Recession Detection using Effective Demand limit

I worked with Dwaine Van Vuuren of Recession Alert a few months ago on using the effective demand limit to detect recessions. He said the effective demand limit gave the best detection of a recession he had ever seen. He said he had seen hundreds. I want to show you one graph he came up with.

This graph is a measure of the gap between real GDP and the effective demand limit. I already showed how this limit will signal a recession in a prior post. This graph from Dwain goes deeper.

The graph shows the “Percent change over one year” of the gap between real GDP and the effective demand limit. The gap was made negative (flipped upside down basically) so that the plot would show a recession down below. (recessions shaded in) (graph up to 1stQ-2013)


Link to Graph

The pink line marks the threshold at 1.6%. In every single case, when the plot broke below the pink line, a recession was currently happening or imminent. The bottom red line was added because of the false positive in 1997. My view is that there was distorted data for 1997.

Apparently from Dwaine, this is a very reliable method to detect a recession. He has since used the effective demand limit to make even better methods to detect a recession, but I will leave it to him to share and publish that work.

The basic explanation of this graph is that the year-over-year gap between real GDP and the effective demand limit narrows above 0% in the graph and gets wider below 0%. An economy recovers from a recession over 0%. (note: When the plot goes back down to 0%, it is concluded that the recovery (within the effective demand constraint) is over.)

A recession is signaled after the gap has narrowed to 0% and then starts widening again. Once the y-o-y “rate” of widening surpasses the threshold, 1.6% in graph, a recession is tumbling in.

As of the 1st quarter 2013, we were not in a recession.

Related reading on business cycles