We’re excited to announce that a new quant research paper, co-authored by Kaiko's Michaël Allouche and Emmanuel Gobet, alongside Stephane Girard, has been accepted for publication in the prestigious Machine/Deep Learning journal, Neural Networks! 💡 Why is risk management crucial in crypto? With the rise of regulatory requirements, such as those from the BIS, understanding and managing risk in the #cryptocurrency market is vital for determining the capital needed for solvency. 📊 How do we measure risk? Metrics like Value at Risk (#VaR) and Expected Shortfall (#ES) are essential tools. While VaR estimates the minimum potential losses within a confidence interval, ES measures the average losses beyond that threshold. 🚩 What about extreme events? In regions with little to no data—where extreme losses can occur—traditional methods often fall short. This paper introduces a well-designed, cutting-edge #NeuralNetwork model, leveraging Extreme Value Theory, to estimate risk measures with greater precision and new theoretical guarantees. Tests on highly volatile crypto data demonstrate its superiority over existing methods. Read the paper for free, here: https://lnkd.in/eM5f5Dbh
