Modeling volatility
Keywords:
Volatility, portfolio management, range of quotations, dynamic linear models, state space, Kalman filter, forecasts, software packages, dlm, forecast, RAbstract
Volatility is a key concept in portfolio management. Due to the fact that it is not directly observable, several other magnitudes have been devised related to it. Here it is used the range, the difference between the high and low daily transaction prices of a stock. Using a sample of 100 observations of the Madrid Stock Exchange Index a model of the state space family is established for the logarithm of the range. After its validation, the model is used to forecast ten future values of the log of the range. The forecasts are evaluated and some conclusions are presented