The article investigates different models used to predict changes in financial market volatility over time. Researchers found that volatility is not constant and can be predicted. They used multivariate GARCH models to forecast correlations and covariances in financial data. A new model called dynamic conditional correlation (DCC) was proposed, which can estimate correlations more simply. The study focused on Engle's DCC MGarch and MGarch BEKK models, showing how they can be used to estimate volatility in financial time series data.