Getting historical intraday financial data can be a pain, especially for non-US markets. If you have deep pockets you can simply buy the data you need, but for retail investors the cost is prohibitive. If you want historical transaction-level data for the Brazilian stock market, for instance,

In this series of posts I’m trying some of the ideas in the book Advances in Financial Machine Learning, by Marcos López de Prado. Here I tackle an idea from chapter 5: fractional differencing. the problem Stock prices are nonstationary - their means and variances change systematically over time.