Gartner hype cycle and robo-advice
The development of the robo advice market has been slower and steadier these months. One of the reasons could be the news of large banks discontinuing their services citing high acquisition costs per customer. AltFi reports that one in four robo-advisors shuttered in two years. In terms of the Gartner Hype Cycle, it seems like the apparent failures of incumbents to leverage emerging technology threw the automated financial advice to the trough of disillusionment, which passes as long as the 2nd and 3rd generations of products appear.
However, we still believe that there are time and space for robo-advisers to gain momentum within developed markets. They don't stand still – Lloyds announced launching the robo-advice platform by the end of 2020 and Wealth Wizards launched fully compliant hybrid robo-advisor in July.
We believe that the core idea of providing automated financial advice should not be reduced only to the change of the supply channel, but it should be intelligently priced as well as significantly improved operationally. Here are a couple of ideas on how it could be done through improving the logic of the offerings: Beyond Modern Portfolio Theory: Expected Utility Optimisation, Hierarchical Clustering: Prediction of Systematic Underperformance.
What does sustainability mean for financial institutions?
The Swedish Supervisory Authority has issued several papers dedicated to achieving sustainability in finance. The regulator surveyed the sustainability work of financial firms and authored the report detailing the role of the financial industry in mitigating the consequences of climate change. Both reports suggest that currently, the financial sector mainly focuses on the ecological and social dimensions of sustainability issues. According to the Materiality Map by the Sustainable Accounting Standards Board (SASB), there are more aspects of sustainability that banks could enhance. The SASB suggests that the most important factors of sustainability for the financial industry are the following:
- Selling practices and product labelling;
- Product design and lifecycles management;
- Systemic Risk Management;
- Business ethics.
This means that to be sustainable, the banks and insurers need to go an extra mile to ensure that the customers receive fair, transparent, unbiased and secure services with reasonable risk management mechanisms in place.
Fighting bias – Methods of overcoming challenges to machine learning models
In the article by Forbes, the authors describe the outcomes of the research of technological adoption within financial services. They compare the leaders and followers within the sector in how they leverage machine learning methods. The results of the study suggest that the leading financial firms are very aware of the challenges that using these algorithms entail. These companies classified the risks associated with malfunctioning systems as of "extreme concern". The article by Banking Dive supports this view by highlighting the issues of explainability, and the potential risks the black box models entail within the lending business.
The concern of the industry is also shared by the UK regulator, which came up with a framework for addressing the problem of interpretability present in some Machine Learning (ML) applications. This method investigates the inputs and outputs of the model, but not its inner workings. It measures feature influences by perturbing input data in order to estimate their Shapley values, representing the features' average marginal contributions over all possible feature combinations.
We believe that sooner or later the flaws of machine learning solutions will manifest themselves in massive reputational scandals. Perhaps that would push the regulators to impose requirements on monitoring systems when using these methods. However, it is possible to defend your institution by recognizing the implied risk of these algorithms entail before any negative events occur and coming up with strategies of managing it.