17 March 2019

What You Quant Is What We Get

Once again, with vigor, we find a pundit making the case that >events drive data, not the other way round. And, once again, the title I see in my dead trees version while I sip coffee differs from the innterTubes. I prefer the dead trees.
Dead Trees: "Recession-Spotting, Beyond the Math"
innterTubes: "You Never Know When a Recession Will Sneak Up on You"

Why the preference? One reminds one of a famous movie (which I've not seen, only read about) about personal catastrophe.
The 2001 recession developed when the internet bubble popped, or at least that's how we tend to remember it. But go back and check the numbers. The internet accounted for, at most, about 2 percent of the economy then. If we use the logic we've been applying to trade wars and government shutdowns, it would seem that popping the internet bubble shouldn't have been enough to cause a recession. But it did.

The problem, of course, is that inflections in the data haven't been predicted (so far) from the up-to-date time series that econ/business/pol types use to inform themselves.
Looking back, for example, we know that a recession officially began in April 2001, yet scarcely anyone understood that then. In June 2001, only 7 percent of economists in the monthly Blue Chip survey believed a recession was underway. In the months before that 2001 recession began, only 16 percent of economists expected that a recession would start within the next year. Now, 25 percent of economists in a Wall Street Journal survey say they expect a recession within the next year, and anxiety seems to be growing.
[my empahsis]

Sounds like a case of once bitten twice shy. On the other hand, there's among the most famous adages of the econ community:
The stock market has forecast nine of the last five recessions.
-- Paul Samuelson/1966

No comments: