EduTrade

Inspiration

“A 2024 report from the FED: most Americans in the lower 90% of income earners feel adrift when managing their investment portfolios. More alarming is that two-thirds of non-retirees say they're not on track in saving for retirement.” We want to build an app that democratize financial portfolio management knowledge by ensuring an easily accessible UI for the older generation and new investors alike to manage their investment portfolios.

What it does

EduTrade is a WebApp that asks the user for basic input: portfolio size and time to retirement. Inspired by Wolfram's quadratic optimization, we implemented quadratic optimization to optimize for select best risk/reward combination on the risk/reward frontier given the user's risk tolerance and present an array of stocks along with their portfolio weights.

How we built it

For the backend, we relied mainly on Python and Flask, translating Wolfram code ot, YFinance ,and Matlab for graph plotting. We also used pandas for the calculation for stocks. As for the front-end, we used React, Tailwind CSS, Vite, firebase for data storage, and Google Gemini API.

Challenges we ran into

Due to time limits, we could not fully sync our statistical models directly to our website. On the front end side, we ran into a lot of bugs when building the queries, and we were forced to leave many features such as good authentication, did not design all of our intended features.

Accomplishments that we're proud of

As first time hackers, we are proud we were able to produce a full-stack project within the 36 hours given. Additionally, we are proud to have used two Postmodern Portfolio theory statistical models for an open-source application. Usually, this information is locked away behind a lofty wealth advisor consulting fee or an investor accreditation to invest in a hedge fund.

What we learned

To make our tool highly relevant, we looked over Wolfram documentation and learned about how mathematical optimization models can be used in statistical analysis. By translating wolfram code to python code, we learned several new languages.