Volatility forecasting model : a risk reduction tool for asset managers

Abstract

An essential element of an investment is its performance. However, understanding volatility is critical when evaluating a future investment. This paper utilizes a regression model aiming to forecast volatility for the S&P 500 Index. It examines the relationship between the Volatility Index, Price-to-Earnings multiple and Asset Class correlations. This paper also evaluates these explanatory variables individually for market forecasting purposes. It also proves that higher volatility corresponds to a higher probability of declining market, while lower volatility corresponds to a higher probability of a rising market. The Parsimonious regression model identifies these three variables as essential predictors to forecast volatility. The results proved to be highly statistically significant and obtained 59.2% level of confidence, which means that 59.2% of the values are correctly predicted in our model. Furthermore, this paper establishes the respective practical explanations for the outcomes provided.