4-10-2003 Lab Two 1I. “ Demonstration” from Eviews (handout)II. White NoiseA. Generate a white noise series of 1000 observations.1. In Eviews, open a new workfile of 1000 undated observations.2. Generate wn = nrndB. Characteristics of White Noise1. Select white noise and open it2. Go to the View menu of the workfile window and select spreadsheet3. Go to the View menu of the workfile window and select line graph4. Go to the View menu of the workfile window and select descriptive statistics5. Go to the View menu of the workfile window and select correlogramC. Having generated your series, how can you convince me that you did it correctly? How do I know it is white noise? How do I know you did not use the uniform or some other distribution?III. Random WalkA. Use your white noise series to generate a random walk.1. Sample 1 1 2. Generate rw = wn3. Sample 2 10004. Generate rw = rw(-1) + wn5. Sample 1 10004-10-2003 Lab Two 2B. Characteristics of a Random Walk1. Select the random walk and open it2. Go to the View menu of the workfile window and select spreadsheet3. Go to the View menu of the workfile window and select line graph4. Go to the View menu of the workfile window and select descriptive statistics5. Go to the View menu of the workfile window and select correlogramC. Once again, illustrate the properties of this series to convince me that you did indeed generate a random walk.IV. Autoregressive of the First OrderA. Use your white noise series to generate an Autoregressive Process of the first order with parameter 0.8.1. Sample 1 1 2. Generate arone = wn3. Sample 2 10004. Generate arone = 0.8*arone(-1) + wn5. Sample 1 1000 B. Characteristics of an Autoregressive Process of the First Order1. Select the autoregressive process and open it2. Go to the View menu of the workfile window and select spreadsheet3. Go to the View menu of the workfile window and select line graph4. Go to the View menu of the workfile window and select descriptive statistics4-10-2003 Lab Two 35. Go to the View menu of the workfile window and select correlogramC. What characteristics or properties would you use to distinguish an autoregressive time series from a random walk?V. Exercise (Due in one week.)A. Use the monthly series of total returns for the Standard and Poor’s 500, beginning January, 1970 and running through February, 2003, and ordinary least squares to fit a linear trend for the period. There are 398 observations. This time series is available at FRED, http://research.stlouisfed.org/fred/ under the data category Business/Fiscal. Select the latter link and scroll to the bottom to find the link for the series.1. interpret the slope2. forecast one period ahead3. what is a 95% confidence interval for this forecast?4. do the estimated residuals meet the assumptions for least squares?B. Use the natural logarithm of the SP 500 for the same period and estimate a trend.1. interpret the slope2. forecast one period ahead3. what is a 95% confidence interval for this forecast? 4. do the estimated residuals meet the assumptions for least
View Full Document
Unlocking...