Oh what a difference three months makes. Fast forward from the time those lines were written to late last week and S&P 500 correlations had risen to 52% from just 9%. As Goldman writes in a new note, that's "the largest and fastest change on record outside of 1987." Here's the rather dramatic visual on that: https://seekingalpha.com/article/4163349-happened-since-1987-s-and-p-correlation-edition
Goldman: The Only Two Times We've Seen A Market Like This Was The Cuban Missile Crisis And The 1987 Crash As Fishman further shows in the chart below, Q1 SPX realized volatility was top-quartile compared with the past 25 years, a sharp contrast to each quarter of 2017 being among the several least volatile quarters of the same period. https://www.zerohedge.com/news/2018...arket-was-cuban-missile-crisis-and-1987-crash
To paraphrase Taleb: "...correlation is charlatanism." However, cointegration might mean something if that data could be presented.
I've found the opposite. Cointegration is usually charlatanism. I hear the term from "pairs traders" who believe "cointegrated stocks" have to "mean revert."
I've found that both breed a false sense of assurance driven by an over-optimistic view of statistical analysis