Podcast: Acerbi on backtesting ES and FRTB’s patchwork rules

Banque Pictet quant explains a new backtesting method for expected shortfall

Podcast 20.02.19 Mauro-Tom
Mauro Cesa and Tom Osborn talked to Carlo Acerbi via audio link

In this episode of Quantcast, Carlo Acerbi, head of valuation and quantitative solutions at Banque Pictet in Geneva, discusses his latest paper written with former colleague Balazs Szekely, an economic adviser at the Central Bank of Hungary in Budapest, which proposes a new backtest for expected shortfall (ES).

The new method, developed when the two quants were employed at MSCI, improves on their 2014 proposal by minimising ES backtesting’s sensitivity to the accuracy of value-at-risk prediction.

The bias to VAR predictions is inevitable, but it can be managed. By applying their method, one can not only calculate the probability of errors in the estimate, but also measure the difference between the predicted ES and the realised ES, allowing the error to be adjusted. 

Acerbi also shares his views on some parts of the Basel Committee on Banking Regulation’s rules, such as the P&L attribution test, which he considers “a Russian roulette for models”. 


00:00 Background history of ES and backtestability

05:55 The new backtest for ES

12:18 As unbiased as possible

15:20 VAR predictions affect ES backtest

18:45 How backtests of VAR and ES compare/sharp backtest

24:10 The P&L attribution controversy

29:55 Is FRTB killing some trading strategies?

To hear the full interview, listen in the player above, or download. Future podcasts in our Quantcast series will be uploaded to Risk.net. You can also visit the main page here to access all tracks, or go to the iTunes store or Google Podcasts to listen and subscribe.

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact info@risk.net or view our subscription options here: http://subscriptions.risk.net/subscribe

You are currently unable to copy this content. Please contact info@risk.net to find out more.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here