Why we need to spend more money fixing the banks.

Last week, the Treasury Department quietly announced it was moving ahead with plans to purchase toxic assets from banks, but in scaled-back form. To my colleague John Judis, this must have been welcome news. For months now, he and I have debated whether President Obama’s efforts would be best spent fixing the financial sector or reviving spending by consumers and businesses. Our debate hinges on a seemingly simple question that turns out to be annoyingly tricky to answer: Is bank lending down because banks are undercapitalized (meaning they hoard cash because they’re not strong enough to absorb losses on existing loans, much less new ones), or because the prospects for economic growth are weak (meaning consumers and businesses aren’t interested in borrowing even if banks are willing to lend)?

John takes the latter view--he believes we’ve done as much for the banks as we can, and that our top priority should be stimulus--while I’m more sympathetic to the former. Now he’s raised the stakes with a persuasive piece on the subject. Suffice it to say, this aggression cannot stand.

John’s first mistake is to understate the seriousness of last fall’s financial crisis. In his telling, the government should have moved on to other economic problems after orchestrating the initial bank bailout. But it’s hard to argue that the initial bailout did the trick. In mid-October, then-Treasury Secretary Hank Paulson forced nine of the nation’s 20 largest banks to accept $125 billion in government money. Despite this increased capital buffer, the five biggest banks cut their lending by 16 percent (roughly $120 billion dollars) during the fourth quarter of last year, according to the Fed. The next 20 biggest banks cut their lending by over 4 percent ($9 billion).

John would probably say this proves his point: It must have been the weak economy, rather than insufficient capital, that led banks to contract credit even after receiving all that bailout money. But the evidence suggests the opposite: During the same period, the country’s small and mid-sized banks (basically every bank outside the top 25) increased lending by about 5 percent ($27 billion). Presumably both sets of banks were facing the same economic conditions; only the biggest ones were hemorrhaging capital.

One reason this was happening is that, in addition to making loans, big banks also run massive investment portfolios, which the financial crisis wreaked havoc on. For example, at the end of last year, Citigroup’s balance sheet showed about $250 billion of investments in various securities (including between $50 billion and $100 billion in mortgage-backed securities) and a “trading account”--basically short-term bets in financial markets--worth about $375 billion ($115 billion of which was invested in derivatives--essentially bets on the price movements of assets like stocks and bonds). Citigroup suffered billions of dollars of losses in each category, perhaps tens of billions--the company used creative accounting to partially obscure the red ink. Smaller banks generally don’t have such investment portfolios.

Similarly, the loan books of the biggest banks were tilted much more toward residential real estate--the epicenter of the financial crisis--than those of small and medium-sized banks. Under the government’s stress-test scenario, about half the losses in the loan books of the five biggest banks comes from residential real estate. The comparable figure for small and midsized banks is about 25 percent, according to a recent Wall Street Journal analysis.

And so last fall, while the Citigroups and the Bank of Americas were teetering on the edge of insolvency, most small banks had yet to feel more than a pinch. In fact, this is borne out in the same Federal Reserve survey (see here and here) John mentions in his piece. While almost every bank that cut its lending cited the economy as a reason--I suspect it’s kind of a catch-all category that serves as a proxy for business confidence--the banks classified as “large” were at least 50 percent more likely to cite concerns about capital than banks classified as “small.” (The gap would almost certainly be wider if we looked at the top 10 or 20 banks; the Fed’s definition of “large” is pretty broad.)

Of course, since early May, when the government announced the results of the stress tests, the country’s biggest banks have been raising capital from private investors at an impressive clip. By this point, it really does look like a second stimulus--say, a payroll tax cut, or additional aid to states--might be a better use of the administration’s mental energy than worrying about the banking system. But, setting aside the nagging risk of a relapse, or the possibility that the banks need more capital than the stress tests indicated, such a conclusion ignores the health of smaller banks, which have been drifting toward insolvency. (The stress tests only covered the 19 largest banks.)

The big banks, with their massive investment portfolios and their large exposure to home mortgages, were most vulnerable to the initial financial crisis and to the residential real-estate bust that precipitated it. It made sense that they felt the strain first. For their part, small banks tend to be heavily invested in commercial real estate, which has faltered only in recent months as the financial crisis has spread through the economy and evolved into a deep recession. In a recent study of small and mid-sized banks, the Journal projected about $200 billion in total losses through the end of next year, with commercial real estate accounting for roughly half of that amount. If that estimate is right, the paper concluded, “more than 600 small and midsize banks could see their capital shrink to levels that usually are considered worrisome by federal regulators.”

Now, even if you combine all the small and mid-sized banks in the country, they only account for about a third of the assets in the banking system. (The 19 stress-tested banks account for the other two-thirds.) The reason they’re worth worrying about is that, relative to other banks, they play an outsize role in supplying credit to the economy. The mega-corporations that do business with the biggest banks have multiple ways to get credit: They can issue corporate bonds, for example, or participate in the commercial paper market (basically a way to borrow money from investors on an extremely short-term basis). These companies tap lines of credit with the big banks that service them only as a last resort. But most of these other options are closed to smaller companies. If a local manufacturer or retailer wants to borrow money, it generally has to hit up a bank, often a local or regional one. (This is one reason the tentative success of the so-called TALF program--which the Fed and the Treasury launched in March to unfreeze the market for securitized loans--is so important. It allows banks to lend money to small businesses and then sell the loans to investors rather than carry them on their balance sheets.)

In fairness, my differences with John are mostly at the margins. Despite my belief that the severity of a recession like this one is closely tied to the health of the banks, I’m the first to concede that bank stabilization isn’t a sufficient condition for recovery. Once a bubble bursts, consumers and businesses will naturally cut back on spending and focus on paying down debt. The only way the economy can rebound is if the government picks up the slack. So John and I agree that, beyond a certain point, an additional dollar of stimulus spending trumps an additional dollar spent fixing the banks. I’m just not nearly as confident as he is that we’ve reached that point.