SUMMARY OUTPUT Regression Statistics Multiple R 0.977099 R Square 0.954723 Adjusted R Square 0.943404 Standard Error 3972595 Observations 16 ANOVA df...

Ted Ralley, the director of marketing research of a large manufacturer of spare parts for automobiles, is working on the sales forecast for the next year. He knows that forecast errors are costly and wants to forecast the sales as accurately as possible. He collected the data on quarterly sales for the previous four years and ran several forecasts using time series forecasting methods, but he was still concerned. Ted believes that economic activity and oil prices has a significant impact of auto parts sales and decided to check if the sales can be predicted better using econometric variables.Estimate a regression model with trend & season dummies to forecast the next year sales:1) Use estimated regression coefficients to compute in-sample forecasts for salesErrorSquared errorAbsolute % errorThe numerator of Thiel’s U statisticThe denominator of Thiel’s U statistic3) Calculate the following error metrics: mean error (ME), mean squared error (MSE), root mean squared error (RMSE), mean absolute percent error (MAPE), and Thiel’s U statistic4) Compute sales forecast in the year 2008Estimate another regression model, but this time use M2 Index, non-farm activity index, and oil prices as predictors to forecast the next year sales. Repeat steps 1) – 4) and calculate error metrics for years 2006-2007 and sales forecast in the year 2008.Compute sales forecasts using additive Holt-Winters model. Repeat steps 1) – 3) and calculate error metrics for years 2006-2007. Calculate the optimal alpha, beta, and gamma parameters of the model using MS Excel Solver so as to minimize RMSE. Compute sales forecast in the year 2008. Compare the error metrics (ME, RMSE, MAPE, Thiel’s U) of the three models. Construct a line plot showing the actual quarterly sales and forecast by these models.Suggest and justify an appropriate decision regarding the sales forecast. Analyze the patterns of the quarterly time series; discuss the properties of each forecasting model and their relevance to predicting the series; select the best forecasting model for the series* show work in excel spreedsheet