Review & Outlook

Our take on the investing, financial, & economic themes of the day

Don’t Fight the Fed, But Don’t Emulate it Either: Why Your Portfolio Shouldn’t be ‘Data Dependent’

Today the Federal Reserve is widely expected to announce that short-term interest rates will remain unchanged.  The Fed says that its decisions about whether or not to raise interest rates will be “data dependent,” but this doesn’t inspire much confidence, because the short-term economic data it depends upon aren’t dependable.  This fact is summed up nicely by James Mackintosh, long-time Financial Times writer who recently moved to The Wall Street Journal:

Not only don’t we know where the economy is going. We don’t even know where it is, and after revisions we frequently find we were wrong about where the economy was, too.

Consider the U.S. employment numbers from last month, which are likely to be cited by the Fed in its rationale to leave rates unchanged.   As CNBC reported those numbers, “Job creation tumbled in May, with the economy adding just 38,000 positions, casting doubt on hopes for a stronger economic recovery as well as a Fed rate hike this summer,” because economists had been expecting 162,000 new jobs.  The market responded to such dire news with fear, and this morning I saw the numbers cited again (along with anxieties about Britain’s possible exit from the EU) as evidence for the global economy’s parlous state.

may 2016 jobs


Mackintosh points out that the purveyor of these numbers, the Bureau of Labor Statistics, says that it is 90% sure that the 38,000 new jobs figure is accurate to within 115,000 jobs.  So while the headline number was +38k, there’s a 9 in 10 chance that the real number of new jobs is between -77,000 to +153,000.  That’s a huge margin for error, to say the least, since the Bureau of Labor Statistic’s 90% certainty is itself a subjective probability.  Even on its own terms, as Mackintosh observes,

One month in ten we should expect the true number to be more than 115,000 away from the figure in the survey. If May was the one in ten, it should be ignored. Worse still, payrolls are subject to big revisions, averaging 43,000 in either direction since 2003 and occasionally, as in November of 2012 and 2014, adding more than 100,000 jobs to the initial report. It is possible the bad May figures could be revised away.

A lesson that most investors should learn from the murkiness of such numbers shouldn’t be obscured by the Federal Reserve’s pretense: it’s foolish to try to manage a “data dependent” portfolio where there isn’t reliable data to be had.  “The social object of skilled investment,” wrote John Maynard Keynes, “should be to defeat the dark forces of time and ignorance which envelop our future.”  Sometimes those dark forces envelop our present to an even greater degree than the future.