Sur Model

SUR Model

To show whether gold derivative positions have considerably less variability in firm's returns, we utilized SUR model from the paper composed by Kim, Nam, and Wynne (2002). The specification of this model, comparing to normal regression model, is that it adopts dummy variables in the model. The dummy variables are set to equal to 1 if the event we chose occurred and equal to 0 otherwise. The coefficients of the dummy variables capture the influence of a sudden huge shock of gold price on the stock return of the gold-hedging and non-hedging firms. Specifically, as the theory suggests, hedging companies will experience relatively lower level of positive abnormal return when a positive shock on gold price occurs, while hedging companies will also experience relatively lower level of negative abnormal return if a negative shock on gold price happens.

In our model, we selected five event days to examine our expectation. The first day we chose was June 13th, 2006, which happened because the combination of a sharply higher US dollar and lower oil prices has the metals markets reeling overnight. The second one was on September 17th, 2008 when investors seek safety amid turmoil in stock markets, spot gold price jumped by 11.29%. The third event happened on October 10th, 2008 when spot gold price plumbed a record one-day loss, -7.43%. The fourth event was that the return of gold price decreased by -5.68% on October 22nd, 2008. The last event included in our model was that the gold price climbed $30.50 to settle at $757.30 an ounce on November 4th, 2008, in addition, the stocks surged as millions of Americans battered by the weakened economy turned out to vote for the next President of the United States.

As the results showed, the averaged coefficient of our hedging group for second event was 0.351%, the day gold price went up 11.29%, was 0.351%, while the averaged coefficient of our non-hedging group was 0.449%. This means a positive shock to gold price does provide a more significant impact on non-hedger than hedgers. Same thing happened when the gold price declined. Our coefficient of hedgers was -2.46% and -3.594% for non-hedgers when the gold price plumbed by 5.68% in one day. This represented a negative shock on gold price does provide any larger influence for non-hedgers than hedgers. Although one or two results did not show a significant matching to our expectation, in general, the results we calculated are consistent with that the gold derivative-hedging firms in our sample exhibited lower levels of volatility.