Here in "Two Myths and a Legend":
'This first myth is embodied in statements like “since 2009, there has been an 85% correlation between the monetary base and the S&P 500” – not recognizing that the correlation of any two data series will be nearly perfect if they are both rising diagonally. As I noted last week, since 2009 there has also been 94% correlation between the price of beer in Iceland and the S&P 500. Alas, the correlation between the monetary base and the S&P 500 has been only 9% since 2000, and ditto for the price of beer in Iceland (though beer prices and the monetary base have been correlated 99% since then). Correlation is only an interesting statistic if two series show an overlap in their cyclical ups and downs. ...
'In the case of quantitative easing, much of what we observe as “causality” actually runs the wrong way. Market declines cause QE in the first place, and the result is a partial recovery of those declines.'
