Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

In this Discussion

Here's a statement of the obvious: The opinions expressed here are those of the participants, not those of the Mutual Fund Observer. We cannot vouch for the accuracy or appropriateness of any of it, though we do encourage civility and good humor.

    Support MFO

  • Donate through PayPal

Can You Tell A Human Financial Adviser From A Robot? : Take The Human Or Robot Quiz

FYI: When was the last time you asked a human for driving directions? One day, the same could be true for financial directions.
Though many already turn to so-called robo advisers for advice about where to put their money to reach their desired financial destinations, MarketWatch wanted to see if these directions are any good. So we asked four prominent robo advisers and four of the traditional, flesh-and-blood variety to construct portfolios for a hypothetical 35-year-old investor with $40,000 to invest.
Regards,
Ted
http://www.marketwatch.com/story/can-you-tell-a-human-financial-adviser-from-a-robot-2015-04-21/print

Comments

  • msf
    edited April 2015
    Spoiler: about 2/3 of the people who took the 8 question quiz (could you tell a robot-generated portfolio from a human one) selected correctly over half the time, only 1/7 did worse than 50/50. FWIW, I got 75% (6/8) right.

    Step right up, try your luck .... :-)
  • Yea, me too (6/8).
  • I got very few correct. I figured the human would pick more funds to show the client they were working hard>Obviously that didn't always work
  • @MFO Members: The Linkster's 8/8

    Result: Scored 100%
    Regards,
    Ted
  • edited April 2015
    My score was a perfect 50%. (Exactly what you'd expect when guessing)

    When these robots sign-up and start posting here, then we got a problem.

    John Hussman might as well be a robot. Wrong now for how many years? Would also explain why his photo on his website hasn't changed a bit over the last 15 years. Robots don't age like we humans.
  • I got 7 out of eight questions corrected.
  • I got 7 out of 8.
  • To be honest, I have no clue as to what sort of a pattern to look for on this thing. I wonder if those of you who did well could share a few thoughts as to what factors you were evaluating here? Also, did you just take a shot by looking at the pie charts, or did you take the time to look up the various recommended funds to get more background detail?

    If you are able to get 7/8 or 8/8 by just looking at the pie charts, I'm truly impressed by you, and depressed with myself.
  • Hey @Old_Joe

    Nope.............could be such a thing as how folks process layout patterns. Folks do see the same data in a visual field that is or is not as comfortable for input.
    What if the data for the test was not pie chart, but bar graph?
    I betcha' the numbers may change for the number of correct answers with a different visual.
    I viewed the first set of data; but did not take the test.
    I'm the head house cleaning person getting ready for guests in a few days.
    A bit short of other time.
    Take care,
    Catch
  • Let me ask one question. If your robo advisor does you wrong, can you take it to court & sue for damages ? Make sure you read the fine print before you sign up !
    Derf
  • if someone programmed the contraption, you can try. If it programmed itself, you might have chartered territory.
  • @hank So, with a solid 50% score, does that mean you're a good guesser, a bad guesser, or just average at guessing?:) And since Ted got a perfect 100%, does that mean he is twice as good at guessing as you are? I dunno, quite the outlier.... almost super-human.
  • edited April 2015
    @heezsafe

    Yes - I'm an average guesser - "perfectly normal" in every respect.:)

    Ted? Ask him. My thinking is he finds hundreds of these quizzes every week. Did well on one of them and decided to post it. We'll know more when he responds to Old Joe's query above.
  • @hank. Ha! Well, Ted can get testy (sorry, @Old_Joe, I know I promised, but couldn't resist). But it's possibly a faux persona, his cover as a NSA proxy: every so often, he offers up an innocent-looking test and then the NSA vacuums the data for our profiles. Probably best to low-ball the self-report, to keep 'em "guessing.":)
  • edited April 2015
    Um, Ted has no "human" frailties. What is the title of this thread again? Can a robot tell a robot from a human better than a human can tell a robot from a human? Oh my! Maybe a robot programmer can tell a program result from a recipe result. But, but, but.... Oh I'm so confused. Which of the scenarios had a glimmer of creativity? Therein lies the dilemma. Who gets the higher fee - a robot or a financial advisor?

    But seriously - do financial advisors employ artificial intelligence?
  • edited April 2015
    The older I get the more that I think that my intelligence may be totally artificial. :(
  • edited April 2015
    Hey OJ - I'm with you in hoping that the 90%+ group share their secret. All I can figure out is that I recognized few, if any, of the fund symbols. That's largely because I haven't had a fresh idea in years. Basically, have stayed with the same 5 or 6 fund houses and pretty much the same dozen funds for well over a decade now. I guess in the past 3 years I've added just one new one (PRNEX).

    Possibly they benefitted from a superior knowledge of those fund symbols and were able to identify funds a computer might overlook.
  • Hi Hank- yeah, that's why I wondered if they actually researched those funds, which I would think would take an awfully long time for each question. Maybe some of our resident geniuses (Ted?) would be able to recognize most of the various funds offered, but I'm not sure, and even if someone did recognize almost all of those fund names, what would be the pattern recognition that would suggest robot vs human?
  • edited April 2015
    Hi OJ - In school I was taught that "correct" answers count for little. It's being able to explain how you arrived at the answer that counts. Still waiting. Maybe I'm just old fashioned.:)
  • Now, now... don't be so cynical. :)
  • @hank, I had lots of issues with explaining my answers. In math for example, I had figured out ways to do the computations in my head using shortcuts. Ex: what is 120x9=? 9x100=900. 9x20=180. Answer is 1080.

    The teachers hated it. I was supposed to use their system to figure out the problem which took much more time than the system I had. Perhaps much if this came about because I was in the generation of students that went back and forth between the old math and the new math.

    Algebra was very hard for me as you might imagine.
  • Although I am just a student, here is my 2 cents.

    Human advisers tend to concentrate on the major categories of US Stocks, Foreign Stocks, and Fixed Income. If a portfolio has 10% or more in the "Other" category, it is certainly a robo adviser. If the portfolio has less than 10% in the "Other", I bet that it is a human. This strategy gets 7/8 correct. It looks like there is one robo that is using a simpler, human-like allocation.
  • edited April 2015
    Hi @JohnChisum
    I still do a lot of simple math as you explained. However, through the 8th grade; when the math teacher would have a challenge in our class, lists of numbers to add were written on the blackboard and two students at a time would have a "race" to obtain the correct addition answer.
    Myself or one neighborhood girl, were the winners most of the time. We both used the method you noted, and I still add what I call "clumps of numbers" together. I am usually able to obtain a correct answer with a good deal of speed.........in spite of my "old" brain cells. My mother also does simple math in this fashion. We have talked about this and she can not recall "teaching" this method. One may suppose this is a genetic trait, eh? The way some brains are wired. I find this method also very useful with percentages. I can't further explain any of this.
    Take care,
    Catch
  • How interesting! I suffer from John's issues as well. My in-head approach to his example is 120 x 10 =1200, less 100 = 1100, less 20 =1080. More than one way to skin el gato.
  • It's actually enjoyable when everyone else reached for the calculators and you say the answer before they can compute it.
Sign In or Register to comment.