GTU Publishes Study Modeling Risks in Crypto Asset Markets

December 1, 2025 - GTU Office of Press and Public Relations
 

The study examining the extreme behavior of Bitcoin returns within the framework of Extreme Value Theory (EVT) is published by Springer Nature.

Assoc. Prof. Erhan Uluceviz, a faculty member at the GTU Faculty of Business Administration, has published an important research study that scientifically models extreme price movements in crypto asset markets. The book chapter titled “Assessing Bitcoin Return Extrema in the Context of Extreme Value Theory,” published by Springer Nature within the series Contributions to Finance and Accounting under the title “Machine Learning in Finance: Trends, Developments and Business Practices in the Financial Sector,” offers a strong contribution to the finance literature by analyzing the tail behavior of Bitcoin returns using the EVT approach.

Central Research Question: How can Bitcoin’s extreme price movements be modeled within the framework of EVT?

Crypto asset markets generally exhibit high volatility and sudden price jumps. Therefore, accurately modeling extreme price movements is critically important for individual investors, institutional actors, and algorithmic trading systems that conduct algorithmic trading operations.

This study examines hourly Bitcoin price data from the period January 2018 – July 2024 and analyzes weekly

• maximum
• minimum
extreme returns.

The research findings demonstrate that both the positive and negative tails of Bitcoin returns can be successfully modeled using Generalized Extreme Value (GEV) Distributions.

Potential Added Value for Financial Markets:

1. A New Framework for Scientific Risk Forecasting
The EVT framework, widely used in fields such as climate science, finance, insurance, engineering, and environmental sciences, can also be effectively applied to crypto asset risk management.

2. Weekly Forecasting Capacity
The approach of constructing a weekly risk horizon by using hourly data can be considered an innovative method for managing short-term volatility in crypto assets.

3. Input for Algorithmic Trading and Portfolio Management Systems
The parameter and distribution estimates obtained through the EVT method are suitable for use in high-frequency trading strategies and AI-driven buy–sell models.

With the introduction of Bitcoin ETF products into global stock exchanges, crypto assets have become an important component of traditional global capital markets. This development further increases the importance of extreme value analyses.

This study serves as a valuable reference for both academic circles and market practitioners in understanding the risks arising from crypto assets.

Publication Details:

Contributions to Finance and Accounting (Springer Nature, 2025)

DOI: 10.1007/978-3-031-83266-6_11
 

Contact

Gebze Technical University – Science Communication Office