During my usual morning reading process I came across a posting on PragCap
by Cullen Roche regarding a chart of U.S. recession probabilities. The
chart can be found at the St. Louis Federal Reserve website and is
derived from a study by J. Piger and M. Chauvet, from the University of
Oregon, which was published in the Journal of Business and Economic
Statistics in 2008. (For other economic geeks the full paper is attached)
Cullen points out that "What’s
interesting about this index is the current reading. At 20%, the index
is at a level that has ALWAYS been followed by a recession. As you can
see below, the index has never approached 20% without a subsequent
recession. All 6 recessions since 1967 have coincided with 20%+ readings
in the US Recession Probabilities index." Currently, that index,
as shown in the chart below, is approaching that 20% level as of August
which is the latest reported data.
The recession probabilities, as stated in the research article, are obtained from a "dynamic-factor
markov-switching model applied to four monthly coincident variables:
non-farm payroll employment, the index of industrial production, real
personal income excluding transfer payments, and real manufacturing and
trade sales." Without getting to far into financial econometrics a
Markov switching model involves multiple equations that can
characterize the time series behaviors between different data sets. By
permitting the switching between equations the model is able to capture
more complex dynamic patterns. The question that we want to answer is
whether the indicator is currently correct in its prediction of a U.S.
recession or "is this time different."
No comments:
Post a Comment