Bayesian statistical analysis of daily returns runs in Brazilian stock exchange

In the financial market analysis, it is usual that the random variations in price movements share non-trivial statistical properties as return distributions, absence of autocorrelations in asset returns, volatility clustering and asymmetry between rises and falls. Most of fluctuations in asset prices have been deeply investigated using time series models to get inferences of interest and forecasting. In this paper, the main goal is, instead of using time series models in the data analysis it is considered the use of distributions for the run lengths and absolute run returns of historical price stated in NYSE stock exchange for three private banks located in Brazil in the period ranging from July,19 2013 to July, 19 2018 using discrete Weibull distributions as an alternative for the exponential law commonly used in this kind of analysis. Under this modeling approach it is possible to get information on the market structure as the probability of the stock-market is equally likely to go up and down everyday and the magnitude of returns, for example.