Page 27 - FINAL CFA II SLIDES JUNE 2019 DAY 3

EXAMPLE: The Durbin-Watson test for serial READING 8: MULTIPLE REGRESSION AND ISSUES IN REGRESSION ANALYSIS
correlation: Suppose you have a regression output which
includes 3 independent variables that provide you with a
DW statistic of 1.23. Also suppose that the sample size is MODULE 8.7: SERIAL CORRELATION
40. At a 5% SL, Are the error terms SC?

Answer:
From a 5% DW table with n = 40 and k = 3, we find:

But DW = 1.23
DW < d l (i.e., 1.23 < 1.34) = =
1.34 1.66

Reject Ho and conclude that the regression has positive serial correlation among the error terms.

Correcting Serial Correlation

Adjust the coefficient standard errors (using the Hansen method), which also corrects for CH (results also called serial
correlation consistent standard errors or Hansen-White standard errors):
• ONLY use the Hansen method if serial correlation is a problem.
• ONLY use White-corrected standard errors if ONLY heteroskedasticity is a problem.
• If both conditions are present, use the Hansen method.

• Improve the specification of the model. The best way to do this is to explicitly incorporate the time-series nature of the data
(e.g., include a seasonal term).