and on market prices. The weaknesses stem from possible noise in the market from conflicting events and questions about market efficiency.
Acquisitions. Seven studies8 find cumulative average residuals (CARs) at the announcements of transactions are significantly more negative for diversifying deals than for focusing deals. These studies suggest that mergers that focus the firm enhance the buyer’s share value by 1 to 3 percent more than diversifying deals. Yet six other studies9 show significantly positive CARs for diversifying acquisitions. Most of these, however, are studies of conglomerate acquisitions in the 1960s (e.g., Hubbard and Palia 1999) or are associated with the relaxation of regulatory constraints on diversifying acquisitions (e.g., Carow 2001). On balance, the event studies of acquisition announcements suggest that focus pays more than diversification.
Joint ventures and alliances. Three studies consider the effect of focusing or diversifying JVs. Ferris et al. (2002) find focus-increasing JVs show materially larger CARs than diversifying JVs. Chan et al. (1997) report that horizontal alliances involving technology transfer have a materially higher CAR. And Gleason et al. (2003) find that horizontal deals in the financial services industry have materially higher CARs than diversifying deals. The event studies of JV and alliance announcements suggest that focus pays more than diversification.
Divestitures, spin-offs, and carve-outs. Generally, divestitures, spin-offs, and carve-outs are good news for investors: Since these deals shed assets, the results would seem to be roughly supportive of focusing. But what matters is the nature of the assets being disposed. Two studies of carve-outs reported in Exhibit 6.18 suggest a materially larger announcement CAR when the carved-out unit is not from an industry related to the parent’s core business (Hurlburt et al. 2002; Vijh 2000). Three studies of spin-offs in Exhibit 6.17 show a materially larger announcement CAR when the transaction is focus-increasing (Veld and Veld-Meruklova 2002; McNeil and Moore 2001; Johnson et al. 1996). Regarding divestitures, Donaldson (1990) reports materially larger positive CARs at the announcement of sale of noncore assets compared to core asset sales. Dittmar and Shivdasani (undated) report that over the year following the divestiture, firms that became single-business firms had a 3 percent higher return than those that remained diversified. In short, the event studies of divestitures, spin-offs, and carve-outs are consistent with benefits from focusing and penalties from diversification.
Q TESTS Tobin’s Q is a measure of economic efficiency estimated as the ratio of the market value of assets divided by book value. The higher the Q, the higher is efficiency. Typically, these studies regress Q against a variety of independent variables, including measures of diversification and focus. Four studies give findings consistent with the benefits of focus. Lang and Stulz (1994) find that diversified firms have lower Qs than single-business firms. Morck and Yeung (1998) find that diversification is associated with lower Q except where the industry is information-intensive. Aggarwal and Samwick (2003) report that diversification has a significantly negative effect on Q.
EXCESS VALUE STUDIES: TESTS FOR A DIVERSIFICATION DISCOUNT A logical test of the impact of diversification on value is to compare the actual market value of the firm with its “sum of the parts” value, where each part of the firm is valued at multiples consistent with industry peers. The “excess value” of a diversified firm is simply the difference between actual and imputed values. Nine studies10 find negative excess values for diversified firms, in the range of 8 to 15 percent—this is the famous “diversification discount”11 that is often cited in debates over the unprofitability of diversification. On the basis of these findings Lamont and Polk (2002, page 75) asserted, “Diversification destroys value.” Yet, more recent studies have challenged the size and even the existence of the diversification discount. The line of attack is that certain data sources contain an inadvertent bias in favor or the diversification discount and that many of the units acquired by diversified firms were already discounted before their acquisition—this means that the existence of a discount has little to do with a strategy of diversification. Finally, some studies use more granular data arguing that business segments are too large to capture the costs or benefits of diversification. Nine studies12 report no discount, or even a diversification premium using these revised research approaches. If one believes in the power of the newer techniques, the excess value studies would suggest that diversification has a neutral or positive effect on value.
PRODUCTIVITY STUDIES Another line of research is to consider the impact of diversification or focus on the productivity of business units and plants. Lichtenberg (1992) found lower total factor productivity with increases in diversification. But Schoar (2002) reported that plants in diversified firms were 7 percent more productive than plants in single-business firms. Nevertheless, increases in diversification are associated with a net decrease in productivity. Plants that had been acquired actually increased their productivity, whereas incumbent plants decreased in productivity—but since there were fewer acquired than incumbent plants, the total effect on productivity was negative. Schoar wrote, “Diversified firms experience a ‘new toy’ effect, whereby management focus shifts towards new segments at the expense of existing divisions. As a whole, these results indicate that diversified firms have a productivity advantage over their stand-alone counterparts. They even increase the productivity of their acquired assets. With each diversifying move, however, these firms lose some of their productivity advantage.” (Page 2380)
PROPENSITY TOWARD DIVESTITURES Scholars have studied the characteristics of those firms that undertake divestment. Porter (1987) and Ravenscraft (1987) found that divestiture follows acquisition: Firms may buy, but are not assured of retaining their purchases. Their reading was that growth through acquisition was not a stable growth strategy. Weston (1989) offered the rebuttal noted earlier. Kaplan and Weisbach (1992) found that firms were more likely to divest unrelated acquisitions than related acquisitions, suggesting that unrelated acquisitions don’t pay.
STUDIES OF LONG-TERM REPORTED FINANCIAL RESULTS Though accounting results are easily “managed” by executives and are vulnerable to exogenous effects unrelated to diversification, they are an ongoing focus of investigation. Four studies13 showed that firms following strategies of unrelated diversification underperform those firms who focus more. Yet four other studies14 found improvements in operating performance following diversifying acquisitions. In addition, Anslinger and Copeland (1996) found that firms pursuing a conscious strategy of unrelated diversification have realized high abnormal returns for sustained periods. Baker and Smith (1998) documented high absolute returns to Kohlberg, Kravis, and Roberts, a well-known leveraged buyout firm and owner of a diversified portfolio of industrial interests. Fluck and Lynch (1999) presented a model of corporate strategy in which both diversifying acquisition and then focusing divestiture create value: This relies on the existence of positive net present value (NPV) projects that are unable to obtain financing in public markets. The large firm acquires, finances, grows, and then divests these businesses profitably. In sum, it seems that diversification or focus may not help to discriminate among firms on the basis of long-term performance. Anslinger and Copeland (1996) argued that it is the postacquisition management strategy and structure, rather than the strategy of diversification or focus, that matters in producing long-term performance.
Practical Implications of the Research Debate over Diversification versus Focus
How are we to make sense of these disparate and