Algorithms for time series forecasting relevant to equity and exchange-traded fund (ETF)

Problem to be solved

STRATxAI Technologies Limited (t/a STRATxAI) specialise in data driven smart portfolio construction, powered by its proprietary Platform – Alana. Alana processes 8 billion+ data points daily, leveraging advanced algorithms and quantitative methods to optimize portfolio construction and power STRATxAI portfolios. Combining their domain expertise in quantitative investing and machine learning, the platform enables financial professionals to leverage the power of data, without having the need to code themselves. STRATxAI, in collaboration with the global investment platform eToro, launched the Cutting-Edge Smart Portfolio, which recently reached 1-year of live performance.
STRATxAI approached the Nimbus Gateway to access their data science expertise. They are interested in applying a series of state-of-the-art data driven machine learning (ML) models to construct smart portfolios across the equity and exchange-traded fund (ETF) space.

Solution

The Nimbus Gateway Data Science Research team recommended and documented a scalable machine learning model solution approach incorporating preprocessing, automatic model hyperparameter optimisation, validation and testing. The solution was implemented in Python with maintainability and extensibility in mind and can be expanded upon in future work.

Impact for the company

This project assisted the company to address challenges associated with AI model portfolio optimisation with the goal of freeing up resources for advisors to focus on client relationships and business growth, all while managing portfolios more effectively.

“The independent analysis and extended validation and development of our artificial intelligence project by the Nimbus team greatly increased our confidence factor and gave us a much-improved understanding on a potential scalable ML approach"
Dr Paul Clifford
CEO StratxAI