On Nov 13th, 2019, Professor Xue-Zhong He from University of Technology Sydney gave a seminar in Room 337, HSBC Business School Building. The seminar aims to answer the question that whether a possible withdrawal from liquidity provision by high frequency traders (HFT) introduce and intensify market fragility during turbulent times.
The research is motivated by flash crashes in recent years which has triggered heated debate about the relation between market crashes and high-frequency trading, in particular through endogenous liquidity provision. In his research, Professor He and his co-authors build a nonlinear rational expectations equilibrium model of high-frequency endogenous liquidity provision to explore fragile liquidity.
Endogenous liquidity provision applies a cut-off strategy either profit from providing or taking liquidity from the market and speculate on the size of information shock based on their signal. And this paper focus on their switching behavior. The market regime works as follows. Without the existence of high frequency trader, the liquidity demand from informed trader is offset by the liquidity supply from designate market maker (DMM). When there is HFT and the information signal is moderate and below the threshold value, HFT also acts as liquidity supplier. However, when the signal is extreme and go beyond threshold value, the HFT switch to be the liquidity demander and generate the liquidity mismatch as the aggregate liquidity demand by informed trader and HFT excess the liquidity provided by market maker. The model applies enforcement learning. The expectation function is updated as a weighted average of prior expectation and realized expectation. The results show that DMM liquidity supply shrinks for increasing adverse selection and the market crash happen when ELP evaporates.
This paper also provides us with some other interesting findings. Price change is more sensitive to small trading volume but less so to larger trading volume. And price function under the taker regime has much higher slope than under the maker regime. Breaks in liquidity are more severe when signals between ELP and informed trader are more correlated. Information latency in segmented market benefits overall liquidity. When the signal is not so accurate, the impact on market fragility actually decreases.
During his seminar, many interesting questions were raised and followed by thoughtful discussions among Professor He and the audience. The faculty as well as the students are quite intrigued by the topic. We thank Professor He for his excellent presentation.
By Linlin Ma, Associate Professor and XinboWang, MA student
微信編輯：程雲