With trillions of dollars in cash sitting on their balance sheets, corporations have never had so much money. How executives choose to invest that massive amount of capital will drive corporate strategies and determine their companies’ competitiveness for the next decade and beyond. And in the short term, today’s capital budgeting decisions will influence the developed world’s chronic unemployment situation and tepid economic recovery.

Although investment opportunities vary dramatically across companies and industries, one would expect the process of evaluating financial returns on investments to be fairly uniform. After all, business schools teach more or less the same evaluation techniques. It’s no surprise, then, that in a survey conducted by the Association for Financial Professionals (AFP), 80% of more than 300 respondents—and 90% of those with over $1 billion in revenues—use discounted cash-flow analyses. Such analyses rely on free-cash-flow projections to estimate the value of an investment to a firm, discounted by the cost of capital (defined as the weighted average of the costs of debt and equity). To estimate their cost of equity, about 90% of the respondents use the capital asset pricing model (CAPM), which quantifies the return required by an investment on the basis of the associated risk.

But that is where the consensus ends. The AFP asked its global membership, comprising about 15,000 top financial officers, what assumptions they use in their financial models to quantify investment opportunities. Remarkably, no question received the same answer from a majority of the more than 300 respondents, 79% of whom are in the U.S. or Canada. (See the exhibit “Dangerous Assumptions.”)

That’s a big problem, because assumptions about the costs of equity and debt, overall and for individual projects, profoundly affect both the type and the value of the investments a company makes. Expectations about returns determine not only what projects managers will and will not invest in, but also whether the company succeeds financially.

Say, for instance, an investment of $20 million in a new project promises to produce positive annual cash flows of $3.25 million for 10 years. If the cost of capital is 10%, the net present value of the project (the value of the future cash flows discounted at that 10%, minus the $20 million investment) is essentially break-even—in effect, a coin-toss decision. If the company has underestimated its capital cost by 100 basis points (1%) and assumes a capital cost of 9%, the project shows a net present value of nearly $1 million—a flashing green light. But if the company assumes that its capital cost is 1% higher than it actually is, the same project shows a loss of nearly $1 million and is likely to be cast aside.

Nearly half the respondents to the AFP survey admitted that the discount rate they use is likely to be at least 1% above or below the company’s true rate, suggesting that a lot of desirable investments are being passed up and that economically questionable projects are being funded. It’s impossible to determine the precise effect of these miscalculations, but the magnitude starts to become clear if you look at how companies typically respond when their cost of capital drops by 1%. Using certain inputs from the Federal Reserve Board and our own calculations, we estimate that a 1% drop in the cost of capital leads U.S. companies to increase their investments by about $150 billion over three years. That’s obviously consequential, particularly in the current economic environment.

Let’s look at more of the AFP survey’s findings, which reveal that most companies’ assumed capital costs are off by a lot more than 1%.

The miscalculations begin with the forecast periods. Of the AFP survey respondents, 46% estimate an investment’s cash flows over five years, 40% use either a 10- or a 15-year horizon, and the rest select a different trajectory. Some differences are to be expected, of course. A pharmaceutical company evaluates an investment in a drug over the expected life of the patent, whereas a software producer uses a much shorter time horizon for its products. In fact, the horizon used within a given company should vary according to the type of project, but we have found that companies tend to use a standard, not a project-specific, time period. In theory, the problem can be mitigated by using the appropriate terminal value: the number ascribed to cash flows beyond the forecast horizon. In practice, the inconsistencies with terminal values are much more egregious than the inconsistencies in investment time horizons, as we will discuss.

Having projected an investment’s expected cash flows, a company’s managers must next estimate a rate at which to discount them. This rate is based on the company’s cost of capital, which is the weighted average of the company’s cost of debt and its cost of equity.

Estimating the cost of debt should be a no-brainer. But when survey participants were asked what benchmark they used to determine the company’s cost of debt, only 34% chose the forecasted rate on new debt issuance, regarded by most experts as the appropriate number. More respondents, 37%, said they apply the current average rate on outstanding debt, and 29% look at the average historical rate of the company’s borrowings. When the financial officers adjusted borrowing costs for taxes, the errors were compounded. Nearly two-thirds of all respondents (64%) use the company’s effective tax rate, whereas fewer than one-third (29%) use the marginal tax rate (considered the best approach by most experts), and 7% use a targeted tax rate.

This seemingly innocuous decision about what tax rate to use can have major implications for the calculated cost of capital. The median effective tax rate for companies on the S&P 500 is 22%, a full 13 percentage points below most companies’ marginal tax rate, typically near 35%. At some companies this gap is more dramatic. GE, for example, had an effective tax rate of only 7.4% in 2010. Hence, whether a company uses its marginal or effective tax rates in computing its cost of debt will greatly affect the outcome of its investment decisions. The vast majority of companies, therefore, are using the wrong cost of debt, tax rate, or both—and, thereby, the wrong debt rates for their cost-of-capital calculations.