(Felix Salmon) Jonathan Stempel of Reuters has a very useful look at what's going on at Berkshire Hathaway, which fell $6,500 per share today to close at its lowest level in over five years. After reading his article, I think we might be one step closer to understanding why Berkshire's CDS are trading so very wide right now. But first, here's David Gaffen:
In recent weeks, the credit-default swaps has seen a marked decline in liquidity and trading, so a smaller amount of insurance contracts purchased can still cause large shifts in prices of a particular credit-default swap. "It only needs a little bit to move the market a long way," says Tim Backshall, chief credit derivatives strategist at Credit Derivatives Research.
Now here's Stempel:
A credit rating downgrade would likely not be material. Berkshire would have to post "nominal" additional collateral on derivatives of "far below 1 percent of assets" if Berkshire lost its "triple-A" ratings, Buffett's assistant, Jackie Wilson, said. It was posting no such collateral as of Sept 30, when Berkshire assets totaled $281.7 billion.
Put those two together, and things start to come into focus. The people who bought derivatives from Berkshire weren't worried about counterparty risk at the time. But now, looking at what's happened to triple-A counterparties such as AIG and the monolines, everybody is worried about counterparty risk. And nowhere is counterparty risk higher than in these derivative contracts written by Berkshire Hathaway -- because even if it's downgraded, it only has to put up a negligible amount of collateral.
Says Whitney Tilson, who knows Berkshire intimately:
I doubt Buffett would write any contracts that would require Berkshire to post collateral in the event of a downgrade.
Which is all you need to know, really: if that's true, then anybody needing to hedge their BRK counterparty risk has no choice but to buy CDS protection, whatever the price.
If the stock is falling based on the idea that the CDS market might be on to something, then maybe Tilson is right that it's a screaming buy at these levels. (Although one wonders whether it wouldn't be more profitable to just write credit protection instead.)
But we're in a Dow 7,500 market right now: there are lots of screaming buys out there. Back to Gaffen:
"The insurance business which is a significant portion of their earnings, is a very cyclical business," says Harry Rady, CEO of Rady Asset Management in San Diego. "With everything else at an extremely significant discount, I see a lot more downside. There is so much opportunity to buy assets for 50 cents on the dollar, why buy a dollar for a dollar because Warren Buffett runs it?"
Or, to put it another way: if Warren Buffett isn't buying back his own shares at these levels, then that must be because he sees better uses for Berkshire's capital elsewhere. Which in turn means that you and I might be better off elsewhere, too.
Of course, Buffett might be buying back his own shares right now, we don't know. But the way in which those shares are plunging would suggest that he isn't. They ended last week at $101,000 apiece; now they're $77,500 -- a drop of 23% over the course of four trading sessions. If they just return to where they were on Friday, an investor buying at these levels would make a profit of more than 30%. But the same thing can be said of many other stocks, too, including ones in Buffett's portfolio such as Goldman Sachs and General Electric -- both of which, unlike Berkshire, have access to unlimited Fed liquidity.
So yes, Berkshire might be looking cheap -- but other stocks are looking cheaper, and always remember that Berkshire, like any insurance company, is very highly leveraged, with contingent liabilities many times higher than its asset base. AIG and the monolines were brought down because they stopped insuring real-world risks and started insuring financial risks. Berkshire's in the same business, with its credit default swaps and equity puts. Which means that in fraught times like these, there will be lots of question marks over its real value.