I have previously written about investing in reinsurance companies, followed up by how to use free cash flow to help analyze these companies, but I intentionally left out one of the behemoths in this space: Berkshire Hathaway (BRK.B). Part of the reason is because BRK.B is a conglomerate that invests in a diverse array of industries, including pipelines, railroad, private jet leasing, candy stores, and jewelry stores, in addition to various lines of insurance. It didn't seem fair or easy to lump this conglomerate into a comparison of other companies that have operations limited to reinsurance and insurance lines. In addition, it is harder to compare the metrics when the insurance lines are buried in such large numbers produced by all the other sectors in which Berkshire conducts business. Well, my curiosity got the better of me, so I wanted to at least see how Berkshire's reinsurance lines compare, in terms of valuation and profitability, versus other companies in that sector. I also want to enlist some help by Seeking Alpha readers, since I know that Berkshire's followers are some of the savviest investors I have come across.
Diving into the Berkshire 10Q
The most recent 10Q at the time of writing this article is for the third quarter of 2012, and the filing is found here. Comparing the overall performance of Berkshire insurance segments is difficult but not impossible. The profitability of underwriting and investment income is broken out by insurance segment. The loss ratio is listed as well, but requires digging into the footnotes. What is not possible is to evaluate the industry standard measurement of combined ratio, although we can approximate that by subtracting out the segment expenses to compute a net margin by insurance line.
The other reason I chose to compare Berkshire Hathaway to reinsurance companies is because I wanted to test the widely-held belief that reinsurance operations are a key driver of Berkshire profits. A nice analysis by Vinod Palikala from 2010 can be found here. (Linked thanks to author's permission)
Insurance Segments in Berkshire
According to Berkshire's filings, the insurance is grouped into four categories (although some of these categories contain multiple brands within them):
- General Re
- Berkshire Reinsurance Group
- Berkshire Primary Insurance Group
For third-quarter 2012, these lines had the following reported revenue and earnings. (Note that the net margin is my own calculation of earnings divided by gross revenue, based on these reported figures. Berkshire does not report combined ratios for any of the segments, so I have to use this net margin as the closest approximation.)
|Revenue (M)||Earnings||Net Margin|
|Berkshire Reinsurance Group||$2578||$-102||-4%|
|Berkshire Hathaway Primary Group||$588||$121||21%|
Next, I graphed each of these four insurance lines according to their relative percentage contribution to overall insurance segment profit (see graph, below). I could have used gross revenue, or gross premium written, but I am more interested in the overall profitability. Note the negative number for Berkshire Re and that using gross premium written would not have revealed this loss. The interested reader can go to the 10Q filing footnotes and identify the reasons behind this loss, and this will be discussed below. The astute reader may also question why the two reinsurance lines, General and Berkshire Re, had such drastically different profit/loss in the same quarter. I can't help but wonder why they are maintained as separate segments on accounting statements. Perhaps that is fodder for future analysis.
This analysis gives some immediate food for thought. First of all, GEICO, the familiar car insurance company, accounts for approximately half of Berkshire's total insurance revenue and over 60% of insurance profits. I like its commercials, but an auto insurance company is not what I want to invest in. Secondly, its two lines of reinsurance, General Re and Berkshire Re, are large but not fantastically profitable. As already noted, Berkshire Re resulted in a loss this particular quarter, and as always, one should go back multiple quarters to see the longer-term trend. One could also look at year to date to get a smoothed number. (This analysis is limited to most recent quarter, as an example of where to get the numbers and how to analyze some of this yourself rather than getting the analysis via sell-side analyst reports, for example.)
In Berkshire's favor is that the above analysis does not include the investment return on "the float" (paid-in insurance premium), and proponents would likely argue the use of this capital is Berkshire's singular biggest strength. I don't disagree. For reference, the most recent 10Q filing footnotes list the current float as $72 billion. No doubt, Warren Buffett can put that money to use for generating significant investment returns.
My analysis is trying to compare its actual insurance business to others in that sector. If I had wanted to compare it to other conglomerates, the closest competitor is probably something like Blackstone Group (BX), which invests in private equity and also engages in asset management, or perhaps a value-oriented mutual fund.
Just to dive a little deeper into the quarterly loss for Berkshire Re in Q3 2012, let's compare with another reinsurance player, Axis Capital Holdings (AXS). This company has a respectable $4 billion market cap, which makes it mid-size among global reinsurance companies. It has a current price to book of 78% compared with BRK.b price to book of 117% (you cannot compute the book value on a specific segment of Berkshire, and even if you could, you still couldn't isolate an investment to just one segment - you have to buy the whole thing!). AXS has a combined ratio of 85%, making it somewhat equivalent to a net margin of 15% versus the loss of 4% that Berkshire Re segment had for the same quarter. Even General Re with a net margin of 11% did not perform as well as AXS. So what gives? Why not buy an apparently better-performing company at a substantial discount to book value, while also getting a pure play on reinsurance?
This comparison of Berkshire Re to AXS shows that the loss was not industry-wide, and that makes it a good reason to dig in deeper into the footnotes to get a more complete story. The Berkshire 10Q lists a few things such as $118 M loss in foreign currency transactions for the Berkshire Re segment, and also some deferred charges being reamortized. Also, the footnotes specifically state there was an absence of large catastrophe losses in the quarter. So, it seems likely that the loss was of a non-recurring and non-operating nature. However, the entire amount of loss that quarter was accounted for the foreign currency loss, and the reader should at least ask why the company had exposure to that large a loss (especially since there are ways of hedging against that), and asking whether similar future risks still lurk within the company. Also, long-term deferred charges and periodic re-amortizations sound like an overhang the segment faces, and is a fair thing to inquire about, given that this analysis is specifically aimed at understanding the reinsurance business.
Proportion of profits from insurance
Now that we have reviewed the four separate insurance lines within Berkshire, we will next look at how insurance as a whole contributes to Berkshire's profits.
|Revenues||Expenses||Net income||Gross Margin|
|Railroad, utilities, and energy||$8,419||$6,369||$2,050||24%|
|Finance and financial products||$1,079||$832||$247||23%|
Finally, I will take the net income from each segment and display in pie chart format to show the relative contribution of each segment to overall profits.
This graphically shows that insurance overall is indeed a major driver of Berkshire's profits, accounting for 61% of earnings in Q3 of 2012. As noted earlier, about 60% of that insurance number was due to GEICO. Reinsurance was much smaller; even if one ignored the quarterly loss of Berkshire Re segment, reinsurance was still about one fourth of overall insurance profits. Thus, while Warren Buffett often talks about how great the reinsurance underwriting is, at least in this most recent quarter I don't see it contributing directly to much of the overall profits. It seems much more accurate to discuss the insurance as being driven by auto insurance via the GEICO segment. While GEICO has cost structure advantages versus competitors and thus a respectable net margin, it is still merely retail insurance and something I tend to avoid, and certainly not a segment I would pay above book value for.
I am still trying to wrap my mind around how to approach Berkshire. If one wanted to invest in only its insurance lines, it seems possible to do a paired trade with long BRK.b / short SPY (short about 40% of the BRK position size, to hedge out the other diversified non-insurance businesses). At a first guess, it seems that Berkshire's diversified businesses might be a pretty good proxy for a broad index such as SPY; it is big enough that it would be hard for it to deviate too far from the overall direction of the economy and the total stock market. As further support of the long side of this paired trade, Berkshire has recently increased the "floor" of its share price by raising the buyback limit order buy price from 110% of book to 120% of book, as analyzed here. The stock currently trades at 117% of book value, so some buybacks may be imminently under way.
However, if I go back to my initial reason for analyzing Berkshire as an insurance company, I reach a different conclusion. My decision is that while Berkshire might be a very well-run, value-oriented company with a superb investment track record, I simply cannot personally invest in it as a reinsurance play. That component of its business is simply too small to be a material part of the investment thesis, at under 2.5% of total profits (worse if you include the quarterly loss in Berkshire Re segment). Furthermore, most of this insurance profit is due to the GEICO division, which while profitable, is not the exposure to reinsurance that I seek.
In summary, I feel that Berkshire is a highly profitable company with a stellar track record at reinvesting its own capital, but it does not meet my basic criteria as a primary reinsurance company and thus I will not be including it within my future reinsurance analyses. Despite this, I will maintain Berkshire on my personal "shopping list," and given an appropriate entry point I might consider a small position in it for the portion of my portfolio designed to deliver market-like returns. I am normally an index investor for the bulk of my portfolio, but would consider Berkshire to be one of the few companies with a chance at delivering some alpha.
So again, a paired trade such as long BRK.b / and partially hedge by shorting half the amount in SPY or other vehicle might hedge out the market risk and give some exposure to any alpha that Berkshire is able to generate - but that's a trade far removed from my initial thesis of investing in it as a reinsurance play.
Thanks for reading, and have a fantastic 2013.