The impact of machine learning in the retail financial market

Otto Hooper


Supervised by Oktay Karakus; Moderated by Bailin Deng

The purpose of this project is to present academic findings based on the possibilities of analysing the financial market, utilising machine learning (ML) tools in favour of the retail investor. Retail investors are defined as non-professional investors who trade financial securities in their spare time as a hobby. The information used to make a prediction in the financial markets in this day and age is not scarce, instead, information is widely accessible and free through services such as Yahoo Finance, Bloomberg.com, Investopedia.com, et cetera. Retail investors, in most cases, are unable to profit based on their own sound judgement and investments. E.g., in 2020, forty-four percent (44%) of new retail investors had their goal fixated on which financial security they could profit from through a short period.

Even if a retail investor conducts proper due diligence and has good sound judgement on an equity, in most cases, they end up exiting or entering the trade too early or too late. Multiple factors such as investing psychology, trends, news, historic price patterns, and so forth often cause poorly executed trades. On the other side, financial institutions such as banks and pension funds, utilise primarily their own ML tools and other expensive platforms such as the Bloomberg Terminal. These institutional investors have large neural networks that are utilised in their own proprietary software to execute millions of trades each day, and scan for potential investments based on multiple different factors; larger trades are executed directly by traders utilising accessible information, often through Bloomberg while following strict guidelines by their managers. Retail investors have little to no access nor knowledge of these tools, therefore unable to utilise them.

The newfound modern problem in the financial markets is the ever-increasing instability caused by a change of market control and authority. Retail traders in January 2020 consisted of 17.1% of the U.S market. Looking back in 2019, retail investors consisted of around 14.9% of all order flow; in 2010, they made up around 10.1%. Now, what if every retail investor had ML tools that both executed and presented trade opportunities? Abstractly speaking this would indirectly cause retail investors to become institutionalised into strategic groups for different investing strategies. Long-term investing would be based on fundamental values, seeking long-term yield and dividends where ML would crawl over every single earnings- and news report for any possible change of true asset value. Short-term investing would rely heavily on the controversial interval patterns of pricing, often known as technical analysis. The true value would reach a point where no one would profit or lose since supply and demand would cease to exist. Conceptually that means the financial market would cease to exist for the purpose of selling for profit and buying for future profit- instead, it would solely focus on future dividend yield. Hypothetically this would provide a more stable financial market. The short-term effects of this can only be described as a gradual transition towards “financial freedom”, increasing the amount of asset wealth held by retail investors.

I would like to conduct research into existing ML models that institutional investors use and examine if there are any accessible tools for retail investors and how well they perform in the market. A case study to see how many people would be willing to invest in financial securities such as commodities, shares, et cetera, if they had access to tools and insights similar to what institutional investors have. Can this validate their investment ideas with the assistance of ML? I would require ethical approval from the university. The outcome of the project would be to produce research and a project that predicts the market to prevent different negative outcomes caused by ML impact over time when gets more adopted by the retail market.

Initial Plan (06/02/2023) [Zip Archive]

Final Report (12/05/2023) [Zip Archive]

Publication Form