Machine Learning Solutions to Challenges in Finance: An Application to the Pricing of Financial Products
32 Pages Posted: 7 Sep 2019
Date Written: August 31, 2019
The recent fast development of machine learning provides new tools to solve challenges in many areas. In finance, average options are popular financial products among corporations, institutional investors, and individual investors for risk management and investment because average options have the advantages of cheap prices and their payoffs are not very sensitive to the changes of the underlying asset prices at the maturity date, avoiding the manipulation of asset prices and option prices. The challenge is that pricing arithmetic average options requires traditional numerical methods with the drawbacks of expensive repetitive computations and simplified models with non-realistic assumptions. This paper proposes a machine-learning method to price arithmetic and geometric average options accurately and in particular quickly. We show the effectiveness of the new method by carrying out comprehensive numerical experiments. Finally, the method is verified by an empirical test. This empirical test actually shows that the machine learning method provides a new model-free method for asset pricing.
Keywords: Machine learning, Finance applications, Asian options, Model-free asset pricing
JEL Classification: G11, G32
Suggested Citation: Suggested Citation