Homework 2

Question 1
The management of a supermarket wanted to check the effect of the number of commercials
broadcast on local radio (X) on the gross sales (in million of dollars) at the store (Y). The
management experimented for eight weeks by broadcasting a different number of commercials each
week on local radio. The following information are given:

X = 183 Y = 29.61 X2 = 4529 Y2 = 111.4837 XY = 701.81 n=8

(a) Find the least square estimates of the intercept and the slope coefficients. Give an
interpretation of the two coefficients.

(b) Compute the correlation coefficient (r) and the coefficient of determination (r2). Give an
interpretation of the coefficient of determination.

(c) Predict the gross sales for a week with 25 commercials.

Question 2
The following table gives data on average public teacher pay (annual salary in dollars) and
spending on public schools per pupil (dollars) in 1985 for 50 states and the District of
Columbia.

Table 5.5
Average teacher salary and per
pupil spending ($), by state, 1985

SALARY SPENDING

19583 3346
20263 3114
20325 3554
26800 4642
29470 4669
26610 4888
30678 5710
27170 5536
25853 4168
24500 3547
24274 3159
27170 3621
30168 3782
 26525 4247
27360 3982
21690 3568
21974 3155
20816 3059
18095 2967
20939 3285
22644 3914
24624 4517
27186 4349
33990 5020
23382 3594
20627 2821
22795 3366
21570 2920
22080 2980
22250 3731
20940 2853
21800 2533
22934 2729
18443 2305
19538 2642
20460 3124
21419 2752
25160 3429
22482 3947
20969 2509
27224 5440
25892 4042
22644 3402
24640 2829
22341 2297
25610 2932
26015 3705
25788 4123
29132 3608
41480 8349
25845 3766

To find out if there is any relationship between teacher’s pay and per pupil expenditure in
public schools, the following model was suggested: Payi = β1 + β2 Spendi + ui, where Pay
stands for teacher’s salary and Spend stands for per pupil expenditure.
(a) Plot the data and eyeball a regression line.
(b) Suppose on the basis of (a) you decide to estimate the above regression model. Obtain
the estimates of the parameters, their standard errors, r2, RSS, and ESS.
 (c) Interpret the regression. Does it make economic sense?
(d) Establish a 95 percent confidence interval for β2. Would you reject the hypothesis
that the true slope coefficient is 3.0?
(e) Obtain the mean and individual prediction of Pay if per pupil spending is $5,000.
Also, establish 95 percent confidence intervals for the true mean and individual
prediction of Pay for the given spending figure. Note that the standard error of the
mean prediction is 520.5117 and standard error of the individual prediction is
2382.337.
(f) How would you test the assumption of normality of the error term? Show the test(s)
you use.

Question 3
To study the relationship between investment rate (investment expenditure as a ratio of the GDP)
and saving rate (savings as a ratio of GDP). Martin Feldstein and Charles Horioka obtained data
for a sample of 21 countries (see the table below). The investment rate for each country is the
average rate for the period 1960 – 1974 and the savings rate is the average savings rate for the
period 1960 – 1974. The variable Invrate represents the investment rate and the variable Savrate
represents the savings rate.

SAVRATE INVRATE
Australia 0.25 0.27
Austria 0.285 0.282
Belgium 0.235 0.224
Canada 0.219 0.231
Denmark 0.202 0.224
Finland 0.288 0.305
France 0.254 0.26
Germany 0.271 0.264
Greece 0.219 0.248
Ireland 0.19 0.218
Italy 0.235 0.224
Japan 0.372 0.368
Luxembourg 0.313 0.277
Netherlands 0.273 0.266
New Zealand 0.232 0.249
Norway 0.278 0.299
Spain 0.235 0.241
Sweden 0.241 0.242
Switzerland 0.297 0.297
U.K. 0.184 0.192
U.S. 0.186 0.186

(a) Plot the investment rate against the savings rate.

(b) Based on this plot, do you think the following models fit the data equally well?

Invratei  1   2 Savratei  u i
ln( Invratei )   1   2 ln( Savratei )  u i

(c) Estimate both of these models and obtain the usual statistics (i.e. coefficients, standard
errors, t-statistics, and r2).

(d) How would you interpret the slope coefficient in the linear model? In the log-linear
model? Is there a difference in the interpretation of these coefficients?

(e) How would you interpret the intercepts in the two models? Is there a difference in your
interpretation?

(f) Would you compare the two r2 coefficients? Why or why not?
(g) Suppose you want to compute the elasticity of the investment rate with respect to the
savings rate. How would you obtain this elasticity for the linear model? For the log-linear
model? Note that this elasticity is defined as the percentage change in the investment rate for a
percentage change in the savings rate.

(h) Given the results of the two regression models, which model would you prefer? Why?

Question 4
What is knowns as the characteristic line of modern investment analysis is simply the
regression line obtained from the following model:
r it   i   i r mt  ut
where r it = the rate of return on the ith security in time t

r mt = the rate of return on the market portfolio in time t

ut = stochastic disturbance term

In this model  i is known as the beta coefficient of the ith security, a measure of market (or
systematic) risk of a security.
On the basis of 240 monthly rates of return for the period 1956-1976, Fogler and Ganapathy
obtained the following characteristic line for IBM stock in relation to the market portfolio
index developed at the University of Chicago:
 ^
r it  0.7264  1.0598rmt
se  (0.3001)(0.0728)
r 2  0.4710
df  238
F1, 238  211.896

(a) A security whose beta coefficient is greater than one is said to be volatile or
aggressive security. Was IBM a volatile security in the time period under study? (i.e.
Test the hypothesis H 0 :  2 = 1 against H 1 :  2 > 1.)

(b) Is the intercept coefficient significantly different from zero? If it is, what is its
practical meaning?