High-Frequency Lead-Lag Effects and Cross-Asset Linkages: A Multi-Asset Lagged Adjustment Model

Abstract

Motivated by the empirical evidence of high-frequency lead-lag effects and cross-asset linkages, we introduce a multi-asset price formation model which generalizes standard univariate microstructure models of lagged price adjustment. Econometric inference on such model provides: (i) a unified statistical test for the presence of lead-lag correlations in the latent price process and for the existence of a multiasset price formation mechanism; (ii) separate estimation of contemporaneous and lagged dependencies;(iii) an unbiased estimator of the integrated covariance of the efficient martingale price process that is robust to microstructure noise, asynchronous trading and lead-lag dependencies. Through an extensive simulation study, we compare the proposed estimator to alternative approaches and show its advantages in recovering the true lead-lag structure of the latent price process. Our application to a set of NYSE stocks provides empirical evidence for the existence of a multi-asset price formation mechanism and sheds light on its market microstructure determinants.

Microstructure and High-Frequency Price Discovery in the Soybean Complex

“…Following Buccheri et al (2019), Buccheri, Corsi & Peluso (2021) and Buccheri, Bormetti, Corsi & Lillo (2021), we assume that this discretized multivariate dynamics is latent in a high-frequency setting and thus inaccurately observed on account of the ubiquitous microstructure noise present in financial markets. A state-space representation is thus fully justified, with the transition equation following from expression (18):…”

“…The state-space representation of the vector error correction model (VECM) described in Seong et al (2013) considers the Expectation Maximisation algorithm proposed by Dempster et al (1977) to cope with mixed-frequency or asynchronous data in cointegrated time-series models. More recently, Buccheri, Corsi & Peluso (2021) demonstrated that this filtering methodology adequately deals with microstructure noise and the information lag that exists among markets at the high-frequency level 5 . Employing a slightly 4 The microstructure frictions can be associated to the bid-ask bounces, the discreteness of the price grid but also the technique used to construct the high-frequency price dataset (Hansen & Lunde 2006).…”

Inference for time-varying lead–lag relationships from ultra-high-frequency data

“…A related model has also been studied in Robert and Rosenbaum (2010) by utilizing the random matrix theory and Ito and Sakemoto (2020) by multinomial dynamic time warping. Other approaches to investigate lead-lag relationships in a continuous-time framework include Hawkes process-based models (Bacry et al 2013; Da Fonseca and Zaatour 2015), a wavelet-based method (Hayashi and Koike 2018), and a multi-asset lagged adjustment model (Buccheri et al 2020). Several empirical approaches have been proposed, as well; see Pomponio and Abergel (2013) and Dobrev and Schaumburg (2016), for example.…”

Comment on: Price Discovery in High Resolution

“…The main difference is that we introduce the lagged adjustment mechanism, which allows to describe lagged absorption of information across different markets. The proposed model belongs to the class of "Multi-Asset Lagged Adjustment" (MLA) models introduced by Buccheri et al (2018a). However, while they considered the case of several assets traded in the same market, we consider a different problem, namely the case of one asset traded on several markets.…”