Private Equity Analysis: Correlation vs Causation
In one of my previous posts about private equity analysis (apparently everything is counter-cyclical now), I talked about approaching potential investees and hearing that almost all of their businesses were counter-cyclical. Of course I was sceptical, and of course, the chickens have come home to roost now.
Implicit in this theme is the concept of correlation vs. causation. That is, the difference between a) two correlated variables, and b) one variable causing another variable. Some business owners would see a drop in economic growth and then an increase in their sales and assume their business was counter-cyclical. This is a simple correlation test. But thinking along the lines of causation, maybe a drop in GDP doesn’t lead to an increase in motor yacht sales or beach homes. Maybe the correlation was simply the result of a lag or government deficit spending.
So today’s hint (to self and others) is to consider causation rather than correlation in private equity analysis. Moreover, don’t be bedazzled by statisticians with regression models claiming causation. Use some common sense and don’t be too afraid of relying on impartial anecdotal evidence. You may have learned to live and die by facts in b-school, but we all know the importance of gut feel.
