Wednesday, December 14, 2011

So What is BRK Really Worth? (Part 3)

In Search of a Premium to Book Value
We looked at the two column method of BRK valuation and saw that yes, BRK is in fact cheap trading at 20-30% less than something like $150,000/share in intrinsic value.

However, we also noticed that if you value the investments held on BRK's books dollar for dollar, the expected return on that piece of BRK is not that particularly high. 

BRK followers often say that BRK is worth more than book value.  Buffett himself has said numerous times that BRK is worth far more than book value as book value doesn't take into account a lot of value created in the operating businesses over the years.  A favorite example is See's Candies, which BRK has owned for a long time and is obviously worth a lot more now than what it is valued on the balance sheet.  Another example is GEICO; Buffett says that GEICO is on the balance sheet at book value, but the goodwill of the business has grown dramatically and that is not reflected on the balance sheet.

However, I never really thought hard about the source of the premium to book value very specifically; for example, which segments deserve to be valued at a premium to book and by how much?

In order to figure this out, I decided to take a look at the ROE of each of the segments to see if in fact any of the segments should be valued at a premium to book, and if so, by how much.

The result of this digging has been a little surprising to me.

BRK, the Insurance Company
Here I'll start with the insurance company.  I have often wondered what the valuation of the insurance business should be if it was treated like any other insurance company.  Most BRK followers would tell you that it's impossible to compare as BRK is a different animal:  the investment side is run by the greatest investor of all time and the insurance underwriting side has managed to earn profits over time which is very rare in the insurance business.

I accept that argument, but let's see what the implied ROE of the insurance business is; how much is that investment prowess and underwriting excellence actually worth to shareholders?  What's the point if it's not going to earn a high ROE?

To figure this out, we can calculate what the expected returns on the investments are, assume some sort of underwriting profit and then see what the ROE of the insurance business is.  If the ROE is higher than 10%, then the insurance business may be worth more than book value, and if less than 10% it may be worth less than book value.  Anywhere around 10% would mean the insurance business is worth around book.

First, let's look at the various pieces of the insurance operation.  It's a little confusing at first because in the annual report balance sheet, the insurance segment is put into a section called "insurance and other" segment, but this includes the "manufacturing, services and retailing" segment.  So in order to look at the insurance business on it's own, we have to deduct the assets and liabilities of this segment in order to get the shareholders equity committed to the insurance operation.

Once we get that, all we have to do is add up the expected investment return on the investment portfolio held in the insurance business, add an assumed underwriting profit (it would be more conservative to assume breakeven insurance operations, but let's keep that in for now) and then see how much that amounts to compared to the insurance segment shareholders equity to get an ROE of the business.

From the 2010 annual report:

Shareholders equity in insurance segment:  $96.4 billion

Total investments
     Cash:       $25 billion
     Equities:  $60 billion
     Bonds:     $33 billion
     Other:      $19 billion
     Total:       $137 billion

Earned premiums in 2010:  $31 billion

So here is a sort of expected return on the investments:
    Cash:           $0  (0% interest rate)
    Equities:      $6 billion (10% return on stocks)
    Bonds:         $660 million (2% yield)
    Other:          $1.9 billion  (10% coupon on Bank of America preferreds etc...)
    Total:           $8.6 billion

BRK's underwriting profits since 1996 has averaged an after tax 2.2% of earned premiums, or a pretax 3.4%.

So using the a $31 billion earned premium figure for 2010, that's an underwriting profit of $1 billion.

Taking the two together, the insurance business would earn around $9.6 billion/year.  Compared to the shareholders equity of $96.4 billion, that's an ROE of 10%/year.  Some will argue that the 10% expected return on stocks is too high as BRK only owns the largest of the large cap names.  A figure of 6-7% is often tossed around.

If we take the expected return on stocks down to 7%, then the ROE of BRK's insurance business would come down to around 8%.

Is this business worth more than book?  Strictly speaking, I would think it's worth around book.  But keep in mind that this is a static analysis; it doesn't take into account things like growth in the business.  If the insurance business is overcapitalized and it can grow float and therefore increase investment and underwriting returns without raising more capital, that would obviously boost ROE.

However, when I value things, I like to look at returns on a static basis and get that 'growth' for free and not have to bake it into the cake.  So from that point of view, valuing BRK's insurance business at book value, to me, is very conservative, but not totally unreasonable.

Munger himself has also said that after growing float at such a high rate in past decades, it is unreasonable to assume that BRK can keep growing float like that.

Insurance Segment Taxes and Shareholders equity
The above analysis is a good sketch of how the insurance business works and how much BRK can make, but there are a couple of points to keep in mind.

First of all, the shareholders equity of the insurance segment would be reported differently at other insurance companies.  The fact is BRK reports $36 billion of deferred tax liabilities for unrealized capital gains on the stock portfolio.  This is not deducted in the above figures because BRK reports this liability outside of the insurance segment.  I think that's because BRK considers this liability an interest free loan from the U.S. government; it doesn't have to be paid until they sell stock with a very low cost basis.

If you deduct this $36 billion from the above segment shareholders equity, it would boost ROE, of course.  The fair value of the insurance business would then be 1.6x book value of the insurance business after deducting deferred taxes.  You can either call it 1.6x book value, or book value plus deferred taxes (which I prefer as this is the way the insurance segment is reported in the annual report; before deferred taxes).

Is it reasonable to not deduct deferred taxes?  Since BRK rarely sells stocks, it is not completely unreasonable.  If that is the case, then the fair value of BRK's insurance business is the segment shareholders' equity as reported on the balance sheet before deducting the tax liability (however, when valuing BRK as a whole, we would then have to not forget to add the taxes back as it is deducted 'below the line' in the annual report balance sheet).

Also, the above figures are pretax.  After tax, 8-10% ROE would translate into 5-6.5% returns.  This may be slightly higher as the equity portfolio would compound pretax as they don't have to pay taxes until they sell.

Either way, I think it is fair to assume that the insurance segment is worth 1.6 book value net of deferred taxes, or 1.0x book without deducting deferred taxes, or around $96.4 billion.

I am sure many will be surprised that it is not worth far more, but it's important to remember that back in the early 1990s BRK's equity holdings were 140%-150% of the insurance segment shareholders equity and even more than BRK's shareholders equity.  That's a lot of upside leverage to the stock-picking skills of Warren Buffett.  Just imagine, whatever he can pull off in the stock market lead to a levered return to BRK shareholders due to the size of the stock portfolio being bigger than the entire net worth of the company!

Now, however, the stock portfolio is only around 50% of the insurance segment shareholders equity (before deducting DT) and 38% of BRK's shareholders' equity.

It's easy to see how the expected return back in the early 1990s, just on this fact alone, was far higher than it is today. 

Conclusion to Part 3
So we decomposed the expected ROE of the insurance segment and we can put a value on it of around segment shareholders' equity of $96.4 billion.    The fair value book value ratio is lower (1.0x before deferred taxes or 1.6x after deducting deferred taxes) than we would have imagined largely due to a smaller stock portfolio as a percentage of net worth (less investment leverage from stocks) than in the past. 

In any case, here we find one source of premium to book value of BRK.  The insurance segment ROE as reported in the annual report is 10% or so, but that didn't deduct deferred taxes.  This means that the deferred taxes deducted outside of the segment should be added back in thus becoming one source of the premium to book value of BRK.   More on this later.


  1. Hi kk,

    I'm beating myself for not buying BRK or MKL a few years back. Well, my perspective and investment philosophy have shifted since then.... Anyway, I'm now re-reading your posts on BRK and MKL to get myself better educated. The prices may one day get back to the bargain bin level.

    A question: you mentioned above "back in the early 1990s BRK's equity holdings were 140%-150% of the insurance segment shareholders equity". How is that possible if Buffett (and insurers in general) keeps the float in cash/bonds for the required liquidity for paying claim? So, when you said "equity" here, did you include cash/bonds?

    1. Hi,

      The biggest item to account for that is deferred taxes. You can see the 1995 balance sheet here:

    2. Right. I can also see the float at the time was much small than shareholder equity.