Yeah but my point is it only looks like a correlation and a divergence because the y-axis is in log. The chart is manipulated to make it look so instead of actually being so. The chart only tracks the prior 15 years before the "divergence" and uses this as the basis for saying these two lines should track? 15 years isn't a correlation, it's a coincidence, especially when the chart is in log and cropped with axes that don't go to zero. If you back this chart out to a bigger timeframe, you'll find that these two lines don't track each other at all, whether the y-axis is in log or not.