Overall a good video but it seems a few updates are required . For instance the PnL attribution - the old rule are mentioned here . Probably that is what is included in FRM syllabus , I don't know. The revised PnL attribution test is based on rank correlation and kolmogorov-smrinov test
Hello Raj, that's right. The video follows the FRM prescribed reading from John Hull's book on Risk Management and Financial Institutions (2018 edition). Will put in updates for PnL attribution in due course.
Very good and succint explanation. Do you do a case example on those calculations anywhere on your channel? it would be really great if you go through the steps with examples.
Hello Anand, 10/20/60/120/250 as per the 2014 consultative document when the idea was proposed to do away with the standard 10-day horizon used by VaR, and then later revised to 10/20/40/60/120.
If backtesting done for 97.5% confidence daily VaR results in more than 30 exceptions, the trading desk will have to switch to standardized approach till the issues with the model are sorted out.
@@finRGB ok so if our losses comes out to be more than VaR for more than 30 times we have to switch to std approach right. But since it is an internal model approach why r we restricted to 97.5% CI, why can't we take hight CI. I understand that it's given by regulator but shouldn't the board of tha bank decide the CI? Btw thanks for replying and also great video man i was reading about frtb but this video was very very helpful great work
@@bharatchoudhary4424 Your understanding is correct. With regards to choice of 97.5%, we are looking for a way to test an internal model that is based on Expected Shortfall prescribed to be calculated at 97.5% confidence. Expected Shortfall is not easy to backtest, so we backtest a proxy i.e. VaR and that too at two confidence levels: 97.5% and 99% confidence.
![[UNCUT] ลูกหมี-ลิลลี่ เหงียน เปิดใจฟาด ปู มัณฑนา I คนดังนั่งเคลียร์ I 15 ก.ค.67](http://i.ytimg.com/vi/FKQ60Y7bJ6s/mqdefault.jpg)
I watched a lot of FRTB videos…. This was succinct, refined and eloquent… The Best…,,Thanks
Thank you for appreciating, Francisco.
Excellent FRTB recap / overview. Very clear and understandable.
Thank you for the appreciation, Philip.
This is amazing explanation.. thanks 🙏
Excellent overview of FRTB chapter, i have went through multiple videos but this one is very good
Could not agree more....Crisp and clear...Well explained. Appreciate the efforts.
Awesome
excellent
This is so well explained.. thanks a tonn
Glad that the video was helpful, Darshana.
Overall a good video but it seems a few updates are required . For instance the PnL attribution - the old rule are mentioned here . Probably that is what is included in FRM syllabus , I don't know. The revised PnL attribution test is based on rank correlation and kolmogorov-smrinov test
Hello Raj, that's right. The video follows the FRM prescribed reading from John Hull's book on Risk Management and Financial Institutions (2018 edition). Will put in updates for PnL attribution in due course.
Very good and succint explanation. Do you do a case example on those calculations anywhere on your channel?
it would be really great if you go through the steps with examples.
Thank you for the kind words, Prof. Hassan. At the moment, no, but will cater to this in one of the future videos.
It should be
Liquidity Horizons => (10/20/40/60/120)
Correct me if I am wrong.
Hello Anand, 10/20/60/120/250 as per the 2014 consultative document when the idea was proposed to do away with the standard 10-day horizon used by VaR, and then later revised to 10/20/40/60/120.
Sir plz share the part 1 of the video
Hello Saurabh, this video relates to Part 2 of FRM. It reviews the prescribed FRTB reading in single video, so there aren't two parts to it.
What do u mean by the thresholds i mean 30 is the limit for what?
If backtesting done for 97.5% confidence daily VaR results in more than 30 exceptions, the trading desk will have to switch to standardized approach till the issues with the model are sorted out.
@@finRGB ok so if our losses comes out to be more than VaR for more than 30 times we have to switch to std approach right.
But since it is an internal model approach why r we restricted to 97.5% CI, why can't we take hight CI. I understand that it's given by regulator but shouldn't the board of tha bank decide the CI?
Btw thanks for replying and also great video man i was reading about frtb but this video was very very helpful great work
@@bharatchoudhary4424 Your understanding is correct. With regards to choice of 97.5%, we are looking for a way to test an internal model that is based on Expected Shortfall prescribed to be calculated at 97.5% confidence. Expected Shortfall is not easy to backtest, so we backtest a proxy i.e. VaR and that too at two confidence levels: 97.5% and 99% confidence.