The article introduces a new method to better estimate daily risk levels in financial markets. By using high-frequency data, the researchers developed a model that considers various factors like volatility and overnight returns. They tested the method on 20 assets and found it improved the accuracy of predicting potential losses compared to existing methods.