OCTOBER 2019CIOAPPLICATIONSEUROPE.COM8The Importance of Risk Management for the New Bank EBBE NEGENMAN, CHIEF RISK OFFICER AND MEMBER OF THE EXECUTIVE BOARD AEGON BANK AND KNABowadays, banking is all about the customer experience. Is your bank giving you the effortless onboarding process? Is the latest technology fully embraced and working on your iPhone? Is your watch an alternative for the banking card? If not, you should be aware that the new entrants in the banking world already offer all these features. Surely this adds to better banking at this moment for the customer. And this is noticed by the incumbent banks. Consequently, these old banks are massively partnering up with Fintechs, are implementing agility in their IT departments, and more often you spot a CEO of an old bank without tie but with sneakers. This movement is great, but there is more risk than ever in the system. The tech banking world is extremely complex and less than a few understand the complex algorithms and the IT systems that are in the heart of the New bank. Consequently good risk management is now evolving in the conditio sine qua non. "All models are wrong, but some are useful" is a famous quote of the British statistician George E.P. Box (1919-2013). The truth of his statement was evidence by the Great Financial crises of 2008. In normal circumstances combining mortgages of averages credit quality in tranches, one tranche with lower than average and another with prime credit quality, provides a useful investment opportunity and is mathematical correct. However, it turned out in the extreme event all mathematically proven un-correlated events were in the reel world behaving completely differently. The model was not even useful anymore, even stronger we discovered it had navigated us in the opposite direction. Basing your investment decision on the outcome of models only is therefore introducing a new risk. The bank should have good risk manager in place for addressing this type of model risk. Reel world is following its own principles. This holds for risk models as well. Even the methods that are in use to determine the likelihoods of extreme events can be completely wrong. A simple VaR model is commonly used by most risk management departments of banks on measuring market risk to survive all normal market volatility. In most banks the models are calibrated to capture about 99.95 percent of market movements. However in the crises we encountered market movements that were far in the tail of the 0.05 percent as predicted by these models. As a NEbbe NegenmanIN MYOPINION
< Page 7 | Page 9 >