The Diversification Discount and Inefficient Investment

Recent work suggests that shares of diversified firms sell at a discount, possibly because managerial self-interest makes it difficult or impossible to direct internal cash flows to their most profitable use.

Nothing demonstrates that business conditions undergo continuous upheaval quite so well as the parade of fashions in corporate structure. In the 1960s, the preference for synergy and diversification fueled an urge to buy up firms and conglomerate. In the 1990s, the preference for tightly focused businesses has fueled an urge to spin-off and carve-out. Research on corporate diversification has followed a similar progression. Early work concluded that a conglomerate, or diversified firm, could increase its profits by pooling cash flows from different lines of business and directing them to their most profitable use. Shares of the diversified firm would sell for more in the equity markets because it generated superior returns. But more recent work suggests that shares of diversified firms sell at a discount, possibly because managerial self-interest makes it difficult or impossible to direct internal cash flows to their most profitable use.

In The Cost of Diversity: The Diversification Discount and Inefficient Investment(NBER working Paper No. 6368), Raghuram Rajan, Henri Servaes, and Luigi Zingales find that the "excess value" of diversified firms relative to single segment firms is, on average, negative at -9.6 percent. There is also a wide variation in excess value, with a number of conglomerates trading at a premium, which the authors try to explain.

The authors describe how a discount might arise in a business in which there are two divisions and headquarters has limited power over division managers. Each division must choose one of two investments. The first takes advantage of cooperative efforts between the two divisions. The Gucci handbag and Gucci perfume lines, for example, make cooperative investments that preserve the brand's reputation for quality. This cooperative investment will raise the profits of the whole company if both divisions choose it. The second investment generates lower returns for the business as a whole because it does not take advantage of the benefits from cooperation. For instance, one of the divisions could run down the quality of the brand name in order to boost its own profits at the expense of the firm's. Though it is a poor choice for the diversified business as a whole, it is attractive to division managers, perhaps because it enhances their outside visibility and increases their value on the job market.

The authors show that managers are more likely to prefer the investment that benefits themselves rather than the company as a whole in firms with large differences in divisional resources or investment opportunities. Headquarters can counter the effects of managerial self-interest by using its discretionary funds to influence managerial investment choices, thus preventing "greater average investment distortions." But peace has its price--headquarters can induce division managers to act in the interests of the firm as a whole only by transferring funds from divisions with more profitable investment opportunities to divisions with less profitable ones. The extent of these transfers "in the wrong direction" increase in the diversity of size-weighted investment opportunities of the divisions in the conglomerate.

The authors use 108,050 firm-segment-years from the 1979-93 Compustat Business Segment Information database as the source of the asset, capital expenditure, and depreciation data forming the backbone of their tests. Financial firms, and those with total sales less than $20 million, are excluded. The book value of assets and the ratio of market value to the cost of asset replacement value are used to track the relative value added by headquarters' allocation of funds to each segment within each firm. When they attempt to explain the value added through allocation using a measure of the variation of investment opportunities in each firm's operational segments, the results suggest that an increase in diversity does indeed have a large and statistically significant negative effect on a diversified firm's market value. It also explains well the direction of apparent transfers within the conglomerate.

-- Linda Gorman