Of course, you have to be able to take some losses. If you had a $500,000 loan go bad, you would have eaten up all your profits and dipped into your capital. Now you would have only $900,000 in capital. That means you could only make $9 million in loans. You'll either have to raise more capital from investors or reduce your loan portfolio, in addition to writing off the bad loan.

Now, what constitutes capital at real-world banks is a very complex thing. How much it costs to find money to lend can vary wildly. Many of us “lend” money to our banks at zero cost to the bank via our checking accounts. They pay more for savings accounts, borrowing from other banks, etc., and charge different rates for different types of loans. It's a very complicated business, but the basics of my simple illustration will get us where we need to go.

When I want to find out something about US banks, I turn to PL Capital. They run a fund which invests in smaller banks, and have been consulting with banks for many, many years. They really know the business. I asked Rich Lashley to give me his thoughts, which I've summarized for you below.

In the US, and in much of the developed world, there are about 8500 banks on 2 tiers of banking. The top 50 banks have more than 80% of the assets; the rest are generally much smaller. Excluding 5 states hit particularly hard by the housing crisis (see more below), it's business as usual for most other banks.

There's money for most smaller businesses at smaller banks except for residential development. The true credit crunch is at the top, for larger projects. Want to do a $3 million deal? If you have a reasonable project, it can get financed. Want to do a $300 million deal? Lots of luck, for reasons I note below.

Most of the smaller banks have plenty of capital and are looking to put it to work. Rich guesses there are about 400 or 500 banks (5% of the total, concentrated in states with bad housing markets) which are in some level of financial distress, meaning they need to raise capital or reduce lending. His guess is that we will see about 100-150 banks fail over this cycle. He'd be surprised to see more than 3 top-50 banks fail. Most of the larger banks, if they get into trouble, would be absorbed by better-capitalized banks in search of market share.

In fact, Rich is quite bullish on selected bank shares. 86% of the banks/thrifts in the US made money in the first quarter of 2008, and almost 50% earned more than they made in 2007's first quarter (i.e., they made more money after the credit crunch than before.) But nearly all bank stocks have declined due to the credit crunch. Over time, growing earnings will allow for a rebound.