Why aren't the banks lending?

We keep hearing that the economy is going south because the banks aren't lending money, and the banks aren't lending money because their balance sheets are clogged with toxic assets that nobody knows how to value.

It is certainly true that the banks aren't lending money, and this is certainly very bad for the economy, but the balance sheet argument is fundamentally flawed. There are two reasons banks aren't lending money:

(1) they think almost no borrower is creditworthy because of the outlook for the economy; and

(2) the market value of almost any loan they make to an entity other than the US government is at best 80% of par (i.e. worth at least 20% less than what they loaned).

The first reason should be obvious. Creditworthiness, and therefore bank profitability, is supercyclical. When the economy is good, companies are making money, unemployment is low, refinancing is easy, and everybody is a good credit. When the economy is bad, it's the reverse. And the fact that asset prices are falling fast just makes the risk of lending even higher because you never know when you might want the money back to buy something really cheap (there is opportunity cost on top of credit risk).

The second reason is more directly tied to asset prices, and it is something I haven't seen discussed much in newspapers, on TV, or elsewhere. The illiquid sectors of the fixed income market are incredibly distressed. The prices of assets are shockingly low relative to fundamental value. People keep blathering on about toxic assets, but there is no such thing as a toxic asset. "Toxic" means poisonous -- capable of causing injury or death -- and certainly implies that you would pay to get rid of it. But the so-called toxic assets have positive value. They are assets, not liabilities. They got the name "toxic" because their market value, and in many cases, their fundamental value (which is quite different as I'll explain), is now much lower than it used to be.

Now, if the fundamental value of an asset is much lower than where a bank is marking it on its books, then that is an issue of dishonest accounting and possibly fraud. I'll concede that a lot of banks have this problem, but I think the vast majority of them are domiciled in Europe and Asia.

If the market value of an asset is much lower than where a bank is marking it on its books (but the fundamental value is in the ballpark), then that is a potential liquidity problem. If the bank is forced to delever (e.g. pay back demand loans like poor George Bailey in It's a Wonderful Life), then it might have to sell that asset in a distressed market and end up with a lot less capital than it and its regulators thought it had.

Banks have a liquidity problem which has been mitigated substantially by government action over the past 6 months, but it is still acute. By making new loans at par that they can't sell for anything remotely close to par (think of buying a new car in the morning, and then trying to sell it back to the dealer in the afternoon), they are just exacerbating their liquidity problem.

I will give just one example of how distressed prices are in the fixed income market. Consider the following mortgage-backed security:

WAMU 2006-AR10 1A1 (CUSIP: 93363EAA3)

This is something that some lazy commentator on CNBC might call a toxic asset. This bond is backed currently by 790 individual jumbo mortgage loans. The average FICO score of the mortgagors is 730, the average age of the loans is 30 months, the average loan size is $700K, and the average original loan-to-value ratio was 68%. All of the loans are so-called 5/1 hybrid ARMs, which means that they have a fixed rate (average is 6.3% for these loans) for 5 years, and then they reset to a floating rate (which is on average Libor + 2.2%) for the remaining 25 years.

These are pretty good loans, solid loans, to wealthy people living in expensive neighborhoods. The current loan-to-value ratio is obviously much higher than 68% right now because home values have declined even in good neighborhoods, but it's certainly not over 100%. The performance of the loans has been deteriorating modestly over the last year, and the percentage of delinquent loans is now 7.3%.

However, the particular bond mentioned above (the 1A1 class) is the supersenior bond in the deal structure. It has over 10% credit protection from other bonds in the deal which are subordinate to it. So for this bond to take its first dollar of principal loss due to foreclosures, the mortgage pool would have to suffer over 10% realized credit loss. Given that the average loss severity on a foreclosure for this deal is less than 50%, you would need at least 20% of the loans to default to touch the supersenior bond. In recognition of this fact, S&P and Fitch still assign this bond a AAA rating, although Fitch has had the bond on negative credit watch since August 2008.

The 1A1 bond is paying a 5.9% coupon and has been averaging about 1.5 pts (i.e. 1.5%) of principal paydown each month. Even in an extremely pessimistic scenario, it won't take more than 10 pts of principal loss total. If it was Fannie Mae guaranteed, this bond would probably be priced higher than 103. How much would you pay for this bond?

Well, I paid 54.5 cents on the dollar for this bond two weeks ago. It was offered at 59, and I bid 52.5 to start. In retrospect, I probably could have bought it lower. If this bond returns less than a 12% per annum yield over the next 4 years, I'll buy a hat and then eat it.

Another point -- it's absolute garbage that nobody knows how to price this ... garbage. This stuff trades. The problem isn't so much lack of transparency; it's that the prices are ridiculously low. It is as if the risk-free interest rate is 12% in the non-agency mortgage backed securities world and 1% everywhere else. If the government really thinks that these assets are clogging balance sheets, then it should hire some smart unemployed Wall Street bond traders and start buying this stuff up. I figure the government will get a couple billion dollars worth before the effective "risk-free" interest rate in this sector collapses to less than 5%. The market will heal, new loans will make sense again, and the government will turn a profit to boot.