|
 |
     |
|
A link to a faculty list at Robert H. Smith School of Business, University of Maryland: here Market Microstructure Invariants ( pdf )
A model of market microstructure invariance is presented based on the intuition that stocks with high and low levels of trading activity differ in the rate at which the time clock generating trading activity ticks. This model as well as two alternative benchmark models are consistent with traditional adverse selection and inventory models. The model is tested using a database of over 400,000 portfolio transition trades provided by a leading vendor of portfolio transition services. In a pooled cross-section of stocks, holding volatility constant, the model of market microstructure invariance predicts that a one percent increase in trading activity leads to a decrease of two-thirds of one percent in mean portfolio transition order size as a fraction of expected daily volume, leads to no change in the higher moments of the distribution of order size, leads to an increase in market impact costs (measured in basis points) of one-third of one percent for trading a given percentage of expected daily trading volume, and leads to a decrease in bid-ask spreads of one-third of one percent. Compared with the predictions of alternative models, empirical results match closely the predictions of the model of market microstructure invariance, with order size conforming closely to a log-normal distribution. The result is a simple empirical transaction cost formula for stocks, which estimates market impact and bid-ask spread costs as a function of dollar trade size, average daily dollar volume, and volatility. Trading Game Invariance in the TAQ Dataset ( pdf )
Our paper tests the theory of trading game invariance using the sample of unsigned trades from the Trades and Quotes dataset from 1993 to 2008. We examine two predictions concerning trading patterns. First, the number of trades should vary across stocks proportionally to their trading activity in 2/3 power. Second, the distribution of trade sizes as a fraction of trading volume should vary across stocks proportionally to their trading activity in -2/3 power. Our results support the invariance theory, though the evidence is distorted by various market frictions such as relative tick size, minimum lot size, clustering of trades, and order shredding. For the number of trades, the power coefficient is equal to 0.69 (with standard errors of 0.002) but, following a reduction in tick size in 2001 and a consequent spread of algorithmic trading, the estimate increases to 0.79 (with standard errors of 0.011). The distribution of trade sizes for individual stocks varies with the trading activity in a manner predicted by the invariance theory as well, i.e., when trade sizes are adjusted for differences in trading activity, their distributions are stable across stocks. These distributions are similar to the distribution of a log-normal variable truncated from below at the 100-share threshold. Portfolio Transitions and Stock Price Dynamics ( pdf ) This paper employs a proprietary data set of portfolio transitions to analyze the short run price-volume relation and its association with the long run performance of newly hired and terminated managers. Unique to our study is its focus on price dynamics that is not affected by potential endogeneity of trading decisions. In the short run, purchases of new stocks induce permanent price increases. Such price changes are especially pronounced for large orders as well as for stocks with a high degree of information asymmetry and negative past returns. In contrast, sales of legacy stocks induce only transitory price declines. In the long run, the evidence shows that institutional sponsors are able to hire managers that are, on average, more skilled than the terminated ones. The apparent benefits of portfolio transitions, however, do not exceed transaction costs. Selection Bias in Liquidity Estimates ( pdf ) This paper studies the trading costs estimated from the data on portfolio transitions. Our estimates avoid a selection bias problem, which is endemic to most estimates based on price changes and quantities traded. Since traders often employ price dependent strategies and cancel expensive orders, the conventional estimates tend to overestimate available liquidity. We find that the liquidity is lower than it is usually believed, especially in high volume markets, as illustrated by the Flash Crash in May 2010. High trading costs have implications for the assessment of the viability of trading strategies, the performance of money managers, and the actual limits to arbitrage in financial markets. Our bias-free estimates also allow us to calibrate the existing measures of liquidity and to examine the degree of non-linearity in market impact functions. Optimal Trading Strategy and Supply/Demand Dynamics ( pdf )
The supply/demand of a security in the market is an intertemporal, not a static, object and its dynamics is crucial in determining market participants' trading behavior. Previous studies on the optimal trading strategy to execute a given order focuses mostly on the static properties of the supply/demand. In this paper, we show that the dynamics of the supply/demand is of critical importance to the optimal execution strategy, especially when trading times are endogenously chosen. Using a limit-order-book market, we develop a simple framework to model the dynamics of supply/demand and its impact on execution cost. We show that the optimal execution strategy involves both discrete and continuous trades, not only continuous trades as previous work suggested. The cost savings from the optimal strategy over the simple continuous strategy can be substantial. We also show that the predictions about the optimal trading behavior can have interesting implications on the observed behavior of intraday volume, volatility and prices. Are Ex Ante Price Impact Functions Linear? |
|
    |
|||
 | ||||||||