Level 1: Post what you can remember here

varun34by02

Member
So could I be the refere :) ?
and the winner is......................................................Aleksander with the long run volatility (not gamma). Don't forget to pay the referee guys
You will definitely get you commission but don't you thnk you need to prove it 1st :p I won't lose $100 so easily ;)
 

varun34by02

Member
I read the question like this:
Garch (1:1) .... what's the long run volatility used in the model. you just have to square root the long run variance given in the information.
Yes. you are totally right,if that is the question.. but the point of contention is the question itself ...
And i think you are refering to a different question altogether here.. the 1 i am talking abt is the 1 where they gace A/B/C/D and asked which 1 has max weight on long-run variance and the options were just A/B/C/D ..no numbers !!

Anywas - Now I hardly see any resolution coming out of it unless GARP testifies :p :p
 

Frm_Indonesia

New Member
Anyway Guys, I still remembered some questions :
1) In normal distribution what is happen if number of sample is enlarged?
a . mean sample will be smaller
b. standard deviation will be larger
c. Mean sample close to 0
d. standard deviation close to 0

2)In Geometric Brown Motion, what is the consistent distribution to determine Future distribution?
a. if return normal dis then future also normal dis
b. if return log normal dis then future normal dis
c. ....
d. b. if return log normal dis then future also log normal dis

what do you think?

Thanks
 

varun34by02

Member
Anyway Guys, I still remembered some questions :
1) In normal distribution what is happen if number of sample is enlarged?
a . mean sample will be smaller
b. standard deviation will be larger
c. Mean sample close to 0
d. standard deviation close to 0

2)In Geometric Brown Motion, what is the consistent distribution to determine Future distribution?
a. if return normal dis then future also normal dis
b. if return log normal dis then future normal dis
c. ....
d. b. if return log normal dis then future also log normal dis

what do you think?

Thanks
1. As per central Limit theorem, I chose D
2. No Idea then.... No idea Now :p :p
 

MaggieMay

New Member
Anyway Guys, I still remembered some questions :
1) In normal distribution what is happen if number of sample is enlarged?
a . mean sample will be smaller
b. standard deviation will be larger
c. Mean sample close to 0
d. standard deviation close to 0

2)In Geometric Brown Motion, what is the consistent distribution to determine Future distribution?
a. if return normal dis then future also normal dis
b. if return log normal dis then future normal dis
c. ....
d. b. if return log normal dis then future also log normal dis

what do you think?

Thanks


2) Log normal return dist then future lognormal dist.
 

Aleksander Hansen

Well-Known Member
2) d. but the question is stupid:
No matter what the distributions of one-period returns may be, the long-term compounded values tend to be lognormally distributed if returns are independent. . This is the case for GBM.
 

Aleksander Hansen

Well-Known Member
Question 1

Fact 1:
If you take a large number of random samples of size n from the same population, and obtain the mean of any variable X in all the samples, these sample means will have a normal distribution around the population mean. (CTL & LLN)

Fact 2) The closeness to the theoretical normal distribution will increase if you enlarge sample size or you increase the number of samples taken.

Fact 3) Note that fact 1 & 2 says nothing about the distribution of variables in the population or within each sample.

Fact 4) When the sample size n is large, the sample average Y_hat is normally distributed with mean μY and standard deviation
σY_hat ≡ σY /sqrt(n); or equivalently, [(Y_hat − μY)/σY_hat] ∼N(0,1)

Fact 5): the precision of Y_hat depends entirely on the absolute size of the sample n, not whether n is large relative to the size of the population.

Fact 6): The spread of the data Y (population) will not change as the sample size grows.
The standard deviation of Y_hat will shrink as the sample size grows because we become more certain about the true value of the mean as the sample size grows.
 

nagesh_frm

New Member
This one was quite complex .. the Y-axis had excess portfolio return given..but i could not crack it :(very good question i must say.. such questions only differentiate classes from masses !!


This is how I solved it:

y = Rp - Rf
x = Rm - Rf
m = beta

The eqn was of the form y = mx + c
On substituting all the terms:

Rp = Rf + beta (Rm - Rf) + c

the first two terms together consitute the CAPM return. Since alpha is the absolute excess over the CAPM return, the answer is the constant value denoted by c. I guess , it was something like 3.x or so.

I think the R2 was a reduntant piece of info..
 

nagesh_frm

New Member
There was also a question about the Jensen`s Alpha and a given Regression. Actually i can`t remember what it was all about...

This is how I solved it:

y = Rp - Rf
x = Rm - Rf
m = beta

The eqn was of the form y = mx + c
On substituting all the terms:

Rp = Rf + beta (Rm - Rf) + c

the first two terms together consitute the CAPM return. Since alpha is the absolute excess over the CAPM return, the answer is the constant value denoted by c. I guess , it was something like 3.x or so.

I think the R2 was a reduntant piece of info..
 

nagesh_frm

New Member
The Y-axis is the excess portfolio return and the X-axis is the excess market (or target) return, and it askes what's the portfolio's jensen's alpha. The regression was like y=0.45x+0.31. Of course I don't remember either b0 or b1 but just made up the numbers.

This is how I solved it:

y = Rp - Rf
x = Rm - Rf
m = beta

The eqn was of the form y = mx + c
On substituting all the terms:

Rp = Rf + beta (Rm - Rf) + c

the first two terms together consitute the CAPM return. Since alpha is the absolute excess over the CAPM return, the answer is the constant value denoted by c. I guess , it was something like 3.x or so.

I think the R2 was a reduntant piece of info..
 

varun34by02

Member
This is how I solved it:

y = Rp - Rf
x = Rm - Rf
m = beta

The eqn was of the form y = mx + c
On substituting all the terms:

Rp = Rf + beta (Rm - Rf) + c

the first two terms together consitute the CAPM return. Since alpha is the absolute excess over the CAPM return, the answer is the constant value denoted by c. I guess , it was something like 3.x or so.

I think the R2 was a reduntant piece of info..

Honestly, with a month gone by the exam day.. i hardly rem a single question... this soln looks plausible :)

When are the results out ?
 

varun34by02

Member
July 6th. I am eagerly waiting.

Any idea, what could be the cut off?. I get mixed answers. Some say 60, some 67 plus....
yes 6th it is..no idea about cutoff though :confused:

We have argued a lot about the possible Integer..but no1 but GARP would know that.. with the quality of paper and considering 45-50% GARP Historical Pass Rate - 60 seems befitting !! but there is a but....who knows.. so lets wait for few more days - 15 to be precise :oops:
 
Top