While Europe is dominating headlines this week, UBS' Art Cashin suggests "mark your calendar and cross your fingers" as he notes the disproportionate prevalence of events that occur in September. Focusing on The Economist's Greg Ip's recent post on a possible seasonal pattern in banking crises, via this recent Reinhart & Rogoff extension paper by Laeven and Valencia [10], he notes: "The frequency with which the world goes to hell in September seems hardly random." Unfortunately the authors provide no explanation for this beyond observing, "An interesting pattern emerges: banking crises tend to start in the second half of the year, with large September and December effects." Ip and Cashin offer some thoughts on why this is so historically, and more importantly why this time is no different, as the avuncular Art concludes with: "try to enjoy your summer".
[11]
Art Cashin, UBS: Mark Your Calendar And Cross Your Fingers
Savvy Fed observer, Greg Ip, now writing and blogging at the Economist has come across a possible seasonal pattern in banking crises. Here’s a bit of what he wrote:
Most readers of this blog are familiar with the path-breaking work of Carmen Reinhart and Kenneth Rogoff on financial crises. Fewer have heard of Luc Laeven and Fabian Valencia, but they too have contributed richly to the raw material of crisis economics. Labouring away at the International Monetary Fund, they have assembled a detailed database of 147 banking crises from 1970 to 2011. They recently published a new working paper [10] detailing some of their latest findings. There's lots of fascinating stuff in here but what leaped out at me was a chart showing the likeliest months for crises to begin.
Ip then shows a chart indicating a disproportionate number of events in September. He continues:
The frequency with which the world goes to hell in September seems hardly random. Unfortunately the authors provide no explanation for this beyond observing, "An interesting pattern emerges: banking crises tend to start in the second half of the year, with large September and December effects."
I recall reading that historically the fall was a popular time for crises because farmers needed to borrow to bring in the harvest, which strained fractional-reserve banking systems. I can't think why that pattern would persist into the post-industrial era. One theory blames the October 30th fiscal year-end of American mutual funds; managers trying to avoid losses or hold onto gains for the year were more likely to succumb to herd behavior as that date approached. But that wouldn't explain why the pattern holds in other countries which, I assume, have different year-ends. Maybe it's because policy makers and bankers don't confront their problems until they get back from vacation, the macro equivalent of doctors scheduling c-sections during office hours.
No matter the explanation, this gives more reason, as if you needed more, to sweat as the fall approaches. Just to elevate the anxiety further: another calendar pattern to which Mr Rogoff first alerted me is the tendency for crises to happen in election years; think Mexico, 1982 and 1994, Korea, 1998, America, 2008, Greece, 2009. The intuition behind this was that crises are the result of imbalances that accumulate over a long time. Politicians have a strong incentive to delay dealing with them until after an election, and often, as was the case with Greece, to actually hide the truth until the polls close. This means imbalances often reach their breaking point right around the election. This has been a busy year for elections: we've already had Russia, France (we're still waiting for the true state of France's fiscal affairs), and Greece, twice. Mexico's is just a few weeks away, and of course, America's fiscal problems come to a (contrived) head mere weeks after November's election. That's also roughly around when China has its version of elections, i.e. the long-orchestrated change in the Communist Party's leadership.
Try to enjoy your summer.