The following regression model is to test whether the Metals Mining industry stock returns are sens
Question: The following regression model is to test whether the Metals & Mining industry stock returns are sensitive to commodity price changes and global market changes for the period from January 2000 to September 2011:
\[{{R}_{t}}=\alpha +{{\beta }_{1}}{{R}_{oil,t}}+{{\beta }_{2}}{{R}_{MSCI\_world,t}}+{{\beta }_{3}}{{R}_{SP1500,t}}+{{\beta }_{4}}{{R}_{Gold,t}}+{{\beta }_{5}}{{D}_{t}}+{{\varepsilon }_{i,t}}\]where:
\[{{R}_{t}}\] : log return on the S&P Metals & Mining Industry index in month t
: log return on the MSCI world market index in month t
: log return on the SP1500 in month t
: log return on the WTI crude oil price in month t
: log return on the Gold price in month t
: is a dummy variable which equals 1 during the recent international financial crisis (August 2007 – September 2011) and 0 before.
| Dependent Variable: METALS_MINING | ||||||
| Method: Least Squares | ||||||
| Sample: 2000M01 2011M08 | ||||||
| Included observations: 140 | ||||||
| Variable | Coefficient | Std. Error | t-Statistic | Prob. | ||
| Constant | 0.270 | 0.644 | 0.418 | 0.676 | ||
| OIL | 0.136 | 0.055 | 2.449 | 0.016 | ||
| MSCI_WORLD | 0.973 | 0.447 | 2.178 | 0.031 | ||
| SP1500 | 0.403 | 0.430 | 0.936 | 0.351 | ||
| GOLD | 0.461 | 0.116 | 3.981 | 0.000 | ||
| DUMMY | -1.146 | 1.089 | -1.053 | 0.294 | ||
| R-squared | 0.696900 | Mean dependent var | 0.463071 | |||
| Adjusted R-squared | 0.685591 | S.D. dependent var | 10.80154 | |||
| S.E. of regression | 6.056662 | Akaike info criterion | 6.482106 | |||
| Sum squared resid | 4915.542 | Schwarz criterion | 6.608177 | |||
| Log likelihood | -447.7474 | Hannan-Quinn criter. | 6.533337 | |||
| F-statistic | 61.61980 | |||||
| Prob(F-statistic) | 0.000000 | |||||
| Additional Tests: White Heteroskedasticity test = 1.9672 (p-value = 0.0875) Jarque-Bera = 43.4157 (p-value = 0.0000) Durbin-Watson stat = 2.0002 | ||||||
You are required to:
(a) Identify the key underlying assumptions of the linear regression model and discuss the available statistical tests for examining these assumptions.
(40 per cent of the marks)
(b) Hypothesise and explain the expected signs of the coefficients and compare the actual signs with your hypotheses.
(20 per cent of the marks)
(c) Critically evaluate the performance of the regression model taking full account of the assumptions underpinning the estimation technique.
(40 per cent of the marks)
Solution Format: Word Document
