Pixelbliss - Fotolia
This could be the year when automated financial guidance – robo-advice – takes off in Europe. A research report published by investment bank Morgan Stanley in November 2015 predicted that several European banks would pilot robo-advice during 2016, many partnering with startups as the most cost-effective way to do this.
By submitting your email address, you agree to receive emails regarding relevant topic offers from TechTarget and its partners. You can withdraw your consent at any time. Contact TechTarget at 275 Grove Street, Newton, MA.
Some US robo-adviser startups have already linked up with established firms. Investment giant BlackRock bought FutureAdvisor, while Betterment is collaborating with financial services group Fidelity on a service for institutional investors.
Robo-advisers depend on their software. “It is core and vital to the business,” says Betterment’s chief technology officer, Dustin Lucien. “We like to say we’re a technology company that happens to be in finance.”
Betterment uses its own code for key functions. “That’s a strategic decision on our part,” says Lucien. “Anything we consider innovation or core to our competitive advantage we want to have full control over, and we also want to build precisely what we need and not configure something that has more general-purpose needs.”
In-house software runs trading systems, record-keeping and the website, and the firm uses the R programming language for its financial models, as well as Python and some open-source software. It uses packaged software for customer relationship management, project management and business functions, generally software-as-a-service.
UK robo-adviser Nutmeg, whose shareholders include 212-year-old finance house Schroders, takes a similar stance. “Nutmeg’s core systems – everything from the website right through to our core trading and asset management platforms – have always been built in-house by our own development team,” says chief technology officer Ewan Silver, adding that this provides greater flexibility.
Among programming languages, the firm mainly uses Java and Ruby along with Go for infrastructure, but is looking at functional languages to build new services – either Clojure or Scala. It has been purely cloud-based since it started, allowing some delivery teams to release several updates a day, says Silver.
“We do have a few packaged applications, most notably Babel Systems, which is part of our core trading infrastructure, but as a general rule we like to retain control of our technical destiny,” he says.
Nutmeg is an IT-dominated business, with the engineering and development teams making up the largest group in the organisation. “Organisationally, we are increasingly creating mixed ‘delivery’ teams that consist of developers, product, designers, operations, investment and marketing all working together to solve individual issues,” says Silver.
These teams, which are expected to try out new ideas to improve things for customers, include a customer outcomes group which combines engineers, designers, marketing and customer service staff. “Their goal is to help our customers make better financial decisions through the use of behavioural economics, design ‘nudges’ and a variety of other ideas,” he says.
Do everything online
Such work aims to let customers do everything online. “The fact that the site is easy to use, draws users’ attention to the right places and that we actively work to explain arcane financial terms in simple language means that the vast majority of our customers are more than happy to simply sign up for our service online and manage their money that way,” says Silver.
However, the firm makes customer service staff available by phone or live chat, and draws on what customers ask about to improve the website further.
Dustin Lucien, Betterment
Betterment also provides human customer service. “People like to be able to reach out and talk to another person and get some guidance,” says Lucien. But its staff are limited to helping customers use the automated service – they do not provide advice.
“The regulated investment adviser within the context of Betterment is actually our application,” Lucien adds.
Betterment is helped in this by its focus on long-term investment, which generally uses a basket of shares rather than individual stocks. To this end, the firm puts its customers’ money into exchange-traded funds, a type of low-cost index tracker. It also rebalances portfolios automatically, minimising taxable capital gains and maintaining the proportions invested in different areas.
Saving for the long term
Customers are often saving for the long term, such as for a pension, but that does not stop them worrying when prices fall – often the worst time to sell, says Lucien. “People don’t always act in their best interests in the market,” he adds.
Although the firm does not stop customers selling when they want, it does encourage them to consider the implications by providing information on the consequences, including through a tax impact preview.
Nutmeg also focuses on exchange-traded funds, but is currently only a “discretionary investment manager” rather than providing independent financial advice. The Financial Conduct Authority, which regulates the sector in the UK, treats advice generated by software in exactly the same way as that from a human, which means a robo-adviser providing advice is liable for the output of its software. Similar regulations apply in the US.
Automation and its limits
Alistair Haig, a researcher at the University of Edinburgh who previously worked as head of investment risk for fund manager Baillie Gifford, says it is relatively rare for UK-based robo-advisers to provide formal advice – although Wealth Wizards is an exception, providing independent financial advice on pensions online.
Instead, most give guidance to a limited set of products, which does not have the same responsibilities attached. They often do this by automating processes that are typically carried out by financial advisers through the use of fact-finding questionnaires. These will ascertain a client’s attitude towards risk, their financial goals – such as how long they are happy for their money to be tied up – and ability to sustain losses, with the answers used to help choose appropriate products.
Nutmeg plans to introduce full investment advice later this year. “You could say our suitability process is the embryo for a future advice offering,” says Silver. “It’s a short, simpler version of what our robo-advice service will deliver – that is, an intelligent, professionally constructed questionnaire that helps an individual to explain their circumstances and preferences to the system, which will then advise on the basis of Nutmeg’s in-house expertise.”
Silver says the technology implications for this “are not necessarily that complicated”, but that there are demanding challenges elsewhere, with the product team having to work with financial advisers to distil their knowledge into a decision tree process. “As we build and later extend the service, I would expect the longer-term implications for the engineering and data science teams to be very exciting,” he says. “It is a hard problem to solve, but one that needs to be dealt with.”
Silver adds: “Longer term, I think there is a huge opportunity to use state-of-the-art algorithms and data science to really understand individual risk preferences and goals, and potentially to take advice beyond what a human can currently offer.”
Ewan Silver, Nutmeg
Lucien also sees possibilities to expand Betterment’s advice. “We could advise around the selection of insurance policies,” he says. “We could help think about home purchases and the timing of that.”
The firm is already planning to do more to persuade customers to think long-term, he says. “The same type of communication you would look to an individual adviser for, we can deliver through our platform.”
Although there are new firms like Nutmeg and Betterment in every market, Edinburgh University’s Haig believes the next decade will see most existing financial services firms incorporate a robo-adviser system at the start of their sales processes. “What you end up with is a proposition to customers that’s part human, part machine,” he says, with customers that have straightforward needs being dealt with digitally.
“If something complicated comes up, then there’s a human intervention,” says Haig. For unusual customers and products, a human adviser will continue to be the better option, he adds.
Haig says starting with an online system has advantages for customers, because it will allow them to explore options without feeling pressured to make a decision in a meeting or a telephone call, potentially improving their financial literacy.
Alistair Haig, University of Edinburgh
This should help fill the ‘advice gap’ caused by the UK government’s Retail Distribution Review, which in 2013 led to a ban on independent financial advisers from earning commission. Although this practice was widely seen as generating biased advice, it also caused human advisers to charge upfront for their time, which means they are now rarely used by people without large incomes or assets.
“Robo-advisers, if they are done well, can provide an improvement for the rest of us,” says Haig.
Natural language generation
Other work in financial services is also being digitised, he says. As well as the use of natural language generation software to produce routine reports, financial research is moving from standard published reports to bespoke work where institutional clients set their requirements digitally. Although this will not spell the end of research analysts, it is changing the nature of their work, says Haig.
“People are having to innovate and provide different things, and are using a lot more technology to do that,” he says. “It could be a really exciting time. It’s going to get churned up in the way that, in the last few years, a lot of independent financial advisers in the older guard have given way, or have had to innovate substantially.”
Haig also thinks digitisation could automate some legal and regulatory compliance work, which has grown greatly since the financial crisis of the late 2000s.
“The more big data we have, the more that’s recorded on emails, the easier it is to process phone conversations, the more we use smart questionnaires, then it may be that headcount can fall,” he says. “That might make financial services safer as well.”