Lucky or Good?

In a paper presented at last year’s Academy of Management meetings, Andy Henderson of the University of Texas, along with Michael Raynor and Mumtaz Ahmed of Deloitte Consulting, investigate companies who have persistently high performance (press release related to the paper here). They tracked the returns of over 20,000 companies from 1965-2005. They took a closer look at 865 firms who had at least 11 years of data, and after controlling for a number of factors outside the direct control of management (industry, time trends, firm size, etc.), they found that 45 ranked in the top decile of annual performance five or more times. Sounds like a good set of companies for another type of Good to Great book.

But, here’s the kicker (and what makes this a pretty cool paper). The authors wondered what we might have seen in the data if success were simply random. After running a bunch of simulations in which success of hypothetical companies was randomly determined, they found that about 37 firms would be expected to hit the top decile of annual performance five or more times. So, it’s quite possible that the title of a book examining these companies should be Lucky rather than Good to Great.

This is a pretty big issue for research that claims to isolate the determinants of sustained superior performance. We may be studying companies who just got lucky rather than companies whose managers have figured out some better way to compete.

See the Deloitte site on the overall persistence of profit project, from which the paper came.