Is there an optimally diversified conglomerate? Gleaning answers from capital markets

Abstract

Motivated by recent productivity-based theories of diversification, we argue that only conglomerates with an optimal degree of diversification can utilize their comparative advantages across various industries and achieve economies of scope by eliminating redundancies. Evidence from both corporate bond and equity markets suggests that optimally diversified conglomerates consist of either (1) approximately five equally weighted divisions, or (2) one large core business segment that roughly accounts for 75 % sales. Moreover, the relative size of divisions has a critical impact on how diversification affects credit spreads and excess values. Nonparity among divisions correlates with greater costs that increase with the number of divisions.

Appendix: Diversification choice model

Campa and Kedia (2002) highlight the endogeneity problem in most diversification studies. This is the question they pose: is it the diversification that causes the diversification discount, or is it the firm characteristics that lead a firm to diversify that causes the diversification discount? They propose an endogenous self-selection model and use Heckman’s correction to control for the self-selection bias induced on account of firms’ choosing to diversify. The goal is to control for the self-selection bias created by firms’ decisions to diversify by integrating the diversification decision. We follow their methodology closely to achieve the same goal. The dependent variable takes the value of one if the firm is diversified and zero otherwise. Then, as the first step, we estimate a probit model of diversification choice. The independent variables were chosen based on the model used by Campa and Kedia (2002). Readers interested in the reasoning for the inclusion of certain variables such as control variables are referred to their paper. We introduce four additional control variables: research and development expenses, market-to-book ratio, dividend paying nature, and industry dummies to capture aspects discussed in Villalonga (2004). We classify the results, shown in Table 13 in Appendix, into four categories: growth, size, industry association, and economic conditions. The first, growth aspect of the firm, which is captured by capital expenditures, negatively influences the decision to diversify. The second, firm size, positively influences the decision to diversify. Total assets, dividends, and membership in a major index are indicators of firm size; all three of them are statistically significant and positively related to the diversification decision. The third aspect, industry association, positively influences the decision to diversify. The proportion of diversified firms in an industry and the proportion of sales in an industry that is contributed by conglomerates provide an idea of the industry concentration. The fourth aspect, economic conditions (captured by the dollar volume of mergers and acquisitions, number of mergers and acquisitions, recession months, and 5-year GDP growth rate) warrant careful interpretation. The mergers and acquisitions activity indicates the current robustness of the economy. We find that robust activity in the economy increases the probability that a firm will diversify, which is not a surprising result. The 5-year GDP growth rate indicates the trend in economic growth over a period of 5 years, and we find that it increases the probability that a firm will diversify. The lagged variable, which indicates the changing trend, imposes a negative impact on the decision to diversify. Consistent with the changing trend argument, the lagged recession months variable also imposes a negative impact on the decision to diversify.

Table 13 Probit model of diversification using debt sample

See Tables 13 and 14.

Table 14 Probit model of diversification using equity sample