Blog

RoboQuestions

February 9, 2018

By STEVE WHITTINGTON, CFA

How will Roboadvisors react when increased market volatility occurs?

In math, and particularly computer science, an algorithm is defined as an unambiguous specification of how to solve a class of problems…algorithms can perform calculation, data processing, and automated reasoning tasks.

Guess who uses algorithms as trading rules for managing a portfolio of assets?

Roboadvisors do.

By now, most of you have heard of roboadvisors.  A roboadvisor, short for “robot” advisor, is a type of financial advisor that provides investment management of assets with little human intervention.  On the surface, they offer cost-efficient portfolio management capabilities, having set up very complex mathematical rules (you know, algorithms) to determine an asset allocation for any profile of client.

If you believe in classic “market efficiency” theory, then markets that are as deep and as liquid as, say, U.S. equity markets can move instantaneously and efficiently on information, and roboadvisors, armed with their powerful algorithms, can react faster than a human portfolio manager will be able to.  And while this is true, technologically speaking, there is one question that still needs to be addressed.  And that is simply:  what happens when the markets go down?

Roboadvisors, and the algorithms that govern them, were created in an unsustainably positive run-up of equity performance, and have never been tested in a prolonged real-time downturn, also known as a bear market (defined as a 20 percent correction from a market’s all-time high).

While roboadvisors were undoubtedly backtested against a robust set of market data and conditions, they are pretty new to the scene in an absolute sense.  They may be able to trade instantaneously, but how do they handle “animal spirits?”  How do they account for investment professionals who lived through lean markets in years past and have anecdotal observations that a roboadvisor does not? How would a roboadvisor handle an increase in volatility, much less an exogenous shock to the market like a 9/11 event, a fiscal cliff, or The Big Short?

On Friday, February 2 and Monday, February 5, we got the inkling of our answer.  On those two successive trading days, VIX did indeed spike 115%, and the Dow Jones Industrial average was down 666 and 1,175 points, respectively.  Yes, part of that market pullback was driven by human beings, in particular the ones who were scrambling to cover their short trades against VIX.

However, I suspect part of the pullback, and in particular the accelerating nature of the pullback, was due to certain safety parameters for these algorithms being breached.  In this case, selling begets more selling, thus the accelerating rate of decline as these algorithms continue to trade into their safety rules, each time trading on deteriorating conditions brought on by their previous sale.

Portfolio management is driven by fundamentals and metrics, but at its core is an activity that is meant to preserve and grow wealth for real human beings.  On a day like this Monday, when those real human beings need to talk to another human being, they are left wanting by the inherent lack of service that a purely analytical function cannot give.  At MONTAG, we offer hands-on client service as well as portfolio management depth and experience.  In times of increased market turbulence, that could be a decided advantage over the roboadvisor path, don’t you think?

Click here for our website disclosure.

get in touch