Modelo de Predicción del Precio de Bitcoin Mediante el Uso de Redes Bayesianas y Simulación de Montecarlo

The objective is to design a model for predicting the price of Bitcoin over time, based on a Bayesian network and Monte Carlo simulation. Bitcoin is an electronic means of payment that allows irreversible person-to-person transactions. Its price varies depending on political, economic, and other factors; these factors, organized in a Bayesian network and iterated in a Monte Carlo simulation, can predict the price. A quantitative approach was chosen, predictive bibliographic research using a deductive method. The sample included the price of 5,189 days; 10,000 iterations were performed. It was validated with experts. The most important variables were analyzed using PESTEL analysis, finding that factors such as US support, Chinese support, speculation, and the use of Bitcoin by the population are very important in determining the asset's price. The resulting model has the potential to predict the price of Bitcoin and can be continuously refined. It can be calibrated in real time using RSS feeds and news readers. Once a model is obtained that is considered theoretically valid, the network must be updated so that it continues to learn the probability values ​​and structure.