Monday, November 5, 2007

Good or bad news about the economy?

Just a few days back, the latest employment data for the US was released, and the data was surprisingly good, as this Bloomberg story reports. But how exactly should we interpret a single piece of evidence? The data is volatile, and as a result, there are lots of conflicting pieces of evidence, as Nouriel Roubini discusses here. The rational thing to do is to take moving averages through the data, to drop out some of the volatility (and if we have 12 month moving averages, drop out any distortions caused by seasonal factors or- more likely- poor adjustment for seasonality), and then interpret the smoothed data, as James Hamilton proposes here. The conclusion? The news is not good.

No comments: