HFT Trader is a high-frequency trading (HFT) system implemented in C++ to analyze and trade on financial market data, specifically forex EUR/USD. The system implements basic strategies, collects market data using the Alpha Vantage API, and executes trades. This project aims to simulate a basic automated trading system with performance benchmarking.
- C++ (C++11): Main programming language for implementing the trading system.
- Google Benchmark: Used for performance benchmarking.
- libcurl: Handles HTTP requests to fetch market data from the Alpha Vantage API.
- jsoncpp: Parses JSON data from market data responses.
- Alpha Vantage API: Provides real-time market data for currency pairs like EUR/USD.
- Market Data Fetching: Continuously fetches live market data using
libcurlfrom the Alpha Vantage API, which provides real-time exchange rates for currency pairs. - Simple Strategy: Implements a basic local minimum strategy where the system buys when a local minimum is detected and triggers a take-profit when a set price is reached.
- Execution: Executes trades based on the strategy and provides buy/sell signals.
- Benchmarking: Performance benchmarking to evaluate the efficiency of the trading strategy using the Google Benchmark library.
- Local Minimum Strategy: The system identifies local minima in market prices. A buy signal is triggered when the current price is lower than both the previous and the next price point. The trade is closed when the price reaches a target take-profit level.
The system's performance was benchmarked using the Google Benchmark framework. Below are the key results:
- Average Time per Iteration: Approximately 60 microseconds.
- CPU Time per Iteration: Approximately 30 microseconds.
- Total Iterations: 23,079 iterations during the benchmark.
These results indicate that the trading system is efficient enough for high-frequency trading scenarios, where speed is crucial.
- Performance: The system's execution time per iteration is optimized for real-time trading environments.
- Strategy Effectiveness: The local minimum strategy performs as expected, generating buy and sell signals based on price behavior.
- Benchmarking: The benchmarking results demonstrate the system's capability to handle multiple iterations swiftly, making it suitable for high-frequency trading.