We consider an ‘all or nothing day’ to be a day where the net daily breadth reading (daily advancing stocks minus declining stocks) for the S&P 500 is above +400 or below -400. While these types of days were practically non-existent in the 1990s, beginning in the early 2000s, their frequency started to rise with the increased popularity of trading in the S&P 500 ETF (SPY). Whereas investors used to buy and sell individual stocks, the increased popularity of SPY moved the market more towards the type of environment where investors were buying and selling the market.

All or nothing days also increase in frequency during periods of increased market volatility, and that trend has been no different this time around either. The chart below shows the 50-day moving average of all or nothing days going back to 1990. Over the last 50 trading days, mot than a third of all trading days have been all or nothing days. The only two other times where the average was higher in the last thirty years were in December 2008 and November 2011. The average got close to current levels back in late 2015 and early 2016 but was never able to quite get above 33%.

(Click on image to enlarge)

Looking at the frequency of all or nothing days on an annual basis shows another interesting trend. So far this year, there have been 19 all or nothing days for the S&P 500. We may be barely a quarter into 2020 so far, but this year’s total already ranks above more than half of the 31 years since 1990. In fact, at the current pace, the S&P 500 is on pace to have 70 all or nothing days in 2020, which would tie 2011 for the most ever in a given year. It’s only April, but 2020 is shaping up to be the year of record volatility. 

(Click on image to enlarge)