The core business of the teams ‘Market, Credit & Actuarial Risk’ and ‘Capital Management, Liquidity & Treasury Risk’ (being part of Financial Risk) consists of helping clients with risk modelling and risk capital challenges. In addition to our client work, all colleagues of Financial Risk work on one or more internal projects to strengthen and future-proofing our services. As such, I am involved in the data analytics & innovation team regarding financial risk management. Innovation and data analytics are of course popular and hyped terms and quite generic, therefore I will give you more context what innovation in our team entails and what my role is.
Setting the scene
Since the financial crisis, regulators have been more stringent to financial institutions and increased the number of regulations. All for the sake of preventing another major crisis, by increasing risk control. We believe that by setting the right incentives and making it easy and affordable to monitor the world around us we can offer an alternative to that and use risk to power performance. Innovation can be used to determine those incentives and monitor the risky world surrounding our clients. The most important solutions we are currently working on are robotic process automation, semantic text analysis and machine learning, which are elaborated below:
Robotic Process Automation
Robotic Process Automation (RPA) is a service with which we can save costs for our clients and make their processes more foolproof and their results more consistent. The challenge is to distinguish the repetitive and structured parts which are suitable for automation. This is especially difficult for financial risk management since the activities are definitely not standard and require a lot of intellect to complete. Together with a small team we developed a demo, (or ‘prototype’), to show our clients what can be achieved with RPA. The conversations regarding the PRA prototype enabled us to test our hypotheses and the feasibility of RPA in risk management and sharpen our storyline and product.
Semantic text analysis
There is a huge amount of unstructured and unused (public) data that can be used to assess different types of risk. With a third party we developed a tool which predicts credit risk for companies using publicly sourced newsfeeds. Language has its own logic and structure and based on that logic entities and semantics are identified. These semantics are then aggregated per company. While scraping and aggregating data is one, making it useful is another challenge. And that is exactly what we have done to make that public data useful for many of our clients. This tool is really promising but requires still some development and further fine-tuning. We work on this tool together with our clients to co-create a service that fulfills their needs!
Machine learning truly is a hyped term and not a day goes by without a newspaper publishing about it. The financial industry is strongly regulated and even the sort of financial models are prescribed, such as the more classic GLM/logit models. Machine learning models (random forest, neural network, etc.), however, have their advantages (and disadvantages) compared to these classic models. We are building different types of models to quantify risks for our clients. Currently, our clients use models based on machine learning techniques as challenger models to better understand on which areas the innovative models differ from the more classic ones. When the added value becomes clearer and the disadvantages can be explained properly, we believe this type of model will show its way to the regulator.
So, as you can understand there are many topics and innovations covered in the data analytics & innovation team and the list could be endless. It is up to you to bring in new innovations so we can explore and assess their value for our clients and ourselves. That amount of freedom and ownership feels rewarding!
Mijn naam is Tim Waaijer, consultant bij Deloitte Financial Risk Management. Als net-af student Finance aan de Universiteit van Amsterdam ben ik sinds 1 oktober werkzaam bij Deloitte. Ik heb als specialisatie Risk Capital. Verder woon ik een aantal jaar in Amsterdam en vul ik mijn vrije tijd met surfen, rennen, boeken lezen en hele goede koffie drinken.