Here at Dock9 we’ve been thinking about robots. Not just our dream of finding one to make our coffees and wash the mugs up though, we work with several mortgage providers and we’ve been looking at robo mortgage advice.
Robo advice hit the mainstream over the past few years providing investment advice at a lower rate to people who wouldn’t normally have knowledge of, or access to, investment vehicles. It was a natural next step for other financial companies to make use of this service and the most recent entrant to this market is Habito, who recently introduced a Digital Mortgage Adviser using a sourcing system with access to hundreds of products.
Customers have the ability to discuss their mortgage requirements from any device, using the AI and an algorithm which combines the key elements of a borrower’s financial lifestyle with real-time mortgage rates to produce an indicative monthly payment.
Although the vast majority of decisions are still currently made with a human broker (70%) this is likely to change in the future.
Certainly the Financial Conduct Authority (FCA) is interested in robo advice, having given its new robo-advice unit a £500,000 budget for its first year in July. The purpose of the unit is to provide individual feedback for innovative firms looking for authorisation, but will also aim to publish general guidance and toolkits for firms looking to develop automated advice models.
Does it mean a trip to the job centre for mortgage brokers? Not yet. The process is still far from fully automated and there is still a lot of human judgement involved. This is where the opportunities exist, if brokers can get in on the act in its infancy they can make themselves useful now and future-proof their jobs. After all, it is the brokers that train the machines at Trussle, another ‘Proptech’ company that uses robo-advisers
So how are the original robo-advisers faring so far in the investment sector? Well among the fans there has been some some criticism from some pretty big authorities. A Bloomberg article recently stated “The robo-adviser trend is too new to have produced definitive research comparing returns to those of human advisers. The U.S. Securities and Exchange Commission in 2015 cautioned investors that an automated program might not fit their needs or might be built on faulty assumptions. More recently, the top securities regulator in Massachusetts questioned whether an automated program could act in a client’s best interest if it knows so little about an investor.” (see the full article here)
One of the problems is AI is only as good as the data that is fed into it, just look at the rather racist result of the AI in the BeautyAI contest - out of 6,000 entries 75% of the winners it chose were white European. Out of 44 winners, nearly all were white, a handful were Asian, and only one had dark skin. That’s despite the fact that, although the majority of contestants were white, many people of colour submitted photos, including large groups from India and Africa. Algorithms can easily perpetuate biases, resulting in fake and vulgar stories trending on facebook and a racist chatbot spewing neo-Nazi views on Twitter
Another concern is companies that are providing the robo-advice AI won’t be the ones providing the financial product at the end of it. The European Banking authority discussion paper states: ‘As a result of increasing disintermediation in financial services, which is facilitated by the use of technology, it may be more likely that different financial institutions can be responsible for different parts of the service offered to consumers online. For example, the tool that offers automated advice to the financial institution’s clients may be provided by a financial technology firm (Fintech), who is responsible for its correct functioning, but may not be the same institution that the consumer interacts with. This increases the complexity of the business operations of financial services firms, and may create new or increased operational risks, that financial institutions are not used to managing.’ If something did go wrong who would be ultimately responsible? Would you blame the bank whose product you bought or the AI that told you to buy it?
The UK Financial Services Panel has pointed out these risks to the FCA: ‘Regulatory scrutiny and assessment of the risk and suitability of tools used by automated or ‘robo-advice’ services will be essential as those services become more widely used by consumers. Tools that are poorly developed and/or monitored have the potential to give rise to widespread systemic misselling and therefore further undermine confidence in the sector.’ No one wants to see a repeat of the endowment mortgage mis-selling scandal.
As much as the robot adviser needs neither remuneration, reward or lunch breaks so is great for the company, will it truly be beneficial for the consumer? Have these companies actually asked their potential customers if this is what they wanted? I’m assuming millenials, being the target market here, are too severely cash-strapped by student loans and high house prices to be the ones using it just yet.
We will be revisiting this topic again next year, until then only time will tell if robo-advice becomes the norm, be it for all aspects or just more simple mortgages/remortgages.
On a final note, as someone who is in the process of selling and buying a house at this very moment (with the aid of a lovely human mortgage broker called Dougie), I wondered - will many buyers be happy to make the biggest purchasing decision of their lives sitting alone at a laptop? With the complexity of my situation there have been lots of phone calls as well as emails and you can guess which one gets a quicker answer to my questions.
If you would be interested in taking part in our next blog post on this subject, or just want a chat over coffee (sadly still made by a mere human) talk to us on 0207 977 9230 or drop us a line at email@example.com