Picture the scene: a driver is stopped for an alleged speeding offence. The police officer checks the driver’s documents and reviews the speeding camera’s telemetry. The driver was indeed over the speed limit, and the officer gets ready to write a ticket. Yet a smirk spreads across the driver’s face – he seems unconcerned about what’s surely about to happen.
Just before the officer puts pen to paper, he double-checks the date. His eyebrows shoot up as he consults his calendar. Then, he lets the driver carry on – no ticket, no warning, no nothing.
As he motors off, nearby pedestrians stare at each other in disbelief. One of them shouts to the officer: “He was way over the limit!”. But the officer simply shrugs his shoulders and leaves.
Now replace the driver with a too-big-to-fail US bank, the police officer with the Federal Reserve, and the pedestrians with, well, observers concerned about financial stability, and you get a pretty good sense of how systemic risk regulations work in the States.
US global systemically important banks (G-Sibs) have capital add-ons assigned using a two-tier system. The first tier uses the the Basel Committee on Banking Supervision’s assessment methodology, known under the US rule as Method 1, to designate a bank as too-big-to-fail. This is done by computing a systemic risk score for each large bank, found by averaging the scores of five systemic indicator categories: size; interconnectedness; complexity; cross-jurisdictional activity and substitutability. Those that exceed a threshold score are slapped with the G-Sib label.
Then, these G-Sibs are assessed using the Federal Reserve’s own calculus, known as Method 2. This uses the first four indicator categories above plus a short-term wholesale funding factor, and multiplies the values recorded in each by fixed coefficients to come up with a different systemic risk score.
Method 2 indicator values are published quarterly. However, only the year-end values, using indicator amounts from the final quarter of each year, are used to produce Method 2 scores, which in turn set G-Sib capital surcharges for the following year but one if they exceed the Method 1 scores – which so far they always have.
This year-end bias makes for some strange incentives. Multiple Risk Quantum analyses have shown that US banks get larger, more complex and further intertwined with other financial institutions during the first three quarters of each year, only to shrink and simplify as the calculation date approaches.
And why not? After all, why should banks curtail their growth from January to September if it is only from October to December when their activities become relevant under Method 2?
But, to turn back to the speeding driver analogy, if banks are not discouraged from becoming more systemically risky over most of the year, might they feel free to engage in the sort of reckless behaviour that necessitated a G-Sib framework in the first place?
Assessing systemic risk should not be a simple matter of ticking the right boxes once a year. The consequences that a number of failing firms would bring about for the financial markets are real, regardless of the date on the calendar.