THESIS
2022
1 online resource (ix, 68 pages) : illustrations (chiefly color)
Abstract
In market microstructure literature, informed trading is a phenomenon where traders with private
information trade profitably against market makers who then provide liquidity at a loss. Developed
by Easley et al. (1996), the Probability of Informed Trading (PIN) is a widely used measure of informed
trading due to its explanatory ability for various market states such as spreads and volatility.
With the increasing prevalence of high frequency trading (HFT), the PIN measure was updated to a
volume-based measure called Volume-synchronized Probability of Informed Trading (VPIN). Despite
their widespread applications, both the PIN and VPIN measure have been scrutinised for their
theoretical and computational problems. This thesis addresses several of these issues and makes
two contributions....[
Read more ]
In market microstructure literature, informed trading is a phenomenon where traders with private
information trade profitably against market makers who then provide liquidity at a loss. Developed
by Easley et al. (1996), the Probability of Informed Trading (PIN) is a widely used measure of informed
trading due to its explanatory ability for various market states such as spreads and volatility.
With the increasing prevalence of high frequency trading (HFT), the PIN measure was updated to a
volume-based measure called Volume-synchronized Probability of Informed Trading (VPIN). Despite
their widespread applications, both the PIN and VPIN measure have been scrutinised for their
theoretical and computational problems. This thesis addresses several of these issues and makes
two contributions. In particular, we use the Geometric Poisson distribution in place of the Poisson
distribution to model trade size endogenously, and we develop a Bayesian-based approach to detect
periods where information events occurred. Doing so is integral to making accurate estimates of
the key parameters in PIN. Our findings also provide insights on the usefulness of the parameters
estimated.
Post a Comment