Standard Price and Volume Bars || Financial Data Structures || Financial Machine Learning

Today we will be exploring the financial data structures as discussed in Advances in Financial Machine Learning by Prof. Marcos Lopez de Prado [2018]. We will create four standard price and volume bars commonly used in financial machine learning within academic literature: time bars, tick bars, volume bars and dollar bars. In this tutorial we are learning how to construct these standard bars in Python from raw trade information on a given product, here we are using Commonwealth Bank of Australia (CBA) equities trading information from my CommSec brokerage. We also discuss the issues of using time bars for financial machine learning in terms of over and under sampling information of low and high-activity periods as well as undesirable statistical properties such as serial correlation, heteroscedasticity, and non-normality of returns. Full code available on my website: ★ ★ 📚 Book Recommendations 📚 ★ ★ 📚 Stochastic Calculus for Finance
Back to Top