wtf is form4lab?

form4lab is a research tool built around one question: when company executives and directors trade their own company's stock, does that trading tell you anything useful? By US law, these insiders must publicly report their trades to the SEC within a few business days on a document called a Form 4. This is the legal, disclosed kind of insider trading, not the illegal kind reported in the news. form4lab automatically collects these Form 4 filings, stores them in a database, and scores the insiders behind them based on their track record. You can then test trading ideas against this historical data through a built in backtester, and optionally run those ideas on a simulated brokerage account using fake money, so nothing real is ever at risk unless you choose to enable it. The project ships with a few simple example strategies to learn from, but it is upfront that these are illustrative only and not expected to make money after costs. Building an actual working strategy is left up to the user. The maintainers are clear this is educational software, not investment advice. Setup uses Docker Compose, running a database, a web dashboard, and a background scheduler that keeps pulling in new filings and updating scores. There is also a path to run it locally with Python 3.11 or newer instead of Docker. Either way, you need to supply an identifying name and email address, since the SEC requires this on every request to its data system. If you want to act on the signals automatically, form4lab connects to Alpaca, a US brokerage that offers free paper trading, letting you watch a full signal to order pipeline play out with simulated money before ever considering real funds. This makes form4lab useful for developers, researchers, or serious individual investors who want to explore insider trading patterns rigorously rather than by gut feeling, while keeping the option to stay purely in research mode with no brokerage connected at all.