Rise of the robots: The man versus machine advisor debate

When it comes to investing, a new article written by Nick Shalek claims that software is better than 99 percent of humans when it comes to financial advising. His post has spread around the industry like wildfire, with wildly mixed reviews.

In 1998, the prevailing wisdom was that e-retailing made sense for every vertical and that all entrenched businesses would eventually fall to a new category-killer, brick-and-mortar was finished forever.  Which turned out to be true for some industries (books, music), half true for others (apparel, autos) and not true at all for a few (supermarkets - anyone seen Peapod lately? Webvan?)

So is financial advice like selling CDs?  Or is it, in fact, about relationships and confidence more than anything else?

I firmly believe that wealthy people will always prefer to get financial advice from living, breathing, thinking people rather than software.  They are happy to pay so as to have real relationships with their estate planners, financial advisors, CPAs and private bankers.  This seems obvious to me and probably to you.  But it is not obvious to many others.

I don't know Nick Shalek but he seems really smart and well-intentioned in what he's trying to say here:

Thankfully, Software Is Eating The Personal Investing World (TechCrunch)

The gist of his article, which is very good, is that software is better than 99% of humans who attempt to invest money.  His post has spread around my industry like wildfire and the reactions to it are all over the map.

I've been asked three times about what I think of his spiel, so I'll give you some bullet points of my reaction to this argument in its entirety, not just Shalek's rendition:

1.  I agree with him that simple is better, all new investors should begin with low cost index funds and set up their accounts to automatically add the same dollar amount to them on a regular schedule - regardless of price or market conditions.  They should certainly not begin by following stock picks from gurus or rodeo clowns on TV.

2.  I agree with him that machines remove the fear-and-greed cognitive foibles from the equation - but what Shalek doesn't get is that investors actually have to learn these foibles firsthand by messing up.  They also need to learn their own risk tolerance by making mistakes with fear and greed and dealing with the consequences.  Otherwise, they'll never improve and be confident in how their money is invested.

3.  Shalek has a great premise but then he completely screws it up by bringing the online investment advisor sites into the discussion as some kind of solution.  He casually mentions them like "these guys are fine, whatever" but the truth is that they are not a solution for anyone in need of real financial advice and they will be pivoting to a more human approach very soon anyway.  I cover this here.  My friend Leigh Drogen smashes them for an entirely different reason - their reliance on MPT and EMH in the strategies they run.  I won't digress here on that, read Leigh's piece for the final word. I'm going to go ahead and assume that Shalek is a VC backer of one of these online advisor sites or some entity he is affiliated with is or his friends are.  I could be wrong, don't feel like looking it up.  His post appears on TechCrunch and they love having authors post "articles" about companies they're invested in, it's practically become an art form over there.  The bottom line is that one or two of them will be successful but most won't and will disappear.  Also, they will never truly compete with financial advisors to the HNW channel who have real account minimums.  They will compete more with the Fidelitys of the world.

4.  He also trots out the whole "market-timing is dangerous" trope that Merrill Lynch and Charles Schwab have been sending you postcards about every month for the past 20 years to justify a buy-and-hold mindset.  And that stupid "if you missed the best 5 days your performance would have only been...".  We've actually done the homework on that ol' chestnut - turns out if you missed the worst 5 days or months or whatever, you were even better off.  And just so you know - those "best 5 days" are usually bear market rallies that occur as corrective counter-trends in the context of vicious bear markets in which they're enveloped.  The Dow doesn't go up 400 points in a day unless it's down 900 over the prior few weeks.  Go look those "best days" up and when they've occurred - usually very nearby the worst days.  Nick, some of us can "time markets" even though we don't necessarily use that term; some of us have the tools or the products or the platforms that allow us to get more heavily or lightly exposed at different points in time.  In a secular bull market this kind of risk modulation is not necessary, but is that what this last 12 years in the market has been?

5.  This is perhaps the heart of the matter: Financial advisory is only partially about running the money and most advisors outsource at least a portion of that particular aspect anyway.  Shalek has never served as a personal financial advisor (I don't believe he has based on his available bio) so I'm not sure he understands how much more is involved than simply the asset management part.  I don't run every penny I manage and software does most of the heavy lifting in terms of posture, stock selection, research, implementation, rebalancing etc.  If only the whole job consisted of simply running the money!  I'd be on a chaise lounge all year!  He seems almost clueless about what clients come to advisors for in the first place.  Can you imagine a wide swathe of wealthy people going through a year like 2008 with their money entrusted to a fucking website?  Emailing some knwo-nothing helpdesk kid in Bangalore who's pretending his name is Ralph while the world and its markets are imploding?  GTFO.

6.  Last thing - at turning points, the machines will miss the most important things.  Many humans will too, don't get me wrong. I spent an hour and half on the phone with a team running tens of billions of dollars for State Street Global Advisors about a new product we may bring on to our platform.  Their offering is quantitatively based but run by guys with hardcore fundamentalist chops and tendencies.  "Why?" is the question we asked.  The answer, which they've likely been delivering to every sovereign wealth fund and pension fund and endowment who've asked them, is that you can't trust machines implicitly in a world where policy and politics have replaced economics - a perfect summation of this era we find ourselves mired in now.  The machines cannot read policy.  It is also crucial to understand that at inflection points, the algo or software is going to probably do the exact wrong thing, because that is the essence of an inflection point - a point at which the way forward is totally obvious to almost all market participants.  At a certain point someone needs to think differently and by definition, the machine can't do it.

Anyway, that's where I stand on Nick Shalek's piece and the Man vs Investing Machine debate in general.  I'd love to be proved wrong and be able to entrust 100% of my clients assets and needs to a software program, but I live in the real world, not Disruption Hippie Land where every single industry can, should and will be broken just for the sake of breaking it.

Software is good, smart people employing software is better.