Many financial time series exhibit volatility clustering and the Baltic Dry Index is not an exception. It has a structure and may depend on the past. Geopolitical events, speculation, unexpected shocks, are among the main factors that produce volatility in the shipping industry. Volatility modeling is crucial for risk management as it helps shipping participants to adjust their positions in a portfolio, especially during market fluctuations, economic downturns, or bear and bull periods. In this chapter we present the Generalized Autoregressive Conditional Heteroskedasticity (GARCH) models. This type of models, used to analyze time-varying volatility in financial time series such as stock returns, were developed to express variance, not only as a function of the past returns, but also as a function of the past variance. We begin introducing ARIMA models, and how to fit them with RStudio. Since ARIMA models assume constant conditional variance, a phenomenon known as homoscedasticity that is not assumable in many financial time series in which volatility is usually clustered in very specific periods of time, GARCH models were subsequently developed.