Despite their willingness and investment, many asset managers are struggling to establish an efficient and programmatic way to incorporate machine learning and big data into their execution strategies. Our contributors share their perspectives.
Trading with Machine Learning and Big Data
Despite their willingness and investment, many asset managers are struggling to establish an efficient and programmatic way to incorporate machine learning (ML) and big data into their execution strategies. As a result, the percentage of trades executed with artificial intelligence (AI) and big data techniques remains small. Contributors from Virtu Financial and Man Group share their perspectives.
In Chapter 7, the global head of analytics client support at Virtu Financial details real-life examples of ML’s application in analytics and trading, including feature importance, transaction cost analysis, normalized and unbiased algo trading, and a trading strategy recommendation model. The author describes the effective combination of data and ML-based techniques used to enrich decision-support processes, as well as some of the current barriers facing buy-side firms as the technology matures. This chapter will help you piece together a picture of ML and big data applications in the trading landscape.
In Chapter 8, a team of experts from Man Group zooms in on price–time priority limit order book markets, the most common market design at major exchanges trading cash equities, futures, and options. Limit order books (LOBs) provide a wealth of high-frequency data that can be used to develop data-driven execution algorithms using ML methods. LOB data can be used to engineer informative features for the prediction of a range of relevant variables: spreads, trade volume, volatility, and short-term price movements. The authors demonstrate how ML models predicting different variables can be combined into a sophisticated execution algorithm that plans an execution trajectory in both volume and price dimensions.