From the 10K:

BHRG periodically assumes risks under retroactive reinsurance contracts. Retroactive reinsurance contracts afford protection to ceding companies against the adverse development of claims arising under policies issued in prior years. Coverage under such contracts is provided on an excess basis or immediately with respect to losses payable after the inception of the contract. Coverage provided is normally subject to a large aggregate limit of indemnification. Significant amounts of environmental and latent injury claims may arise under the contracts. Under certain contracts written over the last five years, the limits of indemnification provided are exceptionally large. In March 2007, an agreement became effective between NICO and Equitas, a London based entity established to reinsure and manage the 1992 and prior years? non-life liabilities of the Names or Underwriters at Lloyd?s of London. Under the agreement NICO is providing up to $7 billion of new reinsurance to Equitas. In 2009, NICO agreed to provide up to 5 billion Swiss Francs (approximately $5.3 billion as of December?31, 2011) of aggregate excess retroactive protection to Swiss Reinsurance Company Ltd. and its affiliates (?Swiss Re?). In 2010, BHRG entered into a reinsurance agreement with Continental Casualty Company, a subsidiary of CNA Financial Corporation (?CNA?), and several of CNA?s other insurance subsidiaries (collectively the ?CNA Companies?) under which BHRG assumed the asbestos and environmental pollution liabilities of the CNA Companies subject to a maximum limit of indemnification of $4 billion. In 2011, BHRG entered into a contract with Eaglestone Reinsurance Company, a subsidiary of American International Group, Inc. (?AIG?). Under the contract, BHRG agreed to reinsure the bulk of AIG?s U.S. asbestos liabilities up to a maximum limit of indemnification of $3.5 billion.

Retroactive insurance is an interesting business, and one that few insurers have as a core skill.? It is my estimate today that BRK is good at it, unlike most.? With Retroactive insurance, typically you are rescuing another insurer from some claim exposure that threatens their existence.? The insurer needs certainty, or something near it, and so they approach a much large and stronger insurer to absorb some of the risk of an exposure that is already incurred, but uncertain to to ultimate payout.

The rescuing insurer will charge a lot, and insist that the rescued insurer still have some risk on the matter, and probably limit its total payout, after which the rescued insurer is on the hook again.? That limit will likely be so high that the rescued insurer will say, “It’s never going to get that high.”? Fine, and maybe true, but this allows the rescuing insurer to have some certainty itself, that it will never pay an unlimited amount in the rescue.

For BRK, there is another angle, and that is that retroactive insurance produces a lot of float, and in most cases (Asbestos & Environmental) the float lasts a long time.? Thus BRK thinks it has an advantage in investing the float.? Together with their size, and the acumen of Ajit Jain, it makes them a unique place for insurers in trouble to seek shelter, for a tidy fee of course.

That doesn’t mean this can’t go wrong, but if properly managed, since BRK is one of the few companies that can do this, they probably make very good money on this.? (Ugh, they have AIG and Swiss Re as clients, which are large savvy firms.? If they need protection from BRK, and are willing to pay up, guess what — BRK is in the driver’s seat, because there is no one in the private sector capable of doing this.)

Insurance subsidiaries? investments are unusually concentrated and fair values are subject to loss in value.

Compared to other insurers, our insurance subsidiaries may concentrate an unusually high percentage of their investments in equity securities and may diversify their investment portfolios far less than is conventional. A significant decline in the fair values of our larger investments may produce a large decrease in our consolidated shareholders? equity and can have a material adverse effect on our consolidated book value per share. Under certain circumstances, significant declines in the fair values of these investments may require the recognition of losses in the statement of earnings.

This is potentially BRK’s largest weakness, and why I would love to see the statutory books for their insurers.? This goes beyond the large public companies that they have purchased.? Where are all of the private businesses lodged on the BRK balance sheet?? They may be there with really low valuations — I don’t know, because I have never looked — and that is why I want to ask BRK for their statutory statements.? I believe it would be intriguing.

Berkshire Hathaway Inc. has guaranteed debt obligations of certain of its subsidiaries. As of December?31, 2011, the unpaid balance of subsidiary debt guaranteed by Berkshire totaled approximately $16 billion. Berkshire?s guarantee of subsidiary debt is an absolute, unconditional and irrevocable guarantee for the full and prompt payment when due of all present and future payment obligations. Berkshire also provides guarantees in connection with long-term equity index put option and credit default contracts entered into by a subsidiary. The estimated fair value of liabilities recorded under such contracts was approximately $10.0 billion as of December?31, 2011. The amount of subsidiary payments under these contracts, if any, is contingent upon future events. The timing of subsidiary payments, if any, will not be fully known for several decades.

Add in $8 billion of holding company debt, and that is the risk that the holding company faces, which isn’t that much for a company the size of BRK.? BRK has bought a series of businesses that produce consistent cash flow, so don’t worry about holding company debt.

BRK has more debt than that, but Buffett lets bond investors take the risk by not guaranteeing subsidiary debt.? In acquisitions, Buffett never guarantees the debt.? But even with new debt issues, many BRK subsidiaries offer their own non-BRK-guaranteed debt, whether it is Burlington Northern, or a Mid-American sub offering debt to back a solar power plant.? Now bond investors know the BRK would never walk away from a subsidiary’s bonds, right?

No, actually they don’t know that, though Buffett’s record has been good with the debts of non-guaranteed subsidiaries.? Buffett is a better risk than most, but at the subsidiary level, you can’t be sure.? Buffett could save money and take on more risk by borrowing at the holding company, but he typically does not do that.

P/B Multiple Compression

Over the last decade, BRK’s P/B multiple has shrunk, and shrunk to the point where Buffett has drawn a line in the sand saying that subject to liquidity and market conditions, he will buy back stock at levels below 110% of stated book value.? (And, I suspect that is one reason why he spent so much time in the recent letter attempting to explain why the goodwill at BRK represented value.)? I think that means that there is now a limit, a floor to the price of BRK common stock, of course, subject to continued adequate performance.

Book Value and Insurers

Some have criticized Buffett’s growth in Book Value vs. the total return of the S&P 500 table, partly because of the declining P/B multiple.? I would simply say that this is endemic to an insurance mindset, where all we do care about is growth in net worth (book value).? Insurance is a mature, stable business.? No one has a way of obsoleting it, so we suspect.? It’s difficult to start a new insurer of any significant size, so we have protected boundaries to a degree.

Thus the focus on growing book value.? If we grow book value, eventually market value will follow, right? Right?!

Ugh, I think so, but sometimes I wonder, particularly with all of the insurers trading under book when they have little risk of insolvency.? Buffett can draw his line in the sand, but what of other insurers, many of which trade well below book?? Should they draw their own line in the sand, and defend a valuation level?

Personally, I would announce a generalized buyback without a lot of hoopla; make it boring, this is insurance after all so that should not be hard; get some actuaries to toss in big words to aid in obfuscation, so few conclude that the company will buy back significant stock.? Then start nibbling; be the bid or just behind it on days when things are weak. Don’t be a pig; rule #1 here is that the market should doubt that you are there.? But subject to that, buy, buy, BUY!!

The great Buffett after announcing his buyback only bought back 0.03% of all shares so far.? Most insurers can do much more quietly, with far less fanfare.? After all, Assurant bought back 14% of its shares last year.? Wow.? (This was the insurer that was offered in full to Buffett, and he said it was too complex… even the great one can goof.)

Now, maybe Assurant is my extreme case in buying back, and doing it at good valuations, but that in my opinion should be the goal of most public businesses with low valuations that are earning a lot, and not getting the proper respect.? Money talks, but be quiet, not brash, and suck in shares quietly at low valuations.? Let the financial statements do the bragging for you when investors realize that you have been building value doubly through operations and buybacks.

Summary

Even if you don’t invest with him, you can learn a ton from Buffett.? He is a consummate investor, businessman, and insurance executive.? Though I have never met him, I consider myself blessed to have learned from him.

Full disclosure: long AIZ