Annual Report Notes and SOTP Valuation

[jay21]

I read the Annual Report this weekend and came away with a few notes and comments on valuation.  The main reason why I am starting this thread is to try to do a SOTP valuation as I think the investments plus 8 to 10 times multiple of earnings is an insufficient valuation method. 

 

The most telling comment in the shareholder's letter was in the Manufacturing and Retailing section, where in reference to the businesses in that segment he wrote: "Furthermore, the intrinsic value of the businesses, in aggregate, exceeds their carrying value by a good margin. Even so, the difference between intrinsic value and carrying value in the insurance and regulated industry segments is far greater. It is there that the huge winners reside."

 

So those are the businesses that we should focus on.  I'll start with the regulated business entities.  There's some background info here:

 

http://www.cornerofberkshireandfairfax.ca/forum/berkshire-hathaway/bnsf-and-midamerican/

 

BNSF made $5,377 million pre-tax and recognized $2,005 million in income tax.  However, only $1,194 million was paid in taxes.  Tax advantages such as these are reasons why I think that the 8 to 10 pre-tax multiple applied may be low.  I think UNP is the most appropriate comparable and is trading at a 16.7 P/E (which compares to the 19x Buffet paid).  This would imply a value of $56,312 million.  I think the railroad industry may be slightly undervalued, but the $56 billion value seems reasonable and I believe Buffet paid $34 billion.

 

MidAmerican enjoys tax benefits as well, but the reinvestment opportunities appear to be worse.  If we were to mark this to market, we would use a pretty high multiple given the low interest rate environment.  I think there is some merit to using a higher multiple given the weaker correlation to economic conditions.  I am going to use a 12x multiple even though the market may use a 15 - 20x multiple.  This gives us a  $15.9 billion valuation.  A MTM valuation using the low multiple would be $19.8 billion. 

 

$56 billion plus $16 billion is $72 billion for the regulated business value as what I would consider a low point.  I don’t think this is much different than a 10x pre-tax number but I also think it is a low point.

 

I am valuing the insurance companies on a P/BV basis.  Using insurance assets of $278 billion minus the $42 billion in cash and liabilities of $120 billion (doesn’t include deferred taxes), leaves equity of $116 billion.  Backing out the $48.6 billion of the Retailing segment’s equity leaves $67.4 billion.  I think that the insurance ops are worth 2x BV given the history of underwriting profits, high auto multiples, and high proportion of equity investments.  That gives a value of $134.8 billions and add back the free cash to get $176.8 billion.  Total value now at $248.8 billion, which is close to the $256 billion in market cap today.  If the retailing and financial products are only worth book, then the value tops $300 billion.  I believe this is a conservative valuation and I think it’s close to the investments plus 10x pre-tax multiple.

[original mungerville]

I agree that Berkshire remains undervalued - more so relative to the market (as I believe current corporate profits margins are unsustainable).

 

Its important to understand that Buffet bought a railroad worth at least 56 Billion for over a 40% discount! That's a lot of value-add for one deal! Soros made $1B shorting the Yen just recently, Buffet made $20 Billion on BNSF!

 

With that kind of deal-making Berkshire still has some decent runway in front of it. Meanwhile Goldman warrants and Heinz deals will keep Warren out of the bars while he waits for the next big deal.

 

 

[kevin4u2]

How do you come up with your multiples? You seem to pull some of them out of thin air. Is their any reason why brk's insurance companies are worth 2x book while the rest of the industry trades at a discount to book. I would add that insurance valuations are quite reasonable given the state of the fixed income market today. What I'm trying to say is why do you use comparables in some instances and not others? Thanks.

[jay21]

How do you come up with your multiples? You seem to pull some of them out of thin air. Is their any reason why brk's insurance companies are worth 2x book while the rest of the industry trades at a discount to book. I would add that insurance valuations are quite reasonable given the state of the fixed income market today. What I'm trying to say is why do you use comparables in some instances and not others? Thanks.

 

I am looking at value my expectations of cash flows and returns on capital. A MTM based upon comps is a datapoint. 

 

I noted some qualitative factors as to why I believe that the insurance co is worth 2x BV (history of underwriting profits, high auto multiples, and high proportion of equity investments).  Some insurance cos do trade at a discount.  However, MKL, WRB, and RLI trade at 1.26, 1.41, and 1.95 respectively.  Given that BRK's investment portfolio will have higher returns due to the equity weighting, a higher than market multiple may be appropriate.  PGR (as a comp to Geico) trades at a 2.55 BV

 

I am open to more criticism as the point of the thread is to look at the actual businesses that are owned by BRK and develop CF and return expectations about them.