Taxation and the Cost of Capital: The "Old" View, the "New" View and Another View

Abstract

IntroductionEconomists agree that the cost of capital is an important analytical tool for predicting a country's intersectoral distortions, its growth performance, or its attractiveness for international capital.However, there is no consensus as to how the tax influence on the cost of capital should be measured.The cost of capital is defined as the minimum pre-tax rate of return an investment project must earn to be profitable.The controversy among tax economists is primarily concerned with the question of how the required rate of return is affected by personal and corporate income taxation, and the double taxation of dividends has received particular attention.While it is obvious that the double taxation creates a substantial tax burden for corporations, there is no agreement as to how much of the burden fails on marginal investment projects.The "old view" is that the total tax burden falls entirely on marginal investment projects and therefore implies a high cost of capital, far above the market rate of interest.By way of contrast, the so-called "new view", which, in fact, is no longer that new, is that only the tax burden on retained earnings matters.The burden of the double taxation of dividends is seen to fall largely on intramarginal investment projects and is believed to not affect the cost of capital overly.This note is an exposition and critical review of some of the arguments exchanged between the members of the two schools.It presents the basic theories, discusses the role of financial optimization, comments on immature firms, and includes an analysis of the role of share repurchases which has recently led to a revival of the "old" view among North American economists.It also presents the idea of a political Miller equilibrium.The crucial reason for the different views on the way taxes affect the cost of capital is that authors make different assumptions about the firms' financial decisions, sometimes without explicitly mentioning them.The holders of the new view have pointed to the importance of these assumptions and they have emphasized that new share issues, debt, and, in particular, retained profits should be distinguished as alternative sources of finance.Three different cost of capital expressions are typically used by them depending on which source of finance is assumed.Equally important, however, is the distinction between alternative uses for profit.These uses are not only dividend and interest payments, as is usually assumed, but also profit retentions and share repurchases.The specification of the use for profits is as essential for the calculation of the cost of capital as the specification of the source of finance.Only when it is dear where an additional dollar used for investment comes from and where its returns are going, is it possible to calculate the tax burden on marginal investment and to find out which minimum pre-tax return is required to make this investment profitable.Distinguishing retained earnings, debt and new share issues as marginal sources of finance and retentions, dividends, interest income, and share repurchases as marginal uses for profit in principle gives up to 12 different expressions for the cost of capital.However, since interest payments are the use for profits only when the source of finance was debt, the number reduces to 7. The possibilities are depicted in Table 1.As will be shown, not all of them will be equally rdevant to an optimizing firm: however, they will turn out to be a useful guide in the course of this paper. The Old ViewIn the good old days economists distinguished just two financial alternatives: equity and debt.With equity finance, the conceptual experiment for determining the cost of capital was that shareholders inject an additional dollar into their firm by purchasing r newly issued shares and compare the returns in the form of dividends with the returns they could have received by investing their money in bonds.Let iv be the annual

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 94 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.