Horizon to adopt innovative methodology for market liquidity detection
Horizon Software, a prominent player in electronic trading solutions and algorithmic technology, has announced it is backing a new methodology to identify liquidity.
The company unveiled a new research paper titled “Uncovering Market Disorder and Liquidity Trends Detection,” spearheaded by Yadh Hafsi, a PhD candidate at Université Paris-Saclay.
This paper introduces an innovative methodology for detecting significant changes in market liquidity, a crucial factor in the trading industry.
Refining the identification of liquidity fluctuations in financial markets
The research primarily aims to refine the identification of liquidity fluctuations in financial markets. Liquidity, the ease or difficulty of buying or selling assets like stocks without causing major price shifts, is a vital aspect of trading. Accurately recognizing changes in market liquidity is key for traders, investors, and financial institutions.
Horizon Software has not only supported Yadh Hafsi’s research but also plans to integrate his findings into their product offerings. This integration will enable end users to see these liquidity signals during their trading decisions and incorporate them into Horizon’s algorithms to enhance execution efficiency. Horizon is actively engaging with clients and users to tailor the research application to their specific needs, enhancing the technological benefits of this methodology.
Olivier Masdebrieu, Horizon’s Chief Technology Officer, expressed enthusiasm about making the paper publicly available and the integration of its findings into Horizon’s products, particularly the ‘Horizon Extend’ platform. Horizon is a cross-asset electronic platform designed for principal and agency trading, and the integration of these new features is expected to significantly improve execution performance.
“It is great to see this paper available to all and even better that Horizon was able to support Yadh Hafsi in his research, utilizing our platform ‘Horizon Extend’, a cross-asset electronic platform for principal and agency trading. We are excited to integrate these features, which will improve execution performance, and are very keen to present the usability of these findings to our global client bases. We’re looking forward to supporting the development of Yadh’s next paper, as part of our commitment to utilizing all avenues, academic and industry-based, to developing forward-thinking market insights to feed into our products,” said Masdebrieu.
Yadh Hafsi, the leader of this research project and a PhD candidate at Université Paris-Saclay, said: “This paper contributes significantly to our understanding of liquidity dynamics in financial markets. My professors and my colleagues have been very helpful, and we will continue working on this domain in the next year.”
Sylvain Thieullent, CEO of Horizon Software, underlined the firm’s dedication to providing advanced solutions that enable market participants to navigate liquidity fluctuations more effectively and confidently. “Our collaboration on this research paper reflects our dedication to providing cutting-edge solutions that empower market participants to navigate liquidity fluctuations with greater precision and confidence. By leveraging advanced methodologies, we aim to equip our clients with the tools they need to make informed decisions and thrive in dynamic market environments.”
Horizon Software, known for its proficiency in Market Making, Agency Trading, and Algo Trading Technology, has been empowering capital market players for over two decades.
The company’s electronic trading platform, certified by B Corp, provides direct connectivity to more than 80 exchanges worldwide. Horizon enables clients to create, test, and implement automated trading strategies in real-time, adhering to its ‘Trade Your Way’ philosophy. The platform’s integration capabilities through rich APIs and the confidentiality of traders’ proprietary strategies further enhance its value in the trading industry.
“Uncovering Market Disorder and Liquidity Trends Detection”
The research paper titled “Uncovering Market Disorder and Liquidity Trends Detection” focuses on developing a new methodology to detect notable changes in liquidity within order-driven markets. This paper is significant for several reasons:
- Objective and Methodology: The primary objective is to develop a method for dynamically quantifying the level of liquidity of a traded asset using its limit order book data. The approach employs Marked Hawkes processes to model trades-through, which serve as a proxy for liquidity. The aim is to accurately identify moments of significant increase or decrease in liquidity intensity, using a minimax quickest detection problem approach for unobservable changes in a doubly-stochastic Poisson process.
- Importance of Liquidity in Trading: Liquidity is crucial for the efficient functioning of markets. It is defined as the ability of an asset to be traded rapidly, in significant volumes, with minimal price impact. Measuring liquidity involves considering time, volume, and price, including aspects like the tightness of the bid-ask spread, the depth of the limit order book, and its resilience. This measure is essential for capturing transaction costs and assessing the depth accessible to large market participants.
- Liquidity Regime Change Detection Methodology: The paper introduces a novel liquidity regime change detection methodology. This method assesses the resilience of an order book using tick-by-tick market data and is aimed at understanding the dynamics of liquidity changes and their impact on the distribution patterns of the liquidity proxy.
- Identification of Liquidity Regimes: After the modeling phase, the proxy is used to identify intraday liquidity regimes. This involves detecting when the distribution of liquidity undergoes changes as fast as possible, a process known as “Quickest change-point detection” or “Disorder detection”.
- Distinguishing Liquidity Regimes: The methodology enables the distinction between different liquidity regimes by detecting changes or disruptions in the distribution of the liquidity proxy, represented by the number of trades-through. This involves a sequential detection methodology that compares the distribution of observations to a predefined target distribution, with the objective to detect changes rapidly while minimizing false alarms.