© 2020 Arizent. All rights reserved.

Tinder Understands its Users Better than Most Robo Advisors

Register now

Fiduciary implies making suitable recommendations based on a comprehensive understanding of the client’s goals and unique situation. If a handful of generic questions are deemed sufficient for Know-Your-Customer, then dating applications like Tinder are better fiduciaries than most automated investment tools today and the risk tolerance questionnaires they employ.

(The number of risk profile questions asked by robos in 2015 to obtain asset allocations ranged from Hedegeable's high of 16 to Betterment's and AssetBuilder's low of 4 questions, according to a recent study by Cerulli Associates. Schwab's Intelligent Portfolios asked 12 questions; Wealthfront and FutureAdvisor (now under BlackRock) asked 7; and SigFig asked 6 questions. Cerulli researchers noted that the number of initial questions might be misleading, since investors are sometimes asked more questions after they sign up, but before they invest.)

It’s strange when questionnaires ask hypothetical questions such as: “What would you do if the stock market drops 20%?” that can count as KYC. Such questions are intended to gauge a client’s psychological tolerance for risk, but what actually happens depends on the situation.

A fiduciary’s job is to de-risk before such a scenario occurs and deter clients from selling in a panic, in other words -- they should help clients act against behavior bias. Yet, if this is their objective -- why would a fiduciary recommend a portfolio that caters to the bias in the first place?

This is the broader issue with placing clients into pre-determined buckets. While the assets under management at robo advisors remain a tiny fraction of the market (barely 1% of overall investable assets in the U.S.), managed accounts have been widely adopted, especially by broker-dealers. Just moving from a commission account to a fee-based account with an additional risk tolerance questionnaire does not equal sufficient KYC or suitability measures.  

As more wealth management firms apply automated investment tools on smaller accounts, we realized that true KYC is still enjoyed by only high-net-worth and ultra-high-net-worth clients. Perhaps the next generation of robo advisors, be it direct-to-consumer or geared toward advisors, will know their customers before asking them to commit.


When building the algorithm for an automated investment platform in the U.K., the portfolio manager and I found that the least scientific part was mapping limited client input into a suitable portfolio. In terms of measuring suitability, KYC is the first part of the equation; without it, suitability can only go so far.

During that experience, I also worked with behavioral scientists to determine how a client’s psychological risk tolerance should impact portfolio recommendation. In other words, if a client is naturally prone to risk -- should we allow her to take more risk? The experts recommended that it should not move the portfolio risk level higher or lower, and should instead be communicated to the client. Given this conclusion, I question the scientific validity of tools that only gauge a client’s risk preferences.   

Tools claiming to be scientifically valid because they contain behavioral models, a mean-variance optimization, or the support of Nobel Laureates do not always produce suitable recommendations. Most U.S. robo platforms are more flexible about which portfolio the client can choose, and risk questionnaires put most clients into moderate and moderately aggressive models.

A proper KYC policy should examine not only a client’s goals and risk tolerance, but also the unique risks she is exposed to in her life and finances, sometimes unknowingly. A strong fiduciary would systematically identify these risks and design a set of customized solutions that target them specifically. As a client goes through different stages in life, the fiduciary should also adjust the investment strategy accordingly.

Risk definition is another aspect of suitability. Every model portfolio provider has a different definition of “conservative” or “moderate.” A model portfolio labeled as moderate can cheat by taking a higher level of risk to show a higher return or ranking during bull markets.

The fiduciary should help clients understand risk not as standard deviation but as the likelihood of permanent loss of capital, i.e. the potential downside, as the market is often driven by human behavior rather than normal distribution as assumed by Modern Portfolio Theory. MPT-based forecasts and portfolios can often surprise the investor by the downside.


The Department of Labor's proposed fiduciary legislation is expected to be a boon for robo advisors -- an assumption helped by U.S. Labor Secretary Tom Perez's endorsement of automated investment services as low-cost fiduciaries.

We’re all for technology lowering cost, though Perez's strong support for the fiduciary capability of robo advisors makes one wonder. Consider that the SEC and industry-funded studies like Melanie Fein’s "Robo-Advisors: A Closer Look," have questioned the fiduciary status of robo advisors, at least in theory.

In addition to suitability, Rule 3a-4 requires human services that distinguish an investment advisor from an investment company by having personnel available for consultation through periodic reviews or timely updates, and allowing clients to opt out of investments, none of which are met by this generation of robo advisors. Well-funded robos can afford to register as an investment company, but advisors who use automated technology or managed accounts may not want to deal with it.

Advisors can still be profitable while staying compliant with the right choice of technology. As regulators flesh out new rules, advisors would be wise to exploit the short-comings of competitors that violate fiduciary standards, and leverage technology to gain a competitive advantage over robo advisors and peers who do not utilize the tools available to them.

Min Zhang is CEO and co-founder of Totum Wealth.

Read more:

For reprint and licensing requests for this article, click here.