Progression To A New Mean

2022 is the year of the return to business as usual. By the end of it, the COVID restrictions, the elevated demand for goods, the low demand for services, the messed-up supply chains, above-trend GDP growth, record-high inflation, uber-accommodative monetary policies, madly expansive fiscal policies, and the absurdly negative 10-year US Treasury real yield will all revert to ‘normal.’

It is statistical. In the 19th century, Sir Francis Galton observed that extremely tall (short) parents tend to have taller-(shorter-)than average children. However, in all cases, the children tend to be closer to the average than their parents. He presented his findings as a ‘regression to the mean.’

In Thinking Fast and Slow(2011), Nobel Prize-winner Daniel Kahneman notes that the power of the regression to the mean tends to be underestimated. He takes the following fake headline as an example: ‘Depressed children treated with an energy drink improve significantly over three months.’ The statement is true, but it is misleading since it would be true without the energy drink, too: As an extreme group, depressed children regress to the mean over time.

Ordinary circumstances follow extraordinary ones. The end of 2022 and 2023 could thus look and feel like 2018, possibly the most ‘normal’ year the world has experienced since the Great Financial Crisis. Following this logic, valuation multiples in the Industrial Tech sector would return to their 2018 level, i.e., c.12x-14x forward EBIT from about 16x throughout 2021 as earnings normalize. The growth/quality premium over value in the sector would narrow from close to 50% to c.30-40%.

These levels would be consistent with the economic regime of the last decade, one defined by anemic GDP growth and ultra-low real interest rates. Because the delta between expected growth and interest rates was smaller after the GFC than before, valuation multiples were structurally, mathematically higher than in the noughties (background in this 2017 Sunday note).

From here onwards, pessimists such as Nouriel Roubini draw attention to the higher post-COVID government debt levels, ongoing fiscal deficits, and the ‘greenflation caused by the implementation of ESG initiatives discussed last week (see also Climate Change = Stagflationlast October). Higher rates would choke growth and potentially lead to another global financial crisis.

I am warming up to a more optimistic perspective, even if it appears to be less fashionable. Its proponents would point to the productivity boom set off by the accelerated use of new technologies. They would observe that individuals adjust to new conditions with impressive agility, like the hysterical Hazan family in Family Business.’ And they would argue that, after two decades of under-investments in fixed assets, investments in sustainability would pay for themselves in record time, freeing up resources including raw materials (e.g., circular and sharing economy), energy, and labor (e.g., increased productivity).

The global economic output potential would rise, with disinflationary consequences. High GDP growth would be associated with low real rates. A new economic regime with potentially higher valuation multiples than under the post-GFC era would emerge. The growth/quality premium would further compress. Europe, a value market, would shine.

Instead of a regression to the mean, there could well be a progression to a new, more sustainable mean.

114 views0 comments

Recent Posts

See All

Nowadays, big problems and their solutions are expressed in trillions. COVID’s global economic cost has been estimated at more than US$ 12 trillion. The Ukraine war may cause US$ 2.8 trillion in foreg

To appreciate the current capital market trends, it is worth reflecting upon the coffee market evolution. For most of the 20th century, the traditional choice of coffee drink could be summarized as sh

The world is only starting to mourn the exceptionally benign macroeconomic circumstances experienced over the last thirty years. It is dawning on it that the future will require a different approach t