Capital Budgeting Theory and Capital Budgeting Practice University of Texas at El Paso Pierre C. Ehe MBA The three articles by Mukherjee posit the idea that inconsistencies exist between capital budgeting theory and capital budgeting practice. Mukherjee attempts to uncover the reasons behind these inconsistencies and gain a better understanding as to why theory does not translate well into practice. He specifically researches the capital rationing decisions and leasing decisions amongst Fortune 500 companies and provides various reasons and conclusions about his findings. As has been taught and enforced throughout the MBA program, the mission and purpose of a corporation is to maximize shareholder wealth. It is the fiduciary duty of corporate agents or management to act in the shareholders interest with every business decision they make. However, Mukherjee finds that this purpose of a corporation is not a sentiment shared by many Fortune 500 companies. The mission of a company is at the top of the corporate strategy pyramid. It may be this prominent disconnect at the top that leads to a disconnect between capital budgeting theory and practice that trickles down throughout the organization. Propositions Mukherjee lays the foundation for his research by reviewing the four stage framework of the capital budgeting process. The four stages are as follows: 1) Identification of an investment opportunity, 2) Development of an initial idea into a 1
specific proposal, 3) Selection of a project, and 4) Control, including post audit, to assess forecast accuracy. In The Capital Budgeting Process: Theory and Practice, Mukherjee proposes that the gap between theory and practice stems from deficiencies in the theory itself. Specifically, Mukherjee suggests the selection stage contains the most inconsistencies between theory and practice. In Capital-Rationing Decisions of Fortune 500 Firms: A Survey, Mukherjee examines the specific component of capital rationing of the capital budgeting process. He proposes that if theory is followed, capital rationing should not exist. He states In theory, a firm should expand to the point where its marginal return is equal to its marginal cost. In other words, firms should accept all independent projects having positive NPVs, reject those with negative NPVs, and choose between mutually exclusive projects the one with the highest NPV. Such decisions would maximize the value of the firm and the wealth of its shareholders. Mukherjee assumes capital rationing is commonplace amongst corporations and seeks to uncover the reasons why it is done. Another specific component of the capital budgeting process Mukherjee examines is the lease versus buy or borrow decision. His main purpose of writing A Survey of Corporate Leasing Analysis was to expand on a previous article by O Brien and Nunnally by asking Fortune 500 companies additional questions about the leasing decision. He proposes that the financing decision can be quantified, a distribution of firms in terms of leasing methods can be obtained, and a favor of either a before or after tax cost of debt can be identified. Method and Data Analysis 2
In The Capital Budgeting Process, Mukherjee does not conduct any new research, but rather compiles many previously conducted surveys and interprets their results as a compilation. He examines over 15 different surveys conducted from 1959 1980. His main purpose is to identify how many companies participate in all four stages of the capital budgeting process. For the first stage identification theory implicitly assumes that all projects are evaluated based on economic merit. Mukherjee s survey analysis shows almost all of the companies surveyed participated in the identification phase. Furthermore, the data indicates that new project ideas are generated at lower levels or operational levels of management. This counters any theory suggesting investment ideas are generated at the top management or executive levels. At the second stage development theory implicitly assumes that all feasible ideas are developed into full blown proposals, and none is accepted or rejected for noneconomic reasons. The survey data shows that most companies screen an idea before a complete project proposal is developed, with the most common screening approach being a review by a non-specialist. Research suggests this screening phase may explain such a high acceptance rate of final project proposals, indicating the final decision makers only see the projects that have passed through other levels of management. The survey data also concludes that the prevalent measure of each project in the screening phase is a traditional cash flow approach. However, financial executives hesitate accepting the traditional cash flows as a true measure of the project s value. These hesitations may come from previous work relationships with the project s main contributor and interdepartmental politics. 3
In the third stage selection Mukherjee s compilation demonstrates that all the companies surveyed participate in the selection stage. In this stage, the data shows that almost all respondents have approved a project that was not quantifiably profitable for public and employee safety reasons and the documentation for each project may vary depending on urgency criteria. Additionally, while it is unclear who is making the final approval decisions, it is clear that finance is not responsible for budget development as one may suspect, but finance does play a role in selection control. Regardless, the data seems to indicate almost all proposals submitted for final approval have already been approved, and those that do not receive the final approval stamp are recycled and eventually accepted, signifying the final approval phase is merely a formality. Under selection Mukherjee analyzes four aspects of capital budgeting that are uniform amongst the companies surveyed. First, theory suggests the use of discounted cash flows (DCF) and net present value (NPV). The use of DCF seems to be widely accepted in practice with internal rate of return (IRR) serving as the favorite DCF technique, followed by NPV. And Mukherjee would argue that while both IRR and NPV are effective, NPV is superior because it conforms to the value additivity principle, its assumed reinvestment rate is consistent with valuation theory and uniform across projects, and it avoids multiple answers for a single project. Most firms also use non- DCF techniques as secondary analysis, techniques such as payback period and accounting rate of return. Second, all firms unambiguously consider risk when evaluating a capital budgeting project. The most popular techniques for considering risk are applying risk adjusted discount rates and sensitivity analysis. Third, internal capital rationing is the norm and is not widely viewed as being a problem, although theory 4
suggests capital rationing should not exist in efficient capital markets and requires companies to invest as long as the expected rate of return of a project is higher than its required rate. The theoretical approach to dealing with capital rationing is linear programming; however, the data shows it is infrequently used. Lastly, there is considerable agreement about using a weighted average cost of capital (WACC) as a hurdle rate. While there is variation in exactly how each company calculates their WACC, there is consensus about not only using WACC in investment decisions, but in lease purchase and bond refunding decisions as well. There is also strong consensus in the fourth stage control indicating most firms perform post audits on their projects. The most common technique used is return on investment (ROI) and research finds that a reward system often affects the success of a project. The success of a project is most often measured by return on assets, concurrence between expected and actual results, and profit and loss. In Capital Rationing Decisions Mukherjee s method of data collection was through a four page questionnaire. It was mailed to the CFOs of all 1992 Fortune 500 companies and a total of 102 surveys were used in the final sample. This sample closely corresponded to the population of the Fortune 500 companies and most of the industries were represented. Through the analysis of his data, Mukherjee found that 36% of companies do not place a limit on the capital available for investment projects and 64% of companies do operate in a capital rationing environment at least some of the time. He also found that the nature of capital rationing was most commonly imposed by management internally, 82%, and not imposed by lenders externally. Among the reasons for capital rationing, companies most often cited a lack of trust in project 5
forecasts and large downside risk for doing so. Other reasons for capital rationing included: it s used to discourage biased cash flow forecasts; it s used when projects are non-routine or unique in nature; and it s used to preserve borrowing capacity to finance potentially high NPV projects in the near future. When determining the investment ceiling, Mukherjee s results show that for half of the respondents the initial amount of available capital is closely tied to the level of internally generated funds and only a quarter consider external sources of financing. Additionally, half of the responding companies indicated they would lower the ceiling when adherence to the original ceiling would require acceptance of low NPV projects or require external financing, and over 40% would raise the ceiling to accommodate high NPV projects. Mukherjee suggests that this willingness to lower or raise the investment ceiling insinuates that capital rationing is soft in nature. In accordance to what we have learned in class and discussed about Mr. Trump, Mukherjee finds that companies overwhelmingly select projects based on their IRRs. A plausible explanation is that percentages are much easier for management to understand and put into perspective than a dollar amount. And the last analysis performed by Mukherjee was the extent to which the respondents agreed with the following statement: Capital rationing should not exist because in efficient capital markets, funds should be available for all projects that provide greater than the required rate of return. Interestingly, almost 70% of the respondents disagree with this statement and almost 30% agree. As one might expect, of the disagreeing, 85% are companies that reported using capital rationing frequently. 6
In Leasing Analysis, Mukherjee reviews various textbook recommendations of the lease versus buy or borrow decision and he also conducts research of his own on the matter through a survey. The main findings of Mukherjee s textbook analysis are that all of the textbooks reviewed agree that the leasing decision is a financing decision (as opposed to an investment decision), agree on the formula or model used to compute the net advantage to leasing (NAL), agree on the rate to be used to discount the depreciation tax shield and lease payments, and agree that a positive NAL may make a previously rejected project acceptable. Mukherjee s method of data collection for the survey was through a six page questionnaire. It was mailed to the CFOs of 455 companies of the 1988 Fortune 500 companies and a total of 83 surveys were used in the final sample. This sample closely corresponded to the population of the Fortune 500 companies and most of the industries were represented. In accordance with the textbook review, Mukherjee found that 88% of the companies surveyed view leasing as a financing decision. Of these respondents, a preference is shown for NAL over IRR. Of the respondents, almost all of them use NAL or IRR in analyzing leases. Some of the stated advantages of NAL, in order, are NAL s ability to express the final number in dollar amounts, the ease of computing NAL, and the ability to use different discount rates in NAL. Among the companies using IRR, the top advantage stated was it is expressed in a percentage rather than a dollar amount, which seems to be more in line with what we know about percentage preferences. The survey findings indicate that almost all of the companies in the sample use the same formula to compute NAL and all employ the same rate to discount all related 7
cash flows, which is most often the after-tax borrowing cost. Additionally, for those companies that use IRR it is computed uniformly. Another important finding of Mukherjee s results was that 70% of companies do not bother computing NAL if a project has a negative NPV. Mukherjee suggests these companies ignore the feedback effect, which is inconsistent with theory, or few assets are actually linked to favorable leases. Of the firms indicating they do conduct NAL after an NPV rejection, the two stated reasons for doing so are: if the capital budget is spent, leasing may be necessary and other non-monetary benefits may exist from leasing. Mukherjee also analyzed the debt-lease relationship and found 47% of the respondents view leasing as a substitute for debt, 22% think leasing complements debt, and 31% believe one has no bearing on the other. It is interesting to note that of the companies believing leasing is complementary or independent of debt, they use NAL and IRR calculations assuming debt-lease substitution. Lastly, Mukherjee examines the reasons for leasing and future leasing activities. The survey data reports 82% of companies consider leasing to avoid the risk of obsolescence, and over half state leasing being cheaper than borrowing as a reason for leasing. It does not appear future leasing activities will be impacted by new tax laws (in 1989) and most companies will keep leasing at a minimum because of its perceived high financing expense. Conclusions In all three of his articles, Mukherjee notes the disconnect between capital budgeting theory and capital budgeting process. In The Capital Budgeting Process, Mukherjee draws conclusions based on his survey research specific disconnects, 8
including: projects are rejected or accepted for noneconomic reasons, cash flows of a project are not always calculated correctly and uniformly, IRR is the chosen acceptance standard and not NPV, payback period is widely used in practice, the cost of capital is misused, companies often change the payback period to adjust for risk, and the contradiction of using ROI with DCF encourages short run profitability and not the long term success of the project. Mukherjee suggests that these differences between theory and practice do not exist at the sole fault of companies, but rather deficiencies may exist in the theory itself. He concludes that theory does not account for organizational structure and behavior in corporate decision making, and therefore, the notion of accepting or rejecting a project for pure economic reasons falls short. Similarly, theory does not integrate corporate strategy into decision making nor does it integrate risk behavior. Corporate strategy and risk assessment are often reasons for the rejection or acceptance of a project and are clearly valid reasons for doing so despite what theory suggests. Lastly, Mukherjee concludes that theory may be too complicated to follow methodically due to unrealistic assumptions about data availability. For example, as his research found in two of these articles, many companies do not use linear programming when dealing with capital rationing because the data collection is cumbersome and management doubts the complete reliability of the data. At the end of The Capital Budgeting Process Mukherjee acknowledges that more surveys need to be conducted and their scope needs to be much broader than what has been done in the past. Mukherjee does just that with his surveys in Capital Rationing Decisions and Leasing Analysis. In Capital Rationing Decisions, Mukherjee concludes 9
that the reluctance to issue external financing serves as the primary reason for capital rationing. However, companies that never or infrequently employ capital rationing are more profitable than those that do ration. This supports the theory that a company should expand to the point where its marginal return is equal to its marginal cost, which maximizes shareholder wealth. Following this theory seems to lead to more profit; however, most companies facing capital rationing do not think rationing runs counter to the value maximizing principle. In Leasing Analysis, Mukherjee concludes that also counter to capital budgeting theory, companies do not conduct lease analyses for negative NPV projects. Mukherjee suggests the explanation for this may lie in the forecasting bias perceived by business executives. If a project has a negative NPV and a positive NAL is obtained, there is still a bias that the project will not be as profitable as one with a positive NPV. This notion may support why companies prefer buying or borrowing when making the lease versus buy/borrow decision. In sum, through Mukherjee s three papers, it is clear that companies follow the methodology of capital budgeting theory, but how management interprets the results does not follow theory. For instance, while there is unanimous agreement on how to calculate NAL, a positive NAL result is often interpreted differently. Rather than relying on theory to dictate decisions management makes, management often relies on their own intuition and experience; which, in my opinion, seems more rational than accepting all positive NPV investment projects. As Mukherjee states, Considering the capital budgeting process as part of overall company operations requires that the models consider a firm s strategic need to grow and innovate. It may not always be that 10
managers are acting selfishly to maximize their pockets, but rather they are acting strategically to maximize the companies growth and sustainability. References Mukherjee, Tarun K. and Henderson, Glenn V. The Capital Budgeting Process: Theory and Practice, Interfaces, Vol. 17, No. 2, March-April 1987, pp. 78-90. Mukherjee, Tarun K. and Hingorani, Vineeta L. Capital Rationing Decisions of Fortune 500 Firms: A Survey, Financial Practice and Education, Vol. 9, No. 1, 1999, pp 7-15. Mukherjee, Tarun K., A Survey of Corporate Leasing Analysis, Financial Management, Autumn (Fall) 1991, pp. 96-107. 11