WORKING PAPER ABSTRACTS - 1989
Listed below are the abstracts for all of the working papers for this year. To see a list of paper titles (with links to available PDF files) click here.
This paper develops a signalling model with two signals, two attributes, and a continuum of signal levels and attribute-types to explain a puzzling phenomenon: the underpricing of new issues. Both the fraction of the new issue retained by the issuer and its offering price convey the unobservable "intrinsic" value of the firm and the variance of its cash flows to investors. Many of the models comparative statics results are quite novel, empirically testable, and consistent with the existing empirical evidence on new issues, as well as the beliefs of many investment professionals. In particular, a closed-form solution for the signalling schedule indicates that the degree of underpricing, which can be inferred from observable variables, is positively related to the firm's post-issue share price.
The existence of nominal wage and debt contracts is a puzzle. In a model with strategic complementarities, where imperfectly competitive firms inefficiently underinvest, nominal wage or debt contracts are shown to be preferred to indexed contracts. Nominal contracts are an optimal arrangement between firms and workers because such contracts can improve on the underinvestment equilibrium by implementing Pareto-improving transfers between agents. Moreover, we show that if there are multiple underinvestment equilibria, then monetary policy can have real effects because the monetary authority can choose a money supply rule to coordinate beliefs and, thereby, select the best equilibrium.
The usual approach to determine if market prices of uninsured bank liabilities reflect the risk of default is to regress the yield spread of bank debt against accounting measures of bank risk. To date these results have been mixed. Here we argue that this is because previous investigations lack a theoretical model of bank liability pricing. Without this, linear regressions have difficulty addressing the question. In this essay we use contingent claims valuation to determine whether implied volatilities embedded in bank liability prices are correlated with accounting measures of bank risk. Observed yields on subordinated bank debt over equivalent maturity treasuries are used to compute implied bank volatility. Accounting measures of bank risk turn out to predict the volatility of bank assets and suggest the existence of market discipline.
If returns on some stocks systematically lead or lag those of others, a portfolio strat-egy that sells "winners" and buys "losers" can produce positive expected returns, even if no stock's returns are negatively autocorrelated as virtually all models of overre-action imply. Using a particular contrarian strategy, we reconcile the strong positive autocorrelation of weekly portfolio returns with the weak negative autocorrelation for individual stock returns. We show that the returns of large stocks lead those of smaller stocks, and present evidence against overreaction as the only source of contrarian prof-its.
This paper examines the stochastic properties of aggregate macroeconomic time series from the standpoint of fractionally integrated models, and focuses on the persistence of economic shocks. We develop a simple macroeconomic model that exhibits long-term dependence, a consequence of aggregation in the presence of real business cycles. We derive the re-lation between properties of fractionally integrated macroeconomic time series and those of microeconomic data, and discuss how fiscal policy may alter their stochastic behavior. To implement these results empirically, we employ a test for fractionally integrated time series based on the Hurst-Mandelbrot rescaled range. This test is robust to short-term dependence, and is applied to quarterly and annual real GNP to determine the sources and nature of long-term dependence in the business cycle.
A comparative static analysis of a competitive equilibrium under heterogeneous earn-ings expectations and constant absolute risk aversions demonstrates that unsystem-atic trading volume in response to a public information release is proportional to the change in relative heterogeneity of beliefs and is unrelated to either the dispersion of prior beliefs or the absolute change in the consensus expectation. In contrast, under homogeneous expectations and non-constant absolute risk aversions, total trading volume is a function of the absolute change in the consensus earnings expectation. Using a large number of individual analysts annual earnings forecasts, the change in relative heterogeneity and the absolute change in the consensus expectation are measured around interim quarterly earnings reports and shown to possess predicted impacts on trading volume.
The empirical implications of the consumption-oriented capital asset pricing model (CCAPM) are examined, and its performance is compared with a model based on the market portfolio. The CCAPM is estimated after adjusting for measurement problems associated with reported consump-tion data. The CCAPM is tested using betas based on both consumption and the portfolio having the maximum correlation with consumption. As predicted by the CCAPM, the market price of risk is significantly positive, and the estimate of the real interest rate is close to zero. The performances of the traditional CAPM and the CCAPM are about the same.
An individual investors demands for risky capital and riskless bonds depend on the investors subjective beliefs about the payoff to risky capital. This paper determines equilibrium asset prices and returns in a capital market in which investors have heterogeneous subjective expectations of the payoff to capital. Increased heterogeneity increases the riskless rate of return and reduces the stock price. Heterogeneity can also dramatically increase the equilibrium equity premium on stocks relative to bonds. Therefore, calculating the equity premium under the assumption of homogeneous beliefs could dramatically understate the equity premium that would prevail under heterogeneity.
An equilibrium pricing model with time-varying conditional moments of consumption growth is used to analyze the behavior of conditional moments of stock returns for long and short investment horizons. We examine the behavior over time of estimates of the conditional means and variances of consumption growth and returns. Business cycles appear to be associated with all of these estimates. The tendency for estimates of the price of risk to be higher during recessions, when coupled with the business-cycle variation in estimates of the moments of consumption growth, appears to be consistent with the pricing model.
This paper analyzes the effects of lump-sum tax policy in an overlapping generations model in which consumers have uncertain longevity. It extends previous analyses by considering the case in which private insurance arrangements are actuarially unfair. In addition, it considers the polar case of actuarially fair insurance and the polar case of no insurance. A general condition for debt neutrality is derived. This condition depends explicitly on the degree of actuarial unfairness in insurance and on the extent to which parents care about the utility of their children.
The paper shows how a cash-in-advance model of money demand can be written in a way that combines a simple, yet empirically defensible, money demand equation with tractability in asset pricing. Return premia are determined as in the standard barter exchange model, except that a short-term risk-free nominal interest rate enters into the first order condition. In special cases, asset prices satisfy the familiar barter-economy Euler equations exactly. Thus, contrary to much of the literature, money may not significantly affect asset pricing. Simple barter-economy Euler equations are approximately valid even in the presence of money.
Population and labor force growth are expected to approach zero in the first half of the next century and this will tend to reduce further an already unsatisfactory level of aggregate saving. This paper investigates the determinants of aggregate saving under these demographic assumptions, with particular emphasis on workers strategies for dealing with the risk of outliving their resources after retirement. The permanent income and life cycle models represent different strategies for addressing this risk and are found to differ substantially in their implications for savings.
Aggregate saving is found to be unfavorably affected by taxes on labor income, while the effect of taxes on property income is ambiguous. Public and private pension plans have a negative aggregate impact, even though employer plus employee contributions fully support workers post retirement benefits. Inheritance taxes have no effect in the cases considered.
We develop a stochastic model of nonsynchronous asset prices based on sampling with random censoring. In addition to generalizing existing models of non-trading, our framework allows the explicit calculation of the effects of infrequent trading on the time series properties of asset returns. These are empirically testable implications for the variances, autocorrelations, and cross-autocorrelations of returns to individual stocks as well as to portfolios. We construct estimators to quantify the magnitude of non-trading effects in commonly used stock returns data bases, and show the extent to which this phenomenon is responsible for the recent rejections of the random walk hypothesis.
We investigate the extent to which tests of financial asset pricing models may be biased by using properties of the data to construct the test statistics. Specifically, we focus on tests using returns to portfolios of common stock where portfolios are constructed by sorting on some empirically motivated characteristic of the securities such as market value of equity. We present both analytical calculations and Monte Carlo simulations that show the effects of this type of data-snooping to be substantial. Even when the sorting characteristic is only marginally correlated with individual security statistics, 5 percent tests based on sorted portfolio returns may reject with probability one under the null hypothesis. This bias is shown to worsen as the number of securities increases given a fixed number of portfolios, and as the number of portfolios decreases given a fixed number of securities. We provide two empirical examples that illustrate the practical relevance of these biases.
Returns computed with closing bid or ask prices that may not represent "true" prices imparts measurement error into portfolio returns if investor buying and selling behavior displays systematic patterns. This paper finds systematic tendencies for closing prices to be recorded at the bid in December and at the ask in early January. After controlling for changing bid and ask prices, this pattern results in large portfolio returns on the two trading days surrounding the end of the year, especially for low-price stocks. Other temporal return patterns (e.g., weekend and holiday effects) are also related to systematic trading patterns.
The implications for takeover activity of certain types of capital market imperfections are examined. Management, "knowledgeable outsiders" and "uninformed investors" are assumed to differ in the extent to which they have access to the total information potentially available at a given point in time. More information means less dispersion in the subjective probability distribution of future cash flows and a lower required risk premium.
Stockholders also differ with respect to their investment horizons and tend to be myopic beyond their own horizon. While they may estimate the expected value of earnings and dividends within their horizon quite accurately (relative to the full-information estimate), they have little information regarding events beyond that horizon and may tend to underestimate the long--term trend in periods when this is steeper than the trend within the relevant horizon. This tendency, in conjunction with the inverse relationship of the required risk premium to information access, implies that in such periods market participants will generally underprice stock relative to its fundamental value as defined in terms of an infinite horizon and maximum information. Unless those with the most information and the longest horizons dominate trading, stocks will be undervalued in the market.
If market price remains below fundamental value for any significant period, a profit opportunity is created for any "knowledgeable outsider" with sufficient information to perceive the discrepancy. The welfare implications of takeover bids motivated solely by such a discrepancy are investigated. Also investigated is the dilemma of outside directors when faced with a takeover bid above market price but below what they perceive as long-term value. Acceptance of these terms may well benefit stockholders with short horizons, while damaging those with longer horizons.
The problems of aggregating divergent individual forecasts of inflation, each subject to error on a number of grounds, into a "market forecast" are considered; and it is shown to be extremely unlikely that such an aggregate will always reproduce precisely the optimal forecast based on all relevant prior events, whether or not these are accurately observable. If this is not the case, then regressions relating the inflation rate to prior values of the nominal interest rate generate parameter estimates that are subject to bias, even if the real rate is constant. Inferences based on those parameter estimates are likely to be invalid.
Transferring physical capital and transferring production and sales activities from one country to the other, typically entails large adjustment costs. The model of this paper features two homogeneous stocks of physical capital located in two different countries, separated by an "ocean." The two physical stocks are optimally invested in a random production process yielding real returns, or consumed by local residents, or transferred abroad. Retro-fitting, transferring and re-building capital equipment, and increasing production and sales abroad either takes time (during which capital is idle) or consumes real resources. Under proportional transfer costs, trade imbalances, consumption imbalances and capital imbalances between the two locations are shown to be persistent. The stochastic process for the deviation from the Law of One Price (LOP) is obtained. By construction, this process is compatible with financial market efficiency and with the possibility of (costly) trade in commodities. Whereas empirical studies have found no evidence against the hypothesis that LOP deviations follow a martingale, the theoretical process which is found differs markedly from a martingale: the drift is non linear and mean reverting. But the behavior of the conditional variance more than offsets the reverting effect of the drift and the conditional probability of a move away from Parity is greater than the probability of a move toward Parity. When some price barriers are reached, however, reversion is triggered. We decompose the real-interest rate differential into an expected price change and a risk premium for which a very simple expression is found. The behaviors over time of the rate differential and of its components are examined.
This paper calculates the stochastic properties of consumption when income follows a fractional stochastic process, and shows how this may explain both the excess sensitivity and the excess smoothness paradoxes. It then uses a recently developed improvement of the Rescaled Range Statistic to find long term memory in consumption. The remaining sections undertake Monte Carlo simulations to assess the finite sample size and power of the test, conduct cross country comparisons (France, Canada, U.K.), and provides a possible explanation.
Under certain conditions, efficient markets imply random walk behavior in real exchange rates. Much of international finance theory, however, is based on the idea of purchasing power parity, which implies mean reversion in real exchange rates. This paper uses variance ratio statistics to test for random walk behavior in real exchange rates. Unlike most previous tests of this hypothesis, the tests do reject a random walk for monthly data; however, the monthly statistics do not provide evidence in favor of mean reversion. Interestingly, tests using annual data for the twentieth century are unable to reject a random walk despite evidence of mean reversion. This appears to be due to the relatively small number of observations available.
Budget deficits can be eliminated in two ways, by increasing taxes or by cutting spending. In recent years, the question which of these two ways is --- or should be -- chosen by the government has received considerable attention.
This paper provides a historical perspective on government reactions to deficits by looking at a long series of US-budget data, from 1792-1988. The main findings are that, on average and in present value terms, 50-65% of a deficit due to tax cuts and about 70% of a deficit due to higher government spending are eliminated by future spending cuts. Only the remainder, about 35-50% or 30%, respectively, is eliminated by future tax changes.
Security baskets and index-linked securities are securities whose values are functions of the cash flows or values of other assets. Intermediaries create security baskets by pooling or bundling more primitive assets such as mortgages, credit card receivables and other loans, or equities as in the case of closed-end mutual funds. Index-linked securities, such as index participations and stock index futures, are created by stock and futures exchanges. Creation of these "composite" securities would appear to be redundant if investors could individually purchase the securities that compose the security basket or index, thus creating their own diversified portfolios. However, we show that when some investors possess inside information, composite securities are not redundant. They provide superior "liquidity" for uninformed investors. By holding these securities, uninformed investors with unexpected needs to trade can reduce their expected losses to investors with inside information. Moreover, the existence of these securities affects real investment decisions and equilibrium rates of return. We provide an application of our model to the problem of international portfolio choice and mutual fund design.
A binomial approximation to a diffusion is defined as computationally simple if the number of nodes grows at most linearly in the number of time intervals. This paper shows how to construct computationally simple binomial processes which converge weakly to commonly employed diffusions in financial models. It also demonstrates the convergence of the sequence of bond and European option prices from these processes to the corresponding values in the diffusion limit. Numerical examples from the Constant Elasticity of Variance stock price and the Cox, Ingersoll, and Ross discount bond price are provided.
We develop contingent claims valuation models for corporate bonds that are capable of generating yield spreads consistent with the levels observed in practice. We incorporate important features in the valuation related to the occurrence of and payoff upon bankruptcy and focus on the default risk of coupons in the presence of dividends and interest rate uncertainty. Numerical solutions are employed to show that the resulting yield spreads are sensitive to interest rate expectations but not to the volatility of the interest rates. Interaction between call provisions and default risk in determining yield spreads is explicitly analyzed to show that the call provision has a differential effect on Treasury issues relative to corporate issues.
Dual trading is said to occur when an entity sometimes trades as a broker for customers, and at other times trades for its own account. Dual trading is quite pervasive throughout the United States securities and futures markets as well as in financial and commodity markets throughout the world. The pervasiveness of dual trading is due to the fact that many of the skills and facilities required to be a good broker are also necessary to be a good trader. Dual trading increases the supply of both brokers and floor traders because a dual trader can earn income from two activities to cover the costs of training, an Exchange seat, and time spent on the floor. He has less idle time and facilities when he can switch from the activity in low demand to the activity in high demand.
When rates of return on bonds are computed over extremely short holding periods, the ex post cross-sectional relationship between realized return and risk is linear. It is therefore possible, at any time, to extrapolate the cross-sectional relationship to a zero risk level, and thus to determine the implied instantaneous riskless rate of interest. We apply this technique to French bond price data. Using this rather unique data set in which prices are sampled daily, we are able to compare the overnight rate implied in bond price data to the actual overnight money market rate. We conclude that the two rates are significantly different, which is evidence of segmentation between the two markets. The institutional set-up prevailing in France during the sample period explains the segmentation result.
As a result of the historical importance of debt and equity, the traditional focus of inquiry into firms choice of capital structure has been "What is the optimal debt/equity ratio?" This approach lead to the Modigliani and Miller theorems and a large body of subsequent work but has not been very successful in explaining firms actual choices of debt and equity. The notion that firms finance their activities with debt and equity is a simplification; corporations have issued securities other than standard debt and equity for many centuries. This fact and the rapid pace of financial innovation in recent years suggests that a more fundamental issue than "What is the optimal debt/equity ratio?" is "What are the optimal securities that should be issued?" This paper surveys recent studies of capital structure that have looked at this question.
We solve for the optimal dynamic trading strategy of an investor who faces two constraints. The first constraint is a limitation on his ability to borrow for the purpose of investing in a risky asset, i.e., the market value of his investments in the risky asset X, must be less than an exogenously given function of his wealth X(W). The second constraint is the requirement that the investors wealth be non-negative at all times, i.e., Wt>O. We assume that the investor has constant relative risk aversion A, and the value of the risky asset follows a diffusion with drift m+r (where r is the risk free rate) and per unit time variance s2. In the absence of the constraints, X &Mac186; (m/s2)*W/A. We prove that in the presence of the above constraints the optimal investment is X &Mac186; Min[(m/s2)*W/a, X(W)]. The coefficient a is not in general equal to A, and represents the extent to which the investor alters his strategy even when the constraints are not binding because of the possibility that the constraints will become binding in the future.
An empirical analysis of the effects of foreign exchange risk on global equity portfolios is performed. For typical levels of foreign equity holdings, exchange risk is not found to be a substantial component of the overall risk of a portfolio and the value of hedging that risk is found to be very sample period specific. During periods of rapid currency appreciation, for example 1980-85, hedged portfolios did outperform unhedged portfolios; however, this result is reversed during periods of rapid currency depreciation. Overall, while hedging does reduce the riskiness of a global portfolio, it also reduces return so that no statistically significant improvement in portfolio performance is obtained.
This paper examines the return characteristics of low-grade bonds using dealer bid prices. The volatility of an index of these bonds is less than the volatility of indexes of higher-grade bonds such as long-term Treasury bonds. This reduced volatility is due in large part to the shorter duration of low-grade bonds. We also present evidence that low-grade bonds are a hybrid security with features of both stocks and bonds. A detailed analysis of the returns realized by all low-grade bonds issued in 1977 and 1978 indicates that any relation between bond age and probability of default does not induce a bias in the results based on our index of lower-grade bonds. Moreover, we present evidence that at least part of the observed tendency for the probability of default to increase with age is due to cyclical factors.
The dynamic behavior of security prices is studied in a setting where two agents trade strategically and learn from market prices. Each trader receives a private signal about fundamentals, the significance of which depends on the signal received by the other trader. In trading, each agent wants to deceive the other trader into revealing his signal, while not revealing his own signal. We show that trade is self-generating because agents learn the value of the asset only through observation of the market price. Uninformed agents, technical analysts, can also trade by charting past prices. These chartists ensure market efficiency. Equilibrium price paths of the model may display reversals in which all traders rationally revise their beliefs, first in one direction and then in the opposite direction even though no new information has entered the system. A piece of information which is initially thought to be bad news may be revealed, through trading, to be good news. This fad-like behavior results from rational strategic interaction and Bayesian inference. In this model security prices do not follow a martingale.
Much of financial theory neglects transactions costs. Perhaps the most successful implementation of it -- i.e. continuous-time portfolio choice and option pricing -- is downright inconsistent with the existence of any transactions cost at all. Nonetheless prima facie evidence from the trade is that transactions costs are a source of concern for portfolio managers. The presence of practically any friction in financial markets qualitatively changes the nature of the optimization problem; for it produces the need sometimes to do nothing and sometimes to act, an issue which, of course, does not arise in frictionless situations.
The investor considered here does not consume along the way. He accumulates wealth until some terminal point in time. At that point he consumes all. His objective is to maximize the expected utility derived from that terminal consumption. We postpone the terminal point infinitely far into the future to obtain a stationary portfolio rule. The optimal portfolio policy which we find is in the form of two control barriers between which portfolio proportions are allowed to fluctuate before some trade is resorted to.
We show how to calculate these two barriers exactly.
Copyright ©1998 The Wharton School of the University of Pennsylvania. All rights reserved.
Site designed by Singularity Design.