The article explores how skewness and kurtosis, which measure extreme variations in financial data, are important for risk assessment. The researchers developed a model that allows for time-varying skewness and kurtosis in financial markets. By applying this model to foreign exchange, stock, and interest rate data, they found evidence of co-variability of moments beyond volatility. The study shows that extreme variations in financial markets can persist over time, and that there is a certain predictability in these moments.