Lawrence Harris

Estimating the components of the bid/ask spread☆

Abstract This paper develops and implements a technique for estimating a model of the bid/ask spread. The spread is decomposed into two components, one due to asymmetric information and one due to inventory costs, specialist monopoly power, and clearing costs. The model is estimated using NYSE common stock transaction prices in the period 1981–1983. Cross-sectional regression analysis is then used to relate time-series estimated spread components to other stock characteristics. The results cannot reject the hypothesis that significant amounts of NYSE common stock spreads are due to asymmetric information.

A transaction data study of weekly and intradaily patterns in stock returns

Abstract Weekly and intradaily patterns in common stock prices are examined using transaction data. For large firms, negative Monday close-to-close returns accrue between the Friday close and the Monday open; for smaller firms they accrue primarily during the Monday trading day. For all firms, significant weekday differences in intraday returns accrue during the first 45 minutes after the market opens. On Monday mornings, prices drop, while on the other weekday mornings, they rise. Otherwise the pattern of intraday returns is similar on all weekdays. Most notable is an increase in prices on the last trade of the day.

Market vs. Limit Orders: The SuperDOT Evidence on Order Submission Strategy

This paper discusses performance measures for market and limit orders. We suggest two measures: one for precommitted traders (who must trade) and another for passive traders (who are indifferent to trading). We compute these measures for a sample of NYSE SuperDOT orders. The results suggest that the limit order placement strategies most commonly used by NYSE SuperDOT traders do in fact perform best. Limit orders placed at or better than the prevailing quote perform better than do market orders, even after imputing a penalty for unexecuted orders, and after taking into account market order price improvement. Unconditional order submission strategies that use SuperDOT to offer liquidity in competition with the specialist do not appear to be profitable.

Transaction Data Tests of the Mixture of Distributions Hypothesis

This paper presents new tests of the mixture of distributions hypothesis. Previous tests examined security prices and volume measured only at daily intervals. Here, differential implications of the hypothesis for transaction data are derived and tested. The new predictions emanate from the assumption that prices and volume evolve at uniform rates in transaction time. The results support this assumption and the mixture of distributions hypothesis in general. In addition, the tests suggest that the daily transaction-count may be a useful instrumental variable for estimating unobserved realizations of stochastic price variances.

Cross-Security Tests of the Mixture of Distributions Hypothesis

New cross-sectional tests of the Mixture of Distributions Hypothesis are presented. The tests assume that the distribution of the mixing variable (often interpreted as the daily rate of flow of information) is not identical for all securities. Cross-security differences in the mixing distribution cause cross-security differences in the joint distribution of returns and volume. The Hypothesis provides predictions about how these differences appear in the joint distribution. The predictions are confirmed in tests based on cross-security correlations among summary statistics that characterize shape and covariational attributes of the joint distribution of returns and volume. The results are consistent with the Mixture of Distributions Hypothesis.

Optimal Dynamic Order Submission Strategies in Some Stylized Trading Problems

Optimal dynamic limit and market order submission and resubmission strategies are derived for several stylized trading problems. Separate solutions are obtained for quote- and order-driven markets. The results provide practical rules for how to trade small orders and how to manage traders. Transaction cost measurement methods based on implementation shortfall are proven to dominate other methods. Since investors demand liquidity when they submit market orders and supply liquidity when they submit limit orders, the results improve our understanding of market liquidity. In particular, the models illustrate the role of time in the search for liquidity by characterizing the demand for and supply of immediacy.

A Day-End Transaction Price Anomaly

A large mean price change is observed on the last daily NYSE transaction. This suggests that closing prices may not consistently represent stock values. Transaction prices are studied to further characterize the day-end price rise and to determine whether it is due to any limited subsample of stocks or dates. The results indicate that the phenomenon is pervasive over most firms and days. Some evidence suggests that it is caused by a change in the frequency of ask prices at day-end.

A Maximum Likelihood Approach for Non-Gaussian Stochastic Volatility Models

A maximum likelihood approach for the analysis of stochastic volatility models is developed. The method uses a recursive numerical integration procedure that directly calculates the marginal likelihood. Only conventional integration techniques are used, making this approach both flexible and simple. Experimentation shows that the method matches the performance of the best estimation tools currently in use. New stochastic volatility models are introduced and estimated. The model that best fits recent stock-index data is characterized by a highly non-Gaussian stochastic volatility innovation distribution. This model dominates a model that includes an autoregressive conditional heteroscedastic effect in the stochastic volatility process and a model that includes a stochastic volatility effect in the conditional mean.

Estimation of Stock Price Variances and Serial Covariances from Discrete Observations

Stock price discreteness adds noise to price series. The noise increases return variances and adds negative serial correlation to return series. Standard variance and serial covariance estimators therefore overestimate the variance and serial covariance of the underlying stock values. Discreteness-induced variance and serial covariance depend on underlying volatility and on the size of the bid/ask spread. Simple formulas for approximating the effects of discreteness on variance and serial correlation are derived and presented. The approximations, which are accurate in daily data, can be used to adjust the standard variance and serial covariance estimators.

Price Inflation and Wealth Transfer During the 2008 SEC Short-Sale Ban

Using a factor-analytic model that extracts common valuation information from the prices of stocks that were not banned, we estimate that the ban on short-selling financial stocks imposed by the SEC in September 2008 led to substantial price inflation in the banned stocks. The inflation reversed somewhat following the ban, but the data are too noisy to conclusively link the reversal to the ban. Other factors such as the pending TARP legislation may also have affected prices, though our results suggest that it was not a significant factor. If prices were inflated, buyers paid more than they otherwise would have paid for the banned stocks during the period of the ban. We provide an estimate of