The Incredibly Non-mean-reverting Corporate Profits

For all of the talk about the necessary mean reversion in corporate profits, let’s start with auto-correlation of Corporate Profits as a % of GDP:


No such mean-reversion.  Future profits are actually well correlated with past profits, but the relationship slips into statistical insignificance – but not into negative (mean-reverting) territory to much degree of statistical significance.

Let’s take a derivative — year/year — to see if _growth_ of profits will demonstrate mean-reversion:


The same story is told here.

Mean-reversion is, at the very least, not supported by historical precedent.

This entry was posted in Uncategorized. Bookmark the permalink.

3 Responses to The Incredibly Non-mean-reverting Corporate Profits

  1. Q says:

    Thank you for presenting an alternative to the profit margin reversion, what would be the response to Dr. Hussman’s profit marign section in this link?

  2. Michael Green says:

    Your analysis is very dependent on time period selected. Mean reverting on a quarterly basis? Of course not. Mean reverting on a decadal basis (approximately a full cycle) — yes. So the clear point is that once profits begin declining, they will keep doing so until stretched on the other side. No permanently higher plateau.

    • Matt Busigin says:


      Thanks for your comment.

      If you look at the graph, however, you’ll see that it is *not* reverting on the decadal basis. The auto-correlation out 40 quarters is about zero — there is no relationship between profitability now and out 10y, let alone a negative one.


Leave a Reply