Accounting data and volatility predictions2018-07-25
Generally speaking, financial research has studied how past equities and options volatility can help to predict future volatility in the markets. However, new research from Suhas Sridharan, assistant professor of accounting, investigates the impact of supplementing past volatility data with actual financial statement information to forecast future realized volatility. Sridharan used a large sample of 47,398 quarterly observations from 3,078 firms taken from 1996 to 2012. Her results indicate that incorporating accounting-based information, such as “standard deviation of the earnings yield, standard deviation of the change in premium of market value over book value, and the covariance of the two,” into forecasting models lowers forecast errors compared to models based solely on past realized volatility. She finds, “Equity returns volatility is significantly positively related to the earnings yield volatility and the volatility of the change in market to book premium. Volatility is significantly negatively related to the covariance of the earnings yield and change in market to book premium.” Sridharan also discovered that using accounting-based fundamental information in trading strategy could help to predict option returns.