Does robo-advice work?

Tech start-ups are hot. Within that, ‘fintech’ is hot. And within that, ‘robo-advice’ is a term that’s starting to be bandied around even in the Indian market. What do these terms mean and why are they relevant to anyone associated in the asset/wealth management industry? (By the way, I have combined the wealth and asset management industries, as well as related service providers like stockbrokers and accountants, to call them money management industry).

Fintech 2.0 will address savings and investments

Until now, fintech, or financial technology start-ups, have been focusing on e-wallets/payment gateways and peer-to-peer lending, leveraging the broader trends in increasing financial penetration and rise of social media. According to an Oliver Wyman report, US$23.5 billion of venture capital investment has flowed into the fintech sector in 2013/14, of which 27% has been in consumer lending, 23% in payments and 16% in business lending.

Indeed, the same Oliver Wyman report states that fintech 2.0 is around the corner; it will encompass some important banking innovations based on ‘Internet of Things’, smart data, distributed ledger and frictionless processes beyond payments and consumer credit, in areas like mortgages and savings/investments. As examples of reducing friction in the savings area, the report cites examples of fintech tools such Mint, Yodlee, and Blueleaf that monitor customers’ spending habits. It also mentions SigFig and Betterment as examples of robo-advisers providing investment recommendations based on stated investment goals.

While the Indian savings/investment industry is relatively underdeveloped, it may well try to by-pass the expensive development process and jump directly to a more cost-effective process, as the telephony industry did. Providers like MyUniverse (promoted by Aditya Birla group), Arthyantra, FundsIndia, Scripbox etc are offering different levels of automation and execution to Indian savers directly, bypassing the expensive personal advice process.

Robo-advice automates part of wealth management

The wealth management value chain can be seen as two distinct sub-processes –

1. Client advisory process – where the adviser 1) understands the clients including their risk profile, goals, constraints etc; 2) then proposes a solution (based on the investment process); 3) executes the solution; and 4) reviews the portfolio on a regular basis, perhaps re-balancing. This is essentially the front office.

2. Investment process – where the adviser sources internal or external expertise on investments including on 1) long-term outlook on capital markets; which lead to 2) strategic asset allocations; 3) sometimes dynamic/tactical asset allocation; 4) decision on active/passive ways of investing based on prevailing opportunities; and 5) selection of funds or stocks.

The client advisory process is similar to being a general practitioner, while the investment process is similar to being a specialist. A specialty is not a trivial pursuit. It takes years of study and experience to master the investment area.

Robo-advice is basically a tool/site that provides portfolio management online with minimal human intervention, using some algorithms. The underlying principle is the use of set rules to segment customers into one, from within a range of, ‘risk profiles’ or ‘goals’, and then recommend the ‘optimal’ solution for that risk profile or goal. Different sites/tools offer different levels of ‘robots’ or automation in the value-chain.

Within this context, it’s useful to think about which points in the value chain automation can help –

• Risk profiling? I am still skeptical of the notion of risk profile, at least as a number like IQ, let alone there being a right method of some questions being able to deduct it. How does anyone know how they will react to an event they have never experienced before? While automated questions are great, I think good advisers use them to guide discussions.
• Goal calculators? Ok, it’s useful to have calculators converting my dreams to dollars – assuming reasonable assumptions have been put in.
• Asset allocation? I am not sure to how to automate working out strategic asset allocations without having an educated and subjective opinion on the choice of asset classes, optimization model (mean-variance anyone?), capital market assumptions, asset class limits etc. Institutional investors around the world use subjective processes or overlays because they know models don’t work. So why subject retail investors to that illusion?
• Re-balancing? Sure, automation is good.
• Fund selection? Haven’t we proven past performance, and therefore star ratings (by whatever name or brand), don’t work? Again, institutional investors and wealth firms select from a qualitatively pre-vetted list.

With all these limitations, it is worth asking what we are automating. Is it even possible to automate? Aren’t we just serving up pre-designed model portfolios?

If that is indeed the case, I believe the least we can do is handhold retail investors who are fearful of markets. I think advisers are very useful to educate and coach…and therefore needed. While I am all for reducing friction, I would be wary of any robo-advice site/tool that claims to offer ‘optimal solutions’ based on ‘sophisticated algorithms’.