Stressing over Bank Stress Tests

If it weren't tacky to say I told you so, I'd refer you to my previous posts and interviews pointing out the dangers and pitfalls of the bank stress tests. They weren't just a mistake; they were a mistake anyone without sleep deprivation should have seen coming a mile away. It's not the biggest mistake or the most important mistake that policymakers have made during this crisis, but it was the easiest to foresee and to avoid, except for the sleep factor. These boys need to get their rest. The economy and the banks will be there in the morning. In fact, a little less regulatory attention might just be what the doctor ordered.

The headline on Saturday's (May 2) Wall Street Journal says, "Citi Said to Need Up to $10 Billion."

"Citigroup Inc. may need to raise as much as $10 billion in new capital, according to people familiar with the matter, as the government continues negotiations with banks over the results of its so-called stress tests." [Underline added]

"The bank, like many others, is negotiating with the Federal Reserve and may need less if regulators accept the bank's arguments about its financial health, these people said. In a best-case scenario, Citigroup could wind up having a roughly $500 million cushion above what the government is requiring.[Underline added]

To the uninitiated-or should I say the unjaded?-it might come as a surprise to read that banks are negotiating with the government over the amount of capital needed and that the range of possible outcomes is as wide as $10,500,000,000. Let me restate that: according to the WSJ, Citi thinks it may have $500,000,000 more capital than required while the Federal Reserve thinks it might need an additional $10, 000,000,000. (Note the extra three zeros in the second number. I know, we're all suffering from zero overload.)

"For Heaven's sakes," you may be forgiven for asking, "why don't they just go in there and count the capital and compare it to what the formula says it should be?" Finding and counting the capital reminds me of my Fed days when I wondered what I'd say if someone came in and asked to see the discount window. Likewise, there is no place to go to find and count the capital. Investors "put capital into" banks by buying stock. But capital isn't tagged and followed electronically. It gets used in making loans and investments; it's a moving target. Demands by some in Congress to know what the TARP banks did with the taxpayers' money ignore the fact that capital, as money, is fungible. 

Capital is what is left when you subtract the value of total liabilities from the value of total assets. The excess of assets over liabilities is capital, or the amount of the assets owned by the shareholders rather than creditors. The problem is that the valuation of many of those assets and liabilities changes constantly. Most of the "negotiations" taking place will not likely be over the percentage of assets or liabilities needing to be "supported" by capital, although that is a legitimate issue for debate, but over the methods used to estimate the value the assets and liabilities, as well as how much capital should be set aside for potential losses in different types of loans and securities in its portfolio.

There are honest differences of opinion about capital requirement. One should not assume the government or the Federal Reserve has good, defensible, and definitive, answers to these questions that are being resisted by banks. In many cases, the regulators' rules and requirements themselves seem unreasonable and arbitrary and, perversely, are often enforced more vigorously for troubled banks in hard times than for super-sound banks in good times. It's like giving a lame horse a heavier load to tote. The result is often pro-cyclical regulatory policies that make a banking crisis worse and can easily lead to the failure of banks that could have survived otherwise. Let's call it death by accounting. A little supervisory forbearance on some of the debatable issues would save a lot of unnecessary grief and do no harm.

I've already discussed at length the pro-cyclical nature of mark-to-market accounting rules and the unnecessary destruction of bank regulatory capital resulting from them. The recent relaxation of those rules by FASB was inadequate, and will therefore be only marginally helpful. Other problems remain in the rules on OTTI (Other Than Temporary Impairment) of bank assets, although FASB was somewhat more helpful on this issue.

I haven't mentioned- and this may be a big part of the negotiations between the government and the Gang of 19 banks- destructive anomalies in what's called "risk based capital" ratios. It's logical to acknowledge that some loans and investments are inherently more risky than others and should require more capital to be set aside for potential losses. So far, so good, for the abstract logic. In practice, however, the supervisors are enforcing risk-based standards that they, themselves, must know make no sense. Some assets, such as mortgage-backed securities that get adverse treatment under mark-to-market rules, also get slammed on the risk-based capital requirements. For example, a MBS that originally was rated AAA, but has since been downgraded by a rating agency, might have a risk weighting of up to 1250 percent (think batting averages where 1000 is perfect) and would therefore require capital of 125 percent of its total value-which is 25 percent above 100 percent. (I'm being redundant and pedantic here to let you know that I know it makes no sense.) This may occur even if the MBS is current, on its principal and interest payments and with significant subordination in the MBS tranches to take the first losses.

It is almost certain that you think I just misspoke; so let me say it again. The rules sometimes require setting aside capital to cover more than the entire value of a total loss. This doesn't pass the snicker test, but a banker hoisted on that particular petard probably isn't laughing. Wouldn't you think this topic deserves more attention than it's getting?

Other examples, of perverse, pro-cyclical, and unnecessarily destructive regulatory policies include:

Limiting capital reserves banks can set aside for losses during good times. Several years ago the SEC sued SunTrust Bank for accumulating too many reserves for loan losses, which effectively limited the accumulation of capital in the banking system. The name given to this new sin was "smoothing." What it boils down to is that you aren't allowed to make hay while the sun shines.

Increasing requirements for loan loss reserves during the bad times. You can't make hay while the sun shines, but you must make additional hay when it rains. Once a bank experiences loan delinquencies or securities write-downs that threaten its required capital ratios, the supervisors increase the ratios. That is, they raise the bar for survival. I'd call that an unprecedented act of [expletive deleted] except it's pretty much what your credit card company does when you miss a payment or your car insurance company does when you have an accident.

Raising bank deposit insurance premiums for banks whose CAMELS rating deteriorates. CAMELS summarizes the major categories subject to bank examination. CAMEL stands for Capital, Assets, Management, Earnings, Liquidity and interest rate Sensitivity. These categories aren't as independent as they might seem. If a real estate bust impairs some loan assets, capital is reduced by extra reserves for loan losses; and not maintaining required capital ratios is, of course, evidence of poor management. Earnings and probably liquidity are also harmed during the process, making for a perfect strike.

The FDIC's insurance fund comes from premiums paid by banks, as a percentage of their deposits. Taxes don't go into the fund, and bank failures have never cost taxpayers a dime, contrary to what we hear every day on the tube. [I'm talking about banks here; the S&L crisis in the late 1980s did exhaust the FSLIC, the Federal Savings and Loan Insurance Corporation, fund and required some taxpayer supplements. The banking crisis did not require taxpayer help, however.]

Following the bank and thrift crises of the late 1980s and early 1990s, the regulators, mainly the FDIC, adopted a philosophy of "early intervention" and "prompt corrective action." Banks used to be taken over when they failed, defined as no capital left, or liabilities exceeding assets. The new practice of taking over or marrying off a still-solvent bank is designed to protect the insurance fund by intervening BEFORE the bank fails. And people wonder why capital constrained banks are reluctant lenders?

Comments (2)

Trackback URL | Comments RSS Feed

  1. Topics about Citigroup | Stressing over Bank Stress Tests says:

    [...] Another fellow blogger placed an interesting blog post on Stressing over Bank Stress TestsHere’s a brief overviewIf it weren’t tacky to say I told you so, I’d refer you to my previous posts and interviews pointing out the dangers and pitfalls of the bank stress tests. They weren’t just a mistake; they were a mistake anyone without sleep deprivation should have seen coming a mile away. It’s not the biggest… [...]

  2. [...] flogging, also.  Here’s a little silliness to chew on which was part of a recent post by Bob Mcteer: Limiting capital reserves banks can set aside for losses during good times. Several years ago the [...]