Just read your paper, the replicating portfolio equation (10), apart from using regression to estimate the parameters beta, how about using Bayesian inference? Then, We can get the probability distribution of beta. What is your opinion on that?