10x Pretax Earnings! Case Studies: KO, BNI etc.

So, one of the great things about writing a blog is that I get feedback from some pretty intelligent people.  Most of us don't have a Munger to call, but a blog works well too.  If I say something wrong, I'm sure someone would jump in to point it out.

10% Pretax for Stocks Too?
Anyway, I have mentioned 10x pretax earnings or 10% pretax yield as Buffett's valuation measure numerous times here and more than once I've gotten a response saying that this hurdle is for private deals and not for pricing listed companies.  The argument, of course, is that if you buy a stock at 10% pretax earnings, you won't actually earn 10% pretax (due to the additional tax at the investee corporate level whereas in a wholly owned business, a 10% pretax return is actually a 10% pretax return).

It is true that when Buffett speaks of returns in the stock market, he uses GDP growth and dividend yields; earnings can't grow more than GDP and stock returns will reflect earnings growth over time plus whatever dividends you get.

Translating that into individual stocks, you will get earnings growth plus dividend yield equals expected return on the stock.

The only problem with this is that it doesn't tell you what the business is worth.  Would you pay 50x p/e for it?  20x?  The above calculation only works if valuation stays the same.

Anyway, my usual response to this is that many value investors (including Buffett) likes to analyze businesses based on what a rational businessperson would pay for the business in a private transaction.

So, if Buffett is willing to pay 10x pretax earnings for Wells Fargo in a private transaction to buy the whole thing, that is a valuation benchmark for me.  I know that this is not actually possible.  There are size and regulatory issues that will make this unlikely.  But in terms of valuing businesses, I think it is still a useful benchmark.

Is this how Buffett thinks about it? If he pays 10x pretax earnings for WFC stock, he will not necessarily earn a 10% pretax yield.  I don't know the answer to that question. Maybe that's a good annual meeting question.

But as long as I know that Buffett  would be totally happy to pay 10x pretax for the whole business, that's good enough for me regardless of whether that will actually happen.

Yes, you can argue that these "private business transaction" valuations are only valid when there is some chance of a private deal occurring.  But I only think of that when the private valuations don't make too much economic sense to me; valuations per eyeball or per POP valuations in the past, for example, or 40x EV/EBITDA for some media assets, or per acre land valuations etc.; just because some people are paying high prices doesn't mean anything unless there is a real prospect that what you are looking at will also be taken out at some point at the same high level.

Is 10x Pretax Reasonable? 
But 10x pretax earnings, even for listed companies, is not unreasonable at all.  You can translate that 10x pretax into a 15x after tax p/e ratio, and that wouldn't be far off from the 100 year or so long term average of U.S. listed businesses.   Since Buffett buys quality, above average businesses, paying 10x pretax is like paying an average price for an above average business.

So even if my view is wrong, it passes the rationality test; why not pay average prices for above average businesses?  And this is not dependent on market p/e or interest rates because you are using a long term average.  We are not increasing valuations due to decreased interest rates.

Case Studies
So, this discussion piqued my interest again so I decided to go back and look at some more of Buffett's big deals.  I use the term "case study", but it's far from it, really.  I'm only looking at one measure; pretax earnings yield or price to pretax profits.  So I apologize for the exaggerated terminology and to folks who come here looking for a 300 page paper on why Buffett bought something; you'll only see one line.  We all know how great the businesses he owns are, so there really is no need to look at that.

So, I looked at the 2005 purchase of Wells Fargo, Walmart and the recent IBM purchases, but what about some of the other older ones?

Again, since the Warren Buffet Library of Corporate Annual Reports doesn't exist yet, I can only look at some of them.

I got lucky and found a 1988 annual report of Coke, so that's good. Let's start there.

Below, let's take a quick look at Coke (KO), American Express (AXP) and Burlington Northern (BNI) (which a prominent value investing academic said was a crazy/insane deal or some such.  We'll see if it really was a bad price) and some others.

Coca Cola  (KO)
From the 1988 KO annual report:

Pretax earnings:   $1,582 million
Net earnings:        $1,045 million
EPS:   $2.85
Shares outstanding: 365 million
Year-end stock price:  $44.63

From the Berkshire Hathaway annual reports, the cost of KO was:

                     #shares owned           cost ('000)       cost/share (my calculation)
1988 AR       14,172,500                  $592,540       $41.81
1989 AR       23,350,000               $1,023,920       $43.85

So with $1,582 million in pretax profits and 365 million shares outstanding, that's $4.33/share in pretax earnings per share.

So it turns out he paid 9.7x pretax earnings as of 1988 and 10.1x pretax earnings as of the end of 1989.  

That's a pretty stunning discovery, even for me.  I think a lot of value investors were puzzled at what looked like a growth stock purchase by Buffett at the time, but it fits right in with the 10x pretax benchmark perfectly.  He didn't pay up because KO was a really high quality business; he paid what he normally pays.


American Express  (AXP)
So this one doesn't quite fit the mold, but let's take a quick look at it (it doesn't fit only because he didn't pay almost exactly 10x pretax earnings, but far less).

The 1994 annual report is the first time AXP showed up in the BRK letter so let's look at that and what he paid for it:

                        # of shares owned          cost  ('000)          cost/share (my calculation)
1994 AR         27,759,941                     $729,919             $26.29

By the way, I know that this is only an estimated cost per share of the stocks.  There may be some adjustments somewhere that might throw this off, but I don't think it would change things materially.

Thankfully, the SEC database goes back to 1994, so let's pull the relevant AXP figures from 1994:

Pretax earnings:  $1,891 million
EPS: $2.75
Shares outstanding:  496 million

So we don't even have to go very far with this one.  It looks like Buffett paid 9.6x net earnings for AXP.

Pretax earnings per share comes to $3.81/share, so he paid a 6.9x pretax earnings.

It looks like he got AXP really cheaply.   It got pretty cheap in 2009 too.

Moving on.

U.S. Bancorp (USB)
It looks like he started buying USB in 2006, but maybe earlier.  It shows up first in 2006 on the annual report.  He bought more in 2007.  This is from the annual reports:

                     #shares owned           cost ($mn)       cost/share (my calculation)
2006 AR       31,033,800                  $969                 $31.22
2007 AR       75,176,026                $2,417                $32.45

And here are the figures for USB in 2006 and 2007:

              Pretax          diluted (mn)                  Pretax
              earnings       shares outstanding        EPS
2006       $6,912         1,804                            $3.83
2007       $6,282         1,758                            $3.57

2013       $7,990         1,849                            $4.32

So in 2006, BRK was paying 8.2x pretax earings, and the total cost through 2007 comes to 9.1x pretax earnings of 2007.

And interestingly, BRK increased shares held in USB from 78 million in 2012 to 96 million at the end of 2013.  The pretax EPS of USB was $4.32 in 2013 and the stock traded in the range of 7.4 - 9.5x that figure throughout the year.



Burlington Northern (BNI)
So this is one of his other major purchases that made everyone scratch their heads.  There are two things to look at here; one purchase when he just bought the shares and then a second time when he bought out the whole company.  Let's take a look.

The first time BNI appeared in the annual report was 2007.  In 2006, he said there were two positions worth $1.6 billion that was not listed, so BNI was probably purchased in 2006 and other times too (could be some before and some after).

Here is what the 2007 BRK annual report showed:

                       # of shares owned          cost  ($mn)          cost/share (my calculation)
2007 AR         60,828,818                     $4,731                  $77.78

And these are the figures for BNI for 2006 and 2007:

                                                   2006               2007
EPS:                                           $5.11              $5.10
Pretax earnings:                         $2.96 bn         $3.0 bn
diluted shares outstanding:           370 mn        359 mn
Pretax EPS:                                $8.11             $8.25

So from this, it looks like Buffett was paying 9.4x - 9.6x pretax earnings per share.  Voila!

And then of course, BRK bought the whole thing in late 2009 (on an announcement basis).  The offer price was $100, so let's see what the BNI figures were for 2009.  Even though the figures haven't come out yet when the announcement was made, most of the year was over, so they would have known pretty much what the earnings were going to be.

Here it is:

BNI 2009
EPS:   $6.08
Pretax earnings:  $3,368 mn
Diluted shares outstanding:  348 million
Pretax EPS:   $9.68

So at $100/share, Buffett paid 10.3x pretax EPS of BNI.

A lot of people thought Buffett overpaid, but it turns out he just paid what he always seems to pay.  I know, I know.  What about capex, maintanence capex / depreciation and all that?   Yes, that was the argument back then.  I don't know.  I'm just looking at this and noticing a pattern.  I don't have all the answers!

BNI Tangent
What's a blog post here without a tangent?  As I was doing this stuff, I just took a quick look at the famous 'projections' of BNI that was included in the merger proxy.  Buffett has said that he ignores these management projections, but these are often done by management / investment bankers in mergers so they can do their cash flow discount model analysis and whatnot.

So here are the various projections for BNI from the proxy dated December 2009:

2010 Recovery Case

		2009E			2010E			2011E			2012E			2013E			2014E			CAGR
		(In millions, except per share and percentage data)
Total revenue		$	14,013		$	14,994		$	16,601		$	17,611		$	18,667		$	19,418		6.7	%
Freight revenue w/o fuel			12,372			12,830			14,063			15,014			15,834			16,558		6.0	%
Operating income			3,204			3,421			4,336			4,921			5,360			5,745		12.4	%
EBITDA			4,737			5,052			6,056			6,746			7,313			7,825		10.6	%
Net income			1,631			1,717			2,224			2,476			2,663			2,831		11.7	%
Earnings per share			4.77			5.04			6.88			8.41			9.71			10.96		18.1	%

2011 Recovery Case

		2009E			2010E			2011E			2012E			2013E			2014E			CAGR
		(In millions, except per share and percentage data)
Total revenue		$	14,013		$	14,254		$	15,436		$	16,629		$	17,839		$	18,877		6.1	%
Freight revenue w/o fuel			12,372			12,424			13,345			14,291			15,244			16,044		5.3	%
Operating income			3,204			3,092			3,638			4,241			4,775			5,209		10.2	%
EBITDA			4,737			4,723			5,357			6,063			6,724			7,283		9.0	%
Net income			1,631			1,515			1,842			2,149			2,386			2,572		9.5	%
Earnings per share			4.77			4.41			5.43			6.74			8.10			9.35		14.4	%

No Recovery Case

		2009E			2010E			2011E			2012E			2013E			2014E			CAGR
		(In millions, except per share and percentage data)
Total revenue		$	14,013		$	14,012		$	14,410		$	14,622		$	14,844		$	15,069		1.5	%
Freight revenue w/o fuel			12,372			12,377			12,736			12,953			13,176			13,401		1.6	%
Operating income			3,204			3,010			3,224			3,324			3,314			3,310		0.7	%
EBITDA			4,737			4,639			4,939			5,138			5,249			5,363		2.5	%
Net income			1,631			1,465			1,607			1,660			1,631			1,610		(0.3	%)
Earnings per share			4.77			4.27			4.65			4.87			4.94			5.07		1.2	%

Deeper Recession Case

		2009E			2010E			2011E			2012E			2013E			2014E			CAGR
		(In millions, except per share and percentage data)
Total revenue		$	14,013		$	13,544		$	13,618		$	14,000		$	14,283		$	14,756		1.0	%
Freight revenue w/o fuel			12,372			12,107			12,147			12,351			12,632			12,929		0.9	%
Operating income			3,204			2,759			2,728			2,778			2,841			2,898		(2.0	%)
EBITDA			4,737			4,387			4,440			4,588			4,770			4,943		0.9	%
Net income			1,631			1,310			1,295			1,326			1,369			1,399		(3.0	%)
Earnings per share			4.77			3.82			3.74			3.80			3.89			4.05		(3.2	%)

And check this out.  These are the figures for 2013 that BNI actually booked (from the BNI 10-K):

BNI 2013 Results
Revenues:             $22,014 million
Operating income:  $6,667 million
Pretax income:        $5,928 million
Net income:             $3,793 million

The most bullish projection in 2009 was for operating income of $5,360 million and net income of $2,663 million.  Operating income came in 24% higher and net came in 42% higher!


Lubrizol
OK, so here's one more acquisition.  This name might not give BRK holders a warm and fuzzy feeling (due to the Sokol incident), but it is a major acquisition so it is a relevant data point.

BRK bought Lubrizol in 2011 for $135/share.

For the 2010 year, here are some figures for Lubrizol:

EPS:  $10.64
Pretax earnings:  $1.00 billion
Diluted shares outstanding:  68.8 million
Pretax EPS:  $14.53

So a $135/share purchase is 9.3x pretax EPS.

Recap
So let's just recap all of this stuff I said in the last post (part 5) and this one.

These are the multiples to pretax earnings Buffett paid in these big deals:

KO in 1988/89:    10.1x
AXP in 1994:         6.9x
WMT in 2005:      10.3 - 12.9x (range of stock price in2005)
WFC in 2005:         9x
USB in 2006/2007:  9.1x
USB in 2013:  7.4 - 9.5x (range of stock price in 2013)
BNI stock purchase: 9.5x
BNI acquisition:  10.3x
Lubrizol:   9.3x
IBM:  9.7x

I exclude Heinz here as it is a different situation and I think he said he wouldn't have done the deal without 3G.  I may be missing some here as I didn't intend this to be comprehensive by any means, but just looked quickly at some of the large purchases he has made over the years and it is very interesting.

Conclusion
It's amazing how so many of the deals cluster around the 10x pretax earnings ratio despite these businesses being in different industries with different capital expenditure needs and things like that.

Even the BNI acquisition, which many thought was overpriced (crazy / insane deal! Buffett has lost his marbles!) looks normal by this measure; a price that Buffett has always been paying.

And yes, right now I'm the guy swinging around a hammer (seeing only nails), but I notice a pattern and think it's really interesting.

Of course, this actually makes no sense as every company has different capital needs (free cash flow / owner earnings etc.)  Of course, what Buffett calls "owner earnings" are more important than pretax profits.  This was one of the arguments about the BNI deal.

And it is silly to think you can price anything and everything at 10x pretax profits.  Buffett obviously looks at everything else and has a deep understanding of the various businesses and is only willing to pay this amount for the very best businesses out there.

Why he says he will pay 9-10x pretax earnings (OK, for private deals) and yet seems to go out and pay 9-10x pretax earnings on stocks is a good and valid question.

It's amazing, though, isn't it?  Even if it is an odd coincidence.

But I don't think 10x pretax earnings for a stock is a bad price if it's a high quality business that can grow over time etc...  (But you still have to answer the question how much growth there will be and how much a shareholder can expect to get back.)