What the Next Recession Will Look Like: More 2000 than 2008
Thread.
Start with the observation: every recession since the late 60s has been (closely) preceded by a year-over-year decline in real (inflation-adjusted) household assets/net worth.
Thread.
Start with the observation: every recession since the late 60s has been (closely) preceded by a year-over-year decline in real (inflation-adjusted) household assets/net worth.
(The two measures are equally predictive; HH liabilities, only ~15% of assets, change slowly compared to assets.) Straightforward story: when people (suddenly) have less wealth, they spend less. And since spending is what causes production…
There have been three false positives for this predictor since 2000: in 2003, 2011, and 2018. So we can currently only view this as a firm necessary condition for an ensuing recession (with no exceptions since the 60s) — not as a sufficient predictor.
Why it won’t be like 2008: That asset decline was broadly based, encompassing real estate assets first, with equities following. 57% of real-estate assets were held by the bottom 90%; they lost $4T out of $14T in RE asset holdings (29%). The bottom 90 only held 15% of equities.
Now add: the bottom 60% of households turns over about 40% of its wealth each year in spending. The top 20% (consistently) turns over only 5%. You could expect the lost RE wealth of the bottom 90 to reduce their spending by maybe $1T ($4T * 25%) — 5% of GDP.
So what would the spending effect be, today, if the equity market dropped 50% as predicted by @hussmanjp, for instance? The top 10%’s $22T in equity holdings would drop by $11T. Their 5%-of-wealth spending over the next year would decline by ~$550B — 3% of GDP.
(Probably less, because the top 10% presumably spends a smaller % of wealth than the top 20%, say 2% or 3% instead of 5%.)
We might say that top wealthholders’ spending is more elastic — opt for a $100K kitchen remodel vs $300K. (Compare: elasticity of necessary spending like health care and education.) But that limited elasticity operates within and is dominated by the consistently low 5% turnover.
So, stake-in-the-ground prediction: the next real-economy recession, caused by an equity-market drawdown like almost every other postwar recession, will look like those recessions — notably like 2000/2001 — not like 2008.
And in fact it’s smelling very much like January/February 2000 out there right now...
HT @ryanlcooper for a link to that last graph, which actually inspired this whole thread.
