With changes abound, risk needs a big innovation that will change the game for good. But what is that innovation? The FutureRiskMinds, tomorrow's risk leaders, explore data and data management to explore the inefficiencies within the risk function.
A recent Netflix documentary ‘The Great Hack’, centred on the use of large-scale data gathering of American voters, to create a digital fingerprint based on individual political preferences, and the impact of targeted campaigning based on those preferences. It outlined how data is now often described as “the most valuable asset on earth” and a commodity that can be used to drive certain outcomes.
It’s undeniable that the breadth, depth and availability of data is increasing exponentially and in turn so does its power and its value. With the rapid evolution of the digital landscape in financial services and initiatives such as open banking, competition has never been greater. Consumer mobility is now actively encouraged and as a result brand loyalty is being eroded. Firms therefore must invest in extracting the maximum amount of information from the data at their disposal, not only to improve services but simply to remain competitive.
Traditionally the role of the risk manager has been to utilise model outputs to inform decisions on subjects such as lending policy, product launches, fraud prevention, and collection strategies. These models often consume a relatively small amount of historical data to project future trends, they generally have linear relationships and require a large investment of time to maintain, review, and analyse.
Machine learning has been used for some years in investment banking to predict patterns and trends in financial markets and provides the ability to consume and analyse vast amounts of data with non-linear relationships in near real time.
When considered in the context of risk management we can use AI to build our own digital fingerprint of each customer, with greater insight into, and understanding of, their individual needs and behaviours, beyond simply analysing life stage, financial commitments, and credit profile. We can begin to understand customer habits, lifestyle, and personal goals to produce a tailored suite of products which enable them to achieve their aspirations, manage their finances more efficiently or simply live more comfortably.
As with any new technology, challenges exist relating to cost of implementation and sourcing skilled labour to maintain algorithms and validate model outputs. This is in addition to ethical challenges around data privacy, security and bias. All of which mean that a large commitment both financially and strategically is required by an institution to drive the technology forward.
Bringing ethics & artificial intelligence into the corporate space
Whilst these challenges may be prohibitive to some, the technology around machine learning continues to evolve and it’s easy to imagine a time in the near future where the skills required to build and maintain AI processes are commonplace, the costs of maintaining a suite of AI analytics tools is far lower and the banking industry more dynamic as a result.
The next big innovation in risk management is capitalising on our deep understanding of the value chain.
A classical problem in risk management is its reactive nature, as the full impact of new risks is difficult to obtain, and instead, risk and profitability are handled locally and only analysed later globally. This results in suboptimal decisions in terms of trading, hedging and sales margins, as the decisions are made without a full understanding of funding costs, capital- and tax implications.
However, with new regulation such as Interest Rate Risk in the Banking Book and Fundamental Review of the Trading Book, where a very detailed understanding of the risk becomes necessary to satisfy regulatory requirements, the initial steps towards a more transparent value chain have started. This should be used as a stepping stone to an improved understanding of the value chain. Here, a deep understanding of the risk will become even more important as risk decisions will impact the return on allocated capital in far more significant way.
Natural extensions for this will be to capture the effects all the way through to the balance sheet and EBITDA, as the additional transparency will support making better strategic decisions.
Hence, improvements in the calculation framework and infrastructure that can spread the understanding of the value chain throughout the organisation; not only to the salesperson or trader, but also upwards in the organisation, which will be necessary. This will result in a more profitable organisation and a better understanding of what is beneficial as a whole instead of mainly focusing on the local trading desks or sales organisations.
Why do we need machine learning to effectively approach regulatory requirements?
An effective system integration and infrastructure implementation will become an even bigger competitive advantage, as an improved understanding of the entire value chain will help identify desirable and undesirable risk and business processes from a global perspective. Furthermore, trades, investments or strategies that may seem undesirable or expensive from a local perspective, may have beneficial effects either impacting capital by offsetting effects, triggering specific regulations or lowering risk for existing operations. The opposite might also be the case, as seemingly profitable decisions may have a negative impact on the organisation.
Furthermore, increased regulation increases the entry barriers for some markets, so a deeper understanding of the risk can become valuable in itself either as a service or through the additional income stream from entering lucrative markets.
Finally, the tax and accounting layer is often very far from actual operation and typically seen as a lower order effect, even though the implications can be significant. Thus, a stronger connection and better feedback loop can significantly improve profitability and stability of an organisation, which is the goal of any risk team. Hence, as a risk manager, this is a crucial part to include when managing risk of an organisation.
There are of course significant challenges in terms of infrastructure, modelling, and increased calculation requirements, but with modern hardware and calculation techniques, solving these challenges will be the next big innovation in risk management.
Read more >>
