Author Archives: Menzie Chinn

“The Election: Implications for Policy Change?”

That’s the title of an informal panel at the UW La Follette School of Public Affairs on Tuesday. Here are the slides that underpin my presentation.

For now, let the following figure summarize the choices.
clinton_trump_macro

Source: “Trump vs Clinton: Polarization & uncertainty,” Research Briefing (Oxford Economics, 19 Sept. 2016) [not online].

The other panelists are Pam Herd, Greg Nemet, Rourke O’Brien, and Tim Smeeding.

Recession Watch, October 2016

The 2016Q3 GDP advance figures were released today, indicating a 2.9% growth rate (SAAR), exceeding consensus (Bloomberg 2.5%). Tomorrow, Jim will report on the recession probabilities based upon the advance release (see last quarter’s analysis here). Until then, given all the discussion of recession (e.g. [0], [1]), it seems useful to show a few pictures of where we stand today, and the outlook going forward, given some standard and non-standard indicators.

Continue reading

The Kansas Economy: Three Pictures

The Philadelphia Fed released coincident indices today. Figure 1 shows state-by-state 3 month trends. Needless to say, the outlook for Kansas — that laboratory for supply side nostrums — is not auspicious.

coincident2016-09

Source: Philadelphia Fed, accessed 26 Oct 2016.

While Alaska seems to be in the running for worst performing, in fact the 3 month (annualized) decline of 4.5% for Kansas is the worst in the 50 states.

Continue reading

The Mexican Peso and Prediction Markets

I was talking about prediction markets and asset prices (the Mexican peso and the Presidential election, and the pound and Brexit) in my classes this week. It struck me a good time to update this post on the peso’s movements as the odds for a Democratic win change.

election_mxn4

Figure 1: USD/MXN exchange rate (blue), and odds of Democratic win in Presidential election, end of day (red). Observation for 10/20/2016 is as of 2:30PM Eastern time. Exchange rate defined so up is MXN appreciation. Source: FRED, Pacific Exchange Services, and Iowa Election Markets.

The adjusted R2 of a bivariate regression of first differences regression (exchange rate in logs) is 0.08, pretty good on a high frequency time series, in my book (t-stat with HAC robust errors is 2.06).