This website requires JavaScript.

Measuring price impact and information content of trades in a time-varying setting

F. CampigliG. BormettiF. Lillo
Dec 2022
摘要
The estimation of market impact is crucial for measuring the informationcontent of trades and for transaction cost analysis. Hasbrouck's (1991) seminalpaper proposed a Structural-VAR (S-VAR) to jointly model mid-quote changes andtrade signs. Recent literature has highlighted some pitfalls of this approach:S-VAR models can be misspecified when the impact function has a non-linearrelationship with the trade sign, and they lack parsimony when they aredesigned to capture the long memory of the order flow. Finally, theinstantaneous impact of a trade is constant, while market liquidity highlyfluctuates in time. This paper fixes these limitations by extending Hasbrouck'sapproach in several directions. We consider a nonlinear model where we use aparsimonious parametrization allowing to consider hundreds of past lags.Moreover we adopt an observation driven approach to model the time-varyingimpact parameter, which adapts to market information flow and can be easilyestimated from market data. As a consequence of the non-linear specification ofthe dynamics, the trade information content is conditional both on the locallevel of liquidity, as modeled by the dynamic instantaneous impact coefficient,and on the state of the market. By analyzing NASDAQ data, we find that impactfollows a clear intra-day pattern and quickly reacts to pre-scheduledannouncements, such as those released by the FOMC. We show that this fact hasrelevant consequences for transaction cost analysis by deriving an expressionfor the permanent impact from the model parameters and connecting it with thestandard regression procedure. Monte Carlo simulations and empirical analysessupport the reliability of our approach, which exploits the completeinformation of tick-by-tick prices and trade signs without the need foraggregation on a macroscopic time scale.
展开全部
图表提取

暂无人提供速读十问回答

论文十问由沈向洋博士提出,鼓励大家带着这十个问题去阅读论文,用有用的信息构建认知模型。写出自己的十问回答,还有机会在当前页面展示哦。

Q1论文试图解决什么问题?
Q2这是否是一个新的问题?
Q3这篇文章要验证一个什么科学假设?
0
被引用
笔记
问答