A popular TV show in the early nineties was “Short Attention Span Theater”. The title itself neatly parodied the growing culture of clips and sound bites to convey both news and entertainment and, sadly, this trend has continued. One result is that today many people obtain nearly all their financial information from television and internet sources, which deal with complex economic issues in a thirty second burst or a couple of pithy headlines.
Devotees of Jim Cramer’s “Mad Money” (assuming there are, in fact, any) receive a portfolio analysis and stock advice in under ten seconds. Even dedicated business programmes such as MNBC usually cover current news in a few minutes. To do this, they need to find a few headline statistics and then form a quick opinion based on small movements and comparisons to expert analysts “expectations”, or, possibly more accurately, “guesses”.
At first blush, it would seem as though US unemployment figures should be reasonably easy to understand and, although usually considered to be a lagging economic indicator, give readers important information about the current state of the economy. The data is certainly closely followed and the subject of numerous commentaries when published. In reality, it is much more complex than that.
The stock markets often react to fairly small changes in the monthly nonfarm payroll statistics, as calculated by the Bureau of Labor Statistics. The 300+ point slide of the Dow on 4 June, 2010 was widely blamed on the US Government’s May jobs report, although if we are to believe the hypothesis of an even remotely efficient market, it seems odd that information available on a Friday morning would not have impacted the markets until later in the day on the next Monday. Perhaps traders needed a weekend to digest the import of the news?
A light-hearted, simplified analysis of this report is as follows:
431,000 new jobs created (HOORAH!)
But 90 per cent of these were for part-time Census workers (expected, but BOO!)
So new jobs were less than analysts’ forecasts (BOO!)
But overall unemployment fell from 9.9 per cent to 9.7 per cent, because the overall workforce dropped by 322,000 (You’re kidding, right?)
The markets concluded the news was bad and they may well be right. But one very minor statistic provided some hope. Average weekly hours worked rose from 34.1 to 34.2. Hardly startling, you may think, but this increase is the equivalent of 315,000 additional jobs. In other words, if employers had hired new workers rather than working existing employees longer, the jobs increase would have blown away expectations!
As any statistician will tell you, analysis based on data is only as reliable as the data themselves. The BLS does a pretty good job on the whole, but their methodology has several inherent flaws, the greatest of which is their jobs “birth” assumption. The BLS collects their data by asking businesses questions about job creation and destruction.
Obviously, they cannot survey businesses which they don’t know exist, so they have to make fairly heroic assumptions about new business start-ups and the people they employ, based on historical data. The snag is that new start-ups do not follow historical patterns in a recession, so that in 2008 the BLS jobs figure included in aggregate over 900,000 jobs indicated by their “birth” assumptions, which turned out never to have really existed, which had to be adjusted in a large annual correction. Against this background, to try to read much into jobs variations in the tens of thousands is clearly perilous.
These problems of meaningful analysis are not confined to jobs data, but apply to almost all published economic data from mortgage applications to credit card defaults.
The real point is that it is impossible to convey what may be happening to US jobs, housing or GDP in a single headline, but week after week news sources seek to do so. Small changes are taken as indicating future trends without consideration of sampling errors or seasonal adjustments (which in themselves deserve a full article).
It is against this background that the Cayman Financial Review seeks to provide in depth coverage of current financial issues, which goes beyond the simplified summations generally found in the daily papers or internet blog sites. Excellent examples in the current issue include “Financial Transaction Taxes: The issues and the evidence“ by Christopher Culp and “Comparing qifs and sifs to Cayman funds: Managing the myth“ by Nick Rogers, which provide the reader with a thoughtful analysis of the effect of proposed transaction taxes and a comparison of alternative investment fund structures in Ireland, Luxembourg and Cayman.
Brevity has its place, but sometimes the longer attention span needed for serious study has its rewards.