Analyzing and exploiting arbitrage opportunities in the cryptocurrency market

The cryptocurrency market has revolutionized finance with decentralized digital assets, creating new investment opportunities and challenges. Among these, arbitrage has emerged as a profitable strategy due to the fragmented nature of cryptocurrency exchanges and the price discrepancies across them. This thesis explores two primary arbitrage strategies: spatial and triangular arbitrage. Spatial arbitrage exploits price differentials across exchanges, while triangular arbitrage identifies cyclical discrepancies among three currency pairs on a single exchange. Our research investigates these strategies by developing a comprehensive framework that utilizes historical cryptocurrency data. Through this framework, we assess the profitability of each strategy across different threshold parameters, optimizing them for maximum effectiveness. Additionally, we created a web application to visualize and analyze profitability metrics, providing a dynamic tool for understanding arbitrage opportunities in both historical and live market conditions. The web application integrates with major exchange APIs, allowing for real-time monitoring and testing of arbitrage strategies in today’s markets. Our analysis reveals that cryptocurrency markets are gradually becoming more efficient, evidenced by a decline in profitable arbitrage opportunities from 2021 to 2023. However, despite reduced frequency, these opportunities persist, particularly during periods of high volatility. This indicates that while the market is maturing, inefficiencies remain exploitable. Real-time testing confirms this, showing continued potential for algorithmic trading strategies aimed at cross-exchange price discrepancies, even when accounting for transaction costs and execution delays. The study concludes that while arbitrage opportunities in cryptocurrency markets endure, the frequency and profit margins are narrowing, highlighting the need for sophisticated, rapid execution strategies and robust risk management. Our live trading platform, developed for real-time application, bridges theoretical analysis with practical use, underscoring the complexities and potential within an evolving financial ecosystem. Future work could incorporate machine learning to enhance predictive capabilities and adapt thresholds dynamically, aiming to capture smaller margins with greater reliability in a market transitioning towards efficiency.