Q: Let's discuss some of your holdings, starting on the short side. A: The first one is Wells Fargo [WFC], trading at 12 times '08 estimates and 2.7 times tangible book; the group trades at less than two times book. The Wells Fargo name has a storied past and gets the Warren Buffett halo effect because he owns a lot of the shares. But if you look back at the last real-estate recession in the early 1990s, the Wells Fargo side, focused on California, had a lot of credit problems in the real-estate area, and the stock underperformed during that period. The Norwest side, which has more exposure to the Midwest, still has a lot of consumer-credit exposure. Of particular concern is the bank's portfolio of home-equity loans.

Q: Let's move on to another short position. A: BB&T [BBT], which operates in the Southeast. The stock trades at 11 times '08 earnings and 2.5 times tangible book. It's bounced about 30% off its lows in January. They've gotten a pass because, to some extent, their core Carolina and Virginia real-estate markets were among the last to roll into home-price depreciation. So their non-performing assets are still low. But we listened to the Toll Brothers [TOL] conference [call] recently. [Chairman and Chief Executive] Robert Toll graded the markets they operate in and he gave Charlotte an F-minus for current home-building conditions and Raleigh a C-minus. We're also concerned that they have 4% of their portfolio in Alt-A mortgages, which are between prime and subprime, and 20% in construction loans. As with many of the financials, there was this big relief rally on first-quarter earnings. The thinking was these results weren't so bad, but we think that more credit losses are ahead of us.

Q: How about a different short holding? A: Hudson City Bancorp [HCBK], which is based in New Jersey. The shares have gained about 60% from their July '07 lows and now trade at 21 times '08 estimates and two times tangible book. They have a wholesale funding and asset-generation strategy, which allows them to keep expenses low. Basically, they borrow funds from the Federal Home Loan Bank of New York and use repurchase agreements. So they, in effect, purchase money, more so than relying on deposits. In addition, they buy most of their assets, usually through brokers. A big chunk of their assets are first mortgages and mortgage-backed securities. So for the most part, they are not a retail originator of loans, and they are benefiting from the steepness of the yield curve. And, even with all that, they will earn less than a 10% return on equity this year, so I just don't get the valuation. Furthermore, when you have a wholesale business model, that means you don't really have a valuable franchise that another bank would pay much for. So a recent stock price represented approximately a 100% premium for their core deposits, which is how bank acquisitions typically get priced. And I don't think their deposits are worth nearly that much to a buyer. Everything looks great for them now, if you a call 10% ROE great. But they are not immune to credit risk in a recession and a weak housing market. I also think their loan-loss reserve at 0.15% is very low, relative to others'. When the Fed rate-cutting cycle is over, I don't want to own a spread play with credit risk that's trading at two times book.

Q: Let's move to the long side of the ledger. What financial firms do you like? A: One is Hatteras Financial [HTS], based in Winston-Salem, N.C. They are a mortgage REIT. We bought it in a private placement last year and it recently went public at $24 a share. They own all agency adjustable-rate securities, so there is no credit risk here. The market capitalization is about $630 million.

Q: What made you switch to the long side? A: The stock's down from north of 70 in early 2007 to around 12, bringing its capitalization down to about $1.5 billion. One reason we like the company is that it was able to raise more capital recently, something its competitors haven't been able to do. In March, they did a common offering that raised about $500 million -- so they've been able to raise liquidity and capital. At the same time, they are raising prices on premiums and tightening underwriting on the business now being written. The new business is being written based on lower home appraisals after the housing bubble burst -- and yet they are still showing good growth despite the fact that the whole industry has slowed down. Our thesis is that once MGIC gets through writing down its old book of business, the new book will be very profitable and valuable. Even applying our bearish mortgage-credit outlook, we don't see more than another $4 or $5 per share of losses in the next two years. So current book value of $24 should bottom in the high teens in '09 and start rising from there. They'll be quite profitable after that, given their better margins and more conservative underwriting of the current book of business. The stock should trade upward of two times book. So we see the stock, currently at around 12, as a double-to-triple over the next few years.