Measuring a firm's economic profitability : a study of the measurement of a firm's economic profitability with proposals for, and evaluations of, an ex post measure, return on total capital employed (ROTCE), and an ex ante measure, a modified version of Tobin's q (modq) employing current earnings in lieu of capital employed
Despite its significance for industrial economics, utility regulation and competition policy, the measurement of the economic profitability of a firm remains a relatively underresearched area. The difference between the Accounting Rate of Return (ARR), measured on a net replacement cost or current cost basis, and a firm's estimated risk adjusted cost of capital is favoured by many economic researchers and is widely employed in utility regulation, but strong claims have been made for Tobin's q (q - the ratio of the market value of a firm's securities to the cost of replicating the firm, often identified with the net replacement cost of its net assets). Both measures have shortcomings. Davis and Kay have drawn attention to, but have failed to fully explain, a bias in ARR when firms buy in goods and services. Bias in q due to the omission of hidden capital can be significant. In this paper, economic profitability is identified with a firm's input-output ratio expressed in present value terms, and with the internal rate of return on a firm's expenditure in the accounting year, both revenue and capital. In the case of ex Post profitability, the last two measures are shown to be equivalent. Departures from the form of these ideal measures explains the biases in both ARR and q. Employing the Capital Asset Pricing Model, two alternative, operational measures of a firm's economic profitability are derived from the ideal measures with a view to eliminating the biases in q and ARR. The ex post measure is called here the Return on Total Capital Employed (ROTCE) and the ex ante measure is called here modified Tobin's q (modq). ROTCE is appraised using data from a simple corporate model. modq is appraised using data extracted from the accounts of companies comprising the Buildings Materials and Food Manufacturing sectors of the FTA All Share Index. In this study, I/modq and 1/q are shown to be significantly correlated at the 95t confidence level, and some 45k of the difference between them can be associated with taxation effects. Associating market power with the product of Beta and the Return on Sales, 1/modq is found to be significantly related at the 95t confidence level with market power and wages deflated by market value.