What Shape the Recovery?

This blogpost is about laying out the motivation for an article I’m planning to research and, if it offers any insights, publish.

That we have entered a brutal, in both speed and magnitude, recession is obvious. The catalyst was a demand shock where four waves of economic pain that followed each other in such rapid succession that they look like one single, confusing, event. The first was the economic sudden stop; when there’s a potentially deadly germ out there and you don’t want to go near it for your and your community’s sake there is only one equilibrium: the cessation of economic activity.

Second, as revenue stops so does the demand for the employees who produce the goods and services no longer being sold. As confirmed by last week’s jobless claims data, massive layoffs ensued. Third, came the wealth effect. As people saw their wealth drop so rapidly in their liquid investments and became uncertain about the value of their illiquid ones (e.g., housing) demand for the goods and services still offered plummeted.

Fourth, came the paralysis in capital spending which makes the potential for future growth uncertain. But that’s history. Everyone knows that we are in going through an unprecedented contraction, both in terms of potential magnitude (sell-side economists are forecasting Q2 output contraction of 25-30+%) and speed (lost output per quarter). But it is a truism that all recessions end: although we don’t have the data, we’re pretty confident that Europe’s GDP today is higher than in the mid-13th century, as the continent emerged from the plague.

So, the question I’m interested in researching and modeling is the timing and speed of the recovery and, since some of the work will need to be empirical, the question arises about which should be my relevant data points. The Great Recession is clearly too different in nature; it was induced by a financial crisis which led to a retreat in spending, creating an output gap. While exotic financial instruments were blowing up and financial institutions were fighting for their life, the economy remained intact. All that was required to restart output was for consumers to start spending again or for the government to do it in their stead. The reason why the recovery that started in 2009 was slow is because the government didn’t spend enough to fill the output gap.

But this recession is very different. It started with a (man-made) stoppage of economic activity that may or may not result in a financial crisis. Over the past weeks I spoke a lot with Nouriel Roubini, who presciently predicted both the Great Recession and the Corona recession and shared my conundrum with him. He answered, without doubting, that I should model this recession as a natural disaster-induced one like, for example, the 2010 Haiti earthquake.

My starting hypothesis is that this data can be fitted by a 3-parameter function that contains both the recession and the recovery The underlying assumption behind this model is that, during recession-recovery periods, one portion of the economy is shrinking exponentially, while the rest is growing exponentially; therefore, my three parameters are simply: GDP pre-natural event, the rate of decline of the declining portion of the economy, and the rate of growth of the rest of the economy. If this model fits the empirical data, it would substantiate my basic hypothesis that the two legs of the GDP curve are inextricably linked and cannot be considered as separate events.