Knowledge that Transforms

[We test the theory of the term structure of indexed-bond prices due to Cox, Ingersoll, and Ross (CIR). The econometric method uses Hansen's generalized method of moments and exploits the probability distribution of the single-state variable in CIR's model, thus avoiding the use of aggregate consumption data. It enables us to estimate a continuous-time model based on discretely sampled data. The tests indicate that CIR's model for index bonds performs reasonably well when confronted with short-term Treasury-bill returns. The estimates indicate that term premiums are positive and that yield curves can take several shapes. However, the fitted model does poorly in explaining the serial correlation in real Treasury-bill returns.]

[I examine a two-period noisy rational expectations model of a futures market and show that the dispersion of expectations about a weighted average of future prices measures both the additional volatility and the additional expected volume of trade associated with noisy information. The role played by dispersion helps clarify several stylized facts concerning volume and price behavior. Specifically, dispersion can be a factor contributing to the positive correlation between volume and absolute price changes, and the positive correlation between consecutive absolute price changes.]

[I discuss a new method for measuring the deviations between actual transaction prices and implicit efficient prices. The approach decomposes security transaction prices into random-walk and stationary components. The random-walk component may be identified with the efficient price. The stationary component, the difference between the efficient price and the actual transaction price, is termed the pricing error. Its dispersion is a natural measure of market quality. I describe practical strategies for estimating these quantities. For a sample of NYSE stocks, the average pricing error standard deviation estimate is roughly 0.33 percent of the stock price. If the pricing error is normally distributed and if it is always a positive cost incurred by the transaction initiators, the corresponding average transaction cost for these traders is 0.26 percent of the stock price. The dispersion of the pricing error is also found to be elevated at the beginning and end of the trading session.]

[We examine the role of market makers in facilitating price discovery. We show that a specialist may experiment with prices to induce more informative order flow, thereby expediting price discovery. Market makers in a multiple-dealer system, unlike a specialist system, do not have the incentives to perform such costly experiments because of free-rider problems. Consequently, the specialist system may provide open markets where competition fails but at the cost of wider bid-ask spreads. We analyze the effect of experimentation on the bid-ask spread and provide an exploratory analysis of intraday specialist data that is consistent with our price experimentation hypothesis.]

[In this article we derive necessary and sufficient conditions that must be satisfied by equilibrium asset price processes in a pure exchange economy. We examine a world in which asset prices follow a diffusion process, asset markets are dynamically complete, all investors maximize their (state-independent) expected utility of consumption at some future date, and investors have nonrandom exogenous income. We show that it is necessary and sufficient that the coefficients of an equilibrium diffusion price process satisfy a partial differential equation and a boundary condition. We also examine how the dynamics of asset prices are related to the shape of the representative investor's utility function through the boundary condition. For example, in a constant-volatility economy, the expected instantaneous return of the market portfolio is mean reverting if and only if the relative risk aversion of the representative investor is decreasing in terminal wealth.]

[In a one-period model of market making with many exogenously informed traders, we first show that the variance of prices and expected trading volume depend on the public information released at the start of trading. This is accomplished by representing beliefs with elliptically contoured distributions, for which the form of optimal decision rules does not depend on the specific distribution used. Second, if the model is altered so that the decision to become informed is made endogenous, then the decision rules of the market-maker and informed traders depend on the public information. Third, in a multiperiod model with many informed traders and long-lived private information, recursion formulas similar to those of Kyle (1985) hold for all elliptically contoured distributions, trading volume is autocorrelated, and, unless per period liquidity trading is bounded away from zero as new trading periods are added, informed traders' profits vanish.]

[We investigate the informational efficiency of mutual fund performance for the period 1965-84. Results are shown to be sensitive to the measurement of performance chosen. We find that returns on S&P stocks, returns on non-S&P stocks, and returns on bonds are significant factors in performance assessment. Once we correct for the impact of non-S&P assets on mutual fund returns, we find that mutual funds do not earn returns that justify their information acquisition costs. This is consistent with results for prior periods.]

[In this article we are concerned with the effect of the number of investment analysts following a firm on the speed of adjustment of the firm's stock price to new information that has common effects across firms. It is found that returns on portfolios of firms that are followed by many analysts tend to lead those of firms that are followed by fewer analysts, even when the firms are of approximately the same size. Many analyst firms also tend to respond more rapidly to market returns than do few analyst firms, adjusting for firm size. This relation, however, is nonlinear, and the marginal effect of the number of analysts on the speed of price adjustment increases with the number of analysts.]

[Corporate debt sales have been regarded as "no news" events because there is no significant price reaction on average to their announcement. We explore the hypothesis that this lack of average price reaction to debt sale announcements is explained by the partial anticipation of debt offers. Theory suggests that the demand for debt capital is fundamentally related to changes in the sources and uses of funds, and we find evidence that earnings are significantly lower, investment growth is significantly higher, and, for some issuers, debt refunding requirements are significantly greater in the period immediately prior to issue than in periods well before and after the issue. We find that this preissue information conditions investors' expectations of issue, thereby affecting the cross-sectional announcement date price reaction to debt sales in two ways. First, announcement date price reactions are negative, on average, for unanticipated offers or for those offers where prior information suggests that an issue is unlikely. Second, holding the probability of issue constant, announcement date price reactions are significantly more negative for offers that raise more capital than investors expected. These results are consistent with cash flow signaling and asymmetric information models of corporate financings.]

[We examine the behavior of measured variances from the options market and the underlying stock market. Under the joint hypotheses that markets are informationally efficient and that option prices are explained by a particular asset pricing model, forecasts from time-series models of the stock-return process should not have predictive content given the market forecast as embodied in option prices. Both in-sample and out-of-sample tests suggest that this hypothesis can be rejected. Using simulations, we show that biases inherent in the procedure we use to imply variances cannot explain this result. Thus, we provide evidence inconsistent with the orthogonality restrictions of option pricing models that assume that variance risk is unpriced. These results also have implications for optimal variance forecast rules.]