It's not "robo advisory" that does anything of what you mention above, just as little as "Computers" or "automobiles" per se caused catastrophies. Already we see all kinds of Robo advisories to all kinds of customer segments, at least in Scandinavia. It's just technology making processes more efficient. Evolution.

Richard, I don't criticize robo advisory per se, I just emphasize how it is sold to the customers (retail investors). Do you know a robo-advisory service, which frankly mentions all ist limitations and doesn't foist (explicitely or implicitely) the buy-and-hold approach as the ONLY correct one? If yes, please send me a link, I will review this service in positive context.

A few comments:

1) "risk profiling in a few easy steps" is just a first stage of robo-advisory; more sophisticated systems are already being developed and are yet to come. Potentially, they will go beyond the frames of an interview with a financial adviser, as they will analyze banking history and even social networks of an investor (with her/his consent, of course).

2) "extreme numerical sensitivity to the market parameters estimation errors" - isn't it the drawback of a traditional passive asset manager as well? There are some extensions to the MPT, which a applied in practice, that will be implemented (if not already) by robos.

3) "minimizing fees and maximizing the wealth is not the same" does not sound like a valid argument. Can you elaborate what you mean, please?

I agree with Vasily that "robo advisors" are just a marketing name to sell something that is old as world. However if we look at it from a more philosophical standpoint we could notice that our world in general is mostly driven by marketing in various forms; therefore when robo advisory's popularity exceeds a certain threshold, it will definitely generate something what I call a psychological feedback loop: the more people believe in a certain market phenomenon, the better it works. Just consider a 200-period moving average over daily bars for example. So, at this moment the quality of robo advisory will surprisingly increase, and its apologists will proclaim the advent of "more sophisticated systems", while it will be not because of the level of sophistication, but simply because of heavy use of whatever it is, be it sophisticated or not.

Ilia, I can see here no contradiction with what I said. You clearly stated it yourself: the client now has a choice between a number of loud marketing names, brands if you want, no more than that. As they (clients) had never understood how funds managed assets manually, now they don't understand what's going on behind the new name as well.

There is another, much more important problem that can be potentially solved by robo-advisors: to lay all the responsibility on client. In light of the latest tendencies in regulations no wonder that now "tech-savvy solutions" are so heavily promoted.

)^30 = 7.8%, i.e. an average annual growth of DowJones. Shall you give away so much for a passive allocation?!

Ilia Drozdov, CFA,

1) It may or may not happen. To put it briefly, what works for credibility analysis (payment history, social network) will not necessarily work for risk profiling.

2) Yes, it is. What I emphasize is that Robo Advisors do not warn their customers about this flaw. IMO they should (or even must). Just like the pharma companies are obliged to specify subsidiary effects of their pills pretty in detail.

As to MPT, one can easily check its sensitivity like I do it for Markowitz in my post: set some "genuine" market parameters, simulate a limited(!) sample of "historical price data" and try to estimate market parameters from this sample. Then compare the estimated and genuine parameters, and what is more important, the differences between genuine and estimated optimal portfolios.

3) Yes! In my paper I provide an example with wealthfront: they take 0.25% fee annually. But what they do?! They just passively allocate your wealth in "optimal" way. And over 30 years (1+0.0025

Alex Krishtop In part, I agree. in particular I believe that moving average and some more technical indicatiors (sometimes) work only because people believe in them.

So if many investor start using Robo-Advisors, it will create a bull market ... which, however, may turn to a bubble. And sooner or later all bubbles burst.

Typical performance of actively managed portfolios (in the past few years) does not support their superiority over passive allocations.

These performances does not even support superiority over index funds. Add to this around 1% of AUM over 20 years for additional fees, and passive starts to look fantastic. I am with Warren Buffet for this one.

That said, the primary target for the Robo Firms in $13 Trillion of not-invested funds and those invested in Mutual Funds (mostly passive with some 2% fees). And theoretically they differentiate themselves from ETF by the fact that they tactically (once every few months) change allocations between ETFs

Homie Malakooti

>Typical performance of actively managed portfolios (in the past few years) >does not support their superiority over passive allocations.

Agree, but I never claim that on average(!) the robo-advisors are worse than the actively managed funds (indeed the converse is likely true).

But many robo-advisors affirm - implicetely or explicitely - that NO active fund can beat them, which is of course not true.

In future I will also critically review the performance of actively managed funds (in Germany there are a lot of cases how NOT to run a fund). And also do my best to explain the retail investors how to avoid such losers.