"You raise some interesting points, but I am mostly unconvinced. I suppose that if you look at the P/B as a valuation measure, the case is less convincing, but frankly, I don't much care to worry about the historical acquisition costs of the assets Buffett manages, when I have access to the amount of earnings they are generating. It is the earnings and their growth rate that will ultimately determine the value of this investment, not the book value.
Your first comment about the necessity of that dry powder: I agree, having the cash and the equity portfolio arguably makes for some (sorry to say the word) synergy, in that they should enhance the reliability of the insurance operations (although Buffett regularly bemoans the fact that BH doesn't seem to be able to get much of a premium for it.)
Clearly, I am not saying that BH SHOULD be broken up into parts, I am only saying that it CAN BE VALUED based on the sum of its parts."
EmbraceableEwe followed with a very thoughtful set of comments.
I will offer my 2-cents worth on some of both of their points:
1. Yes-Book value is important as an indication of valuation, but less so than in the past when the value of the company was dominated by the insurance business and its associated investments. Earnings power from the non-insurance businesses is now more important than in the past, accounting for by my estimate approximately 30 percent of BRK's current value, either market price or intrinsic.
2. WEB or Charlie said about 2 years ago that BRK needs to hold $10B of its cash hoard for working capital. Based on the company's growth since then, I estimate this needs to be about $12B by the end of this year, the rest being available for suitable investments. I see nothing that indicates we need to hold most of the cash to shore up confidence. After all, the fixed income and equity components of book value are probably the biggest reason for the AAA credit rating.
3. I too like the "sum of its parts" way of thinking about valuation, but I think a P/E is only good for the earnings associated with the non-insurance businesses. The earnings for the insurance businesses are too volatile because of pricing, catastrophes and realized capital gains. I like to look at the 2 columns, investments and earnings of the non-insurance businesses that WEB emphasized in the 2005 Annual Report. Since that time I have been tracking what I have been calling the 2-column metric, i.e., investments plus 15 times trailing 12-month net earnings of the non-insurance businesses (15 being the P/E I am applying to the earnings of this collection of businesses). The value of this 2-column metric seems to come in at 80-90 percent of the various estimates people do of BRK's intrinsic value. It underestimates the value of the insurance businesses by implicitly assuming a zero rather than positive underwriting profit and it gives no value to WEB's investment acumen, i.e., his ability to deploy cash into acquisitions or public investments. I find the 2-column metric a little more intuitively satisfying in helping me think about BRK's value than valuation models that mix the insurance earnings with those of the non-insurance businesses. Of course, there are multiple ways to do valuation estimates and each can offer insight, but I have gravitated mainly to this 2-column metric since the 2005 report.
To calculate the value of the metric I estimate the value of investments in three different ways from the quarterly reports and calculate a range of values. The range is about +/-1 percent around the mid-point, so what follows below are the mid-point results. The results per BRK A share are as follows:(Sorry about any formatting difficulties in the table below)
So, over the past 21months, 7 quarters, investments have grown a total of 26 percent, Non-Insurance earnings 65 percent, and the 2-column metric 36 percent to $134.7K per share. This implies a typical calculation of intrinsic value will come in at about $149K to $168K.
During the 7 quarters, (end date to end date), the S&P500 has grown, including dividends, about 25 percent. BRK's growth in intrinsic value has exceeded the S&P 500 by 11 percent over these 7 quarters.
A few other observations:
4. The growth in earnings is higher than what is a reasonable estimate of the organic growth rate of the non-insurance businesses. The past growth rate has been helped a lot from business acquisitions. This acquisition rate has slowed substantially since mid 2006. Until/unless it picks up again, I expect the non-insurance EPS growth rate to slow. However, my choice of a P/E of 15 in calculating the 2-column metric in essence gives no value to potential acquisitions. They get included only after they occur. So, this situation does not detract from the current value of the 2-column metric. It simply says it may not grow as fast as it has recently unless we get some more good acquisitions.
5. With all the subprime turmoil and massive actual and anticipated write-downs so far across the financial services sector, nothing in BRK's results indicates any adverse effect whatsoever or any impairment to its financial health from the credit crisis. Even AIG is suffering. (The housing slowdown did have an effect on the housing related businesses.) This probably contributed to the recent run-up in BRK's stock price.
6. Everything still looks quite good to me, certainly relative to alternative equity investments available in the public market to which I am willing to commit a substantial fraction (25+ percent) of my portfolio.