I was deeply disappointed by the May 25th New York Times article on Fairfax Financial (NYSE: FFH) by the Pulitzer Prize-winning business columnist, Gretchen Morgenson, Gauging the Big Bets of a Hot Hand. We usually enjoy her work, but this story on Fairfax is certainly an anomaly: it’s a smear and a hatchet job, filled with innuendo, inaccuracies, and misleading statements.
As one of our largest positions, we know Fairfax well and, having once been short the stock, we’re very familiar with the short thesis, which we (obviously) believe is now outdated and wrong. We welcome contrary points of view and, in fact, when we disclosed that we were long the stock last August in a Value Investor Insight article (the stock’s up 22% since then), we heard from some investors who were short the stock and had insightful conversations with them.
Morgenson’s article, however, sheds no insight whatsoever and, in fact, leaves the reader with a picture of Fairfax that is the polar opposite of reality. Let’s go through it carefully:
1) It is a big player in credit derivatives — instruments that have recently burned larger and more established rivals like A.I.G. and Warren E. Buffett‘s Berkshire Hathaway. Indeed, while both of those companies have lately recorded huge losses on credit default swaps — insurance contracts that allow investors to bet for or against a corporation’s bonds — Fairfax has reported immense gains in this arena.
The way Morgenson writes this makes it seem like Fairfax is doing something suspicious or maybe lying about their gains, since AIG and Berkshire are reporting big losses, yet Fairfax is reporting big gains on credit derivatives. Three milliseconds of analysis, however, explains the apparent disconnect: Fairfax bought credit-default swaps, while AIG and Berkshire sold CDSs, so of course Fairfax’s results would be the opposite of AIG and Berkshire! 2) During 2007 and the first quarter this year, Fairfax said it realized $420 million in gains on credit default swaps, helping to propel earnings to records.
I don’t know where Morgenson gets this number, but she’s off by a factor of 5. Fairfax’s filings, summarized on page 29 of the slide presentation from Fairfax’s recent annual meeting (posted here), show that the company has had gains of more than $2 billion on its CDSs. In addition, why did she write “...Fairfax said it realized...” rather than “...Fairfax realized...”? It’s a subtle but, in the context of the rest of the article, no doubt deliberate implication that the company might be lying.
3) The insurance business hasn’t been hot lately for Fairfax, so its investment returns have come in handy. Citing extreme competition in the industry, Fairfax reported consolidated net premiums of $4.5 billion in 2007, a decline of 6.1 percent from 2006. And in the first quarter of 2008, the company reported an underwriting loss of $7.7 million versus a profit of almost $50 million a year earlier.
While factually correct, Morgenson presents Fairfax’s insurance results in the most negative way possible. In fact, in recent years through 2007, Fairfax’s insurance businesses have been doing great. As the charts on page 4 of its 2007 annual report show (also see pages 7-11 of annual meeting slide presentation), Fairfax’s three major insurance subsidiaries have been growing 19-22% annually from 2001-2007, their ROE in 2007 ranged from 23%-26%, and their combined ratios fell. Yes, premiums declined in 2007, but this is smart, as Fairfax wrote less business as pricing got poorer – exactly what Buffett does. Morgenson fails to mention that Fairfax’s combined ratio fell from 109.4 in 2005 to 95.5 in 2006 to 94.0 in 2007, driving a 32.3% gain in underwriting profit in 2007. She also fails to mention that in Q1, net premium written declined a mere 0.5% and that the underwriting loss she highlights was impacted by “a $25.5 million pre-tax charge associated with the settlement of an asbestos lawsuit by Crum & Forster”. 4) Plenty of derivative bets remain on the company’s books, even after recent sales. Unrealized gains on the company’s credit default swaps stood at $960 million at the end of 2007, and in the first quarter, mark-to-market gains on its credit default swaps were $467 million.
That’s a lot of lettuce. Still, credit default swaps are notoriously volatile, and whether the company will be able to cash in these chips is an open question. Indeed, on May 1, Fairfax cautioned that swaps worth $991 million at the end of the first quarter had fallen to $685 million as of April 25. All of this is true, but if Morgenson is going to raise the question “whether the company will be able to cash in these chips”, why doesn’t she note exactly how much of those chips Fairfax has already cashed in? The answer (from Fairfax’s filings, noted on page 96 of our mortgage presentation, available at www.valueinvestingcongress.com) is that Fairfax has harvested nearly $1.1 billion in cash from its CDS portfolio (all in the last three quarters, an extraordinary $885 million in Q1 alone!). 5) Sean Egan, managing director of Egan-Jones Ratings, an independent ratings firm, said investors should be wary of counting on these gains because it is not clear who is on the other side of Fairfax’s swap trades.
"The visibility on counterparties has been and remains very poor,” Mr. Egan said. “You don’t know whether there are hedge funds on the other side with minimal capital levels or whether they are solid credible counterparties. These are massive positions, and Fairfax should provide additional information to give comfort to investors.”
Egan hasn’t done his homework on Fairfax. The company has, in fact, disclosed a great deal of information regarding the counterparties to its CDS positions. One of its subsidiaries, Odyssey America Reinsurance, whose CDS holdings mirror Fairfax’s, makes quarterly filings with the NAIC with precise details about every CDS contract it holds. Attached is the latest filing, with information as of 3/31/08. As you can see on pages 31-32, column 6, the counterparties are primarily Citibank Canada, Deutsche Bank and Barclays, hardly “hedge funds…with minimal capital levels”. In addition, on a recent conference call, Fairfax confirmed that its primary counterparties were Citibank and Deutsche Bank.
As noted above, Fairfax has collected nearly $1.1 billion in cash from its counterparties to date, which certainly speaks to how well Fairfax has handled this risk.
Finally, Fairfax has negotiated very favorable terms on its CDS contracts such that its counterparties must post cash collateral as spreads widen. From page 12 of Fairfax’s Q1 interim report:
The company endeavours to limit counterparty risk through the terms of agreements negotiated with the counterparties to the credit default swap contracts. Pursuant to the agreements governing the credit default swaps as negotiated by the company with the counterparties, the counterparties to these transactions are contractually required to deposit government securities in collateral accounts for the benefit of the company in amounts related to the then current fair value of the credit default swaps. The fair value of this collateral at March 31, 2008 was $737.5 ($886.0 at December 31, 2007). Fairfax has not exercised its right to sell or repledge $152.5 (nil at December 31, 2007) of this collateral.
6) Given the volatility in these instruments, he added, “investors would have to assume that investment gains realized by the company are unlikely to remain at the same level that have been reported to date.”
Simply because these instruments are volatile says nothing about whether the volatility in the future will be in the upward or downward direction. With roughly $18 billion of notional exposure, valued at only $685 million as of April 25th, there’s plenty of room for more gains. If Egan thinks that CDSs on financial companies exposed to the mortgage mess and credit crunch are likely to decline in value, he should say so and defend this point of view.
7) Absent Fairfax’s investment income, he noted, the company’s pretax earnings would have totaled $520 million in 2007, not the $2.16 billion it recorded — quite a swing, but not one that seems to worry the folks at Fairfax.
What an obnoxious way for Morgenson to phrase the end of this sentence. Here’s what she’s really saying: “Fairfax’s gains on CDSs are unlikely to be repeated, so therefore Fairfax’s management should be worried. They don’t appear to be, so therefore they must be stupid or crooked.”
But why should Fairfax’s management be worried about huge gains? (I wish I had this worry!) They made a fantastic bet, making more than 10x their money in less than one year, have already harvested more than half of it in cold, hard cash, and are keeping some of the bet on, believing (correctly in our opinion) that more gains are to come.
But they are certainly not promising this to investors, nor are big CDS gains built into the stock price, as it’s trading at a mere 10% premium to book value. In fact, even if Fairfax’s entire current CDS position goes to zero, the company has already booked enough cash gains that this will have been a highly profitable investment.
8) …even with the juicy derivatives gains, at the end of the first quarter Fairfax’s cash and cash equivalents had fallen to $3.1 billion, down 46 percent from the end of 2006.
This is the worst smear in the article, as it is technically correct, but leaves a completely wrong impression on the reader that Fairfax’s financial condition is worsening. The truth, by any measure, is exactly the opposite, thanks in large part to huge gains in the CDSs Morgenson decries.
Let’s look at the facts: according to Fairfax’s cash flow statement in its 2007 annual report (the bottom of page 25), the company’s cash declined from $5.764 billion at the end of 2006 to $3.113 billion at the end of 2007 (it was flat to $3.109 billion at the end of Q1 08).
So why did cash fall so much ($2.651 billion) in 2007, a year in which Fairfax had strong underwriting profits (up 32.3%) and huge gains from its CDS portfolio? Once again, a three millisecond analysis of the previous page shows that Fairfax invested its cash in “available for sale securities” (a net investment of $1.420 billion) and “short term investments” ($1.538 billion), a total of $2.958 billion, more than explaining the decline in cash. This isn’t a bad thing – insurance companies are supposed to invest their cash in securities and bonds!
To get a good picture of Fairfax’s financial health, here are some relevant metrics (from page 9 of the annual report and page 12 of the Q1 interim report):
In short, completely contrary to Morgenson’s implication, Fairfax’s financial health has improved dramatically since the end of 2006 – interest coverage, for example, has gone from 5.2x to 28.9x in only five quarters!
9) Fairfax is on the hook for whatever claims Sphere Drake has to pay, yet details relating to that particular possibility aren’t disclosed in Fairfax’s filings.
Mr. Taylor said that was because the proceedings have no impact on his company’s operations, adding that the claims will be immaterial. He said he expects the claims consolidated in the most recent solvent scheme of arrangement to be paid later this year or in early 2009.
Shareholders will have to hope the Mr. Taylor is right and that the claims don’t turn out to be whoppers. If they are, Fairfax’s finances might get squeezed.
Morgenson throws in everything but the kitchen sink here – offshore entities, asbestos, litigation, etc. – but fails to present any evidence that calls into question Fairfax’s CFO’s statement that future losses here will be immaterial. Prem Watsa echoed this in his 2007 annual letter, writing:
As mentioned last year, our runoff operations have stabilized, and they did not need any cash from Fairfax in 2007. We continue to feel that our runoff operations will in the future no longer need any cash that will be significant in relation to holding company cash. In fact, Dennis Gibbs is so comfortable with our runoff operations that he has delegated the day to day runoff operations in the U.S. to Bill Gillett and in Europe to Nick Bentley. Dennis will continue to monitor the runoff operations on a strategic basis.
We’ve looked into these exposures and believe that Fairfax is telling the truth. Morgenson makes a big deal about “solvent schemes of arrangement”, but noted that the first one led to losses of only $35 million, so why does she believe that the next one might turn out to be a “whopper”? And even if it’s 10x larger ($350 million), it would barely make a dent in Fairfax, which is holding $1.155 billion in cash at the holding company level as of the end of Q1.
10) Indeed, while some liken Mr. Watsa to Mr. Buffett, the Berkshire Hathaway model consists of a company generating profits through careful insurance underwriting and perspicacious investing. At Fairfax, the underwriting cushion would seem not so plump should Mr. Watsa’s hot investment hand go cold.
Morgenson is correct that Fairfax and Berkshire are very different companies, but not for the reasons she cites. In fact, Fairfax is very similar to Berkshire in its careful underwriting and perspicacious investing. Rather, the main differences are: A) Berkshire has diversified far beyond insurance; and B) Over time, Fairfax has acquired a number of distressed insurance companies and turned them around –like what Jack Byrne does at White Mountains, but not what Buffett does (at least not intentionally).
But in terms of careful underwriting, look at the table at the bottom of page 7 of Fairfax’s annual report, which shows that Fairfax’s cost of float since inception in 1986 has been minus 2.5%, meaning that the company has shown underwriting profits on average – a fabulous record (and this includes some years of big losses from acquired businesses – i.e., Fairfax was not responsible for the bad underwriting; instead, it was able to acquire a company cheaply because of it and then fixed the problem).
As for perspicacious investing and Watsa’s “hot investment hand” (a phrase that makes it seem like he’s just been lucky recently), Fairfax’s investment track record has been consistently phenomenal over a very long period of time. As the chart of page 10 of Fairfax’s 2007 annual report shows, the results are as follows:
To summarize, we find no merit in Morgenson’s obviously biased hatchet job and continue to believe that Fairfax, trading at book value, is a steal. Even if one subtracts its entire CDS portfolio, the stock is trading at 1.27x book value. We think Fairfax’s core business is worth 1.5x book value, so at today’s price, the stock does not fully reflect this value, plus we’re getting a free call option on Fairfax’s CDS portfolio.
Disclosure: Author holds a long position in FFH