Chapter11: CALCULATING THE COST OF CAPITAL
11.2 Firm WACC versus Project WACC
The answer is: It depends. If a new project is similar enough to existing projects, then yes, the firms WACC can be used as the projects WACC, too. But say that a firm is contemplating undertaking a significantly different project—one far different from any project that the firm is already engaged in. Then we cannot expect the firms overall WACC to appropriately measure the new projects cost of capital. Let your intuition work on this: If the new project is riskier than the firms existing projects, then it should expect to be “charged” a higher cost of capital. If its safer, then the firm should assign the new project a lower cost of capital.p. 366
For example, consider a U.S. firm—lets call it GassUp—that currently owns a chain of gas stations. Firm management is considering a new project: opening up a series of gourmet coffee shops. Given the demand for upscale coffee in the United States, as well as the historically volatile oil markets, its probably difficult to say exactly whether the coffee shops will be more or less risky than gas stations. We can probably say, though, that the two enterprises will face different risks. For example, if the coffee shops are located within the busiest and most stable gas stations—say the ones that lie along freeways—then the firm faces remodeling of existing buildings rather than starting from scratch. Locating the coffee shops within existing structures would likely mean that the risks will be lower than building new stations or new coffee shops.
So, does this mean that GassUp should calculate a heterogeneous WACC for each new project using purely project-specific numbers? Well, not exactly. As well discuss below, some inputs to WACC should be project-specific, but others should be consistent with the firmwide values used in calculating a firmwide WACC.
Project Cost Numbers that We Should Take from the Firm
Its tempting to argue that all component inputs for a project-specific WACC should be based on the specific project attributes, but if we created all project-specific numbers, we would be ignoring a fundamental issue pertaining to all bond and preferred stock indenture contracts. Both bonds and preferred stocks create claims on the firm, not of any particular group of projects within that firm. Furthermore, debt claims are superior to those of common stockholders. Thus, if the new project does significantly increase the firms overall risk, the increased risk will be borne disproportionately by common stockholders. Debt holders and preferred stockholders will likely face minimal impact on the risk and return that their investments give them, no matter what new project the firm undertakes—even if those claimants own bonds or preferred shares that the firm issued to fund the new project.
For example, suppose GassUp decides to build entirely separate facilities for its coffee shops, which theyll name “Bottoms Up.” They partially finance their expansion into coffee shops with debt, and the project turns out to be more like “Bottoms Down”—far less successful than the firm had hoped. Though this would be an unfortunate turn of events for GassUps shareholders, the firms creditors would likely still collect their usual interest and dividend payments from GassUps gross revenues from gas station operations.
Creditors understand that their repayment will likely come from continuing operations and take current cash flows into account when a firm comes seeking funds. For example, if a small firm approaches a bank for a loan to finance an expansion, the bank will normally spend more time analyzing current cash flows to determine the probability that they will recoup their loan than it will analyzing the potential new cash flows from the proposed expansion.
Note that this situation holds true only as long as the “new” projects represent fairly small investments compared to ongoing operations. As new projects become large relative to ongoing cash-flow producing activities, creditors will have to examine the likelihood of being repaid from the new projects much more closely. New projects, however great their potential, inherently carry more risk than do established current operations. Changes in the proportion of new projects relative to ongoing operations will thus translate into increased risk for the creditor, who will ask for a higher rate of return to offset the risk.
Since most firms tend to grow incrementally, we will assume (unless otherwise indicated) that were examining situations in which the number of new projects is small relative to ongoing operations. We can therefore assume that using the firms existing WACC for both debt and preferred stock is appropriate.
Project Cost Numbers that We Should Not Take from the Firm: The Pure Play Approach
Since weve decided not to adjust the firmwide costs of debt or preferred stock for the risk of a project, where should we account for the new project risk brought to the firm overall? As with several other questions associated with risk- and profit-sharing that well discuss in Chapter 14, the answer lies with the residual claimant: equity.
The firms risk will obviously change when it takes on a project thats noticeably different from its existing lines of business. Since debt holders and preferred stockholders wont notice this change in risk, the firm spreads the risks that it takes with new projects by transferring most of the risk to the common stockholders.
In response to such a change in the firms risk profile, the stockholders will adjust their required rate of return to adjust for the new risk level. Absent any alteration to the firms capital structure,3 changes in the firms risk profile are due to differences in the firms business risksThe risk of a project arising from the line of business it is in; the variability of a firms or divisions cash flows. based on the mix of the new and existing product lines. The firms beta will reflect those differences in each product line.p. 367
Obviously, no proposed new project will have a history of previous returns. Without such returns data, neither analysts nor investors can calculate a project- specific beta. But to the extent that we can find other firms engaged in the proposed new line of business, we can use their betas as proxies to estimate the projects risk. Ideally, the other firms would be engaged only in the proposed new line of business; such monothemed firms are usually referred to as pure plays, with this term also in turn being applied to this approach to estimating a projects beta.
An average of n such proxy betasThe beta (a measure of the riskiness) of a firm in a similar line of business as a proposed new project. will give us a fairly accurate estimate of what the new projects beta will be.4
This average will be an estimate, in the strictest statistical sense of the word. You might recall from your statistics classes that we will need to be careful to get as large a sample as possible if we want to get as much statistical power for our estimate as possible. Ideally, we would like to find at least three or four pure-play proxies to ensure that we have a large enough sample size to safely make meaningful inferences. In reality, however, two (or even one) proxies might represent a suitable sample if their business line resembles the proposed new project closely enough. In particular, we may want to use betas from industry front-runners, and rely less on betas of any firms that the company really doesnt want to emulate.
What shall we do if we cant find any pure-play proxies? Well, in that case, we may want to use firms that, while not solely in the same business as the proposed projects venture, have a sizable proportion of revenues from that line. We may then be able to “back out” the impact of their other lines of business from their firms beta to leave us with a good enough estimate of what the new projects beta might be.
We should also ensure that we use weights based on the projects sources of capital, and not necessarily the firms capital structure. If the new project is going to use more or less debt than the firms existing projects do, then the risk- and reward-sharing are going to vary across the different types of capital (as discussed in Chapter 14), and we will want to recognize this in our WACC computation.
Finally, we need to consider the appropriate corporate tax rate to use in calculating the WACC for a project. That marginal corporate tax rate will be the average marginal tax rate to which the projects cash flows will be subject. Referring back to our previous example about the appropriate marginal average tax rate to use in computing a firms WACC, assume that the same firm with $400,000 of EBIT from current operations is considering a new project that will increase EBIT by $200,000. Since this $200,000 increase will keep the firms marginal tax in the fifth bracket of Table 11.1, the appropriate tax rate to compute the projects WACC will simply be 34 percent.p. 368
To summarize, the component costs and weights to compute a project-specific WACC should be as shown below in equation (11-7), with the source of each part indicated by the appropriate subscript:
3In reality, new projects are often financed with different proportions of equity, debt, and preferred stock than were used to fund the firms existing operations. As well discuss in Chapter 14, such a change in capital structure will result in a change in financial riskThe risk of a project to equity holders stemming from the use of debt in the financial structure of the firm; refers to the issue of how a firm decides to distribute the business risk between debt and equity holders., with increased leverage magnifying βE.
4As well also discuss in Chapter 14, we will be able to take a straight average of the proxy firms betas as the estimate of our beta only if the capital structures of all the proxies are identical to each other and to that of our proposed new project. If not, we will need to adjust the proxies estimated betas for differences in capital structures before averaging them. Then we will need to readjust the average beta for our projects capital structure before using the estimate.