Have you ever considered ditching your human investment advisor and going with an online automated service instead? In the last few years we have witnessed the rise of automated algorithmic financial advisors, also known more colloquially as robo-advisers. Being initially introduced to the general public in 2008 by start-up companies that offer automated financial advice and wealth management services, such as
Betterment, Wealthfront and Personal Capital, such services have gained greater market acceptance. Two main factors explain the excitement around these automated algorithmic financial advisers – performance and accessibility. In terms of performance, robo-advisers have proved to be very efficient and highly accurate. In terms of accessibility, robo-advisers are more accessible than traditional financial advisors. Robo-advisors never sleep, so they are available online 24/7. Fully automated algorithmic financial advisors are also cheaper than traditional human financial advisors, so more people opt to use them, given that they charge minimal to no fees at all. And lastly, because robo-advisers are cheaper to use, the minimal amount of money they require their users to invest is much lower than the amounts required by human advisors, and by doing so robo-advisers enable more individuals – even ones with very little money – to use their services. Given these advantages, it is not all that surprising that by 2020, robo-advisers are expected to manage $2.2-$3.7 trillion in assets.
Nevertheless, despite the popularity of fully automated financial advising services, their initial appeal has started to wear-off. Recently, the financial industry has witnessed a shift towards adopting a hybrid model that includes some human intervention in the services. Typically, such a hybrid model entails an automated service that offers periodic access to human advisors or in which human financial advisors add robo-advisers to their services, thereby providing some of the advantages that a traditional human advisor and a robo-advisor have over one another. And, it seems only logical that the hybrid model would become the preferred model, as long as implemented properly, for several reasons. First, while a new study shows that Americans are “more likely to automate their day-to-day finances than other activities,” not everyone is ready to do so just yet. Some still want to deal with humans when they need help with their financials. Second, robo-advisers are algorithms that are based on artificial intelligence technology, which is typically binary in nature, strictly rules based with respect to decision-making and lacks some of the flexibility that humans have when approaching making a decision. So, while it may be convenient and cheap to only use automated algorithms as financial advisers such technology has yet to make human financial advisors and their creative and innovative nature obsolete. That said, a hybrid model would still offer significant advantages. For example, robo-advisers, by design, have arguably fewer conflicts of interest, including financial incentives to attract clients, than traditional human advisors. Therefore, using a hybrid model might reduce some of the potential conflicts that might arise when using just humans, while still gaining the algorithmic advantage.
Algorithmic bias cannot be fully eliminated. Moreover, technology alone can help, but will probably never solve societal injustice or fairness debates. But, the negative impact of algorithmic bias can be reduced, by creating and applying industry standards regarding the levels of transparency needed in order to better understand what algorithms are based on and what information they use. Doing so, would help to better protect consumers, just like adding some human intervention in a hybrid model would add value as well. Additionally, such transparency could help prevent financial institutions from continuing to disclose only general products and service information without any relevant details, and actually make the human element in the hybrid model be effective.
Similarly, another example as to why proper implementation of the hybrid model is key, relates to the arguably high level of human reliance on algorithms. Originally, financial advising services were marketed to human advisors as a way for them to integrate digital advice into their services, automate some processes, increase efficiency, and free up time to counsel clients, while lowering their historic client fees. But, human advisors are likely to be influenced by our code-dependent society’s increasing reliance on advanced algorithms, as we tend to view algorithmic results as if they were objectively scientific truths. And so, much like jurors that get exposed to inadmissible evidence during a trial, it would be very difficult for human advisors to ignore and not be influenced by algorithmic results after they got exposed to them, especially, if one believes that we should let computers do what they do best – manage data and crunch numbers at high speed.
Human advisors might then view their role in the hybrid model as one that basically marries the efficiency of a computerized investment manager’s analysis with the comfort of having them (i.e., the human advisors) answer clients’ money-related questions, but not much beyond that. Yet, we should not permit for this to be the case. In order to fulfill fiduciary duties, it is imperative for human financial advisors to roll-up their sleeves and do the extra work necessary to understand true economic earnings and models rather than rely on purely algorithmic conclusions without the conceptual understanding that a human could impart to drawing such conclusions. Technologies like robo-advisers’ big data algorithms support this, because they make it much more possible for human advisors to apply the value investing principles, which have a critical role in offering high-level investment advice. After all, financial advising is more than crunching numbers and using computing power. The human financial advisors of the future will need to harness artificial intelligence to supplement their own human intelligence, not substitute it by blindly relying on algorithms. That’s why we still want humans, and not just robots, to invest and manage our money.