Daily Fintech Conversations

Are robo's just robo's on good days? The Brexit/Betterment Incident

What’s your view of Betterment’s decision not to trade for several hours Friday following the Brexit vote?


Fact no.1 - Betterment as a policy delays trading at the start of any day by 30 minutes.
Fact no. 2 - Betterment notified B2B clients (fin advisors) of the suspension. But of course not the duration.
Fact no. 3 - Betterment did not notify retail clients about the suspension.

How can all this make sense? One explanation is that algorithm is off when bid-ask spreads widen.
Another explanation is huristics combined with algos, are the strategy.

I have no explanation and I see no plausible one, for not notifying retail.

1 Like

What is odd is that they did not do it during other recent crashes eg Jan this year and August last year. Looks like a bad decision to me.

Maybe they did but not with such a long duration.

This is where one realizes that signing up to “Discretionary management” means you hand over the decisions, whether it is 99% to an algorthim and 1% to a human, or some other mix.

Accessing the service in the end, will include both components:

  • Did the algorithm work well? Goog allocation and good execution?
  • Did the human management intervene appropriately?
1 Like

The CEO got grilled on CNBC and handled himself with aplomb, but it still feels like a mistake. Impact on market is not meaningful (as AUM by Robos is still small) but it is wake up call that these services are not totally automated as advertized.

You can be sure, also here history will be repeating. In general clients are very often not ready to lose money. Time will come when they will be bashing robos. Not really much has changed since to tulip crash. We stay human, human brain will evolve but not within a few hundred years.


Our emotional intelligence is not that advanced. We are full of behavioral biases.
Therefore, when profiled and asked “How will you react if your portfolio drops 20%?”; we will give one answer today and then, change our mind when the market actually gaps 5% in one day.


It might be a consequence of general “weakness” of risk management models behind robo-advisors. Financial markets are more and more complex and volatile. The robustness of the risk management machine is probably less visible but not less important than a nice UX. That is what my partner Raffaele Zenti has written on the topic https://www.linkedin.com/pulse/roboadvisors-like-commodore-vic20-apparently-according-raffaele-zenti?trk=prof-post

1 Like

Betterment’s decision to wait was an as it seems was a discretionary one. A compliance issue. The Risk Tolerance Questionnaire used by most if not all so called Robo “Advisors” fail to be quantitative and serve similar to most Brick an Mortar brokerages the purpose to have a checkmark for compliance. The client does not really understand the expected loss he/she may expect if the market repeats a past financial crisis.

There is no value or a new invention. We’ve seen them in the US in the .com times as FinTechs (consumer was not ready) and in supply chain optimization, Travel… It’s a tech model not a business model. It will change the business. In the US driven by DOL regulation and in Europe due to MifiD II. Thats a good thing.

Communication by FinTechs, mostly weekly, is similar as it was 20 years ago by the wire house. It creates excitement and induces speculation. This will burn a new generation of investors during the next crash and makes investors take more risk on the market up move.

Mutual Fund sell past performance (they can’t repeat - Morningstar Study),
Robo’s sell lower cost. Neither gives a investor a better chance to reach his/her goals.


Is Discretionary the right analogy? Or is is like a Wealth Manager picking a fund and then getting concerned about style drift? We pick Robo Adviser asset allocation because we don’t believe in market timing and this move looks like Betterment trying to do market timing.

The good news that Jon Stein of Betterment reiterated when getting grilled on CNBC is that the customers are smarter than that and that they had record customer allocation on Brexit Friday. Retail customers have got the memo that you cannot win with market timing (actually humans cannot but AI based risk management aka Smart Beta maybe can)

@BernardLunn the Betterment halt is not about timing the market. It was triggered because the normally liquid low cost financial instruments used (country or sector ETFs) were MISPRICED & ILLIQUID suddenly.

ETF Repricing becomes problematic under certain market conditions. Not reflecting value if the underlying holdings and/or very wide bid ask spreads.

Bill Winterberg, the king for the digitization of US financial advisors, explains this very well in his latest fppad broadcast, titled SUSPENDED.

He informs us that on page 65 of the Betterment retail agreement, this issue is adressed.

For me this shows that technology hasnt helped resolve basic old issues: Customer signs 80 pages agreement but learns the hard way what he-she agreed for.


“In a time of extreme uncertainty, we wanted to be very careful about how we handled customer orders,” Stein said. “Even if you have the best technology, you don’t want to go out sailing into a hurricane.”

Not stating that we have an algorithm that evaluates … prior executing client orders. So it is at its best Market timing. True, not intended and the client’s best interest in mind. Still its timing the market as the floor broker would do with a @ your discretion order.

In Betterments defense, it makes sense to have experienced groups of humans make such decisions. Computers left alone can do funny things to the market and liquidity.

A good read: https://www.thefinancialist.com/are-you-smarter-than-an-algorithm-2/

1 Like

Thanks that is good read. If the algo is wrong, it should be fixed. Humans not trusting the algos and intervening occasionally on “gut” invalidates the idea of automation. Betterment was cleally allowed to do this and wanted to protect customers - legally correct and good morives. Not sure it sends the right signal about trusting tech to do its job.
Did any other quant funds or Robos do this?

Thanks @Efi that video is a good explainer. Market events like this help educate people about how markets really work.

The market is talking about the UK REITs that announced NO LIQUIDITY to investors.

Bond diversification isnt working. Corporates are becoming also less liquid.

Real Estate funds prove illiquid.

Are robos using ingredients that may not be good diversifiers and turn illiquid in volatile markets?

Maybe we should expect more breaks in trading from robos.

1 Like

They focus a lot on simplicity and transparency in their brand and
experience. A missed opportunity, if well-handled, to communicate.

I agree @amy.radin this is a brand opportunity but that takes courage and leadership in a market environment that currently has more fear than greed.

Betterment really missed an opportunity to innovate in the way robo-advisors communicate in stressful market conditions.
Betterment, one of the top Fintechs in the robo-advsior space, deployed old-fashioned finance practices when the market got volatile and choppy.

Thank you @marclussy @serena.torielli @amy.radin @afaessler @BernardLunn for pointing out the issues at heart, that were summarised on Daily Fintech in

An update on what Betterment is doing after the Brexit halting of trading.

Kimberly adds that Betterment has been engaged in an ongoing internal discussion about its controversial trading suspension, and has sought the input of advisers. The firm is putting final touches on a memo about its trading policies, he says.
“This is something we gave a lot of thought to,” he says. “We spoke to a lot of advisers, some of whom were perfectly happy with the decision, some who provided important constructive feedback.”

What about retail?