Local-first multi-broker portfolio rebalancer

The challenge

A portfolio spread across brokers, accounts, and currencies is difficult to compare with a single target allocation. The practical question is not only what is out of balance? but also where should the next contribution go?

The approach

This local-first Python application consolidates manually entered, Interactive Brokers, and Questrade portfolio data. It normalizes holdings across accounts and currencies, measures current allocation drift, and calculates buy-only contributions that move the overall portfolio closer to its target model.

The application includes:

• a tested domain layer for portfolio and rebalancing rules;
• SQLite persistence that keeps financial data local;
• command-line and browser interfaces over the same core behavior; and
• optional synchronization of foreign-exchange rates.

Engineering focus

Portfolio calculations live outside the interface and persistence layers, making them easier to test and reuse. Broker-specific inputs are normalized before they reach the domain model, and optional network synchronization does not prevent the application from working locally. That separation keeps a growing set of integrations from leaking into the rebalancing logic.