Summary

Jeffrey began by addressing the first major supply side constraint on everyone’s lips: labour. Even if people want jobs, which there's some doubt about, do they have the right skills? Since the 1990s’ recessions it has become common for there to be a mismatch between labour supply and demand as those who’ve worked in specialized firms that become unemployed during a recession do not have the skills for jobs created as the economy emerges from it with new technologies. Using a data source that scrapes the HR websites 30,000 US companies DeepMacro can estimate for how long an opening is on a website before they can estimate it's been filled. At the height of the pandemic this number soared up to 60 - 70 days. That came down to normal levels of 45 - 50 days by August of 2020. Looking at it by sector the worst numbers are in retail, transportation & warehousing, hotels & restaurants, and then arts, entertainment & recreation. The most recent numbers in October estimate 80 or 90 days for many of these sectors. Transportation / warehouse has seen a significant increase throughout the year, having started at 50 days it has gone above 80 - a month longer to fill a job. It does look like there's some spill over to manufacturing – but it hasn't been critical yet.

On the second major supply constraint, ports, how do we measure these blockages? As their primary data source on this DeepMacro uses activity levels as indicated by mobile phone activity in and around the major ports, in particular the ports of LA and Long Beach, part of the same complex and when added together make up close to 30% of all inbound port traffic to the United States. Looking at the number of mobile phones in the area around the port from May until now there has been a steady increase of people when they hadn't been before. Moreover, how long people are waiting near the ports - or “linger time” - a catch all for a lot of the different supply side constraints. For Long Beach since the beginning of May linger time has gone up by about 50%. It was okay through June, but from summer this number has dramatically risen. A similar story in the port of Houston up about 50%, and in Newark the level has doubled. The top five ports in the US are all experiencing a lot of congestion at the gates.

During the pandemic there was a huge reduction in the diversity of what we were doing and spending money on resulting in reduced diversity of firms and more concentration among the survivors, e.g Amazon. The survivors sucked up all of the revenues and the inputs, especially labour. The diversity of activity returned but unfortunately the ecosystem has lost a lot of the speciation and specialisation that was needed to capture revenue. Independent truckers who used to move stuff from the port of Long Beach to the intermediate points of transit, or the people who picked it up at the transit centre that's 50 miles from Long Beach are now driving minivans for Amazon and are probably making more money. It's knocked huge holes in the very fragile ecosystem and what it can produce.

Wrapping this into an overall economic view; on economic growth, 10 out of 10 of the G10 are above trend. DeepMacro’s view is above consensus on all 10. That all 10 out of 10 are above trend is remarkable. On inflation, again, above trend in all 10 and accelerating in nine out of 10. DeepMacro is mixed versus consensus; five versus five. Looking at satellite information on air pollution, industrial chemicals to help estimate industrial activity; global manufacturing is still booming but interestingly enough not in China. China's weak but the US is super strong. Labour demand in China declined by about 20 or 30% when looking at online job offers.

Topics