Augmented Intelligence will make Financial Planning accessible to all

Photo by Olu Eletu on Unsplash

72% of Singaporeans are “very or somewhat concerned about being able to live comfortably in retirement” and target an annual return of 8.4% across all their savings and investment products according to a survey from Blackrock in 2016. But they hold about 48% of their savings in cash. In France, 70% of the respondents surveyed by Deloitte think that the current retirement funding system based on inter-generational solidarity will not last and that their pension will only cover 2/3 of their needs (2016). In the USA, the same year, a GoBankingRates report reveals that 56% of Americans have no- or less than $10,000 retirement savings.

Different countries, different policies, different cultures. Same aspirations. But same conclusion: when it comes to retirement, governments fail to address their citizens’ needs. Therefore, many are left on their own when planning for their future, not to mention various additional financial uncertainties: purchasing a house, children’s education costs, health issues, accidents, unemployment…

The private sector has its own flaws: opaque, complicated and often biased. Financial Advisors are mostly commission-based, “free” to the consumer, but in fact acting as the sales arm of financial institutions which are incentivized to push their own financial and insurance products that are not always aligned to the customers’ goals and needs. Fee-based advisors’ services remove this bias but few people are willing to pay for independent advisory.

What it tells us is that:

  1. Investing is a necessity for all,
  2. Most people do not have a plan to address a growing financial gap,
  3. There is a lack of financial literacy but also an opacity that benefits the sector,
  4. There is a necessary human element in the advisory process.

Since Artificial Intelligence in the true sense does not exist yet, a better choice of word at this stage is Augmented Intelligence: technology, data and algorithms can be used to enhance our abilities as humans to comprehend, interact with and decide upon the world around us.

So how can Augmented Intelligence make financial planning more accessible to all?

Robo-advisory consists of the automation of the financial planning process: customer profiling, goal setting, risk assessment, insurance policies recommendations, portfolio optimization and periodic rebalancing based on market conditions to stay aligned with the financial goals.

Automating this process will need to rely on more than simple hard-coded rules to be relevant and personalized. It will have to continuously learn to match customers goals and risk aversion to protection options and investment portfolios in the customer context; it will be able to “nudge” the customer to trigger action, as well as anticipate and respond to market events. In short: robos bring intelligent automation to the advisory process.

According to Statistica, the average assets under management (AUM) per user managed by “Robo-Advisors” in 2018 amounts to US$40,420 in the USA, $15,803 in France and $14,379 in Singapore. To put things in perspective, by 2025 the total market size of hybrid robo services will increase to USD 16,300 billion, which constitutes just over 10% of the total investable wealth.

Not all robo-advisory companies are made equal though: depending on their license, their operating model may go from a technology enabler (white labelled to financial institutions), to distributors (selling products from other financial institutions), to fund manager (ability to create and sell their own funds). Hence, some companies target consumers directly and attempt to break into the market with a super-low cost, fully automated offering, while others act as intelligent support systems to Financial Advisors and Wealth Managers -giving them evaluation and mapping tools to manage more customers better. Customer profiling and Portfolio Management are two areas of financial planning that can be significantly improved with Augmented Intelligence.

Today’s customer profiling and risk assessment is ridiculously lengthy, unsophisticated and inefficient and barely protects the novice investor from him/herself: a typical client questionnaire includes more than fifty questions and relies on the judgement of a human advisor to assess the customer risk appetite or aversion in order to provide the portfolio that best suits him/her.

As inventor Henry Ford put it: “if I had asked my customers what they wanted, I would have built a faster horse.” Similarly asking someone if he/she is ready to lose 5% or 90% of his/her investment value may not be the best way to get the full extent of the risk profile of an individual. But here is what can be done.

First, build quantitative behavioral economic models using decision games that reveal true preferences and calculate risk tolerance and loss aversion scores. Second, identify patterns in financial plans from people with similar socio-economic background so they can benchmark theirs against their own peer group. Third, make this process better over time with machine learning by correlating the decisions made by each customer along the advisory journey.

These intelligent and personalized nudges along the investment journey can make a tremendous difference in building trust and credibility while creating a sense of urgency and action.

Portfolio management can also be enhanced and optimized: assets allocation is based on client risk profile and interests: geography, currency, industry or even thematic such as environmental, social and corporate governance — ESG, China’s one-belt-one-road, autonomous vehicles, etc. Whereas passive investors favor “lazy man” portfolios made of simple three or four tier index funds, active investors want more granular control over their investments.

Such portfolios composed of diverse securities need to be tracked against market events which represents tremendous challenges today. As Blackrock puts it: “in a world that now sees an average of 4,000 brokerage reports a day comprising 36,000 pages in 53 languages, advanced text analysis is a necessity:” In particular, neural networks used in natural language processing provide new opportunities to augment traditional quantitative financial analysis with qualitative insights derived from the ingestion and processing of market news at scale such as sentiment analysis, institutional investors decisions patterns and market trends. This creates unique differentiation opportunities for both incumbents and new businesses by providing hedge fund grade tools directly to end-consumers -or to financial advisors and wealth managers to help maximize profits for their customers.

It also opens a new door for advisory services pricing models based on true investment performance as opposed as commission based, asset under management based or fee based models.

Moving to hybrid or fully automated advisory services with augmented intelligence requires removing some barriers and revisit regulations and licensing schemes:

  • Incentives and commissions should be revisited to encourage performance-based pricing and direct sale of products in a more fair, transparent and open manner.
  • The duty of care should apply to robos and algorithms which requires what some calls an “Explained AI”: automated tasks requires little explanation but automated reasoning requires explanations to be trustworthy, traceable and auditable.
  • Investors will have to overcome a psychological barrier to invest their life long savings in some obscure start-ups, or may favor instead sometimes less sophisticated but more reputable institutions.

In short, technology should not be replacing advisory human capital but rather augment its intelligence, abilities and productivity to serve more customers better, making financial planning more accessible to all.

Damien KOPP, Head of Products BlueFire AI & Live Withe AI Board Member

About Live With AI:

Live with AI is a non-profit foundation based in Singapore. The foundation gathers thought leaders, decision-makers and French, Singaporean, and international researchers to lead working groups and research projects on the positive impacts of artificial intelligence to our society. The Live with AI community takes advantage of a presence at the heart of the South-East Asia region and an access to several research laboratories to issue recommendations which can be immediately applied and tested among very diverse communities looking for technology disruption. Live with AI is an independent initiative created at the occasion of the France Singapore year of Innovation 2018.