t just own the building. They’d own land around the building that seemed kind of excessive. I got to thinking that the company was probably understating its earnings and its assets by owning a lot of stuff that they were depreciating rapidly and then keeping in use. There’s a note in the financial statements that actually told me the dollar value of assets still in use that had been fully depreciated. It was a big number. Accumulated depreciation was also a big number.
So I looked at the properties the company owned and I went online to find satellite photos of where these properties were (this takes 60 seconds to do nowadays). I also found what counties the properties were in and quickly Googled how to look up property tax records for those counties.
At this point, I was off on an investigative reporter type hunt. I’d invested less than an hour into reading and Googling at this point. It might have been a wild goose chase. But it might be something else.
In this case, it was something else.
I found a property record on one building that showed they were almost certainly carrying the land, the building, improvements and the equipment inside that building for a small fraction of the appraisal value of the land for tax purposes. What the land is appraised at for tax purposes doesn’t prove anything. But I was looking at a picture of the building and it wasn’t old. It was worth a lot more than what it was carried for. Now, normally that’s not interesting. But, remember, this stock was already selling below book value when I first found it. And now I’ve learned that the fair value of its assets is way above book value. When looking at the record and the satellite photo I noticed something odder than that. There seemed to be two buildings on this property. The company had – in a previous year – described an “improvement” to the property but not another building. They were depreciating a second building as an improvement and doing it as rapidly as they were allowed to depreciate an improvement. As a result, I was looking at a building that despite not being terribly old was clearly there in the photo but was no longer on the company’s books at all. And, of course, the cost of that building had passed through previous years of income statements as a depreciation expense. This makes it look like today’s assets are lower than they’re really worth. And it also makes it look like yesterday’s earnings were lower than they really were.
Everything I’ve described for you here was done in one morning sitting in one seat in front of a computer. I didn’t start my look into this company before 9 a.m. I thought at first it’d just be a typical net-net. And then, I was done looking at it by noon. And what I found was a pattern of behavior that understated assets and understated past earnings. Ben Graham would’ve bought this stock. Walter Schloss would’ve bought this stock. It was something you’d expect to find in the 1950s in the U.S. or maybe in Japan in the 2000s. But not in the U.S. in 2010.
It’s all the kind of thing a reporter could easily do without leaving their desk. And they would do it because they think like a reporter not like an investor. I didn’t even call anyone up, email them, etc. All I did was use one photo from Google, what the company filed with the SEC and online land records.
The surprises in that case were all positive.
Sometimes, they’re negative.
I don’t short stocks. And I don’t write negative posts about stocks where I learned something unsettling. I’m not looking for controversy. So if I’m going to say bad things about a company – as I am here – I won’t mention the specific name. Both of these stocks were Chinese companies that had gotten on U.S. exchanges through a reverse merger of some kind. In both cases, a shareholder who owned the stock was emailing me to please assure him it wasn’t a fraud. In both cases, I couldn’t assure him it wasn’t a fraud because – again, within three hours or less – I was pretty sure I had found facts that were only consistent with a fraud.
Now, it helps to know accounting. I’m not an accountant. I really didn’t go to either high school or college. I was basically a dropout after eighth grade. So I don’t have any formal training in accounting. But I’ve been reading 10-Ks since I was 14. And almost every initial lead I’ve gotten on a company comes from something accounting related in the 10-K. It’s usually something in the footnotes. Or, more accurately, it’s a combination of more than one item that seems weird. For example, in the case of the company that was overdepreciating everything – it was just all over their financial statements that they were doing the opposite of what you wanted to do if you were trying to report the highest possible EPS. It was like they were trying to report the lowest possible EPS. That’s odd for a public company. So I focused on that topic.
Here, I’d been brought two potential frauds. So I decided I would focus on those things that could quickly establish a fraud. So I scanned the 10-Ks looking for any hard facts that the company might either forget to fraudulently adjust or literally couldn’t fraudulently adjust. In other words, I assumed that there definitely was a business here doing things. But it’s possible the company was inflating everything 1,000 times or something like that. My assumption was: they aren’t insane – it’s not like they are going to make up the fact they operate a paper plant if in reality they run a landfill. So I thought that at least some nonaccounting disclosures would be accurate. Even a liar wasn’t going to lie about stuff they didn’t think needed to be lied about. It would be pointless and complicated to do that.
Anyway, in one case I found reference to a tax the company was supposed to pay. It was based on the number of units of something. If you multiplied this per item tax by the number of units the company said it had sold – it didn’t match. Something was off. And it was off by a few zeroes. Once you find something like that, I don’t think you can buy the stock. Now, when I tell a shareholder of a company something like that they often say, “Isn’t it possible that” and it’s true that there are other possibilities. If I really was a reporter and not an investor I’d try to talk to someone at the company and try to talk to someone in the Chinese government agency responsible for collecting this tax and see whether the company was: misreporting the tax per unit, cheating the government rather than cheating investors, or there was some complex explanation where a tax benefit of some kind I couldn’t find anywhere in the statements was muddying the waters so bad it offset what I expected to find. As an investor, I can keep the story to myself. So I don’t go asking for management’s side of the story. For me, the story ends there.
The other Chinese fraud I was brought had a strange discrepancy with time. Basically, you can use financial statements to come up with figures like inventory turnover. I’m sure you know that. But what you might not have thought about before is that metrics like receivable turns, inventory turns, etc., are giving you more than just usual financial stats, they are actually telling you something about the time it takes to: sell inventory once you take possession of it, pay bills, collect payments, the length of employee pay periods (a company that cashes out its staff daily looks totally different from a company that pays salaried employees just once a month), etc.
Now, with that in mind. Here’s what I found. This company described in a note how and when it recognized revenue and elsewhere it described its business operations in terms of what it actually did. The description of what it did made sense to me (I wasn’t knowledgeable about the product/process to know for sure whether the description was accurate), but the inventory turns were implying the process happened very, very rapidly (impossibly rapidly in my uneducated opinion) when the business description was implying the process happened at what seemed a normal and reasonable speed. I checked everywhere to see if I was misunderstanding when they actually put stuff into inventory or if somehow they were keeping things on a supplier’s books till almost the moment they passed that thing on in finished form to the customer. Again, a shareholder in this fraud could say “Isn’t it possible” and come up with an explanation for how this company could have made the mistake of incorrectly describing when it recognizes sales, puts things received from suppliers on its books, etc. It’s possible. But, to me – you forget about the stock right there. If everything the company was saying was true, it was a fraud. If it wasn’t a fraud, there had to be something I was misunderstanding or they were explaining badly in how they prepared their books. And it would have to be something very basic to the essential functions of the business. For example: they could have been saying they owned something that in reality they contracted with someone else for and never took possession of. Then, the stock wouldn’t be a fraud in the way I thought it was a fraud. But, it would still be painting a picture for investors that was inaccurate.
Of course, if I was a reporter I would have talked to management, talked to suppliers and talked to ex-employees.
I’m not a reporter. And I’m not a short-seller. So when I see a case like this I just move on. I’m only writing about it here – and not mentioning the name of the company in any of these cases – to encourage people to go out and act like a reporter when it comes to researching a stock.
There are tons of stocks where you can do this even today. Anything related to real estate where you can find comps is really easy. You can do it from behind a desk. You can probably do most of it with just Google – not ever needing to talk on the phone or email. So not quite like a real reporter.
In a recent blog post, I mentioned Ingles Market (IMKTA, Financial). It’s a supermarket, but it owns a lot more assets than a normal supermarket would. Why not assign yourself the story of what Ingles Markets is really worth? What is all the PP&E on its books worth if sold off? There’s a company called J.W. Mays (MAYS, Financial). It owns real estate in Brooklyn. I’ve been to Brooklyn. And I’m guessing Brooklyn’s a nicer place today than when Mays put that property on its books. Perhaps there’s a discrepancy between fair market value and book value. Assign yourself the story. Staying in real estate, how about Seritage (SRG, Financial)? We know Buffett (personally, not through Berkshire [BRK.A][BRK.B]) owns this stock. It was spun off from Sears (SHLD, Financial). A lot has been written about Sears, and there’s a lot of public information available about why Sears would separate Seritage, what rates Sears is paying to rent from Seritage, and also what rates another tenant might pay for the same space. Assign yourself the story: what is Seritage’s property worth?
You can do this with companies that aren’t involved in real estate. I’ve mentioned Ark Restaurants (ARKR, Financial) before. In an earnings call earlier this year, the CEO said:
“We are restaurant guys. I (don’t) know what the valuation on our company should be. I just know that we are trying to build value, but the way I always look at this in my own way was that we have now some $25 million plus from coming from restaurant operations. We got $11 million corporate overhead, part of that’s attributable to what’s going into Meadowlands, which now is probably two years from something happening, but we have $25 million of corporate overhead for restaurant operating profit and $11 million of corporate overhead. We find the right guy to sell this to, looking at $25 million worth of restaurant operating profit, not the $40 million we expect. I don’t know what the multiple should be on that. And I don’t know how Wall Street looks at that. But we look – if we were to sell, we would look to pair off with somebody who mirrors our locations with their locations and doesn’t need a corporate overhead.”
He may not know how to value his company, but you can find out how to do it. You can research the referendum on the Meadowlands that failed last year and learn whether it might be back on the ballot in New Jersey in 2018. You can Google if anyone ever tried to buy Ark (they did; Landry’s made a hostile bid years ago) and at what price. So you can research what the chances of there being gambling in the Meadowlands one day are. And you can research why Landry’s would make a bid for Ark and what kind of people run Landry’s and how they think and so on.
That’s what a reporter would do. It’s also what Buffett did in the 1950s and 1960s especially. If you read “The Snowball,” you’ll see that Buffett started with getting a lead from the Moody’s manual or the annual report of a company combined with his knowledge of accounting (like I suggested you do even today), but then he assigned himself the story of figuring out what the business was worth. He got friends of his to do some of the leg work for him. Sometimes he went out into the world and talked to people who’d know the answer to questions he had. And sometimes he did it in libraries. You can do almost all of it from behind a desk using just an internet connection today. But you still have to do it.
You have a lot of money riding on your investment. Treat your stock research as seriously as you’d treat any project your boss assigns you at work. Think like a reporter and learn to “investigate” a stock instead of just reading what others have written about it.
Talk to Geoff about acting like an investigative reporter
Disclosures: None.
