Introduction
Randolph Corporation is the producer of abrasive products, industrial grinders and sharpeners and coated ceramics for use in aerospace. The company is furthermore divided into four divisions, namely the Home Products Division, the Equipment Manufacturing Division, Ceramic Coatings Division and the Real Estate Division. This strategy has worked well until recently when its stock market performance went down compared to its competitors. According to Randolph’s financial vice president, Gianneti, the main problem with the company is the way risk is incorporated into the financial planning process. In this report, several problems of Randolph Corporation’s capital budgeting process are discussed and solutions suggested.
Divisional Hurdle Rates
The current problem of overinvesting into value destroying projects of the riskier divisions and underinvestment into the less risky ones calls for differential costs of capital for the four divisions. To arrive at the individual hurdle rates a cost of debt of 11% and a tax rate of 40% were assumed to calculate the weighted average cost of capital. These hurdle rates, based on the assumption that a 45% debt ratio is used for all divisions, are reported in Appendix 1. The introduction of these hurdle rates leads to evaluating projects by their respective divisional WACC instead of the corporate WACC when calculating NPV, IRR and MIRR, and consequently different investment decisions for some projects. The capital budgeting process at the moment neither includes these divisional hurdle rates, nor a process of setting a higher cost of capital for riskier projects than for less risky projects within a division. To account for the difference in risk, the proposition was made to multiply a division’s beta with 0.9, 1 and 1.2 for projects of low, average and high risk, respectively. The resulting hurdle rates for the three categories are shown in Appendix 2.
To date, there is no scientific basis for a systematic approach of adjusting projects’ cost of capital according to their risk. Since it would be too costly and sometimes impossible to accurately estimate the exact risk of every cash flow, one has to find a simpler albeit less accurate approach. The somewhat arbitrary method of grouping all projects into three risk categories and multiplying their respective division cost of capital with a factor is an acceptable compromise between accuracy and practicality. It is important to note that also no other choice of factors, besides 0.9, 1 and 1.2 will exactly represent the risk of all projects within the three risk groups. However, if the factors are well chosen, they resemble the average risk of projects in these categories as closely as possible to lead to good investment decisions.
Shifts in the Corporate Beta and the Cost of Capital
According to Kravitz’ analysis, the company has increasingly invested into the Ceramic Coatings division because it used the corporate hurdle rate to calculate the NPV and IRR of projects. If this trend continues, the Coatings division will receive increasing funds and grow at a faster pace than the average division of the Randolph Corporation. Over time, the current weights of 10, 10, 30 and 50 per cent would change towards relatively more weight on the Ceramic, and less on the other departments. Since the weighted average of the division betas form the corporate beta and Ceramic Coatings has a higher-than-average beta, the corporate beta will increase. Looking at the CAPM formula, one can see that a higher beta will lead to a higher cost of capital, ceteris paribus. To illustrate, if the Ceramic Coatings division grew by 50% while the other divisions remain constant, its percentage of corporate assets would rise from 10 to 14.3 per cent, increasing the corporate beta from 0.795 to 0.8023 (the corporate beta was falsely reported as 0.895 due to a false Ceramic Coatings product of 0.195 instead of 0.095). Although this change might look insignificant, over the long term it will have considerable effects on the beta and therefore the cost of capital.
A similar argument as the previous one can be made if a division invests more money than usual into high-risk projects. For instance, if the Equipment Manufacturing division, which has the highest beta of all divisions, invests only into high risk projects over the next years, their division beta would converge from 1.05 towards 1.26, thereby increasing the corporate beta and the cost of capital. If Randolph Corporation made public announcements about their increasing risk appetite, this information would be included in their reported beta in a very short time. If they decided not to announce it, it might take a longer time to be incorporated in broker’s beta estimates for two reasons. Firstly, if firms undertake new projects, it usually takes at least several months until positive cash flows are realized, even if the project is going very well. Therefore, the absence of expected cash flows will only be noticed at a later point in time. Secondly, as will be discussed later in this report, betas are usually based on many periods, so a change in a company’s beta might be concealed by previous periods. To summarize, in the first described case the weight of a high-beta division increased, in the second case the beta itself increased while the relative weight stayed constant. Both scenarios have the effect of increasing the corporate beta, and therefore the corporate cost of capital. Capital structure
Currently, the suggestion is to establish a target capital structure of 55% equity and 45% debt. The real estate division leader, however, uttered her concerns about the division’s competitiveness under this practice. She argues that if the division’s capital structure does not imitate that of stand-alone corporations in the industry, Randolph will be at a disadvantage. Her approach is that each department should be treated as wholly owned subsidiary, establishing separate debt ratios and also issue their own debt. According to Kolasinski 2006, in this case debt holders of the subsidiary have a senior claim for the subsidiary compared to parent-company debt holders. Therefore, if every department is considered separately the debt ratios will and should differ, because a division should be able to capitalize on the tax shield of its own debt. Also, there is higher probability of use of subsidiary debt if the parent has a junk rating or if the division operates in a less volatile industry than the parent. On the other hand, if the company backs the subsidiary debt, the industry standard might not be appropriate, since the debt is less risky than stand-alone debt. In this case, the subsidiary debt holders also have a claim to the parent company’s assets. Additionally, parent backed subsidiary debt occurs especially often if there are better investment opportunities in a division than in the company.
Nevertheless, it might be the case that the lower rates of backed debt lead to the inappropriately low hurdle rates that Continental foods used for its division. This argument is based on the assumption that if a certain division has a lower cost of debt than stand alone industry competitors, the other departments pay for the lower cost of debt of one division by taking over default risk. Hence, a division should not be able to take advantage of the full tax shield if the parent backs the loan, and this needs to be accounted for by adjusting the capital structure. Finally, when considering debt issued at the corporate level and distributed through the headquarters, it seems intuitive to have the same capital structure throughout the company’s divisions, to achieve an equal distribution of tax protection. Moreover, when considering the cost of debt, no division should be at a disadvantage for being part of a corporation with many divisions. Therefore, a department must not be allowed to capitalize on an excessive tax shield by using a greater debt ratio than other parts of the company, while it does not have the risk inherent in high leverage since the company takes on the debt. This could yield a great disadvantage for the other divisions.
Market Beta Estimates
For financial structure decisions, the values of beta coefficients are important in many situations, including the calculation of the WACC and divisional adjustments. Unfortunately, multiple factors have a negative impact on the validity of beta estimates used in the analysis. As an initial concern, beta is calculated using historical returns, but it is the basis for estimating marginal cost of equity, which incorporates future expectations about the firm’s volatility. Moreover, when measuring historical betas, time is important in two ways. First beta changes if returns on different holding periods are used; secondly, it is sensitive to the number of years of data. Furthermore it becomes increasingly difficult to estimate the beta in less developed economies, due to the less consistent data coverage. Therefore, market coefficient analysis of multinational companies, which raise equity across borders, is affected in a similar manner. Also, the measurement of beta becomes virtually impossible if one considers that the real market portfolio cannot be observed, since it would include returns on every asset. In calculations, the market is always approximated by the use of indexes, like S&P 500. Nevertheless, it is possible to arrive at an appropriate estimate using historical analysis and subjective adjustments when necessary. This reflection highlights the fact that the real beta cannot be measured but only estimated, and that estimates differ, depending on the employed techniques.
Assuming that Kravitz was not comfortable with beta analysis is not an abstract thought. Estimating beta is a highly complex matter, but even if calculated correctly it might be a faulty measure. Modern Portfolio theory states that an asset with a higher beta is riskier. This however does not take into consideration that an asset can amplify positive returns of the market more than it amplifies negative periods, and vice versa, which is not reflected in its beta. It might be reasonable to focus attention on the movements on the negative side of returns, arguing that investors perceive risk much more in terms of losses than of gains. As Estrada J. (2006) suggests, the concept of semideviation can be used as a substitute for the market beta, and provide another way of estimating the risk of a company. Assuming the CAPM equation, it is possible to use semideviations as a proxy for the specific risk of a company and replace beta withΣBi / ΣBM. This equation measures the downside volatility of the corporation compared to the downside volatility of the market with respect to a benchmark B. For example, the volatility of a stock below the risk free rate can be estimated to determine whether the stock suits one’s risk appetite. Using this model, the cost of equity is affected by the downside risk of a company, not by its estimate of as reported by the beta.
Incentive scheme
The current incentive scheme is faulty in various ways, and could lead to serious trouble in the future behaviour of management. First, as Johnson 2012 states, ROE is not an optimal measure since equity only accounts for a small portion of the balance sheet, and the employed leverage should also be accounted for. Obviously, shareholders expect an appropriate return on their investments, but the incentive scheme guides the actions of managers. It needs to take into account how risk taking is affected, as well as the fact that all metrics can be cheated. Consistently, a focus on earnings per share growth can foster aggressive accounting practices, or lead to the deferral of crucial investments. In fact, the infamous decisions made by ENRON’s finance department were based on EPS incentives. The current incentive plan at Randolph Corporation could foster a defensive stance against Kravitz’s recommendations for the divisional hurdle rates. The ceramic and equipment departments now have to carry the risk inherent in their operations via a higher hurdle rate, and are therefore expected to argue against the recommendations. In particular, fewer projects will pass their new, higher hurdle rate leading to less growth, for which management was compensated till now.
The opposite holds for the managers of the real estate and home product departments, who will reasonably back the recommendations since they allow the divisional management to undertake more projects leading to more growth and thereby higher remuneration. From the standpoint of maximizing Randolph’s stock price, the reaction of the ceramic and equipment departments would hurt the long term stock price of the company, because these departments are currently using a low hurdle rate leading to value destroying investments. Overall, the changes recommended by Kravitz’ should be implemented on their own, and compensation plans need to be adjusted. The current incentive scheme leads managers to invest excessively into projects which are value destroying because investments lead to growth, and therefore additional pay. On the other hand, ROE only increases if a company either undertakes good investments, or particularly risky ones. Thus, it is important to develop a compensation scheme that also rewards projects with lower risk and consequently lower return, because it lowers the company’s cost of capital.
Concluding the reflection, it is suggested to implement divisional hurdle rates for the calculation of costs of capital and to conduct further research if divisional debt ratios should also be constructed. Moreover, the compensation plan needs to be adjusted, to yield ideal allocation of capital and to minimize value destruction.
