Knowledge that Transforms

In this paper, I set up scenarios where the mean‐variance capital asset pricing model is true and where it is false. Then I investigate whether the coefficients from regressions of population expected excess returns on population betas, and expected excess returns on betas and size, allow us to distinguish between the scenarios. I show that the coefficients from either ordinary least squares or generalized least squares regressions do not allow us to tell whether the model is true or false.

This paper investigates Euler equations involving security prices and household‐level consumption data. It provides a useful complement to many existing studies of consumption‐based asset pricing models that use a representative‐agent framework, because the Euler equations under investigation hold even if markets are incomplete. It also provides a useful complement to simulation‐based studies of market incompleteness. The empirical evidence indicates that the theory is rejected by the data along several dimensions. The results therefore indicate that some well‐documented asset‐pricing puzzles do not result from aggregation problems for the preferences under investigation.

We study the causes and effects of the competition for order flow by U.S. regional stock exchanges. We trace the origins of competition for order flow to a change in the role of regional exchanges from being venues for listing local securities to being more direct competitors for the order flow of NYSE listings. We study the way regionals competed for order flow, concentrating on a series of stock‐exchange mergers that occurred in the midst of this transition of the regional exchanges. The merging exchanges attracted market share and experienced narrower bid‐ask spreads.

The practices of preferencing and internalization have been alleged to support collusion, cause worse execution, and lead to wider spreads in dealership style markets relative to auction style markets. For a sample of London Stock Exchange stocks, we find that preferenced trades pay higher spreads, however they do not generate higher dealer profits. Internalized trades pay lower, not higher, spreads. We do not find a relation between the extent of preferencing or internalization and spreads across stocks. These results do not lend support to the “collusion” hypothesis but are consistent with a “costly search and trading relationships” hypothesis.

Are nominal bonds appropriately discounted for taxes? Empirical estimates of the response of nominal interest rates to changes in inflation, the Fisher effect, have failed to produce a definitive answer. Four reasons have been put forward as possible explanations: (i) Tobin effects, (ii) fiscal illusion, (iii) peso problems, and (iv) different estimators. Utilizing data on taxable and tax‐exempt bond interest rates and several different estimators, we find that the Fisher effect estimates are always larger for the taxable bond relative to the tax‐exempt bond, suggesting that fiscal illusion and different estimators cannot account for the previous results.

We investigate Grossman and Laroque's (1990) conjecture that costs of adjusting consumption can account, in part, for the empirical failure of the consumption‐based capital asset pricing model (CCAPM). We incorporate small fixed costs of consumption adjustment into a CCAPM with heterogeneous agents. We find that undetectably small consumption adjustment costs can account for much of the discrepancy between the observed variance of nondurable aggregate consumption growth and the predictions of the CCAPM, and can partially reconcile nondurable consumption data with the observed equity premium. We conclude that the CCAPM's implications are nonrobust to extremely small adjustment costs.

We present a model of Nasdaq that includes the two ways in which marketmakers compete for order flow: quotes and direct payments. Brokers in our model can execute small trades through a computerized system, preferencing arrangements with marketmakers, or vertical integration into market making. The comparative statics in our model differ from those of the traditional model of dealer markets, which does not capture important institutional features of Nasdaq. We also show that the empirical evidence is inconsistent with the traditional model, which suggests that preferencing and vertical integration are important components in understanding Nasdaq.

Using a sample of NYSE‐listed equities from 1992, this study examines whether market maker performance differs across specialist firms. We find that spreads and depth differ across specialist firms, but the competitiveness of NYSE quotes relative to other exchanges does not appear to be affected by these differences. Differences are also evident in measures of transitory volatility and in the frequency and duration of order‐imbalance trading halts. The results suggest that specialists have a significant effect on execution costs, liquidity, and noise in security prices and that these effects are not completely eliminated by competition or the NYSE's monitoring mechanisms.

How does increased noise trading affect market liquidity and trading costs? We use The Wall Street Journal 's “Investment Dartboard” column, which stimulates noise trading, as a natural experiment to evaluate models of the bid‐ask spread. We find that substantial increases in trading volume and significant but temporary abnormal returns occur when analysts recommend stocks in this column, especially when recommendations come from analysts with successful contest track records. We also find an increase in liquidity and a decrease in the adverse selection component of the bid‐ask spread.

In this paper, we point out that the widely used stochastic discount factor (SDF) methodology ignores a fully specified model for asset returns. As a result, it suffers from two potential problems when asset returns follow a linear factor model. The first problem is that the risk premium estimate from the SDF methodology is unreliable. The second problem is that the specification test under the SDF methodology has very low power in detecting misspecified models. Traditional methodologies typically incorporate a fully specified model for asset returns, and they can perform substantially better than the SDF methodology.