1

Ethiopian Civil Service University

College of Finance, Management and Development (CFMD)
Applied Econometrics for Accounting and Finance – Assignment # 1

Instruction: All discussions should be hand-written, while EViews outputs should be print-outs.

1. The data ‘MSE data for assignment May 2021’ pertain to several characteristics of a random
sample of Micro- and Small Enterprises (MSE). The variables considered are:

Variable Description Coding

business_license Do you have business license or certificate? 0 = No, 1 = Yes
Have you obtained loan from any micro finance
0 = No, 1 = Yes
loan institution?
Have you ever received land or working
0 = No, 1 = Yes
land_premise premises for your enterprise?
util_rate Enterprise's current capacity utilization rate Actual data
current_capital Current capital (in birr) Actual data
initial_capital Initial capital (in birr) Actual data
years_establish Years since establishment Actual data
no_paid_workers Number of paid workers Actual data
1 = Metal and woodwork
2 = Construction
3 = Agro-processing
4 = Textile and garment
subsector Sub-sector of MSE
5 = Leather and footwear
6 = Retail
7 = Urban agriculture
8 = Others

Each group is expected to work on the data corresponding to its section and three sub-sectors as
shown below:

Group # Sectors
Group 1 1, 2 and 3
Group 2 1, 2 and 4
Group 3 1, 2 and 6
Group 4 1, 3 and 4
Group 5 1, 3 and 6
Group 6 2, 3 and 4
Group 7 2, 3 and 6
Group 8 1, 3 and 5
Group 9 1, 3 and 7
 2

Estimate a multiple linear regression model in which the dependent variable is current capital
(CC) and the independent (explanatory) variables are: initial capital (IC), sub-sector dummies,
business license (BL), access to loan from micro finance institutions (LOAN), access to
land/working premises (LAND), capacity utilization rate (CUR), years since establishment
(YSE) and number of paid workers (NPW).
a) Write down the population (true) regression model.
b) Write down the fitted regression model.
c) What percent of the variation in current capital is explained by the independent variables?
d) Discuss the ANOVA result (the F-test). Express the necessary hypotheses symbolically.
e) Which of the explanatory variables are significant predictors of current capital? Briefly
discuss each of them (level of significance, direction of influence, the interpretation of the
estimated regression coefficients, etc.).

EViews procedure
Import the data into EViews. First specify the subsectors to be included in the analysis. For
example, if the subsectors your group is going to work on are 1, 3 and 6, click on sample and
type the following in the box under ‘IF condition (optional)’:
subsector = 1 or subsector =3 or subsector =6

To construct a dummy variable for subsector 1, click on Genr and type the following:
Dummy1 = @recode(subsector=1, 1, 0)

Also construct a dummy variable for subsector 3 as:

Dummy3 = @recode(subsector=3, 1, 0)

Subsector 6 will be the reference category.

 3

2. Consider the data on the excess returns of Ford, General Electric, Microsoft and Oracle
together with the excess returns on Standard & Poor’s (S&P 500) stock market index from
February 2002 to April 2013 (‘Asset returns data for assignment May 2021’). Each group is
expected to construct a simple linear regression model (CAPM) for the specified assets and
time periods shown below.

Evening section 1 - Ford Evening section 2 - General Electric Evening section 3 - Microsoft
Group 1 Feb. 2002 to April 2013 Feb. 2002 to April 2013 Feb. 2002 to April 2013
Group 2 Feb. 2003 to April 2013 Feb. 2003 to April 2013 Feb. 2003 to April 2013
Group 3 Feb. 2004 to April 2013 Feb. 2004 to April 2013 Feb. 2004 to April 2013
Group 4 Feb. 2005 to April 2013 Feb. 2005 to April 2013 Feb. 2005 to April 2013
Group 5 Feb. 2002 to Dec. 2011 Feb. 2002 to Dec. 2011 Feb. 2002 to Dec. 2011
Group 6 Feb. 2002 to Dec. 2010 Feb. 2002 to Dec. 2010 Feb. 2002 to Dec. 2010
Group 7 Feb. 2002 to Dec. 2009 Feb. 2002 to Dec. 2009 Feb. 2002 to Dec. 2009
Group 8 Feb. 2002 to Dec. 2008 Feb. 2002 to Dec. 2008 Feb. 2002 to Dec. 2008
Group 9 Feb. 2002 to Dec. 2007 Feb. 2002 to Dec. 2007 Feb. 2002 to Dec. 2007

Evening section 4 - Ford Weekend section 1 - General Electric Weekend section 2 - Microsoft
Group 1 July 2002 to April 2013 July 2002 to April 2013 July 2002 to April 2013
Group 2 July 2003 to April 2013 July 2003 to April 2013 July 2003 to April 2013
Group 3 July 2004 to April 2013 July 2004 to April 2013 July 2004 to April 2013
Group 4 July 2005 to April 2013 July 2005 to April 2013 July 2005 to April 2013
Group 5 July 2002 to Dec. 2011 July 2002 to Dec. 2011 July 2002 to Dec. 2011
Group 6 July 2002 to Dec. 2010 July 2002 to Dec. 2010 July 2002 to Dec. 2010
Group 7 July 2002 to Dec. 2009 July 2002 to Dec. 2009 July 2002 to Dec. 2009
Group 8 July 2002 to Dec. 2008 July 2002 to Dec. 2008 July 2002 to Dec. 2008
Group 9 July 2002 to Dec. 2007 July 2002 to Dec. 2007 July 2002 to Dec. 2007

Weekend section 3 - Oracle Weekend section 4 - Oracle Weekend section 5 - Oracle

Group 1 Feb. 2002 to April 2013 July 2002 to April 2013 July 2002 to June 2012
Group 2 Feb. 2003 to April 2013 July 2003 to April 2013 July 2003 to June 2012
Group 3 Feb. 2004 to April 2013 July 2004 to April 2013 July 2004 to June 2012
Group 4 Feb. 2005 to April 2013 July 2005 to April 2013 July 2005 to June 2012
Group 5 Feb. 2002 to Dec. 2011 July 2002 to Dec. 2011 July 2002 to June 2011
Group 6 Feb. 2002 to Dec. 2010 July 2002 to Dec. 2010 July 2002 to June 2010
Group 7 Feb. 2002 to Dec. 2009 July 2002 to Dec. 2009 July 2002 to June 2009
Group 8 Feb. 2002 to Dec. 2008 July 2002 to Dec. 2008 July 2002 to June 2008
Group 9 Feb. 2002 to Dec. 2007 July 2002 to Dec. 2007 July 2002 to June 2007

a) Plot asset returns versus market returns. Does the plot indicate an approximate linear
relationship between the two?
b) Estimate a capital-asset pricing model. Write down the fitted regression model.
c) Is there a significant linear relationship between asset returns and returns on the market?
Interpret your result.
d) Is the CAPM ‘beta’ significantly greater than one? State the necessary hypotheses
symbolically and test the same. What is the implication of the result?