Two unrelated “Data” items today:
#1: What the CBOE VIX Index is saying about last week’s selloff:
- While Thursday was a highly unusual day for the S&P 500, as described above, the VIX only closed at 40.8.
- That is not even 3 standard deviations (8.1 points) from the long run daily mean back to 1990 (19.3), or 43.6. Friday’s close of 36.1 is further still from that level.
- Volatility traders have plenty of near-term experience with 5-standard deviation VIX prices (above 59.9), with 11 closes over that level this year. A point of reference here: 2008 saw 28 days where the VIX ended the day over 59.9.
- There’s a good reason why the VIX hasn’t responded in a more pronounced way to Thursday’s move: actual price volatility over the last 30 days is still less than half what it was in early April. Back then, it was 85% and accurately predicted by the VIX close of 83% on March 16th. Over the last 30 days, including Thursday, actual volatility is 28%.
You can see the difference then-to-now in this chart from www.ivolatility.com, with the blue line showing actual S&P 500 volatility and implied volatility in beige:
Takeaway: the VIX only measures market sentiment about prolonged volatility, not the odds of a one day move even if (like Thursday) that decline is statistically very rare. So don’t assume the VIX is “broken” if it doesn’t spike this week on further US equity market volatility. After all, the chart above shows it nailed April’s volatility as it peaked on March 16th.
#2: Before the economic effects of the COVID Crisis skews another much-watched dataset, we wanted to show you one that underpins a topic of considerable political interest: US wages as a percent of GDP from 1947 to 2019:
What we see:
- The period from 1947 – 1974 saw labor’s percentage of GDP hover right around 50%. It would fall in recessions but then bounce back during recoveries.
- Starting in the late 1970s, however, labor’s share of GDP declined through every cycle. For a brief moment in the late 1990s it looked to be making a comeback and in 2000 it was 47%. It had not been there since 1982.
- But after that one false start wages relative to economic output started to fall again. Even 2019, long into an economic recovery, labor share of GDP was lower than in the 2002 – 2008 expansion.
What the data actually means:
- There is little consensus in economic circles about the causes of declining labor share of GDP.
- The most widely cited paper (Karabarbounis and Nieman, 2013, 1678 citations) shows that labor share has been declining around the world and roughly half of the move comes from cheaper “investment goods” like new technologies.
- The second most widely cited paper (Elsby, Hobijn, Sahin, 2013, 650 citations) says a third of the decline is merely accounting and the leading cause for lower US labor share is the offshoring of labor-intensive parts of the US supply chain over the last 25 years.
- The third-most cited paper is newer (2017, Autor, Dorn, Katz, Patterson and Reenen, 380 citations) and proposes that the rise of “superstar firms” has made for lower labor share of GDP. The effect here is twofold: first, leaner organizations (less labor) are more likely to grow and second, competitive industry dynamics pressure companies competing with “superstars” to either emulate their practices or sell to industry consolidators.
Why all this matters: the COVID Crisis has temporarily taken US wage levels/growth off the political radar screen, but we have no doubt it will return as election day draws near. We find it remarkable that there is so little consensus in the academic community regarding labor’s ever-declining share of GDP. Explanations literally range from “it’s just accounting” to “it stems from a litany of factors”. As you evaluate potential policy responses, from changing tax rates to universal basic income, that is a useful fact to keep in mind.
Free resource with implied volatilities for US listed stocks and ETFs (link to IVs for the IVV S&P 500 ETF): https://www.ivolatility.com/options.j?ticker=IVV:NYSEArca&R=1&period=3&chart=02&vct=4
Google Scholar listing of widely cited articles for “labor share decline”, which includes links to the papers mentioned above and several others: https://scholar.google.com/scholar?hl=en&as_sdt=0%2C33&q=labor+share+decline&btnG=