EXAMINATION

AFF700
Advanced Research Methods in Finance

Date 2021-12-10
Time 14.00 – 18.00
Examiner Urban Gråsjö
Teacher Urban Gråsjö
Visit No
Phone 0704614753
Aids Calculator, Textbook: Brooks, Introductory Econometrics for Finance (a printed
version of the textbook is allowed)
Questions 6
Max points 60
Grade Levels A: 54p, B: 48p, C: 42p, D: 36p, E: 30p
Date for Result latest 2022-01-04

Note: Write your anonymous code on each page!!

Good Luck!

1
Problem 1: 18p
This problem consists of 18 multiple choice questions. Each correct answer gives 1 point,
wrong answers give 0 points. There is only one correct choice per question. You do not need
to motivate the answers!

1. Assume that the relationship between a company’s stock price (y) and dividends paid per
share (x) is linear. If the slope of the equation is 0.60 and the intercept is 40, what would be
the expected stock price if the dividend paid was 3?
(a) 41.8
(b) 43.6
(c) 40.6
(d) 40.

2. Suppose you have calculated the following regression results:

=yˆ t 1.25 + 0.64 xt . The standard errors of α̂ and βˆ are 1.22 and 0.25, respectively. Using the
test of significance approach, what is the test statistic value of a hypothesis to test whether the
true value of β statistically different from zero?
(a) 1.10
(b) 0.95
(c) 2.56
(d) Cannot say without more information.

3. In a time-series regression of the excess return of a mutual fund on a constant and the excess
return on a market index, which of the following statements should be true for the fund manager
to be considered to have ‘beaten the market’ in a statistical sense?
(a) The estimate for α should be negative and statistically significant
(b) The estimate for α should be positive and statistically significantly greater than the risk-free
rate of return
(c) The estimate for β should be positive and statistically significant
(d) The estimate for α should be positive and statistically significant.

4. Which one of the following statements must hold for EVERY CASE concerning the
residual sums of squares for the restricted and unrestricted regressions?
(a) URSS > RRSS
(b) URSS ≥ RRSS
(c) RRSS > URSS
(d) RRSS ≥ URSS.

2
5. Suppose that the value of R2 for an estimated regression model is exactly one. Which of the
following are true?
(i) All of the data points must lie exactly on the line
(ii) All of the residuals must be zero
(iii) All of the variability of y about its mean has been explained by the model
(iv) The fitted line will be horizontal with respect to all of the explanatory variables.
(a) (ii) and (iv) only
(b) (i) and (iii) only
(c) (i), (ii), and (iii) only
(d) (i), (ii), (iii), and (iv).

6. Which of these is NOT a reason for adding a disturbance term to a regression model
yt =α + β xt + ut ?
(a) Some determinants of the effect variable may be omitted from the model
(b) Some determinants of the independent variable may be omitted from the model
(c) Some determinants of the effect variable may be unobservable
(d) There may be errors in the way that the dependent variable is measured which cannot be
modelled.

7. Consider the following two regressions

y t = β 1 + β 2 x 2t + β 3 y t −1 + u t
∆y t = γ 1 + γ 2 x 2t + γ 3 y t −1 + u t
Which of the following statements are true?
(i) The RSS will be the same for the two models
(ii) The R2 will be the same for the two models
(iii) The adjusted R2 will be different for the two models
(iv) The regression F-test will be the same for the two models.
(a) (ii) and (iv) only
(b) (i) and (iii) only
(c) (i), (ii), and (iii) only
(d) (i), (ii), (iii), and (iv).

8. What is the most appropriate interpretation of the assumption cov ( ui , u j ) = 0 concerning the
regression disturbance terms?
(a) The errors are linearly independent of one another
(b) The errors are linearly dependent of one another
(c) The errors are nonlinearly independent of one another
(d) The covariance of the errors is constant and finite over all its values

3
9. Standard errors
(a) Give us an idea of the deviation of the errors from their mean
(b) Measure the reliability of the independent variables
(c) Give us an idea of the precision of estimates of α and β
(d) Measure the reliability of the dependent variables.

10. Which of the following would you expect to be a problem associated with adding lagged
values of the dependent variable into a regression equation?
(a) The assumption that the regressors are non-stochastic is violated
(b) A model with many lags may lead to residual non-normality
(c) Adding lags may induce multicollinearity with current values of variables
(d) The standard errors of the coefficients will fall as a result of adding more explanatory
variables.

11. Which of the following would NOT be a potential remedy for the problem of
multicollinearity between regressors?
(a) Removing one of the explanatory variables
(b) Transforming the data into logarithms
(c) Transforming two of the explanatory variables into ratios
(d) Collecting higher frequency data on all of the variables.

12. Assuming a researcher runs the following regression= ut ρ ut −1 + vt where ut is residual

from a regression. If the researcher conducts a hypothesis test with null hypothesis of
H 0 : ρ = 0 against an alternative hypothesis of H1 : ρ ≠ 0 , what type of test is he or she
conducting?
(a) Test for heteroscedasticity
(b) Test for autocorrelation
(c) Test for non-normality
(d) Test for homoscedasticity.

five
Use the following to answer Questions 13 and 14. Suppose that you have estimated the first
autocorrelation coefficients using a series of length 81 observations and found them to be
Lag 1 2 3 4 5
Autocorrelation coefficient 0.412 -0.205 -0.332 0.005 0.543

13. Which autocorrelation coefficients are significantly different from zero at the 5% level?
(a) The first and fifth autocorrelation coefficient
(b) The first, second, third, and fifth autocorrelation coefficient
(c) The first, third, and fifth autocorrelation coefficient
(d) The second and fourth autocorrelation coefficient.

14. What is the appropriate Box–Pierce test statistic?

(a) 4.78
(b) 47.83
(c) 59.05
(d) 5.91.

4
15. Is the following process yt =
3 yt −1 − 2.75 yt −2 + 0.75 yt −3 + ut stationary?
(a) Yes
(b) No
(c) Partly stationary
(d) Cannot say without more information.

16. Assume that you are trying to model the relationship between house prices and rents. If
you find that both series are non-stationary and a linear combination of the two series is
stationary, which of the following is true?
(I) Regressing the levels of house prices on the levels of rents could lead to spurious
regressions
(II) House prices and rents are cointegrated
(III) An appropriate linear combination of house prices and rents is I(1)
(IV) House prices and rents are not cointegrated.
(a) I only
(b) I and II only
(c) I, II, and III only
(d) I, II, III, and IV only.

17. Suppose that a researcher wanted to obtain an estimate of realised (‘actual’) volatility.
Which one of the following is likely to be the most accurate measure of volatility of stock
returns for a particular day?
(a) The price range (high minus low) on that day
(b) The squared return on that day
(c) The sum of the squares of hourly returns on that day
(d) The squared return on the previous day.

18. Which of the following is the most plausible test regression for determining whether a
series y contains ‘ARCH effects’?
2 2 2 2 2
(a) yt2 = α 0 + α1 y t −1 +α 2 y t − 2 +α 3 y t − 3 +α 4 y t − 4 +α 5 y t − 5 +ut
(b) yt2 = α 0 + α1 y t −1+α 2 y t − 2 +α 3 y t − 3 +α 4 y t − 4 +α 5 y t − 5 +ut
2 2 2 2 2
(c) yt = α 0 + α1 y t −1 +α 2 y t − 2 +α 3 y t − 3 +α 4 y t − 4 +α 5 y t − 5 +ut
2 3 4 5 6
(d) yt = α 0 + α1 y t −1 +α 2 y t − 2 +α 3 y t − 3 +α 4 y t − 4 +α 5 y t − 5 +ut .

5
Answers, Problem 1

a b c d
1. A
2. C
3. D
4. D
5. C
6. B
7. B
8. A
9. C
10. A
11. B
12. B
13. C
14. B
15. B
16. B
17. C
18. A

6
Problem 2: 9p

You estimate a regression of the form given by (1) below in order to evaluate the effect of
various firm-specific factors on the returns of a sample of firms. You run a cross-sectional
regression with 200 firms

ri = β0 + β1Si + β2MBi + β3PEi + β4BETAi + ui (1)

where: ri is the percentage annual return for the stock

Si is the size of firm i measured in terms of sales revenue
MBi is the market to book ratio of the firm
PEi is the price/earnings (P/E) ratio of the firm
BETAi is the stock’s CAPM beta coefficient

You obtain the following results (with standard errors in parentheses)

𝑟𝑟�𝚤𝚤 = 0.080 + 0.801Si + 0.321MBi + 0.164PEi − 0.084BETAi
(0.064) (0.147) (0.136) (0.420) (0.120) (2)

a) Calculate the t-ratios.

b) What do you conclude about the effect of each variable on the returns of the security?

c) Based on your results, what variables would you consider deleting from the
regression?

d) If a stock’s beta increased from 1 to 1.2, what would be the expected effect on the
stock’s return? Is the sign on beta as you would have expected?

Explain your answers in each case.

a) 1.25 5.45 2.36 0.390 -0.700

The t-ratios are given in the final row above and are in italics. They are calculated by dividing the
coefficient estimate by its standard error.

b) The relevant value from the t-tables is for a 2-sided test with 5% rejection overall. T-k = 195; tcrit =
1.97. The null hypothesis is rejected at the 5% level if the absolute value of the test statistic is greater
than the critical value. We would conclude based on this evidence that only firm size and market to
book value have a significant effect on stock returns.

c) If a stock’s beta increases from 1 to 1.2, then we would expect the return on the stock to FALL by
(1.2-1)*0.084 = 0.0168 = 1.68%

This is not the sign we would have expected on beta, since beta would be expected to be positively
related to return, since investors would require higher returns as compensation for bearing higher
market risk.

7
Problem 3: 8p
Consider the following system of two equations

y1t = α0 + α1 y2t + α2X1t + α3X2t + u1t (1)

y2t = β0 + β1 y1t + β2X1t + u2t (2)

a) Explain, with reference to these equations, the undesirable consequences that would
arise if (1) and (2) were estimated separately using OLS.

b) What would be the effect upon your answer to (a) if the variable y1t had not appeared
in (2)?

c) State the order condition for determining whether an equation which is part of a
system is identified. Use this condition to determine whether (1) or (2) or both or
neither are identified.

Solution

(a) A glance at equations (1) and (2) reveals that the dependent variable in (1) appears as an
explanatory variable in (2) and that the dependent variable in (2) appears as an explanatory
variable in (1). The result is that it would be possible to show that the explanatory variable y2t
in (1) will be correlated with the error term in that equation, u1t, and that the explanatory
variable y1t in (2) will be correlated with the error term in that equation, u2t. Thus, there is
causality from y1t to y2t and from y2t to y1t, so that this is a simultaneous equations system. If
OLS were applied separately to each of equations (1) and (2), the result would be biased and
inconsistent parameter estimates. That is, even with an infinitely large number of
observations, OLS could not be relied upon to deliver the appropriate parameter estimates.

(b) If the variable y1t had not appeared on the RHS of equation (2), this would no longer be a
simultaneous system, but would instead be an example of a triangular system. Thus, it would
be valid to apply OLS separately to each of the equations (1) and (2).

(c) There are 2 equations in the system of (1) and (2), so that only 1 variable would have to be
missing from an equation to make it just identified. If no variables are absent, the equation
would not be identified, while if more than one was missing, the equation would be over-
identified. Considering equation (1), no variables are missing so that this equation is not
identified, while equation (2) excludes only variable X2t, so that it is just identified.

8
Problem 4: 7p
Assume that the econometric model is of the form: yt = β1 + β2x2t + β3x3t + β4x4t + β5x5t + ut

You decide to investigate the relationship given the null hypothesis H0: β3 + β4 = 1 and β5 = 1.

a) What would constitute the restricted regression?

b) The regressions are carried out on a sample of 96 quarterly observations, and the
residual sums of squares for the restricted and unrestricted regressions are 102.87 and
91.41, respectively. Perform the test. What is your conclusion?

Solution: (5p)
a) The null hypothesis is: H0: β3 + β4 = 1 and β5 = 1
The first step is to impose this on the regression model:
yt = β1 + β2x2t + β3x3t + β4x4t + β5x5t + ut subject to β3 + β4 = 1 and β5 = 1.
We can rewrite the first part of the restriction as β4 = 1 - β3
Then rewrite the regression with the restriction imposed
yt = β1 + β2x2t + β3x3t + (1-β3)x4t + x5t + ut
which can be re-written
yt = β1 + β2x2t + β3x3t + x4t - β3x4t + x5t + ut
and rearranging
(yt – x4t – x5t) = β1 + β2x2t + β3x3t - β3x4t + ut
(yt – x4t – x5t) = β1 + β2x2t + β3(x3t –x4t) + ut

Now create two new variables, call them pt and qt:

pt = (yt - x3t - x4t)
qt = (x2t -x3t)
We can then run the linear regression:
pt = β1 + β2x2t + β3qt+ ut ,
which constitutes the restricted regression model.
b) The test statistic is calculated as ((RRSS-URSS)/URSS)*(T-k)/m
In this case, m=2, T=96, k=5 so the test statistic = 5.704. Compare this to an F-distribution
with (2,91) degrees of freedom, which is approximately 3.10. Hence we reject the null
hypothesis that the restrictions are valid. We cannot impose these restrictions on the data
without a substantial increase in the residual sum of squares.

9
Problem 5: 10p
Assume a VAR(2) with two stationary time series variables, Xt and Yt.

Yt = α 1 + γ 11Yt −1 + γ 12Yt − 2 + β11 X t −1 + β12 X t − 2 + δ 1t + e1t .

X t = α 2 + γ 21Yt −1 + γ 22Yt − 2 + β 21 X t −1 + β 22 X t − 2 + δ 2 t + e1t .

a) Describe if and how this VAR(2) can be used to test for Granger causality.

b) How would your answer to part a) change if X and Y had unit roots and were
cointegrated?

Solution
a) Since we assume that all the variables in the VAR(2) are stationary, estimation and
testing can be carried out in the standard way. That is, you can obtain estimates of
coefficients in each equation if you regress each equation using OLS. P-values or t-
statistics will then allow you to ascertain whether individual coefficients are
significant.
All of the variables we are using to explain the current value of the dependent variable
occurred in the past (e.g. in the first equation the explanatory variables are all dated t -
1 or earlier, whereas the dependent variable is Yt). It is possible that the past might
influence the present, but it is not possible for the present to influence the past. Hence,
in the VAR model, the explanatory variables might influence the dependent variable,
but there is no possibility that the dependent variable influences the explanatory
variable. Problems of interpretation that arise with the regression of Yt on Xt do not
arise in the VAR case. So, in this example we can test if X Granger causes Y, i.e., if Xt-
1 and/or Xt-2 affects Yt and if Y Granger causes X, i.e. if Yt-1 and/or Yt-2 affects Xt. If the
coefficients of these lagged variables are statistically significant then we have proven
Granger causality.

b) Then an error correction model with two equations, i.e. a Vector Error Correction
Model, VECM(2),, would be the appropriate model. The difference between a VAR
and a VECM is the inclusion of an error correction term in the latter model. The error
correction term (et-1) is constructed from the residuals of the regression of Y on X and
should be included in each equation in the model. Moreover, all variables, except the
error correction term, are in first differences in a VECM. By using a VECM it is
possible to determine both long term and short term effects. The long term effect is
determined by the coefficients of the lagged differences and the short term effect by
the coefficient of the error correction term.

10
Problem 6: 8p

A researcher wants to test the order of integration of some time series data. He decides to use
the DF test. He estimates a regression of the form

∆yt = μ + ∅yt−1 + ut

� = −0.20 with standard error = 0.15.

and obtains the estimate ∅

a) What are the null and the alternative hypothesis for this test?

b) Given the data, and a critical value of −2.88, perform the test.

c) What is the conclusion, and what should be the next step?

Solution

(a)The null hypothesis is of a unit root against a one sided stationary alternative, i.e. we have

H0 : yt ∼ I(1)

H1 : yt ∼ I(0)

which is also equivalent to

H0 : ∅ = 0

H1 : ∅ < 0 (2p)

� /𝑆𝑆𝑆𝑆(∅
(b) The test statistic is given by ∅ � ) which equals -0.20 / 0.15 = -1.33 (2p)

Since this is not more negative than the appropriate critical value (-2.88), we do not reject the null
hypothesis.

c) We therefore conclude that there is at least one unit root in the series (there could be 1, 2, 3 or
more). What we would do now is to regress ∆2yt on ∆yt-1 and test if there is a further unit root. The null
and alternative hypotheses would now be:

H0 : ∆yt ∼ I(1) i.e. yt ∼ I(2)

H1 : ∆yt ∼ I(0) i.e. yt ∼ I(1)

If we rejected the null hypothesis, we would therefore conclude that the first differences are
stationary, and hence the original series was I(1). If we did not reject at this stage, we would conclude
that yt must be at least I(2), and we would have to test again until we rejected. (4p)

11