E-piphany

I know that today, and probably for a little while, investors are focused on Ukraine and Russia. I am gratified that for what seems the first time in many years, notes about the conflict tend to include some form of the addendum “and its effect on domestic inflation,” albeit in many cases this is from the perspective of how this engagement will damage or burnish President Biden’s poll numbers at home and the prospects for his party in the midterm elections. How self-absorbed we Americans are! To be fair, in my opinion the importance of the US policy-response operetta was always less about Ukraine than about Taiwan. I hope that doesn’t turn out to be right.

However, today I want to talk about the re-onshoring trend in manufacturing, and the significance of this for inflation going forward.

One of my 2022 themes so far is that the conventional expectation for inflation to peak soon and ebb to a gentle 2% over the next 12-18 months is mostly predicated on the idea that the extraordinary spikes we have seen in certain categories (see: motor vehicles) will eventually pass, and inflation will return to the underlying trend. The simpler observers see it as 12 months since (mechanically) the spikes will all be out of the y/y number in 12 months. Some forecasters are giving themselves a little wiggle-room by saying it will take 18 months as the ports unclog and ‘other knock-on effects’ wash through. But in my opinion, the evidence is strong that the underlying trend is no longer 2%, but more likely 3-4% or higher. Part of that evidence is the great breadth that we have seen in the recent inflation numbers, which suggests either a riot of unfortunate coincidental events all in the same direction, or else a common cause…say, the rapid growth rate of the money supply, which as of the latest report is still growing more than 12% annualized over the last quarter, half-year, and year.

The forecasts of sharply decelerating inflation expect the parade of “one off” causes to end – and, crucially, to be replaced by unbiased random events that are equally likely to be up or down. This is ‘assuming a can-opener,’ and is economist malpractice in my opinion. Because of the continued rapid growth of money, and until that rapid growth slows drastically or reverses, the surprises are mostly going to be on the high side. That’s why I expect inflation to be lower at the end of the year than it is right now, but not lots lower.

All of this, though, obfuscates a trend that had started prior to COVID but has gained great momentum since. When President Trump was first elected, we’d suggested in our customer Quarterly Inflation Outlook that one of the following winds which had kept inflation low despite loose monetary policy throughout the 1990s and 2000s was in the process of stopping and potentially reversing. That following wind was globalization. I eventually ended up talking a lot about de-globalization. Here’s one article from four years ago. I really love the Deutsche Bank chart in it.

In a nutshell, the argument was that domestic goods prices had been kept abnormally low despite strong economic growth and loose monetary policy through the prior quarter-century because businesses had gradually over time offshored production and extended raw materials and intermediate-goods supply chains to cheaper manufacturing locations outside of our borders. But that’s a trick that can only be turned once. When most production is overseas and most intermediate goods imported from the Pacific Rim, costs will resume rising at the rate of inflation in the source country, adjusted for FX changes. For decades, we’d seen core goods inflation near zero despite services inflation in the 2-4% range, as this dynamic played out, but there was no reason that goods inflation should permanently be zero.

So I thought that in 2016 we were already coming slowly to a point where similar monetary policy going forward was going to result in less growth and more inflation because that trick had been used up. The election of President Trump merely accelerated that timeline and increased the probability that the trend wouldn’t only stop but could reverse, causing the division of growth and inflation for a given monetary policy to be distinctly bad and requiring much tighter policy.

COVID-19, and the global response to COVID-19, has more or less totally reversed the arrow of global trade. Businesses are pulling manufacturing back to the US and pulling supply chains back to the Western Hemisphere as much as possible. Geopolitical tensions between the US and Russia, and the US and China, combined with the increased appreciation of the optionality of inventories and the cost imposed by long and variable lead times, which is partly reflected in the need to hold more inventory. And that, in turn, drastically decreases the attractiveness of a long supply chain, especially with global tensions, the rise of democratic populism (“we want what’s ours, not some global citizenship award!”), and the persistent rise in energy and other costs of transportation (driver shortages, etc).

All of which arguments I’ve made before. But I’m not sure I’ve drawn the line clearly enough that the net effect of this changing dynamic – which results in manufacturers choosing higher costs rather than lower costs – is that goods inflation is unlikely in my view to return to being centered around zero. While core services are a bigger chunk of the consumption basket than are core goods, that’s mostly because of shelter services. Core goods is 22% of the consumption basket; core services (less rent of shelter) is 25%. So this is not something that can be idly dismissed. If the mean of the distribution moves from 0% to just 3%, that moves the “normal” level of inflation up ~0.66%. Obviously, I think in the medium-term the number is a lot larger than that, but the key is whether the effect is going to be persistent over a long period of time (think years or decades, not months). I believe it will far outlast COVID, because the causes go far beyond COVID.