Home
My five-year journey to develop BTQuant began during a difficult personal period when I channeled my energy into exploring automated crypto trading. Having briefly mined Bitcoin in late 2013, I was curious about automation possibilities in this space.
With no interest in manual trading, I began researching and discovered "BlueBot." I spent considerable time learning and optimizing settings, eventually achieving some positive results—including a trade that turned a 3% target into over 450% by following market momentum for over a week. Later, I learned that BlueBot was actually an abandoned project commonly used to attract beginners.
This experience led me to explore leveraged trading and test various trading bots. Each platform I tried had limitations—some were overly complex, others responded too slowly to market conditions, and many had implementation issues.
I experimented with Freqtrade, but despite creating detailed strategies, I found the system's response time insufficient for my needs. Hummingbot showed promise with its Marketmaker Strategy on spot markets after considerable experimentation, but its increasingly complex codebase presented challenges even for its developers.
Backtrader came next—initially daunting but impressively efficient with minimal code. During a personal setback, I temporarily returned to Freqtrade but remained concerned about its performance for high-frequency trading due to REST API dependencies. My attempts to build from scratch repeatedly encountered obstacles with CCXT and other middleware components. Troubleshooting these issues consumed more time than actual trading development.
A turning point came when I discovered a GitHub repository connecting Backtrader with CCXT, inspiring me to create something similar but customized to my specific needs. After developing several iterations over the years, BTQuant emerged. The goal was straightforward: develop a system combining dollar-cost averaging with high-frequency trading that offered accurate backtesting and could respond to market changes quickly. I wanted reliable results accounting for practical considerations like slippage and fees without requiring extensive additional configuration. Trading inevitably involves losses—something I've experienced firsthand. Each setback provided valuable lessons that informed subsequent development decisions.
Backtrader's decade of development by experienced professionals, its adoption by professional trading firms, and its use in numerous quantitative trading studies suggested it would be a dependable foundation requiring less troubleshooting than alternatives. Im pleased to share BTQuant—a framework designed to enhance quantitative trading capabilities. Im grateful for the contribution of Jackrabbit Relay, which was instrumental in realizing this vision by providing the reliability and performance needed. BTQuant remains under active development, with several features planned for the future:
-
Support for alot of additional exchanges ✅ Done.
-
Implementation of MicrosoftSQL database for improved data handling ✅ Done.
-
More comprehensive examples for new users ✅ Done.
-
Prebuilt solutions for common trading configurations (Simple to Machine Learning) ✅ Done.
-
Tweaked for Speed, get the maximum, for minimum hardware consumption ✅ Done.
-
Missing something along the list? Want something? Let me know!
As development continues, BTQuant will evolve to offer increasingly robust solutions for quantitative traders seeking reliability, responsiveness, and accuracy.
