top of page

Case Study: Defining the Future of Trading with Generative AI

In today's fast-paced world of financial trading, having advanced analytical tools can make all the difference. This is precisely what Mindmode aimed to showcase with Finmind, a prototype developed for the Swissquote Trading Day on January 25, 2024, at the Kongresssaal in Zürich.

Financial GPT build with chatgpt
Finmind user interface

Finmind was designed to analyze various financial data sources from APIs. We used news articles, annual reports, and four-week chart histories in this case. As AI tools already exist for different markets, such as the US market, we wanted to focus on data from the Swiss Market Index (SMI), offering users a comprehensive stock movement forecast for companies listed on the Swiss market.

a prototype for the financial market build with chatgpt
Finmind analysis four-week chart histories, news articles, and annual reports

Swiss market index financial tool with gpt-4
Choose a company from the SME

The heart of Finmind's analysis capabilities was GPT-4 Turbo from OpenAI, enabling the prototype to deliver positive developments, potential concerns, and predictions on stock movements with extensive explanations. Our inspiration for building the protoype came from FinGPT which is a large language model designed for the finance sector and is open-source, you can read more about FinGPT on Hugging Face. One of Finmind's standout features was its ability to compare companies, for example, a comparison between Zurich Insurance and AXA Insurance, providing users with a clearer perspective on investment choices.

analysis report of Zurich insurance created with generative AI
An example of a possible analysis report

The presentation at the Swissquote Trading Day was more than just a demonstration of Finmind's current capabilities. It was a glimpse into the future of stock market predictions, highlighting the ongoing improvements and innovative features we will see in the next years to come.

We wanted to show our audience that Finmind should be envisioned as a personal trading assistant capable of handling high complexity, integrating unstructured data, and reducing human bias in investment decisions. One could also imagine giving the model more data to analyze, such as company patents, press releases, SEC/FINMA filings, academic studies, economic reports, and community feeds to get an even more extensive context understanding

of the current situation to make an even better prediction.

possible data sets AI can analysis

In conclusion, with the utilization of generative AI, the future of trading looks promising, giving traders a better tool to support their decisions. Finmind is just an example of how this technology can be harnessed to provide data-driven insights for enhancing investment decisions. If you want to see the full presentation of the Swissquote trading day you can rewatch it on the Swissquote Youtube Page and if you have further questions regarding the prototype feel free to contact us under or book an online appointment through our website.

Generative AI used for trading as a copilot
Generative AI for trading - a personal trading assistanta

Disclaimer: It is important to note that Finmind is not a tool for financial advice. Actual trading should always be approached with caution and, ideally, with the guidance of a financial expert.


bottom of page