Exam Feedback May 2017 Part 2 Exam Feedback

[vistag]

I selected - customers are shielded, so they aren't withdrawing the cash.

Yup

 

[bpdulog]

What was the answer for Vasicek model ??

I used this formula:



Forgot what I put down as the answer, maybe 7 something?

 

[KA1991]

Calculation of CCP's equity after exposure of its member.
Did you all get 250M?

 

[KA1991]

Volatility smile in case of price jumps.
This was a frown?

 

[KA1991]

-Best expected MVAR to Expected return (no portfolio var was given) - which portfolio will be increased.
This one I calculated Expected return/BETA = with highest getting most investment

-Step that will reduce undercollateralization
Decrease MTA or something.

-Best practice for small insurance company - fundamental principles of operational risk management.
Senior management create the governance structure for the firm

-About HFT - majorly from risk mitigation perspective.
Ability to track intraday risk exposure with controls.

-Value addition due to use of KRI metric.
This one was monitoring the business or something.

-How operational risk is managed when there's a Vendor.
BCM planning

-french fama - given the regression - to assess where the investment is focused (SMB, HML, Market)
this was value focused.

-how to correct biases caused by infrequent trading
unsmoothening process

Any agreements here?

 

[ashish_pandey]

W


Gev is a severity distribution right?

There were 2 parts to the question..1) threshold (so it cant be gev bit has to be POT or GPD) 2) pot, gpd and gev are probability dist functions with a location or threshold parameter and scale and tail index..they model the probability function of tail events..so i went for PARETO and FREQUENCY

 

[ashish_pandey]

Volatility smile in case of price jumps.
This was a frown?

Bimodal like graph with 2 peaks

 

[ashish_pandey]

Yup

Yes..retail depositors were shielded

 

[ashish_pandey]

True - I think I've done an error there. I selected GP with frequency.

I think thats right..GP with frequency..i can be wrong but i have done basis the logic i quoted in the comment above

 

[bpdulog]

-Best expected MVAR to Expected return (no portfolio var was given) - which portfolio will be increased.
This one I calculated Expected return/BETA = with highest getting most investment

-Step that will reduce undercollateralization
Decrease MTA or something.

-Best practice for small insurance company - fundamental principles of operational risk management.
Senior management create the governance structure for the firm

-About HFT - majorly from risk mitigation perspective.
Ability to track intraday risk exposure with controls.

-Value addition due to use of KRI metric.
This one was monitoring the business or something.

-How operational risk is managed when there's a Vendor.
BCM planning

-french fama - given the regression - to assess where the investment is focused (SMB, HML, Market)
this was value focused.

-how to correct biases caused by infrequent trading
unsmoothening process

Any agreements here?

I agree with most, not sure about MVAR and BCM planning. I think I put they had to look at all external audits of the vendor

 

[ashish_pandey]

I

-Best expected MVAR to Expected return (no portfolio var was given) - which portfolio will be increased.
This one I calculated Expected return/BETA = with highest getting most investment

-Step that will reduce undercollateralization
Decrease MTA or something.

-Best practice for small insurance company - fundamental principles of operational risk management.
Senior management create the governance structure for the firm

-About HFT - majorly from risk mitigation perspective.
Ability to track intraday risk exposure with controls.

-Value addition due to use of KRI metric.
This one was monitoring the business or something.

-How operational risk is managed when there's a Vendor.
BCM planning

-french fama - given the regression - to assess where the investment is focused (SMB, HML, Market)
this was value focused.

-how to correct biases caused by infrequent trading
unsmoothening process

Any agreements here?

I guess to have have the same answers as yours

 

[KA1991]

Bimodal like graph with 2 peaks

I think bimodal is the distribution, not the implied volatility. More of implied density

 

[ashish_pandey]

I think bimodal is the distribution, not the implied volatility. More of implied density

In john c hull..i wrongly used the word bimodal but incase of jumps you will not have smooth volatility smile but 2 peaks

 

[ashish_pandey]

In john c hull..i wrongly used the word bimodal but incase of jumps you will not have smooth volatility smile but 2 peaks

I believe to be sure on this as i knew a problem in john c hull specifically on this

 

[ashish_pandey]

I agree with most, not sure about MVAR and BCM planning. I think I put they had to look at all external audits of the vendor

For MVAR i remember to select the option with asset 1 and asset 3..but there were 2 options with both these assets so i dont remember..Low beta was the biggest clue..and i guess i have also gone for external audit but not sure..

 

[KA1991]

Here is the extract from one of the presentation on John C Hull Volatility smile, it is actually a “frown” with volatilities declining as we move out of or into the money? This is my understanding. I may be wrong.

 

[KA1991]

For MVAR i remember to select the option with asset 1 and asset 3..but there were 2 options with both these assets so i dont remember..Low beta was the biggest clue..and i guess i have also gone for external audit but not sure..

True, however this still required the expected return to beta calculation which would change the risk adjusted outcome. Even I remember asset 1 and asset 3

 

[Roddefeller]

I think bimodal is the distribution, not the implied volatility. More of implied density

I agree 100% with that explanation and so does GARPs official reading (Market Risk, p244-245)

 

[ashish_pandey]

I agree 100% with that explanation and so does GARPs official reading (Market Risk, p244-245)

With your snapshot i guess i got it wrong here

 

[Roddefeller]

True, however this still required the expected return to beta calculation which would change the risk adjusted outcome. Even I remember asset 1 and asset 3

As MVAR = VAR Portfolio/Total Assets Portfolio * Beta and VAR P/Total Assets is constant for all three portfolios, I think Beta equals MVAR, so expected return over Beta all day long in my humble view.