I'm interested in analyzing the risk for cryptocurrencies and a commodity such as gold. Is there an optimal amount of variables to input into these calculations? I want to quantify the risk of the discount being positive or negative for a commodity, its corresponding future contract, and a cryptocurrency. Additionally, would it be appropriate to then use the Poisson probability distribution and traditionally conditional probability to quantify the risk?
You can try to divide the volatility in two components, the first part you can describe with stochastic process (Wt) + Poisson process(Jump)
As Fabrisio said, one way is to separate(decompose) the process into pure jump process (like classical Poisson) and continuous Gaussian-Wiener process - with stochastic volatility and estimated trend (- semimartingale). The book of Cont, Tankov "Financial modelling with Jump Processes" is a decent readable book if you want to use jumps.
- Badges
- Science topic
- Similar topics
- Econ
- Business Economics