Deployment

The deployment of this project utilized Streamlit. To deploy with Streamlit, you’ll have to access either the local-storage or cloud-storage directory through Terminal, depending on your chosen setup configuration. Within this directory, you’ll find a file named app.py, which serves as the deployment script for the application.

Using the Portfolio Optimizer

Execute the below in Terminal to deploy the application:

streamlit run app.py

You should see an application interface that looks like the below.

User Input

This application accepts four key inputs from the user (see About > Stock Prediction Model > Input Parameters for more details):

• List of Stock Tickers (manual or CSV):

  • Manual input - list of tickers seperated by a comma and a single space

  • Upload CSV file - this file must have one column named tickers that contains the stock symbols

• Portfolio Settings:

  • Risk-free rate - users can be enter a value by typing in a number or by using the plus(+) and minus(-) buttons to increment up and down

  • Shorting allowed - users can choose True or False by selecting an option from the dropdown menu

  • Maximize returns or minimize risk - users can choose maximize returns or minimize risk by selecting an option from the dropdown menu

OPTIONAL: Database Settings - when activated, this feature enables users to fetch fresh data from APIs, import it into the database, and execute transformations to generate dbt models, all without performing portfolio optimization

Generated Output

Once you’ve entered the essential parameters, you should choose Run - Manual Input to input tickers manually or Run - CSV File to load tickers from a CSV file. Upon successful execution, you’ll observe the asset allocation outcomes for each ticker along with a plot demonstrating Markowitz’s Efficient Frontier utilizing the Sharpe Ratio and Capital Allocation Line (CAL).