You are on page 1of 5

Lizi Zhu Stats Homework 1

For some reason, the layout of a word file goes misshapen when the file is converted

into PDF.

Q1

generate date_ = date(date,"YMD") // a little different than the example we

learnt in class

tsset date_, daily

describe date_

generate dow= dow(date_) //from the dates of data, we see that

holiday's prices have been screened out.

generate pricem1 = adjclose[_n-1] //unlike the adjusted prices in the example,

here the adjusted close prices look about right, so I choose to use them instead.

generate delta = adjclose - pricem1

generate R_t = delta/pricem1

hist R_t //draw histogram

summarize R_t, detail //compute 3rd and 4th moments

From the histogram, the return distribution doesn’t look like a normal one because

it is skewed towards right. Its skewness is -1.00276 , different from zero. A normal
distribution is always symmetric, meaning its skewness is zero. The return

distribution also has a kurtosis of 14.14506, in excess of 3,so it can’t be a normal

distribution or a near normal distribution.


15
10
Density
5
0

-.4 -.2 0 .2
R_t
Q2

drop if price==. //remove missing data as required

generate pricem1 = price[_n-1] //generate a time series of prices one period

forward

generate delta = price - pricem1 //price differences

generate R_t = delta/pricem1 //simple returns

hist R_t

summarize R_t, detail

We find from the histogram that the distribution does not appear to be a normal

distribution. The third moment and the fourth moment are -0.3813 and 2.6828.

Neither of the numbers is typical of a normal distribution. Without further

conducting a proper hypothesis test of normality, we cannot say in a statistical sense

whether the return distribution is a normal distribution.

You might also like