Realized volatility measures from high-frequency stock data show that stock returns have a lot of unpredictable volatility risk. This risk makes it harder to predict future returns accurately. A new model called asymmetric realized volatility accounts for the fact that volatility is higher during high volatility periods. This model shows that stock returns can have changing levels of volatility, skewness, and kurtosis over time. The model was tested using data from the S&P 500 index and other stocks, showing its advantages in capturing the complex nature of stock market volatility.