To compute our index, we extract the first principle component from these seven time series, using the sample from January 2008 to present. We scale our index to four-quarter GDP growth, so a reading of 2 percent in a given week means that if the week’s conditions persisted for an entire quarter, we would expect, on average, 2 percent growth relative to a year previous. The chart below plots the index based on data through March 21, 2020. The trough of the Great Recession is clearly visible, as well as the following recovery. Developments in the past week saw the index fall to a level unseen since 2008.
In the interim, fluctuations have been smaller, but correspond well to the paths of important macroeconomic aggregates. The top left panel of the following chart plots the index against the twelve-month percentage change in industrial production (IP). The index tracks IP growth very closely. The top right and bottom left panels do the same for capacity utilization (CU) and the ISM manufacturing index. Finally, we plot four-quarter GDP growth. It is clear that despite the noise inherent in the raw, high-frequency data, combining it to construct an index, as we do, provides an informative signal of real economic activity.
Our index is also quite robust to changes in the way it is constructed. Subtracting or adding individual series has little effect on the overall path; the same is true for estimating the weights on each series using only more recent data. The top left panel of the chart below compares our baseline index to one without fuel sales, and the top right panel to one without consumer sentiment, while the bottom left adds data on mortgage applications for purchase. Clearly, the common signal is not driven by the precise choice of series. The bottom right panel plots our baseline against a series computed with weights estimated using only data from 2015 onward, showing that the relationship between these series has been fairly constant during and after the Great Recession.