Skip to Content
Rekenthaler Report

How Smart Should Robo-Advisors Be?

Answer: As much (or little) as they wish.

Kicked About
Earlier this year, one of the leading robo-advisors, Wealthfront, adjusted its investment strategy. (“Robo-advisor” being the argot for companies that sell automated investment advice.) Henceforth, the company reported, it would place some of its larger clients’ taxable assets into a risk-parity fund--“risk parity” being a hedge-fund-inspired technique that adjusts stock/bond exposures in response to movements in the financial markets, while employing leverage.  

This announcement did not go over well. In Wired, Felix Salmon penned “Beware of Roboadvisors Bearing Low Fees.” Financial Planning published “Wealthfront’s mutual fund spurs industry critiques,” with former Morningstar-ite Bob Clark stating, “This is a sales contest golf trip short of a regulatory violation.” The headline for The Wall Street Journal’s “When Your Investing Robot Has a Mind of its Own” (paywalled) was kinder, but the story, from Jason Zweig, was not. The critics spoke--by tossing tomatoes.

I would have, too. To use a soccer analogy, Wealthfront scored an own goal. Adding an in-house, leveraged strategy that has a significantly higher expense ratio (although halved from its original level) to an investment approach that previously held only outside index funds would always be difficult to explain. Doing so while lauding the virtues of indexing, and criticizing the investment choices made by traditional financial advisors, renders the task impossible.

(Whether risk-parity strategies qualify as active investment management is up for discussion. For example, Wealthfront and risk-parity inventor Bridgewater say not, whereas GMO and Morningstar vote yes. What is not open to question is that risk-parity funds feel different--and are priced differently--from conventional indexing.)

Considering Complexity
But enough about Wealthfront; it has already received its drubbing. Onto a follow-up question: Should robo-advisors add complexity to their investment processes?

The Journal’s Zweig thinks not. By adding the risk-parity fund, he writes, Wealthfront defeated its original purpose:   

“[The strategy’s] prospectus says the new fund should be used only by investors who understand complex securities, are highly risk-tolerant and who 'intend to actively monitor and manage their investments in the fund.' The whole point of using an automated online firm like Wealthfront, however, is that you don't intend to actively monitor and manage your investments. You want its computers to do that for you.”

My response requires a bit of backtracking.

Robo-advisors, as with all companies, require revenues. In achieving that goal, simplicity is not always a virtue. It may be that most investors require no more than a static blend of two funds, one being of the global stock markets and the other being of its bond markets. (The latter fund does not now exist but will shortly be launched by Vanguard.) Be that as it may, such advice cannot be sold. There’s no business to be had.

Which means that robo-advice must achieve a certain level of complexity--or “sophistication,” as the providers would state the matter. Robo-advisors can avoid adding complex funds, but they cannot avoid elaborating their asset allocations. If they were to use only a small number of funds, with an easy-to-explain allocation, many potential buyers would forgo paying the advice fee by duplicating those portfolios on their own.

For that reason, Wealthfront and its chief rival, Betterment, employ not two funds when forming portfolios but instead roughly a dozen. Significant differences exist between those companies’ portfolios--and those of the field’s other entrants. Prospective buyers must either conduct research to understand the effects of the various providers’ decisions or choose a provider on faith.

Considering Target-Date Funds
Of course, using a somewhat complex asset allocation is not the same as selecting a complex investment strategy. The decisions underlying the former may not always be clear, but the underlying pieces are transparent. An index fund moves precisely in tandem with its market. In contrast, funds with complex investment strategies sometimes deliver large, unpleasant surprises. They can lose more money than their owners thought possible.

That is the case against robo-advisors trying to be clever: They might get burned. There is, however, a counterargument. If anybody is to use complex investment strategies, it seems that robo-advisors should be pretty high up that list. They are institutions, with institutional buying power and institutional research. They have Ph.D.s at their beck and call. Let the robo-advisors bring those institutional strengths to the retail investor.

These words may sound familiar because similar discussions occur with target-date funds. Some target-date funds keep their portfolios simple--simpler, in fact, than robo-advisors can do because target-date funds need not worry that 401(k) participants will attempt to circumvent their structures. Other target-date series apply institutional techniques. Why invest one way for pension funds and another for 401(k) shareholders?

There is something to be said for the idea that the whole point of investment “solutions”--either structured as funds, as with a target-date series, or as with portfolios, as delivered by robo-advisors--is to ease an investor’s burden. Create the simplest possible solution, given the business constraints, and attempt no more. The best solution is that which never makes an avoidable error.

On the other hand, if investors these days prefer vanilla--as is indeed the case, with conventional index funds attracting most new inflows--then there should be investment opportunity for the other flavors. To be sure, the main attraction of robo-advisors is convenience (along with some bonuses, such as tax-loss harvesting). That will not change. But why not use some of that Ph.D. horsepower to advantage?

Labeling the Tins
Thus, I can accept either approach. Basic robo-advice (or target-date funds) is fine. So is the (allegedly) “smarter” version. However, I would caution that the two approaches be carefully differentiated and labeled. Mutual fund history suggests that the biggest investor disappointments come not from bad performance alone but rather from bad and unexpected performance. Make sure that robo-advisor customers realize what paths they have selected.

John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.