La vérité existe. On n'invente que le mensonge. —Braque (Truth exists. One only invents falsehood.)
Am I the only one who noticed? It’s been two days. . . I haven’t read anything. . . no search results on Google. . .
The topic is the chart introduced as the day’s “Single Best Chart” during Bloomberg Surveillance on the morning of Tuesday, 5 June 2013 (video link). It was titled “A Return to Normality” and purportedly showed 6-month global equity correlation against time over the period 1998-2013. The graphic was attributed to Andrew Garthwaite of Credit Suisse Research.
The levels shown ranged between approximately +0.6 and +2.5, which is a problem. The correlation function’s definition is such that values can range from -1 to +1. It is likely the chart showed a graph over time of a linear asset-pricing model’s first-order coefficient, what is popularly known as ‘beta’. But wait, there’s more. . .
The title of the chart is also troublesome. With levels having recently decreased from 2 to approximately 1.1, the title possibly implies that a value of one is associated with a normal distribution (of what)? Whatever level beta (positive/negative or small-/large-magnitude), whatever level correlation (the real one). . . neither parameter provides evidence for or against the validity of a normality assumption. But wait, there’s more. . .
It is useful to know that correlation (the real one) is a measure of the model’s validity, including the validity of the ‘beta’ parameter estimate. (Yes, we are aware of ‘R-squared’.)
I guess its safe to say that we’ve had our fun?