Comment on: Price Discovery in High Resolution

Abstract

This note is commenting on Hasbrouck (2018). The paper investigates the problem of price discovery on markets with trades recorded at sub-millisecond frequencies. The application of the popular information share measure of Hasbrouck (1995) to such data faces several difficulties, as the underlying VECM would need a huge number of lags to capture dynamics at different timescales. The problem is handled by imposing a set of restrictions on parameters inspired by the Heterogeneous Autoregressive model for realized volatility. We illustrate some potential drawbacks of the information share measure adopted in the paper and propose a modelling strategy aimed at dealing with such limitations. In particular, we introduce a structural multi-market model with a lagged adjustment mechanism describing lagged absorption of information across markets. The advantages of the method are shown in simulations.

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):…”

“…Our model can be cast in a state-space formulation of the VECM model, whereby the idle prices are considered as missing data, unlike inBuccheri et al (2019),Buccheri, Corsi & Peluso (2021) andBuccheri, Bormetti, Corsi & Lillo (2021) 10 . Initially proposed 10 For the sake of completeness, a potential missing-data modification for their algorithm is mentioned in the technical appendix ofBuccheri, Corsi & Peluso (2021).…”

“…as described byBuccheri et al (2019),Buccheri, Corsi & Peluso (2021) andBuccheri, Bormetti, Corsi & Lillo (2021). Nevertheless, our model should not be confused with the model proposed in the latter two contributions, as the information associated with zero returns is treated differently in our model, allowing us to tackle in a different manner the inference biases stemming from the price idleness described inBandi et al (2020).…”