| | It’s no secret that a number of key banking regulations and accounting policies are pro-cyclical. Which is to say, they have the effect of driving earnings higher than they’d otherwise be during the good times, and lower during the bad ones. As the experience of the current downturn shows, this is not always a good thing. Pro-cyclicality was a hot topic at the Federal Reserve Bank of Chicago’s annual meeting last month. Most attendees seem to want to some kind of overhaul of the rules that would dampen their pro-cyclicality or even (as was intended once, long ago) make them counter-cyclical.
I say, good luck. Regulatory and accounting changes that get debated at economic conferences aren’t always so easy to implement in the real world. In Chicago last week, for instance, three highly pro-cyclical rules and practices came in for a lot of discussion: the setting of capital standards, loan loss reserving, and pricing deposit insurance premiums. In all three cases, sensible-sounding fixes aren’t hard to come up with. But as a practical matter, only one of the three practices could easily be modified. Capital Standards. Many attendees supported the idea of changing how regulators set minimum capital standards, to let regulators raise minimum capital ratios during periods of growth and high profitability in order to create a buffer that can be run down during periods of economic weakness.
A great theory! So good, in fact, that Ben Bernanke endorsed the concept during Q&A. (Which is ironic. Bernanke made his remarks just hours before the Fed announced the results of its stress test, an exercise that led to banks adding capital at the bottom of the cycle—the exact opposite of the idea Bernanke endorsed.)
But in the real world, flexible capital standards wouldn’t fly. Let’s suppose, for instance, regulators adopted a policy that raised the Tier 1 “well-capitalized” standard to 8% from 6% during an expansion, with the intention of cutting it back down to 6% during a slowdown. Do you really think that once the slowdown arrived, and bank earnings collapsed, that regulators would follow through with their plan to weaken capital standards? No way! Regulators don’t make career risks like that.
The capital standards that were put in place in 1988, and subsequently adjusted, are sufficient. Capital standards could never be raised high enough to prevent bank failures during downturns, without making lending uneconomic.
But even though the standards are sufficient, regulators this cycle enforced them in a highly pro-cyclical manner via the stress test and its associated capital “buffer” requirement for the 19 largest banks. The test and action plans were indeed “asinine:” they increased banks’ capital requirements at almost precisely the moment that adding capital was most difficult and expensive. Luckily, signs of economic improvement began to appear before the release of the tests’ results.
Loan Loss Reserve Accounting. Under GAAP, loan loss reserves are determined each quarter in a systematic, highly quantitative way, to prevent bank managements from manipulating earnings by arbitrarily setting reserves.
In practice, though, this means that reserve levels fall during good times and rise during times of increasing loan stress—which is the reverse of how loss-reserving was originally supposed to work. Many analysts want to change the reserve accounting rule so that reserves rise during the good times and then are used up as loans go bad. Again, it makes sense in theory. But I just can’t see how this would happen without unintended consequences. For one thing, many managements would have trouble resisting the temptation to use reserves to micromanage earnings.
Rather, I’d do away with loan loss reserve altogether, and would instead simply have banks book credit costs when they charge off bad loans, once the loss is probable and can be estimated. Loan loss reserves are not pools of cash set aside for a rainy day. They’re simply accounting entries. Think of them as a form of capital, only booked on area of the balance sheet other than equity. So do away with loss reserves and raise the capital standards by a comparable amount.
That would be my suggestion--but it wouldn’t be GAAP. Thus we’re left with loan loss reserving system that has a pro-cyclical effect on bank earnings. At this point in the cycle, by the way, the effect is near its peak. It should reverse as bank earnings recover (and reserve additions decline and ultimately reverse) in 2010 and beyond.
Deposit Insurance Premiums. The FDIC insures $4.8 trillion in deposits, held by 8,200 institutions that have $13.5 trillion in assets. The FDIC maintains a “reserve” for losses it will incur by charging banks a premium that is based on the institutions’ risk.
When economic times are good and FDIC losses are low, the premium the FDIC charges banks is also low, or even non-existent. And when times are bad and losses rise, the agency raises premiums to keep its reserve fund replenished. Sure enough, it just announced an emergency assessment of banks to raise $5.6 billion.
But the FDIC doesn’t really have its own segregated pool of funds to back its reserve. That’s an accounting fiction. Rather, the FDIC simply passes along its premiums to the general Treasury coffers.
Many people (and I’m one of them) think it’s crazy for the FDIC to cut premiums when times are good only to raise them at a time when banks can least afford added costs. Better to put in place a consistent, risk-based premium. If the insurance fund then falls into a deficit during a recession, let the FDIC draw on its line of credit with the Treasury, then repay it once the economy recovers.
Of the three key pro-cyclical banking rules and regulations, this one’s the easiest to fix. The other two could stand some revising, as well. But, as I say, that won’t be as easy to do as a lot of people seem to think. |
