Webinars This Blog

Predicting Stock Trends, Statistical Arbitrage & Sentiment Analysis | Algo Trading Projects

4 min read

Complete Recording

Check out the complete recording of the webinar here:


About the Presentations

There has been limited research done on the application of machine learning models in the South African Market. Studies in this market have primarily focused on the more traditional fundamental and technical analysis. This project looked to expand on previous work by utilizing six common technical indicators along with a machine learning algorithm to the top ten constituent stocks in the South African Top40 Index.

The aim of the study was to investigate whether using a machine learning model, for a long only trading strategy, was able to produce a superior return to that of the market. In addition, we also compare the returns produced from our model to that of the more conventional technical indicators, namely: Bollinger Bands, RSI and MACD.

Using ten years’ worth of daily stock price data along with the resulting technical indicators, we utilized the first 7.5 years as training data in our Random Forest Classifier. Following the trade signals generated by the Random Forest Classifier, we looked at two long-only trading strategies. The one strategy looked at a daily rebalancing of our portfolio and the other weekly. Our findings show that over the last 2.5 years (test data), both long-only trading strategies outperformed the SA TOP40 index (benchmark) as well as the traditional technical indicators.

About the Speaker: Desigan Reddy (Fixed Income Dealer)

Desigan has been working in the financial services industry for about 12 years. He began his career in 2007 as a market risk analyst at Nedbank Capital where he spent the better part of 6 years. He then left Nedbank in 2014 to take up a quantitative analyst position at Absa Capital. In 2015, he joined OMSFIN (Old Mutual Specialized Finance) in Cape Town as a risk analyst before joining Futuregrowth in 2016 to take up his current role as a Fixed Income Dealer.

During his time in the industry, Desigan has managed to obtain both his Chartered Financial Analyst (CFA) and Financial Risk Management (FRM) designations. In addition to this, he has also obtained his ACI Dealing Certificate and more recently he has completed his Master’s Degree in Commerce (specializing in Finance and Risk Management) from the University of Cape Town.

Desigan completed the EPAT programme in April 2019 and although his project remains the highlight of the course, he cites the knowledge he gained from learning to code in python as one of his most valuable skill sets and one that he has not only developed but has stated implementing in his current role at Futuregrowth.

On a more personal note, Desigan enjoys going to the gym and is an ardent football and rugby fan. He enjoys action/comedies and his favourite genre of music is pop/rock.


Project 2: Statistical Arbitrage: Pair Trading In The Mexican Stock Market by Javier Cervantes

The Mexican stock exchange has a relatively small market capitalization given the size of the economy. It has very few issuers and is highly concentrated. As a direct consequence, there are relatively few investors involved in this market.

Javier's project posits that this particular set of conditions might provide the ideal environment for an arbitrage strategy. He puts this hypothesis to the test by using a Statistical Arbitrage model. This strategy seeks to profit from the mean reverting nature of two otherwise similar companies whose prices have deviated significantly from their mean.

About the Speaker: Javier Cervantes (Corporate Bond Trader)

Javier Cervantes has over 8 years of experience trading in the credit markets, specializing in MXN-denominated corporate debt. He is currently a corporate bond trader in the MXN desk at BCP Securities in Mexico City.

Javier is also a CFA charterholder, having obtained his charter in August 2017. He holds a BS in Economics from Instituto Tecnológico Autónomo de México (ITAM). Javier enrolled in QuantInsti's EPAT programme in 2016 looking for new tools to enhance his expertise in the ever-evolving world of financial markets.


Project 3: Using Sentiment Analysis To Trade Equities by Siddhant Vaidya

The market is driven by people, and people are driven by emotions. We come across numerous events where sentiments have been more influential in driving a stock up or down than any other factors pertaining to the fundamental or technical aspects of the stock. A few examples of such events: the launch of a new iPhone, cryptocurrency mania, Tesla launching new cars, senior executives resigning or getting fired, etc.

I believe at any point in time if we are able to gauge the sentiment of an asset in the market, we will be able to successfully trade that asset profitably. The best way to assess the market sentiment is by analyzing the main source of information that people consume; news.

About the Speaker: Siddhant Vaidya (Research Analyst)

Siddhant is working as a Research Analyst intern while he completes his MS Finance program at the University of Rochester. He also leads the student-run Meliora Fund at his school, where he and his team are working on Automated Forecasting and DCF Valuation tool. His affinity for solving business problems by combining technology and finance is what drove him to a career in the quantitative finance domain.

Prior to this, Siddhant has more than 2 years of experience as an Equities Trader and a year’s experience as a Data Analyst. He holds a Bachelor’s degree in Electronics and Telecommunications Engineering. Siddhant is passionate about robotics, automobiles, and aviation, and enjoys playing poker in his free time.


Read more EPAT Projects here!


#10YEARSOFQUANTINSTI

As they say, knowledge is the greatest gift in life. While celebrating our 10 years of existence, we’ve planned this series to thank our community. We are thankful to you from the bottom of our heart for showing the love & support in our journey over the years!


The algo trading course by QuantInsti covers training modules like Statistics & Econometrics, Financial Computing & Technology, and Algorithmic & Quantitative Trading that equip you with the required skill sets for applying various trading instruments and platforms to be a successful trader. Be sure to check out the Executive Programme in Algorithmic Trading (EPAT) today! Enquire now!


This session was conducted on:
Thursday, October 15, 2020
9:30 AM ET | 7:00 PM IST | 9:30 PM SGT

Live Q&A | Skills to Get Quant Jobs