All Models are wrong, but some are useful! Why good Model Risk Management matters

[seidu]

All Models are wrong, but some are useful! Why good Model Risk Management matters

This article may interest some of us!!

https://www.linkedin.com/pulse/all-...fa?trk=v-feed&trk=hp-feed-article-title-share

Thanks

 

[David Harper CFA FRM]

@seidu great share! Thank you. We decided to add a new board called "Risk (general)" which is below the board = About FRM but above board = P1.T1. Foundations of Risk (20%) precisely in order to support sharing of such links. So I moved your share here.

I like this VaR example and I think it's a good list of market moving events:

In January 2015 market risk of the EURCHF currency pair expressed in terms of a historical Value-at-Risk model was very low given a lookback period of, say, 2 years. Historical movements can only reproduce a restricted set of scenarios, none of them related to the unlikely but still possible unpegging event. Here, corrective actions such as a model add-on taking into consideration stress periods of currencies that had an unpegging event can be the right way to go.

According to UBS, besides Brexit, market-moving events of 2016 will be the Fed interest rate decision, ECB communication, China’s foreign exchange trajectory and the US presidential election on November 8

 

[seidu]

Hi @David Harper CFA FRM I was wondering where to post it!!
Thanks for adding the new board. I hope we can share similar articles to keep us updated.

Seidu.

 

[n_asnani]

Interesting article!

 

[Dr. Jayanthi Sankaran]

Just one more thing I would like to add here is a book by Emanuel Derman 'Models Behaving Badly'. Just bought it!