At the launch of FinTECHTalents in 2018 – Riskinnov introduced us to risk analysis, assessment and forecasting that helps you assess the risks you are and/or will be facing. Riskinnov returns for FinTECHTalents 2019. We caught up with the team to hear about ‘the story so far…’
What’s new? A new global challenge to wealth management?
The development of wealth management markets in recent years, whether for private banks, specialized departments of commercial banks or family or multi-family offices, have witnessed a new crucial variable emerging: “client-risk”. The growth in amounts under management slowed down partly as a result of increased competition, slowdown in stock market valuations and, more generally, a worsening of market sentiment. The real change, however, is that the evolution of these amounts has become more volatile than was the case prior to the great recession of 2008-2010, thereby affecting the long-run profitability of the sector, already restricted by new regulations (ban of retrocessions by MIFID-II, for example). However, the true importance of this substantially increased volatility is that it has settled almost permanently and it periodically affects the results of many players in the sector. In the fourth quarter of 2018, even well-established private banks have suffered outflows of funds which, in each case, amounted to several billions, while the beginning of the year had recorded net inflows of funds. 2019 has turned out to be no different. Riskinnov’s focus has therefore adjusted and we now propose a proactive client-risk management.
Where should digitization of wealth management extend?
This new challenge is global in the sense that it affects all dimensions of the business. To be sure, searching for new funds to manage and new clients is nothing new for all institutions of the sector, already mentioned above. Digital marketing is today an active area in wealth management. But every new client onboarding is not necessarily a long-term profit enhancing bargain, depending on its future behaviour regarding the nature of its supply of funds brought under management. Whether that supply is likely to be stable or not is now a crucial question. Traditionally, this issue has been dealt with ex post, by statistically analysing client’s behaviours. The increased importance of this client-risk calls today for a more proactive risk management approach. What this means for digitization is a new step forward.
While information treatment in the back and middle office has been a first step of digitisation, a new way of making use of it has appeared in recent years, i.e. to improve information circulation between clients and managers and bring closer to each other various categories of stakeholders. Online easy contacts, blogs, interactions, transparency, have brought clients closer to the centre of the business. A new step has eased information circulation, but also produced new information from this recent evolution. This is what the new version 8.2.7 of NERP is about. In particular, it is a solution to the pro-activity of client-risk mitigation.
Is digitisation enough to produce relevant information?
In itself, digitisation has already transformed wealth management, as has just been argued. However, it needs to use specific techniques and models to proactively produce the information needed to manage client-risk. In particular, Interactive Learning Elicitation (a yin–yang approach, as our logo symbolises), Behavioural Finance as revisited through Mathematical Psychology of Risk and, above all, an adapted Risk Modelling. Contrary to current practice in Finance, we use non-additive, non-linear, multidimensional and discontinuous risk measurement, challenging the idea that Brownian motion is a universal basis for risk modelling.
Why is standard risk-profiling practice insufficient?
Art. 25 of MIFID-II calls for risk-profiling and more specifically for an explicit assessment of risk-attitude. However, usual risk-profiling – including smart questionnaires – resorts to digital scales which basically map orders into numbers. Such qualitative scores are highly unstable (NASA has dealt with the issue) and clearly insufficient as a ground to management efficiency, but even more distant to the type of information needed to mitigate client-risk. From the regulator’s point of view, MIFID-II will continue to be a most incomplete success – not to say a failure – as long as risk-profiling will be based on present techniques. Actually, as is found in literature as well as in all studies conducted by several Financial Authorities in Europe, individual portfolios are more and more similar (and underweighting risky assets) than prior to MIFID-II, which is exactly the opposite to what regulators were looking for. Needless to conclude that present practice of risk-profiling forbids in most cases any meaningful client-risk management.
What kind of risk modelling should relevant robots be based on?
In risk-mitigation, one should look for efficient allocation means between various levels of client-risks. This in turn requires a two-dimensional view of risk and a way of measuring the levels of risk which entails some precise mathematical properties. One should moreover distinguish between (i) the intrinsic individual risk-attitude, traditionally put forward in the financial literature and in economic analysis, on one hand; and on another hand (ii) the contingent individual view of market sentiment, which emerges as a relationship between the individual investor and the evolution of markets. Many practical examples come to mind.
The first part of the risk profile of an individual – his or her risk-attitude properly said – varies very moderately for any given individual (as recent research confirms), except for extremely rare disasters (large military conflicts, events like September 11, major economic crises like the great recession of 2008-10, etc).
The second part of the risk profile adjusts relatively frequently to substantial changes in the general mood of investors and more generally of market operators. It causes very substantial variations in funds investing – the discontinuity of risk modelling is here the image of the « substantial » character of funds movements mentioned above.
The relative weight of the second part of the individual risk profile with respect to the first one is the true source of individual’s unstable behaviour. This individual view of market’s sentiment is the crucial variable on which to ground a client-risk management practice. Riskinnov’s research has been able to grasp and measure this crucial factor, the most important output of NERP 8.2.7
Does Riskinnov’s NERP-robot ignore more traditional requirements of wealth managers?
What has been developed above shows that NERP, by allowing to mitigate client-risk and hence to decrease the volatility of funds supply, enhances client’s loyalty, which belongs as much to risk management as to (i) strategic marketing. But there is of course no reason why it should have no impact on other issues in the business. In particular, NERP has been designed in such a way as to foster: (ii) Internal coordination between portfolios managers and clientele officers (presently very difficult…); (iii) marketing of the image of companies using NERP instead of boring and utterly non-convincing questionnaires (experience has shown NERP forerunners have been attracting UHNWIs); and (iv) a very strong defence against potential court actions, whether individual or class-actions, against wealth management institutions.
Does Riskinnov foster digitisation in other domains of financial activity?
Riskinnov offers decision-aiding regarding market volatility forecasting – this is of important use as well in commodities management – but also a multi-goal bespoke decision-aid under risk, in a sense defined in one of our articles (published with Oxford’s IMA jl of Management Mathematics). The latter applies as well to non-financial sectors. But the effort has been laid in 2019 on developing NERP 8.2.7 which is evoked in this report.
We have just been joined by a new young partner. We hope to be able to raise some funds in the coming year.
Visit Riskinnov’s FinTech of the Week page and see them tell their new Story at FinTECHTalents this November.