Project valuation is a notoriously imprecise art. “Two of the ubiquitous tools in capital budgeting are a wing and a prayer,” economists Eugene Fama and Kenneth French once famously observed. But by using a new method for calculating the cost of equity capital, financial managers can put valuation decisions on more solid ground.

That's the central claim of a June 2001 article in the Journal of Accounting Research. As co-author Charles M.C. Lee explains, many common methods for calculating the cost of equity, including the venerable capital-asset-pricing model (CAPM), share a significant flaw. The underlying theory focuses on expected returns, an imprecise measure. CAPM, for example, predicts the cost of equity as a function of beta — the relationship between the return that investors expect from a company and the return they expect from the market as a whole.

To gauge expected returns, companies typically rely on historical or realized returns. In principle, that shouldn't matter. If markets are efficient, realized returns over a sufficiently long period should approximate expected returns. But in practice, realized returns turn out to be “a pretty lousy proxy,” Lee says. “You wouldn't have invested in many stocks in the past few years, if you used returns over the last 20, 30 or 40 years to calculate the cost of equity.”

Consequently, Lee and his colleagues suggest that managers look to the future, not the past, for guidance — by using the implied cost of equity to make investment decisions. The implied cost of equity is the internal rate of return that sets a company's market capitalization equal to the present value of all future cash flows to common shareholders. Less technically, i 'st the discount rate that investors implicitly are using to value the business. It's not a new concept. Analysts have long used discounted-cash-flow models. But the authors believe that their study represents the first systematic effort to develop a model that predicts the implied cost of equity.

They begin by calculating the implied cost of equity for more than 1,000 companies between 1979 and 1995. Their analysis reveals that average discount rates vary significantly by industry — a pattern obscured in studies that use realized returns. What's more, which industry a company is in turns out to play a leading role in predicting the company's implied cost of equity. Industry membership and three other variables — a company's book-to-market ratio, its long-term growth expectations and the amount of dispersion in analysts' forecasts of future earnings — accounted for 60% of the variation in implied-risk premiums among businesses.

However, Lee cautions against attaching too much importance to the precise magnitudes the paper reports for each of the effects. Instead, he suggests that financial managers use the results heuristically, starting with the average excess return that investors expect in an industry. (That's the return in excess of the expected market return.) He recommends capping that figure to be no higher than 3% or lower than –3%, adding it to the market-risk premium, and adjusting it up or down, depending on how a company compares with its industry peers. A high book-to-market ratio, for example, might add as much as 1.5% to a company's implied cost of equity, whereas below-average growth expectations and dispersion in forecasts could lower it by up to 0.5%.

Like any valuation method, the procedure has potential weaknesses. Perhaps most important, the initial estimates of the implied cost of equity leave room for error. According to Larry Kolbe, a principal at Brattle Group, an economic consulting firm in Cambridge, Massachusetts, the discounted-cash-flow method on which those estimates are based works best with stable companies. Like other equity-valuation methods, the technique is least precise for companies that have a lot of their value in growth options — a category that includes most high-tech enterprises.

Lee and his co-authors acknowledge those limitations. Nonetheless, they feel strongly that traditional CAPM-based methods have failed to develop reliable cost-of-capital estimates. They believe that managers will make better decisions if they learn to estimate the cost of capital without relying on current market prices or realized returns. The study, they say, is a first step toward that goal.

Lee, a professor of accounting and finance and head of the Parker Center for Investment Research at Cornell University in Ithaca, New York, wrote “Toward an Implied Cost of Capital” with William R. Gebhardt, director of weather derivatives at Axia Energy Europe in London, a trader and marketer of energy, and Bhaskaran Swaminathan, an associate professor of finance at Cornell.