Low production, steady demand in early 2020 depleted inventories
As the pandemic began in early spring of 2020, the U.S. economy faced significant supply disruptions: Many local and state governments mandated the shutdown of non-essential production of goods and services to various degrees.
Although production declined, American households still demanded goods as part of their daily lives. Income support from the federal government also bolstered the purchasing power of many households, and thus there was much less of a contraction in total demand than would have occurred otherwise.
So, where do the goods that are consumed in a given month, quarter, or year come from, if not from production? Here we focus on one channel: inventories—specifically, the change in inventories, also called inventory investment. This is the difference between production and sales over an interval in time. Inventory investment includes materials, works in progress, and finished goods.
Early in the pandemic, there was a tremendous drawdown of the nation’s inventories. The change in inventories (which is negative when inventories are falling) is one component of a nation’s investment, which in turn is one component of that nation’s gross domestic product. The change in inventories (seasonally adjusted annual rate) went from a modest –$20.6 billion in the first quarter of 2020 to –$289.9 billion in the following quarter. This was the all-time low for this measure. As is often the case as an economy recovers from a recession, the change in inventories bounced back quickly into positive territory in the third quarter of 2020.
How this graph was created: Search for and select the “change in private inventories” series from 1947 to the present. Adjust the time interval using the options available (the slider below the graph or the boxes in the upper right-hand corner) to zoom in on the pandemic period.
Suggested by Bill Dupor.
Comparing crypto prices with data from Coinbase
FRED has listed the prices of certain cryptocurrencies for some time now. The FRED graph above shows four of them. Note that their prices are measured through an index number, normalized to 100 on 01-01-2018, so we can compare their growth, not their levels in actual dollars and cents.
Just eyeballing the graph reveals that they seem to largely move in unison, which is remarkable given that these prices are highly volatile. So, a significant component of their price variations must be common across all these cryptocurrencies. This commonality may come from news that pertains to all of them, such as regulation, adoption by some large player, or fiscal rulings, for example. Changes to the relative value of their counterpart, the U.S. dollar, can also play a role.
Obviously, there’s also a significant part of these price changes that is idiosyncratic to each cryptocurrency. After all, FRED’s source for these data, Coinbase, lists 5,329 different cryptocurrencies at the time of this writing; something must be differentiating them. They may have different protocols, different underlying assets (if any), different constraints on supply, and different purposes.
How this graph was created: Search for the Coinbase source in FRED, select all series, and click “Add to graph.” From the “Edit Graph” panel, change units to 100 for 2018-01-01.
Suggested by Christian Zimmermann.
Average time to fill an open job rose from 20 to 50 days
It should be no surprise that the job market has had some ups and downs during this pandemic, and one related measure is how long it has taken to fill an open position. FRED, with the help of the Job Openings and Labor Turnover Survey (JOLTS), gives us the tools to look into this.
JOLTS, among other things, provides monthly data on the number of job openings and how many openings have been filled during that month. A simple ratio of these two numbers tells us how many months it takes to fill an open position, on average.* And that’s exactly what we show in the FRED graph above. For example, in early 2011 (the start of this data series), it was taking less than a month to fill an open position.
The pandemic created special circumstances: First, job openings dropped like a stone, so these few remaining open positions could be filled quickly. Soon, conditions reverted. And now we’re in a situation where there are many openings and relatively few of them are being filled. In a matter of months, the time to fill an open position went from 20 days to 50 days, which may go higher still.
*A ratio of 1 means open jobs are filled within a month, on average. A higher ratio means it takes longer.
How this graph was created: Search FRED for “job openings” and use the non-farm series. From the “Edit Graph” panel, add a series by searching for “hires.” Finally, apply formula a/b.
Suggested by Christian Zimmermann.
View on FRED, series used in this post: