Use the data in QLEISCOIN.XLS and statistical software to do the following problems.

COIN—Percentage change in the composite index of coincident economic indicators.
LEIS—Percentage change in the composite index of leading economic indicators.
(A) Run a regression to determine if COIN is affected by LEIS and LEIS lagged 1 through 8 quarters. Do you think LEIS 8 quarters ago helps predict COIN this quarter? Explain.
(B) How many lags of the LEIS should be used to explain COIN?
Explain how you came to this conclusion.
(C) Does the regression you settled on in (B) suffer from serial correlation according to the Durbin-Watson test? Show the five-step procedure.
(D) Does the regression you settled on in (B) suffer from serial correlation according to the Lagrange multiplier test? Show the five-step procedure.
(E) Does the regression you settled on in (B) suffer from multicollinearity? On what do you base your response?
(F) Employ a Koyck lag structure in the COIN/LEIS regression.
Does this regression suffer from serial correlation according to the Lagrange multiplier test? Show the five-step procedure.
(G) Employ a Koyck lag structure in the COIN/LEIS regression.
Does this regression suffer from multicollinearity? On what do you base your response?
(H) According to the Koyck model, a one-unit increase in LEIS implies COIN will be how much higher or lower in the short run?
(I) In the long run, the one-unit increase in LEIS implies COIN will eventually change how much?