A popular method of investigating the market effects of multibank holding company (MBHC) affiliation involves regression of banks' local market share on a dummy variable for MBHC affiliation. The usefulness of this procedure is called into question by means of a theoretical counterexample.