Subprime: Gross Losses are Much Bigger Than Net Losses

I'd like to thank Floyd Norris for responding to my blog entry last week about one of his columns. Norris is a blogger in his own right, of course, but it's always good to see venerable journalists venturing off their own turf and commenting elsewhere, especially when what they say really does clarify matters considerably.

Norris had written this:

One of the more remarkable facts about the subprime crisis is that total losses to the financial system may be about equal to the amount of subprime loans that were issued.

Paul Krugman and I took "total losses to the financial system" to mean net losses; Norris has now explained that in fact he was talking about gross losses.

If the total amount of bets is huge (and it is) then the losses from defaults will also be huge. Sure, others will profit, but the total losses (of those who lose anything) will be greater than the underlying loans...
In a friendly poker game, there are no net losses. Yet some people may end up broke.

That's true, and it's one reason I'm skeptical of people who are itching to buy bank stocks because they "look cheap". So far, the number of banks who have demonstrated that their prop desks can be trusted not to lose vast amounts of money on mortgage-related derivatives is exactly one: Goldman Sachs. Given the size of the mortgage-related derivatives market, and the fact that the likes of Goldman and John Paulson were making billions of dollars going short, we know for a fact that someone else, somewhere else, was losing billions of dollars by going long.

Norris also clarified his comments on currencies, where he worried that "some emerging markets could run into big problems because many borrowers there have taken out loans denominated in foreign currency and could be devastated if local currencies lose value". It turns out he wasn't worried about the dollar strengthening, so much as the Swiss Franc:

In Hungary, one of the countries I was thinking of, I understand that most mortgage lending now is denominated in Swiss francs, whose interest rates are lower than the Hungarian forint. If the forint plunges against the franc, there will be a big problem for Hungarian consumers, whatever happens to the dollar.

Norris is quite right about SFr-denominated mortgages in Hungary: see this FT article, for instance. But as the article points out, the dangers aren't enormous:

OTP Bank, Hungary's leading retail lender, says it could manage a significant softening of the forint in two ways. If the weakening were temporary, it would grant overdraft facilities, essentially allowing customers to refinance a chunk of their loans. If the weakening were more long-term, it would allow customers to extend their repayment schedules.
Gavin Friend, strategist at Commerzbank, plays down the risk of a large unwinding of Swiss franc denominated mortgages. "A depreciation of the forint would be limited as the Hungarian central bank would react with rate hikes," he says. "Paradoxically, the then higher interest rate level in Hungary could even lead to more mortgages being taken out in Swiss francs."

In any event, Hungary and other small European nations are increasingly becoming "euroized" even if they haven't officially adopted the euro. There's not much value to being in charge of a small currency like the Hungarian forint, which means that the central banks in such countries generally work very hard to keep their currencies pegged, either loosely or tightly, to something more solid, like the euro. Given that Hungary is almost certain to join the euro at more or less its current exchange rate at some point in the future, the currency risk involved in taking out mortgages in euros is slim. Obviously, there's greater currency risk in taking out mortgages in Swiss francs, but in reality the Swiss franc has been weakening, not strengthening, against the euro, which is great for those Hungarian borrowers. So I still don't think there's much to worry about on the emerging-market-currency front.