You are on page 1of 2

1

ECO347-FAC407
HW 2

NOTE: For some of the problems in this HW, you’ll need to look up some probability values. You can use
EXCEL, R, Python, STATA, SAS, MATLAB, or just your best friend – GOOGLE!

To refresh your memory


The standard normal CDF is
=

can be evaluated using tables or more easily using one of the software packages mentioned above. is the
standard normal PDF:
1
= , −∞< <∞
√2
The standard normal has mean 0 and SD 1.

Q1. (20 points)


A simple gross return is 0, 0.1 . Compute 1+ < 0.9 .

Q2. (10 points)


A simple gross return is 0, 0.1 . Assuming returns are an independent sequence, find the
probability that a simple gross 2-period return is less than 0.9.

Q3. (10 * 3 = 30 points)


Consider the !-period return: 1 + " ! = 1 + " … 1+ " $%& . Assume the following:
 ln 1 + ) is *, + , ∀-.
 The ) ’s are mutually independent.
01 $2
a) Show that .1 + " ! < /= . /. REMEMBER: is the standard normal CDF.
3√$

b) Given that the skewness of *, + to be 3 + 2 4 3 − 1, find the skewness of " ! (or,


which is the same thing, 1 + " ! ). NOTE: The parameter * is just a scale parameter; it has no
effect on skewness.
c) Comment on the above result. What’s the behavior over a long holding period?
2

Q4. (40 points)


The PDF of a Pareto income distribution is given by
67 8
5 = 89: ,
where > < > 0 and = > 0. The parameters of the model are <, which is the minimum income, and =,
which is called the tail index.
Take >& , … , >? to be all =@ AB =, < . Find an estimate =C of = by the maximum likelihood (ML)
estimation technique.

NOTE:
 The Pareto distribution was originally discovered and applied by Vilfredo Pareto to characterize
income distribution in a society. It captures the trend that a large proportion of the income is
enjoyed by a small proportion of the population. It’s also called the 80-20 rule or the Matthew
principle. A rule of the thumb is, 80% of the income is enjoyed by 20% of its population in a
society. It’s a power-law probability distribution.
 The maximum likelihood estimation (MLE) technique estimates the parameters of an assumed
probability distribution, given some observed data. This is achieved by maximizing a likelihood
function (or better, a log-likelihood function) so that, under the assumed model, the observed
data is most probable. If you have not seen this before, consult any stat book in the library or just
Google it.

You might also like