Friday, 28 January 2011

CFA Level 2, Reading 12.h, Dummy variables

I guess what makes CFA difficult is the fact that one also have other responsibilities!
2 weeks with little progress. Not good. Any way..



The LOS reads:
h. formulate a multiple regression equation by using dummy variables to represent qualitative factors, and interpret the coefficients and regression results


What is a dummy variable?
dummy variable is one that takes the values 0 or 1 to indicate the absence or presence of some categorical effect that may be expected to shift the outcome. 
For example, in econometric time series analysis, dummy variables may be used to indicate the occurrence of wars, or major strikes. It could thus be thought of as a truth value represented as a numerical value 0 or 1. (1)
(This reminds me a lot of using 0 or 1 to do computer programming at the machine level.) 

First time to remember!
If there are 4 variables, you select only 3 dummy variables (n-1)


An example dummy regression analysis might look something like this:
Earnings per share t = b0 + b2Q2t 


Where Earnings per share relate to a semi-annual observation of earnings per share
b0 = Average value of the EPS in the first 6 months of the year
Q2t = 1 if the period t is the second 6 months, otherwise 0


And a possible equation might look like this:
Earning per share = 2.3 + 4.5(Q2t )


Therefore, if you are requested to calculate the Earnings for the first 6 months:
Earning per share = 2.3 + 4.5(0) (because the dummy is 1 if the period t is the second 6 months)
Earnings per share = 2.3


If we want to calculate the second 6 months, it is straight forward, just include 1
Therefore:
Earning per share = 2.3 + 4.5(1 )
                           = 6.8
Source

No comments:

Post a Comment