ESRC/CEPR
Corporate Finance, Investment and Taxation

The pace of change in financial markets throughout the world has focussed increased attention on corporate financial structure and its relation to financial markets and the tax system. A very successful conference focusing on this subject took place from September 2 to September 6 at Wadham College, Oxford, under the auspices of the Economic and Social Research Council and CEPR. The conference was organized by Jeremy Edwards (St. John's College, Cambridge and CEPR), Julian Franks (London Business School), Colin Mayer (St Anne's College, Oxford, and CEPR), and Stephen Schaefer (London Business School and CEPR). Financial support for the conference was provided by the Economic and Social Research Council and by Drexel Burnham Lambert.

The conference brought together an international group of about forty-five scholars working in the area of corporate finance, investment and taxation. A major objective was to attract participants with varying perspectives, such as financial economists and public policy specialists. The papers chosen reflected these varied perspectives, and the discussions which followed the presentation of the papers clearly illustrated the different research paradigms currently in use. Approximately a quarter of the authors of the papers presented and of the participants were young scholars who had recently entered the field; their participation was also an objective stressed by the organisers and the ESRC.

The first paper, by Alan Auerbach (Pennysylvania), was entitled 'Tax Reform, Investment and the Value of the Firm'. It provided an empirical application to US data of Abel's work, which showed how temporary and anticipated changes in the tax code would affect short-run investment behaviour and the value of 'marginal q', which represents the net cost of new capital goods. Auerbach used simulations to calculate effective tax rates and the value of q for two types of fixed investment (structures and equipment) for the period 1953-1990, assuming the Bradley-Gephardt 'Fair Tax Plan' were implemented in 1985. Auerbach's calculations showed effective tax rates on equipment falling to about zero in 1981, as a result of the introduction of the Accelerated Cost Recovery System (ACRS). With the reduction in inflation in 1982, effective tax rates fell even further, to about -20% on equipment (although rates on structures remained positive). Values of q have also declined over time, as the distinction made by the tax system between old and new capital has widened. Average q throughout the 1950s and until 1962 was about 1. It then fell to 0.8 by 1981 and has since stayed there. Implementation of the Bradley-Gephart proposals would cause it to rise again, as the tax distinction between old and new capital was reduced. Auerbach also addressed the issue of how changes in the tax code create windfall gains or losses on old equipment, as well as how such changes affect incentives to invest in new equipment. The Bradley-Gephardt proposals, for example, seem to provide windfall gains to old capital and reduced incentives to make new investments!
In the second paper Jim Poterba (MIT) asked 'How Burdensome Are Capital Gains Taxes?' It is often claimed that the effective tax rate on capital gains is low, and a rate of between 4 and 5% is often quoted. There are two straightforward ways of avoiding such taxes, it is argued, by postponing realization or by offsetting gains by realizing losses. Poterba took issue with this argument. Evidence from the tax status of individuals in the US for 1982 indicated that a substantial proportion of investors who pay tax choose to realize their capital gains and do not appear to make any effort to realize losses in order to offset these gains. Yet 11% of investors not only realized losses in this way but also claimed the maximum allowable loss offset. Clearly, at least some investors avoid capital gains tax through the loss-offset provisions. Another group also avoids capital gains tax, but by the simple device of not reporting them to the IRS! Data provided by brokerage houses and the IRS suggest that a much higher proportion of transactions recorded by brokerage houses results in gains for investors, compared with transactions recorded by the IRS. For trades lasting between six months and a year, for example, the brokerage firm data suggest that nearly 60% of all transactions yield gains, while the IRS tabulations suggest just over 30%. Poterba may be right in saying that non- reporting is more popular than death as a means of avoiding capital gains taxes. An important puzzle raised by Poterba remains: the evident failure of so many investors to minimize their tax liability. The finance literature often refers to the dividend puzzle; Poterba provided a capital gains puzzle.

Julian Alworth (Bank of International Settlements) presented a paper on 'A Cost of Capital Approach to the Taxation of Foreign Direct Investment Income'. Using the approach pioneered by King and Fullerton, he compared different types of double taxation agreements, as well as the integration of corporate income taxation with that of dividends. He then used this analytical framework to examine how alternative systems of corporate taxation and double taxation relief affect the financial policy of multinationals through, for example, their choice of dividends versus new equity issues. Alworth's results suggest that attempts to raise or lower corporate taxes in an individual country, intended to affect the incentive to invest, may be frustrated unless they take account of the residence of the parent company, the legal form of the affiliate, and the type of financing involved.

Most models of international capital market equilibrium suggest that an investor's optimal portfolio should consist of a world- wide spread of risky assets (the 'world market' portfolio), regardless of the particular currency in which the investor ultimately wishes to consume his income. In practice, however, investors' portfolios are heavily weighted in favour of domestic assets. In the final paper of the conference's first day, entitled 'Costs to Crossborder Investment and International Equity Market Equilibrium', Ian Cooper and Evi Kaplanis (London Business School) attempted to provide a theoretical rationale for such observed portfolio behaviour.
The authors developed an international Capital Asset Pricing Model (CAPM) which incorporated barriers to crossborder investment. Such barriers consist of a mixture of measurable costs, such as witholding taxes, and intangibles such as the extra risk of expropriation and the information gathering costs suffered by foreign investors. How can we estimate the level of these intangible costs? Suppose that we assume investors hold portfolios which are optimal given the returns on assets and the quantifiable and unquantifiable barriers to crossborder investment. Then given data on investors' portfolios and asset returns, we can deduce the level of (implicit) costs faced by investors, i.e. the 'shadow prices' associated with the crossborder barriers. Using data from 14 countries on holdings of domestic and foreign securities, Cooper and Kaplanis estimated the costs to investors of crossborder investment and compared these implied costs with the explicit and readily quantifiable costs of barriers such as withholding taxes and safe custody fees. The implicit costs estimated by the model appear to be somewhat high when compared with the quantifiable costs. Investors in the UK and Germany appear to bear costs of between 5 and 8%; these seem high compared with actual costs which can be identified, although the model's estimate of about 2.2% for US and Canadian investors seems about right. The model developed by Cooper and Kaplanis is useful not only in international portfolio management, but also in corporate finance, when estimating costs of capital for risky foreign investments. For example, the analysis suggests that the cost of capital for a UK investor investing in the US is 3.4% less than that for a US investor investing in the same asset.

The Capital Asset Pricing Model (CAPM) acquired popularity in the 1960s and 1970s as an explanation of financial asset prices. Assets are assumed to carry uncertain returns, while investors are risk-averse utility maximizers; the theory suggests that the expected return on an asset should be positively related to its 'beta', a measure of non-diversifiable or 'market' risk. Tests of the CAPM have necessarily involved the use of a proxy variable to represent the return on the market portfolio of risky assets. In 1977 Roll first drew attention to the difficulties involved in testing the validity of the CAPM using such proxies. It is possible, he noted, that the theory might be rejected though true because the proxy is inefficient, or that the CAPM is false even though the test using the proxy supports it; conclusions drawn from tests of the CAPM are therefore ambiguous.

On the second day of the conference Jay Shanken (University of California, Berkeley), in his paper 'Proxies and Asset Pricing Relations: Living With the Roll Critique' analyzed the conditions under which the inefficiency of the market portfolio might be correctly inferred, even though a proxy has been used because the behaviour of only a subset of the market's assets can be observed. Shanken's procedure requires that the investigator specify a lower bound on the adequacy of the proxy employed; more precisely, a lower bound on the proportion of the variation in the full range of market assets which is accounted for by the proxy. Given this assumed lower bound, inferences can be made about the validity of an equilibrium pricing model, even though the full market portfolio is unobservable. Shanken's empirical results suggest that the CAPM can be rejected, conditional on the assumption that the correlation between the proxy and the true market portfolio is at least 0.7.

The second paper of the day, entitled 'Portfolio Choice in Research and Development' by Sudipto Bhattacharya (University of California, Berkeley) and Dilip Mookherjee (Stanford) addressed the question whether the 'winner-take-all' feature associated with an R & D patent system generates significant distortions. Their model focuses on how different knowledge bases affect the riskiness and correlation of R & D strategies, under different assumptions concerning private and public access to these knowledge bases. Bhattacharya and Mookherjee found a large class of situations where the winner-take-all system does lead to socially optimal outcomes. Exceptions occur when the knowledge base is not commonly available to competitors and where the strategies of competing firms have significant effects on the riskiness of their performance. Even this may not be inconsistent with a social optimum if the parties concerned are not too risk-averse and the distributional characteristics of different knowledge bases are well-behaved.

The efficient markets hypothesis forms the basis of much research in financial economics, and indeed has been termed 'the best established empirical fact in economics'. In the third paper of the day Larry Summers (Harvard) asked 'Do We Really Know that Financial Markets are Efficient?', thereby provoking the most heated discussion of the week. Summers argued that belief in the efficient markets hypothesis was a shared act of faith, with little in the way of theoretical or empirical support. In markets with long horizons, such as the equity market, the market valuation of an asset can differ substantially and persistently from the present value of expected future cash flows from the asset, even though investors form their expectations rationally. Conventional statistical tests on holding period returns are not sufficiently powerful to detect this difference, and investigators may conclude incorrectly that the market is efficient.

Summers argued that tests of 'weak-form' market efficiency demonstrated this problem clearly. Such tests are based on the argument that in an efficient market, differences between asset valuations and the present value of future cash flows should display no pattern of correlation over time. Suppose the market really were inefficient, Summers argued, to such an extent that market valuations frequently differed from rational expectations of future cash flows by more than 30%. Conventional tests of 'weak-form' efficiency would require over 5000 years of monthly data in order to have even a 50% chance of rejecting the hypothesis of market efficiency (as they should, given Summers's assumption). Why then are such persistent inefficiencies, if present, not traded away by speculators? The answer, according to Summers, is that since they do not have statistical tests more powerful than those of academics, speculators cannot detect inefficiencies either! Summers cited other well-known anomalies, such as the discounts on closed-end mutual funds, to reinforce his argument that the evidence for efficiency is weak. He concluded that we should be more 'catholic' in explaining the behaviour of speculative markets, and that more powerful tests were needed in order to detect market departures from efficiency.

There are a number of methods by which firms can raise new equity finance. New equity issues may be sold to an underwriter at a fixed price; the underwriter then resells the shares to investors at that price plus the underwriter's fixed expenses. The firm may alternatively make a rights offer to its existing shareholders. This rights offer may be uninsured, in which case the firm sets the subscription price at the time the rights offer is announced. An uninsured rights offer is less costly than an underwritten offer of new equity, but it may fail if the firm's stock price falls subsequently, and the firm may incur costs as a result. Very frequently firms issuing rights make an arrangement with an underwriter by which the underwriter agrees, for a fee, to purchase the rights issue should the stock price fall.

Why does such a large proportion of new equity finance take the form of fully underwritten new equity issues, rather than less costly rights offers to existing shareholders? The first paper on Wednesday, by Eduardo Schwartz (University of British Columbia), entitled 'Rights Versus Underwritten Offerings: An Asymmetric Information Approach' addressed this paradox. Schwartz and his co-author, Robert Heinkel (University of British Columbia) constructed a model with asymmetric information, in which managers of issuing firms know more about the quality of their firm (i.e., its future stock price) than do potential investors. The equilibrium outcome, even when expectations are rational, is one in which all three forms of financing are observed, each firm's choice of financing depending on its 'quality'. Higher-quality firms choose an insured or standby rights offer. They choose to incur the investigation fee associated with the standby offer because they know that the investigation will reveal their high quality and the 'insurance' on the rights offer will therefore carry a low price. Intermediate quality firms choose uninsured rights offers; the subscription price set by the firm in the rights offer acts as a signal of its expected stock price; higher quality firms set higher subscription prices. The lowest quality firms prefer 'pooled' sales of equity at a common offer price through underwriters who are uninformed as to the firm's quality; the offer price reflects the average quality of the firms concerned. Lower quality firms use the apparently more costly underwriting method because both the costs of signalling and the costs of underwriting in a rights offer would be even higher. In addition to suggesting why all three forms of equity finance might exist simultaneously, the model also explains why firms that make uninsured rights offers do not set arbitrarily low exercise prices to ensure the success of the offer, since the subscription price serves as a signalling device.

An enduring anomaly in financial economics is the reliance of firms on internally generated funds as their chief source of equity financing and their corresponding reluctance to issue new common stock. This behaviour is less surprising to financial practitioners, who argue that selling equity causes a firm's stock price to fall. There are no close substitutes for the shares of a particular firm, it is argued, and the increase in the supply of shares causes a decline in the firm's stock price because the demand curve for shares slopes downward. In contrast, finance theory suggests that the price of a security is determined solely by the risk and expected return associated with the security's future cash flow. New issue price effects will not occur in efficient markets unless they are based on changes in expected cash flows from the security.

Paul Asquith (Harvard Business School) presented an empirical analysis of this anomaly in a paper co-authored with David Mullins (Harvard Business School), entitled 'Equity Issues and Offering Dilution.' They examined the 'announcement day' and 'issue day' price effects of over 500 equity issues. The announcement of equity offerings reduces stock prices significantly, and Asquith and Mullins argue that this cannot be explained by changes in the capital structure associated with the equity offerings. Their regression results for industrial issues indicate that the price reduction on the day the new issue is announced is significantly related to the size of the equity offering. Although the price fall causes only a small percentage reduction in the equity value of the firm, the decrease in the value of the firm on announcement day is a substantial fraction of the proceeds of the stock issue. The results suggest that there is an average 31% loss of value of the funds raised by the rights issue, and almost 25% of primary issues produced a dilution greater than 50%. These results may explain why firms are reluctant to issue new equity, and confirm the view of many practitioners that large equity issues depress stock prices. It may be, however, as Asquith and Mullins noted, that large equity issues act as a signal and are more frequently associated with bad news than are small issues. The results also confirm a view expressed by other researchers that firms tend to issue new equity after a rise in the general level of stock prices, and perhaps more important, after a rise in price of that particular firm's stock relative to the market.

The third paper of the day, entitled 'Credit Rationing', was jointly presented by Joseph Stiglitz (Princeton University) and Andrew Weiss (Columbia University). In it they extended the analysis of their earlier paper in the American Economic Review, in which they showed that a bank's profits per dollar loaned could decrease as it raised the interest it charged borrowers or increased the amount of collateral it required. This surprising inverse relationship between the loan profitability and the interest rate charged could arise from the 'adverse selection effects' of increasing interest rates: as interest rates rise, the borrowers who are discouraged are likely to be those who intended to invest in relatively safe projects. It might also be caused by incentive effects: at higher interest rates borrowers choose riskier projects. Higher collateral requirements could also cause loan profitability to decline because of adverse selection. If borrowers are worried less by risk as their wealth increases, raising collateral requirements would discourage less wealthy borrowers and raise the average riskiness of the bank's loans. Stiglitz and Weiss found that the resulting equilibrium may be one in which there is an excess demand for loanable funds at a particular interest rate or collateral requirement and that banks do not have an incentive to raise the interest rate or collateral requirement to eliminate the excess demand; banks therefore ration credit to particular borrowers.

In the paper presented at the conference, Stiglitz and Weiss go beyond their earlier work by analysing a model in which banks can alter both interest rates and collateral requirements simultaneously; in this situation the loan contract acts as a self-selection device. If banks offer several different contracts to borrowers, Stiglitz and Weiss show that there may be credit rationing for each of the contracts; merely changing the model so as to increase the variety of contracts and the number of borrowing groups does not therefore reduce the numbers that are credit-rationed. In addition the equilibrium may be characterized by different types of borrowers choosing the same contract, a result which Stiglitz and Weiss argued accords with casual empiricism.

Textbooks include the transformation of illiquid assets into more liquid liabilities as one of the reasons for the existence of the banking system. In the final paper of the day, entitled 'Competitive Banking in a Simple Model', Luca Anderlini (Wolfson College, Cambridge) discussed theoretical models which might capture this function of a banking system. Banks serve to transform illiquid assets into liquid ones; the need for this arises in his model because depositors are unsure when they will want to withdraw their funds, and production technology is such that early disinvestment is penalised.

His paper makes use of these features in order to gain further insight into the liquidity transformation process and its consequences. Anderlini shows that the combination of an illiquid technology with consumers' 'desire for flexibility' is enough to create the need for a financial intermediary with illiquid assets and more liquid liabilities; this arises even though the model contains no uncertainty concerning the production technology. Banks are risk-pooling institutions, transforming illiquid assets into more liquid ones for their depositors, and this leads to a Pareto-superior equilibrium for the economy. It is precisely this liquidity transformation role, however, which creates the possibility of an unsatisfactory equilibrium in which there is a run on banks. If the liquidity transformation is disturbed by a bank run, the result can be worse for society than the equilibrium without a banking system! This inevitable 'instability' of the banking system seems to provide a good argument in favour of its regulation, Anderlini argues, although regulations which guarantee that the banking system will always run smoothly are difficult to devise in the context of his model.

The first paper on Thursday, presented by Oliver Hart (MIT and CEPR) was entitled 'The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration' (available as CEPR Discussion Paper No. 70). In this paper, written jointly with Sanford Grossman (Princeton University), firms' integration is analyzed in the light of the theory of 'costly' contracts. When it is too costly for one party to specify a long list of particular rights that it desires over another party's assets, it may be optimal to purchase all the rights except those specifically mentioned in the contract; this effectively gives ownership over the asset. Grossman and Hart show that such acquisitions can have harmful effects. For example, a firm may purchase its supplier, thereby removing residual rights of control from the manager of the supplying company. By doing so it can distort the manager's incentives sufficiently to make ownership harmful.

Grossman and Hart apply this model to the insurance industry, where some firms have their own sales force and other firms sell insurance through independent agents or brokers. In the first case the insurance company owns the list of customers and in the second case the agent does. In the second case the insurance company has no control or rights over the independent agent's customer list; this is the essential difference between the two situations. Grossman and Hart show that there is an incentive for the firm having control over the list to over-invest and for the other party to under-invest. They then ask when the insurance company should have control over the customer list. Their model predicts that in products where the renewal of contracts is not guaranteed and is sensitive to the agent's own actions, the agent will be more likely to own the list, whereas in products where renewal is more certain and is less sensitive to the agent's actions, the company will more likely own the list. The most important implication of Grossman and Hart's model is that the distribution of ownership rights has consequences for economic efficiency. These arise from the impossibility of ex ante bargaining over all aspects of the product to be delivered, that is, from the incompleteness of the contract. Integration alters but does not eliminate the incentives for opportunistic and distortionary behaviour.

In recent years, employee compensation plans have increased in variety and complexity. They may include direct salary and bonus plans, deferred salary and bonus contracts of various forms, stock purchase plans, stock appreciation rights, performance units, dividend units and share plans. These plans are often combined with fringe benefits, pensions, and a variety of plans featuring insurance, annuities and loans at various rates of interest. Two distinct literatures have evolved to explain the motivation for the variety of observed compensation contracts. Tax literature has emphasised the virtues of group term life insurance, pension plans, deferred compensation plans, and stock plans that can convert ordinary income into capital gains (although at a tax cost to the corporation). The 'agency' literature of economics, finance, and accounting recognizes that employees cannot always be relied on to take managerial decisions that are in the best interest of the firm's owners. Unless owners can somehow devise and implement effective and costless monitoring systems, the fortunes of employees must be tied in a nontrivial manner to the fortunes of the firm in order to provide the proper incentives. Employees, however, generally prefer less risky compensation schemes. The complexity of compensation packages reflects this conflict between efficient risk-sharing and effective incentives.

Myron Scholes (Stanford) endeavoured to integrate these two approaches in his paper 'Employee Compensation and Taxes: Links with Incentives and with Investment and Financing Decisions,' co- authored with Mark Wolfson (Stanford). They attempted to distinguish those features of compensation plans that appear to be motivated by tax considerations from those that appear to be designed to create appropriate incentives within the firm, and to understand the interaction of these two motives. Scholes argued that with few exceptions, previous analyses have concentrated on either the tax aspects of compensation planning or on the incentive aspects. However, both effects are pervasive in compensation planning, as they are in the firm's capital structure and investment policy. Indeed, Scholes argued, the capital structure question subsumes the employee compensation question. The optimal capital structure cannot be chosen independently of employee compensation planning (and vice versa). Similarly, tax planning cannot proceed independently of the employee incentive structure. As a consequence, it may be difficult to distinguish those compensation contracts that are tax-motivated and those that are incentive-motivated. This identification problem has important implications for empirical work. In the paper presented at the conference, Scholes and Wolfson attempted to predict the forms of employee compensation encouraged by the array of differing employee and employer tax positions. Observed compensation packages that differ systematically from these predictions should then provide important clues to other motivations for the plans that are chosen, such as problems of incentives and of contracting under conditions of moral hazard or adverse selection.

The final paper of the day, presented by Paul Grout (Birmingham and CEPR), was entitled 'Employee Share Ownership Schemes'. His paper analyzed how negotiated wages and the firm's investment decisions are affected by share ownership schemes, not only those involving all employees but also schemes that involve only 'key employees' of the firm. Grout shows that the impact of employee share ownership on investment and employment depends critically on whether employees make the decision to purchase shares jointly or independently of each other. The tax consequences of the share purchases are also important; Grout considers both the case in which the purchases have no tax consequences, and one in which employees' gains from share ownership are taxed as capital gains. This gave Grout four cases to consider in his analysis.

Grout found that certain features are common to all four cases. For example, employees will want to sell their shares at the first opportunity. But employee share ownership schemes can be profitably introduced in three of the four cases; the exception is the joint purchase - no taxation case. If a scheme can be profitably introduced, Grout finds it is because it alters the wage relationship by encouraging employees to have greater concern for the longer run; this may encourage higher levels of investment and employment by the firm.
Grout used simulation techniques in an attempt to quantify the effects of employee share ownership schemes. The simulations revealed significant effects on investment and employment, even when employees own only a small proportion of shares and management ignores these shares in investment decisions. In the case of individual purchase decisions, for example, if the marginal rate of income tax is 40% and employees own only 2% of the company, capital employed by the firm can increase by 6.7% and wages fall by less than 1.5%. The case in which the purchase decision is made jointly by employees always shows a smaller impact on investment, Grout found.

Tax literature traditionally suggests that firms under-invest in risky assets because of the asymmetric nature of the tax system, where profits are taxed in the period earned but losses must be carried forward and their present value thereby reduced. On the final day of the conference, the first paper, presented by Eli Talmor (Tel Aviv) and entitled 'The Structure and Incentive Effects of Corporate Tax Liabilities', attempted to re-examine this issue from a new perspective. The paper, written jointly with Richard Green (Carnegie-Mellon), analyzed the tax liability of a corporation with unused tax deductions as if the tax liability were a 'call' option, in which the exercise price of the call is the allowable deduction. Talmor and Green used this analogy to evaluate the effects of the tax system on corporate investment incentives. Their analysis suggested that the tax system created an incentive to underinvest in risky projects and an incentive for conglomerate merger. Talmor and Green also examined how the firm's investment in risky assets is affected by the tax system, both when the firm is wholly equity-financed and when it is financed by a mixture of debt and equity. In the debt financing case, they find that conflicts of interest can arise between debt and equity holders, and the results they obtain differ from those obtained by other authors. Jensen and Meckling have shown, for example, that an unanticipated increase in the riskiness of the firm will always result in a loss of wealth to bondholders, in favour of shareholders. In the presence of taxes Talmor and Green find that these results can be reversed.

The corporate tax law in the United States is asymmetric in its treatment of operating gains and losses: income is taxed at the statutory tax rate only when positive. Although losses up to the amount of taxes paid in the previous three years qualify for an immediate credit against current taxes, losses in excess of that must be carried forward to be credited against future gains. The 'effective tax rate' on these 'carry forwards' is therefore less than the statutory rate. Conventional capital budgeting rarely recognises this asymmetry. Tax shields are sometimes discounted at a higher rate to allow for the risk that the firm may not have sufficient taxable income to make full use of the shields, but in practice such adjustments are made on a very ad hoc basis. In the second session of the final day Saman Majd (Wharton School) presented a paper co-authored with Stewart Myers (MIT) and entitled 'Valuing the Government's Tax Claim on Risky Corporate Assets'. They outlined a general procedure for calculating the impact of this tax law asymmetry on the after-tax values of firm's real assets. Such a procedure would, they noted, be a valuable aid to managers making capital budgeting decisions, and it would also allow us to evaluate better the effects of tax policy. Majd and Myers's procedure applied concepts used in the analysis of option pricing to establish the value of the government's tax claim on risky corporate assets. No simple formula was available, so the tax claim values must be computed by numerical methods. Majd presented after-tax values for a realistic 'standard project' under various assumptions about project profitability, risk and the tax position of the firm owning it. The results, he concluded, demonstrated that asymmetric taxation of operating gains and losses can significantly affect the after-tax net present value of corporate investment opportunities.

There is strong evidence that the right to control a large corporation is valuable. Incumbents therefore use a variety of devices to maintain control, including changes in the corporate capital structure. The final paper of the conference, presented by Arthur Raviv (Northwestern) and entitled 'Corporate Control Contests and Capital Structure', explored the choice of corporate takeover methods, the determinants of their outcomes and the use of capital structure as an anti-takeover device. Because common stock carries voting rights while debt does not, the firm's debt- equity ratio may affect the outcome of corporate votes and thus may partly determine who controls corporate resources. Incumbent managements can use short-term financial restructuring to influence the form of any possible takeover attempt and its outcome. By examining this relationship, Raviv and his co-author Milton Harris (Northwestern) were able to obtain new insights into the method by which changes in corporate control are carried out and the price changes which result. They found that firms with higher operating risk will tend to have higher debt-equity ratios. The model predicts that there should be a relatively low frequency of proxy fights for control of firms which are heavily regulated and in which the ability of those in control has little effect on the firm's performance. In addition, Raviv and Harris found that among firms in which incumbents are not vulnerable, debt is positively associated with the importance of the ability of those in control. Thus new 'venture capital' firms should have relatively less debt. Finally, their model suggests an explanation of the somewhat puzzling observation that stock values increase in the event of proxy fights by about the same amount, regardless of who wins.


A selection of the papers presented at the conference will be published next year by Cambridge University Press. Conference organisers Colin Mayer and Jeremy Edwards, together with Research Fellow Margaret Bray are launching a major research project on the provision of finance for industry, which they will discuss at a lunchtime meeting on 6 December.