The Risk Management Seminar 2020,
a Zanders production
As with famous movie franchises like Star Wars, The Avengers and James Bond, many risk management professionals have had to wait a few years for the next episode of the Zanders Risk Management Seminar. On Thursday 23 January 2020, however, the seminar rolled out the red carpet once again for its visitors at Pathé Tuschinski in the heart of Amsterdam. At the event, experts explained how to deal with current and future developments in the financial services industry.
In the large, Jugendstil- and art deco-styled movie theatre, Rob Naber welcomed all participants with a KYC (know your customer) analysis: the process of verifying customers’ identities and assessing the risks in the business relationship. He compared two opposites who were both defending their positions at the World Economic Forum (WEF) in Davos: American president Donald Trump and Greta Thunberg, the 17 years old climate activist.
“Although they both are world famous, political exposed persons (PEPs) and top speakers at the WEF, the difference couldn’t be larger," Naber noted. "How would a financial institution treat these two in KYC? If a bank or insurer puts all their activities and data into a computer, in terms of KYC it will probably say ‘no’ to Greta and ‘yes’ to Trump. What does that say? It means that not everything fits into a model; there’s always a need for a human interface.”
How to detect negative surprises
The first speaker, Lex Hoogduin, then took to the stage. As professor of Economics of Complexity and Uncertainty in Financial Markets and Financial Institutions he stated that we must take radical uncertainty seriously. After explaining concepts like uncertainty, surprise, risk and complexity, he illustrated his ‘Framework for Acting under Uncertainty and Complexity’ (FAUC).
“There are two key traits for successfully dealing with negative potential surprises: being alert and resilient,” Hoogduin explained. The framework is applicable to uncertainties like climate change, Brexit or new regulation. Answering the question how the last financial crisis came about, Hoogduin referred to a book called Stress Test, written by Timothy F. Geithner, stating that it was due to a lack of imagination. Apart from data, cultural elements are important too when it comes to scenario thinking, he stated: “The use of narrative techniques creates data to detect and to anticipate negative surprises. Rumor, gossip and stories are much faster than quantitative information.”
A practitioner’s view on model risk
Following the key presentations, seminar participants could attend one of the three parallel sessions. In one such session, Martijn Habing, Head of Model Risk Management at ABN AMRO, presented some practical dilemmas on model risk. He started by asking the audience to share their views on two contradicting statements. First, he focused on the truth factor of models (“To what limit can it predict the impact of an event?”), then he gauged the audience’s opinion on the scope of model risk management (dilemmas in soundness and efficiency) and the pros and cons of regulation.
In the audience someone stated that financial institutions “sometimes have to deal with conflicting regulations”. It was also recognized that modeling long horizons is difficult. Many changes in society, economy and regulation won’t be reflected in models predicting horizons of 10 years and beyond. “Another problem with long-tern modelling is that most models are typically static, while reality is typically dynamic,” Habing added. His last dilemma was on the experts’ role and whether we should let models force decisions. He ends his session by a crash course ‘How to be an instant model-risk critic’.
Gunter Lemmen, Data manager at VIVAT, led a parallel session on ‘Ethics and Artificial Intelligence (AI)’. “Your next Instagram post could determine your life insurance premium,” he stated, noting that AI is already here in the form of Bol.com, Deliveroo, Uber and many others. And not everyone seems confident about the way companies use AI.
“Of the Dutch population, 45% trust algorithms," Lemmen said. "Only around 25% of the population, however, trusts financial institutions with algorithms.” He then presented some interesting and confronting dilemmas following a video about self-driving cars. For example: “If a collision is unavoidable, should the autonomous car rather crash, or ram a motorcyclist next to you to save the passengers inside the car?” The audience was divided: would you sacrifice one life to save many others? It showed that AI contains more moral dilemmas than many realize. The bottom line of Lemmen’s session was to have good internal communication: “Discuss all viewpoints to know the benefit for the company, customers and society,” he concluded.
A regulator’s digital challenge
Jan Berndsen, Head of Retail Supervision for Insurers and Pension Funds at the AFM, started his session by explaining the differences between the two main regulators for the Dutch financial sector: DNB (the Dutch Central Bank) is responsible for the prudential supervision and the AFM (the Financial Markets Authority) is responsible for conduct supervision. Berndsen highlighted digitalization, macroeconomic developments and European regulation, political uncertainty (Brexit) and the transition to sustainability as main trends. Focusing on digitalization of the financial services, he gave some examples about ethical use of client data, noting: “Openness about challenges is important in supervision.”
Berndsen then presented some dilemmas for the AFM, concerning rules for information websites (e.g. Independer) to protect objectively offered advice, how to supervise an algorithm, the balance between consumer freedom and consumer protection, responsibility when algorithms make the choices, and the use of unstructured and behavioral data (e.g. driving behavior) in pricing and targeting clients. “In order to be a ‘smart supervisor’ we recognize the benefits of digitalization, but it also presents new risks," he commented. "Our standpoint is to be open but alert.”
Towards a platform in banking
After the interactive sessions, Joe Katz, Chief Risk Officer at ING Bank Netherlands, gave a presentation on ‘Transforming from pipeline to platform’. He first showed the audience a horizons model. “It was made by McKinsey, the Zanders of strategy consulting,” Katz quipped. The model showed three horizons in time and value. The first is to defend and extend current business: For example, what can banks do to maintain a healthy RoE in the current interest rate environment? The second horizon is to drive growth in new business: the focus here is on managing what Katz called “a tsunami of regulation. Every twelve minutes there is a new regulatory change in the banking sector.” The third one is to ‘set options for future growth business’: “The challenge is to become a platform instead of a pipeline”, said Katz, while showing that 6 out of the 7 largest companies in the world by market value have a platform business model. “A platform business does not create anything but connects different parties.”
He then showed how ING builds its platform by putting all services worldwide on the same interface. “But it may very well be that platforms are not the future or that they don’t work very well in regulated and fragmented industries like financial services," Katz said. "The only thing that I know is that we’re going to try, because it’s the dawn of a new era.”
Let's talk about numbers
As science journalist, mathematician and Professor of Science Communication, Ionica Smeets knows how numbers can sometimes help many but confuse even more. Although the seminar audience contained many quants, she made everyone aware that it easy to misinterpret them.
“I will give you some examples to show you how difficult it is for people to understand numbers,” Smeets revealed. She used the example of a virus outbreak, noting that how people interpret numbers depends to a large extent on how the numbers are presented to them. She then showed how easily people get confused by numbers, for example leading them to a false believe that regularly eating meat massively increases the risk of bowel cancer. She also gave an example of a good way to present numbers: “A visualization finally made me see the different opinions about Brexit in the British parliament within Conservatives, Labour, Lib Dems and the Green Party," Smeets said. "Yet it also made me realize how complicated Brexit is.”
Many people also get confused by the difference between English and American number scales, she said, showing a newspaper header incorrectly listing a trillion with 18 instead of 12 zeros. In the end, the audience had a fresh perspective on numbers and enjoyed a (number of) drinks while exchanging all new insights with each other.
Despite not finishing on a cliff hanger, most visitors indicated that they look forward to the ‘sequel’ of the Zanders Risk Management Seminar.