644  Capital Structure, Dividends, and Share Repurchases Leverage should be lower for companies with lower returns, higher growth potential and risk, or highly specific assets and capabilities. This is the case in sectors such as software, biotechnology, and high-tech start-ups. Potential tax savings are small, because their taxable profits are low in the near term. Management needs more financial freedom, because investments are essential to capture future growth. In contrast, the costs of business erosion are high, because these companies would quickly lose valuable growth opportunities, and any remaining assets have very little value to third parties. For the same reasons, companies with more volatile earnings and higher advertising and R&D costs are generally financed with less debt.18 Leverage also tends to be low for companies producing durable goods, such as machinery and equip- ment, requiring long-term maintenance and support. The highly specific capa- bilities of these companies make financial distress costly for their customers.19 Although some finance textbooks show a high potential tax benefit from higher leverage, the benefit is usually limited for large, investment-grade companies. To illustrate, consider a simple example. Exhibit 33.5 shows how the multiple of enterprise value over earnings before interest, taxes, and amor- tization (EBITA) for an average company in the S&P 500 would change along with the amount of the company’s debt financing, as measured by the EBITA- to-interest coverage ratio. The EBITA multiple is estimated using the basic value driver formula, presented in Chapter 3, and applied using an adjusted- present-value (APV) methodology.20 We assume a long-term ROIC of 14 percent and an unlevered cost of capital of 9 percent—typical scores for a middle-of- the-road S&P 500 company. As the exhibit shows, tax-related benefits from debt do not change enterprise value dramatically, except at very low levels 18 M. Bradley, G. Jarell, and E. Kim, “On the Existence of an Optimal Capital Structure: Theory and Evidence,” Journal of Finance 39, no. 3 (1984): 857–878; and M. Long and I. Malitz, “The Investment- Financing Nexus: Some Empirical Evidence,” Midland Corporate Finance Journal 3, no. 3 (1985): 53–59. 19 See Barclay and Smith, “The Capital Structure Puzzle”; and S. Titman and R. Wessels, “The Determi- nants of Capital Structure Choice,” Journal of Finance 43, no. 1 (1988): 1–19. 20 Applying the APV methodology to the value driver formula and discounting the tax shield on inter- est at the unlevered cost of equity results in the following formula: Value NOPAT ROIC = − −           + × × + = ∞ ∑ 1 1 1 g k g k T D k u D t u t t ( ) where ku is the unlevered cost of equity, Dt is the debt in year t, kD is the cost of debt, T is the tax rate, and all other symbols are as defined in Chapter 3. If we make the additional assumption that companies finance with debt while maintaining a stable interest coverage ratio, the formula can be simplified as follows: Value NOPAT ROIC Interest EBITA = − + −     −           1 1 g T T k g u where EBITA/Interest is the target coverage ratio. Setting a Target Capital Structure  645 of interest coverage (below 2) rarely seen for large, investment-grade compa- nies.21 Compare that with the much bigger impact on shareholder value of key value drivers such as return on invested capital (ROIC) and growth. In contrast, losses in flexibility from higher leverage do translate to signifi- cant value destruction. John Graham and others examined listed U.S. compa- nies over a period of more than 25 years and analyzed the loss in a company’s value due to deviations of its leverage from what was estimated as its theoreti- cal optimum.22 The analysis offers two key insights, illustrated in Exhibit 33.6. First, it confirms our analysis that value at stake is limited to no more than a couple of percentage points for a fairly wide range of leverage around the theoretical optimum. Second, it shows that there is a lot more downside from having too much debt than from having too little. In other words, the losses due to diminished flexibility tend to outweigh the gains from tax benefits and management discipline. Credit Ratings and Target Capital Structure Difficult as it may be to determine an optimal capital structure, it is much easier to find an effective structure—that is, one that cannot clearly be improved upon in terms of shareholder value creation because it is somewhere in the rela- tively flat range of the valuation curves of Exhibits 33.5 and 33.6. EXHIBIT 33.5  Capital Structure’s Limited Impact on Enterprise Value 0 5 10 15 20 0 2 4 6 8 10 12 Interest coverage1 Enterprise value/EBITA 3% sales growth 6% sales growth Approximate coverage range for A+ to BBB– rating 1 EBITA/interest. 21 Note that at such low levels of coverage, the expected value of any tax savings will itself decline be- cause of the growing probability that the company will not capture these savings in the first place. As a result, the true curve would be even flatter than shown here. 22 See, for example, J. Van Binsbergen, J. Graham, and J. Yang, “The Cost of Debt,” Journal of Finance 65, no. 6 (2010): 2089–2136.