The Kansas Palmer Drought Severity Index, Unit Roots, and Snakes in a Room

In my post on modeling the Kansas economy, Rick Stryker takes me to task for modeling the Palmer Drought Severity Index (PDSI) for Kansas as an I(1) process:

I wouldn’t bother with the ADF test, since its null is non-stationarity and it has low power to reject. I would focus on the KPSS, since its null of stationarity is almost certainly what’s really true.

Continue reading

Post-Brexit Out-of-Sample Forecasted Electricity Consumption

Take England+Wales log daily electricity consumption, detrend using Christiano-Fitzgerald band pass filter, and then regress on seasonal (calendar) terms, and temperature/wind/rain factors (Kirchmaier and de Guana de Santiago, 2016, h/t Simon Kennedy at Bloomberg), through April 2016. Then forecast out of sample; the residual looks like this:


Figure 3 from Kirchmaier and de Guana de Santiago, (2016).

Electricity consumption is way under what would be expected from historical correlations, suggesting a decline in economic — particularly industrial — activity.

While monthly estimates of November GDP are up 1.1% relative to June, industrial output is down by 1.2%, according to NIESR (Dec. 7).

It is always useful to keep in mind that economic statistics are sometimes revised by large amounts.

Back to normal?

A year ago, the Federal Reserve decided to raise its target for the fed funds rate by 25 basis points above the floor of 0-0.25% at which we’d been stuck for 7 years. FOMC members indicated at the time that they were expecting to end 2016 at 1.4%, or four rate hikes during the last year. We started this December at 0.41%, and the first hike of 2016 didn’t come until last week. Now FOMC members say they are expecting to end 2017 at 1.4%, or three more hikes from here during the next year. The January 2018 fed funds futures contract is currently priced at 1.23%, suggesting that the market is buying into two, not three hikes during 2017.
Continue reading