The file m-pgspabt-6211.txt contains the monthly simple returns of Procter & Gamble stock, S&P composite index, and Abbot Laboratories stock from January 1962, to December 2011. Compute the log returns. Is there conditional heteroscedasticity in the log return series? Fit a Go-GARCH model with ICA estimation method to the log return series. Write down the fitted model. Is the model adequate? Why?

Data from Example 7.6

Figure 7.17

Transcribed Image Text:

Example 7.6 As a second example, we consider the daily log returns, in per- centages, of EXXON-Mobil stock, S&P composite index, and Apple stock from September 4, 2007 to September 28, 2012 for 1280 observations. The simple returns are from CRSP. The left panel of Figure 7.16 shows the time plots of the daily log return series. As expected, the volatilities of the three assets were high during the 2008 financial crisis and the variability of the Apple stock returns appears to be higher than those of the other two series. Let z denote the daily log returns, in percentages. Analysis of individual series indicates some minor serial correla- tions in the Apple stock returns. However, for simplicity, we continue to assume that the mean equation of z is just a constant. This provides an estimate of the (d) -15 Inrtn 0 10 (b) -10 05 (c) 2008 2009 2010 2011 2012 Year -20 -5 2008 2009 2010 2011 2012 Year 2008 2009 2010 2011 2012 Year TTT 02 -4 eta_t A 2008 2009 2010 2011 2012 Year TTTTTTTT 2 1- -4 eta_t -4 0 4 2008 2009 2010 2011 2012 Year TTTTT 2008 2009 2010 2011 2012 Year