[Algo Trading Projects] Predicting Market Movement and Backtesting Trading Strategies

2 min read

Tuesday, January 19, 2021
8:30 AM ET | 7:00 PM IST | 9:30 PM SGT

About the Presentations

Project 1: Predicting Bank Nifty Open Price Using Deep Learning

With the advent of several machine / deep learning models, there have been several theories emerging in applying these techniques for stock market prediction because of the difficulty and complexity it involves. In this project we’re trying to solve the problem using a classifier to predict whether the Bank Nifty index listed in NSE will go up or down, on the next day open.

Project 2: Strategy Backtesting Environment With Interactive Brokers

An important limitation of retail ‘algo’ packages is their inability to model and constrain complex portfolios. The ‘Portfolio Maestro’ software offered by TradeStation helps in this regard, but the portfolio constraints are somewhat rigid, and the user is restricted to US traded securities.

Given these constraints, this project was developed to test and then implement portfolio-based strategies using Interactive Brokers via their IB Gateway portal. Of critical importance to the author was the ability to test and visualize the impact of variable optimizations, ‘alpha’ filters and portfolio constraints.

This project does not seek to present an optimized strategy, rather it provides a framework to implement existing strategies and further explore the many concepts covered by the EPAT programme. Of key personal interest will be the further evaluation of machine learning techniques to optimize input variables and filters, and advanced statistical techniques to refine the portfolio mix.

About the Speakers

Balamurugan Ganesan
(Lead Analyst - Product Management)

Balamurugan Ganesan is a product professional with an MBA in Finance from SP Jain, B. Tech from NIT Calicut and completed CFA Level 3. He has around 12 years of experience as a Business Analyst / Product Manager with a proven track record in building products in the Fintech industry. He has extensively used the latest technologies in building products such as implementing deep learning models in tensorflow / keras, machine learning in scikit learn, statistical analysis tools using Python.

He is very passionate about the Algorithmic Trading domain where he can put all his skill sets to good use and completed the EPAT programme. In EPAT, he was trained by industry experts in the domain of Algorithmic Trading and gained a solid understanding of various real-world implementation techniques in quantitative finance and high frequency trading. Currently he works as a Feature Lead with Bank of America.

Mark Rendle
(Engineering Fellow)

Mark Rendle is a Engineering Fellow at a US fortune 500 company, based in Houston. His focus is on complex problem solving, decision quality and integrated planning. He holds a bachelor’s degree in Mechanical Engineering and a master’s in Business Administration. He has 20+ years of algorithmic trading experience, generally using the TradeStation platform.