AI for Advisory: a friend more than a foe

By - 3 October 2018

The slow but steady permeation of automation and machine learning in financial advisory and wealth management has been giving the jitters to some professionals. However, there is increasing evidence suggesting artificial intelligence (AI) can empower human advisors more than hurt them as it is far from being able to replace them. Moreover, unlike the retail sector where customers are curious to experiment with new AI-powered solutions, most investors still have a traditional mindset and are resisting change which raises questions as to what will be the best approach to educate the market in order to adopt innovation.

 
So, what can AI technology do for the investment management sector? What opportunities and threats lay ahead for robo-advisory? Who are the newcomers to what in this space? Read below as we unpack this topic.

 
What can AI do for investment managers

 
Automation and machine learning can be beneficial in multiple ways for the investment management and advisory industry. Among the many possible applications, here are the three main areas where AI can help:

 

1. Compliance support
Enforcing Know Your Customer (KYC) rules, Anti-Money Laundering (AML) rules, and tax reporting are all areas where automation is flourishing. Indeed, having algorithms that can process large amounts of data and scan past events in a very short time can enable finance professionals to spot fraud and irregularities much quicker than with traditional methods.

2. Portfolio optimization
Predictive analytics and machine learning represent very promising opportunities for the wealth management sector as they can be used to improve and facilitate portfolio optimization. Predictive analytics are run by algorithms that can analyse large sets of past data to highlight potential future trends and outcomes with high accuracy. On the other hand, machine learning is a feature of an AI algorithm which implies it can learn and adapt based on any type of data – even disorganised or unrelated datasets – without the intervention of a human. This empowers the human advisor to give a better service and it also allows the client to have more control in tailoring their portfolio to match their goals and risk appetite.

3. Customer service
Natural language processing (NLP) and natural language generation (NLG) are types of AI which can understand a query written in a conversational form and reply in a similar fashion, as a human would. For instance, some banks are already testing chatbots which would be able to understand and answer questions such as: “How is my portfolio doing today?” or “How much money can take out of my ISA?”. This type of solution can significantly increase efficiency in customer management and free up time for financial advisors, who can then focus on more complicated tasks.

Limits faced by robo-advisory

 
When we think of AI for business, generally the first thing that comes to mind is IBM Watson. Indeed, the product created by the tech giant has already been adopted by some very large financial corporations – such as Deutsche Bank and AZN – to support their investment and client management departments. It is therefore set to become a leading solution for robo-advisory. However, it’s important to note that although the IBM Watson Client Insight for Wealth Management product is clearly directed at financial companies, the tech behind it is still designed to work for a lot of different sectors. Considering the level of complexity of the financial sector, this lack of specificity can be a real disadvantage and a key obstacle to the integration of AI into financial services. Furthermore, IBM Watson requires significant investment in time and training for employees to be able to use it which may be another deterrent in its adoption. We argue that vertical and specialised AI solutions will take the lead in the long run as they will be easier to use and will be more able to respond to the complex demands of the wealth management and advisory sectors.

 
Another difficulty faced by the robo-advisory subsector is the lack of user trust. A study conducted by YouGov and commissioned by the trading platform IG in January 2018 found that only 23% of investors were confident they could trust a robo-adviser to deliver good investment returns. Moreover, the participants in the study also stated robo-advisory was their least favourite source of financial guidance. Therefore, it’s clear that the sector will have a hard time innovating fast if a mindset shift doesn’t first take place for investors. Nevertheless, maybe it should be considered as the responsibility of the finance professionals to educate their clients regarding the opportunities represented by new technologies. There again a gap in the technical skills of the financial advisors and wealth managers may be to blame as well.

 

Newcomers to watch

 
Among the 2000 startups that have been rated by Early Metrics, two in particular stand out for integrating AI technology into the investment management sector: Bambu and Euklid.

 
The latter is based in London’s Shard but led by a team completely Made in Italy, with Antonio Semeone at its head. Euklid creates a range of AI algorithms (aka algos for the techies) for wealth management based on biocomputing, a scientific approach mixing maths, physics and psychology. The team explains: “Our algos are specifically trained to recognize conservative and innovative behaviours, selecting from chaos and randomness, catching the result of mutual dynamic interaction.” Their algorithms are intelligent in the fact that they can learn from disorganized data and from the behaviour of other economic agents and then adapt the investment recommendations in real time – a feature they refer to as Genetic Logic.

 
On the other hand, Bambu offers both solutions for retail financial advice and for more affluent wealth management. More specifically, on top of AI algorithms for wealth managers, it also delivers white label solutions for financial businesses who want to offer a robo-advisory service directly to their clients. Their solution allows for a better tailoring of the portfolio to the individual’s risk profile, rather than basing the investment decisions on standard risk profiling. The startup was founded in 2016 in Singapore and has already expanded to Malaysia, Hong Kong and the UK. Moreover, its initiatives attracted the interest of big corporates resulting among other things in a series A funding round at $3 million in July 2018, with Franklin Templeton Investments as lead investor.

Press coverage