A new study proposes a novel method for pricing options on Bitcoin using neural networks.
The model incorporates Bitcoin (BTC) price dynamics and sentiment data to capture cryptocurrency markets’ frequent jumps and volatility clustering.
Researchers developed a bivariate jump-diffusion model based on past prices and Google Trends sentiment to describe Bitcoin price evolution. They derived an extended Black-Scholes equation for valuing Bitcoin options. The pricing partial differential equation was then solved numerically using neural networks.
“Neural networks provide a flexible parametric approach based on their universal approximation theoretical results.”
An extract from the study
The model was tested on highly volatile stocks like Tesla since active crypto options markets are still developing. Results showed average absolute pricing errors of around 3%, demonstrating viability.
This paper aims to provide the initial modelling foundation, which can be improved incrementally as research progresses.
An extract from the study
The jump-diffusion model enables financial applications like risk management, derivatives pricing, and portfolio optimization in the emerging cryptocurrency space. The researchers note arbitrage opportunities and market inefficiencies in crypto may require departing from traditional models that assume efficiency and no arbitrage.
As the crypto space evolves, enhanced models can capture microstructure details empirically. Nonetheless, this jump-diffusion methodology offers a starting point for financial engineering tailored to cryptocurrencies.
We welcome future research to build on these initial techniques for pricing and valuation of cryptocurrency-denominated assets.
An extract from the study