BUILDING A STOCK TRADING SIMULATION SYSTEM WITH AUTOMATED TECHNICAL ANALYSIS

Abstract

The complexity of algorithmic trading often precludes individual investors from effectively validating strategies without extensive programming expertise. To address this, this thesis develops a comprehensive stock trading simulation system utilizing a modern technology stack, including Next.js, NestJS, and machine learning algorithms for trend prediction. The methodology integrates a drag-and-drop strategy builder with automated technical indicators, enabling users to execute asynchronous backtesting on historical data via background job processing. Key results indicate that the platform successfully delivers real-time market visualization and detailed performance metrics, such as profit/loss and maximum drawdown, through a scalable and interactive interface. Furthermore, the system incorporates workflow automation and a multi-agent architecture to enhance data processing efficiency and system modularity. The integration of Redis caching and WebSocket communication ensures high performance and real-time user notifications during intensive computational tasks. Ultimately, the project demonstrates a practical application of contemporary software engineering practices—including microservices communication, event-driven design, and containerized deployment—to solve a real-world need in financial strategy testing. Conclusively, this system effectively democratizes access to advanced financial analysis, serving as a robust foundation for future educational and commercial investment applications.