Michael Mauboussin: “The Success Equation:Untangling Skill and Luck” | Talks at Google

MALE SPEAKER: Hello, everyone. Thank you all for coming. And folks who are joining
remotely– welcome, as well. Today we have a very, very
special speaker– someone whose work I deeply admire
and have been fond of. I was thinking about what is
the best way to introduce him? Should I say he is a Managing
Director at Credit Suisse? Should I say he’s an Adjunct
Professor at Columbia? Usually not all practitioners
are good academics or appreciate academia. And not necessarily
all academics are good practitioners. At lunch today our speaker said,
I love teaching and writing. And those are the two things
that he’s most fond of. So I think that’s the
best way to introduce him. He’s a teacher and writer. So without further ado,
ladies and gentlemen, please join me in welcoming
Thank you, [INAUDIBLE]. And good afternoon. It’s a real pleasure to
be here, and I really look forward to
today’s discussion. The topic today is an
unendingly fascinating topic of skill and luck. I’ll tell you my
own story of luck. When I was a senior
in college I had no idea what I wanted
to do with my life, but I did know I needed a job. I went to school
in Washington, DC and one of the firms
that came to recruit was Drexel Burnham Lambert. Now this is probably
before most of your time. But it was a fairly hot
investment bank at the time. On campus interview–
did sufficiently well that I was invited to
New York for the big day. So I get my very best
suit out, polish my shoes, go to New York. The day of the big
interviews, we’re sitting around a conference
room and they go, here is the basic setup. You’re going to have six
interviews with various people from our training
program, and you’re going to get 10 minutes
with the head guy. So of course, you want
to be good all day. But for those 10 minutes make
sure you have your A game. So I go through
my six interviews. They go reasonably well. I get my 10 minutes. I’m ushered into this
guy’s huge office, and I see peeking out
from underneath his desk a Washington Redskins trash can. Now I’ve been going to
school in Washington. The Redskins were
good back then. So I said to the
guy off-handedly, that’s an awesome trashcan. This hits the guy in
the emotional seat. And my 10 minute interview
becomes 15 minutes of him going on about
the virtues of athletics and how much the football
team is a metaphor for life. And I’m basically going
like this the whole time. So I go back to school. Couple weeks later I
get the coveted letter in the mail– you’ve
been offered this job. This is great. I’m gainfully employed,
I move to New York, and I start my job. About three months into
it, one of the senior guys in the program pulls me
aside, puts his arm around me, says, hey, kid you’re doing OK. Everything’s fine. I just want you to know. But I have to tell
you something now. The six people you interviewed
voted against hiring you. Very reassuring, right? So I’m like, OK,
so why am I here? And he said, well,
the senior guy came in and overrode
all their decisions. And he insisted we hire you. So I like to say my career
was launched by a trash can. And that was pure luck. And thankfully there are
no Laszlo Bock algorithms involved with that, either. Otherwise, I’d
still be unemployed. The topic today
is skill and luck. I really want to talk
about three things. First I want to
define my terms a bit and talk about what I call
the three easy lessons– three things you can take
away fairly quickly. In the middle part– the
meat of the discussion– I want to talk about the
complexion of skill and luck– what they look like and
how they change over time. And then finally I’ll wrap up a
bit with what to do about this, but also why we struggle
so much in our lives to understand the role of luck. One way to think
about this is to think about a continuum of activities. On the far left–
pure luck, no skill. You might think lotteries
or roulette wheels. On the far right–
pure skill, no luck. Maybe running races or
chess would be over there. It’s very important
to define terms. So let me just take
a moment to do that. I’m going to define
skill right out of the dictionary,
which is the ability to apply one’s knowledge readily
in execution or performance. So you know how to
do something and when you’re called on to do
it, you can do it on cue. Now luck– as you
might imagine– is much more
difficult to define. It actually spills into moral
philosophy quite quickly. But I’m going to
say luck exists when three conditions are in place. Number one is it operates for
an individual or organization. So it happens to you, or your
favorite team, or your company. The second is it
could be good or bad. And I don’t mean to
suggest it’s symmetrical because we’ll see in a few
moments it’s certainly not symmetrical. But there’s a plus sign
and a minus sign possible. And third is it is
reasonable to expect a different outcome
could have occurred. So if we rewound the tape of
time and we played it again, it’s reasonable to expect
a different outcome could have occurred. Now what I’ve done here for fun
is arrayed professional sports leagues based on one
season of performance. So this is where
they actually lie on the luck/skill continuum. That’s the NBA, Barclays Premier
League Soccer, Major League Baseball, the NFL, and
then finally the NHL. Now I’ll mention one other
thing on luck and skill, which is kind of fun. There’s a really
interesting test which I learned from
the poker people about figuring out if there’s
any skill in an activity. And that is, ask if you
can lose on purpose. If you can lose on purpose,
there must be some skill. If you can’t lose on purpose,
it’s basically all luck. So that’s another nice little
litmus test to figure that out. I’m going to go through this
very quickly, mostly as a set up for the next discussion. You might imagine that
your outcome in life is drawing from a luck
distribution and a skill distribution. So I take the two
outcomes and combine them, and that’s going
to be my outcome. So in the far left it would
be a luck distribution, and you’re drawing
only zeros for skill. So only luck is going
to make a difference. On the right it’s
going be drawing from a skill distribution
and zeros from luck. So only skill matters. And then everything
in life is going to be some
distribution for each. So we’re drawing
these distributions. So this very simplistic
set up allows me to proceed with what I
call the three easy lessons. So easy lesson
number one– whenever you see an extreme
outlier– and by the way, most of the outliers
we observe are positive outliers because negative
outliers almost always die, either literally
or metaphorically. So whenever you see
a positive outlier, it’s always the combination of
great skill plus great luck. And if you think
about it for a moment, that really has to be true. It’s a right hand draw from
both of those distributions. Now there are many ways
I could demonstrate this. One of the convenient ways
is the world of sports. This, of course,
is Joe DiMaggio, who hit in 56 straight
games in 1941. DiMaggio was a
.325 career hitter, one of the best of all time. And it turns out if you look
at all the players with 30 or more hitting streaks,
their career batting average is over .300. They’re about one and a
half standard deviations above the average. So saying this very
differently– not all skillful players
have streaks, but all the streaks are
held by skillful players. And it make sense, right? Because skill is the
prerequisite, and then luck comes on top to get
to be an outlier. The second observation–
second lesson– is about reversion to the mean. By the way,
reversion to the mean is a fascinating
concept because I think most people have a sense
that they know what it means. But if you actually observe
the behavior of most people, they don’t act as if
they understand reversion to the mean. So let me just be technical. You guys all understand this. Reversion to the mean says
that an outcome that is far from average will be
followed by an outcome with an expected value
closer to the average. Right? Now here’s the
classic example, which you may remember from
your stats class years ago– the heights
of fathers and sons. Very tall fathers
have tall sons. But the heights of the sons
are close to the average of all the sons. And likewise, short
fathers have short sons. But the heights of the sons
are close to the average of all heights of the sons. Now here’s the one
thing I think is very interesting
and practical about this reversion to the mean
concept and the skill/luck continuum, which is it
turns out where you lie on the luck/skill continuum
defines the rate of reversion to the mean– not
just that it occurs, but the rate of
reversion to the mean. So for example, if
you’re on the pure luck side of the continuum, there’s
complete reversion to the mean. In other words, expected
value of the next outcome is some measure of the
average, mean, or in some cases the mode. If you’re on the pure skill
side of the continuum, there’s no reversion
to the mean at all. We run a sprint against
Usain Bolt, he wins. We run again, he wins again. No reversion to the mean. So if you know where
your activity is, you automatically have a sense
for the rate of reversion to the mean. So if you’re thinking
about things like business performance or markets
or what have you, it’s a very, very helpful
and useful heuristic. Now the third lesson is
the one that’s probably garnered the most
interest from this book. It’s a concept I call
the paradox of skill. I want to be clear
that it’s not my idea, but I gave it that name. And the paradox of skills says
in activities where both skill and luck contribute to
outcomes, it is often the case that as skill increases–
skill improves– luck becomes more important. So how can we have
more skill leading to more luck being important? Now I learned about this
concept from Stephen Jay Gould, a very eminent
biologist who wrote a lot about evolutionary
theory and also liked to write about baseball. And one of my favorite
essays he wrote was about Ted Williams,
who you see here. Ted Williams was the
last player to hit over .400 in Major League Baseball. He also did it in the year 1941. And so the question
Gould was pondering is why has no one
been able to replicate this feat over the
ensuing 70 plus years? He said maybe it’s because the
players play at night, maybe because there are
relief pitchers now, maybe– none of these
things actually checked out. The answer was, it turns out
the standard deviation of skill for Major League Baseball
players has gotten narrower, which is to say the difference
between the very best players and the average
players is less today than it was a generation ago. So if you accept that
batting average is some skill plus
some luck combined, if the standard deviation
of skill goes down, that means the standard
deviation of batting average should go down. And that’s, indeed,
precisely what we’ve seen. So the standard deviation of
batting average in the 1940s was .032. And the standard deviation
of batting average today is about .027. So saying this
differently, Ted Williams was a four standard
deviation event in 1941. And if you were a four standard
deviation event in 2011– let’s say 70 years later–
you would hit .380. Now .380 is obviously fantastic. But it doesn’t get you over
that threshold of .400. Now here’s the thing
I want to emphasize. It turns out the paradox of
skill is everywhere we look. It’s very visibly
true in the world of investing where the standard
deviation of excess returns has been coming down
steadily for 50 years. It’s true in the
world of sports, as I’ve already mentioned. It’s also true in the
world of business– for example,
quality of products. Now it’s much more uniform
than it was a generation or two before. I want to leave you
with this one thought. The distinction
I’m making here is that absolute skill
has never been higher. And that’s true
everywhere we look. But relative skill has
never been narrower. That means more is
being left to luck. And I want to just
keep that in your mind. More is being left to
luck in our modern world than it has been in years past. We’ll come back to
that in a few moments. So those are the
three easy lessons. The paradox skill says as
absolute skill improves, if relative skill
comes down, more is actually going
to be left to luck. The paradox of skill actually
makes a very specific prediction which
is, in realms where there is no luck we should
see two things happening– if my description is correct. One is an absolute improvement–
in this case grinding to some physiological limit. I’m using an athletic
endeavor here. And the second is
clustering because I’m suggesting the right
tail is coming in. That means the performance is
getting clustered of people. So there are many
ways I can show this is actually occurring. One convenient example is
Olympic men’s marathon time. So the white line
there is the time of the guy who won
the gold medal. And you can see that
guy ran– in 2012– about 23 and a half minutes
faster than the guy who won the gold medal in 1932. So not a huge surprise. But if you’re a runner, it’s
about a minute a mile pace pick up over 80 years. The more interesting
line is the blue one, which is the difference
between the guy who won the gold medal and
the guy who came in 20th. Now mind you, this is supposed
to be some of the best, fastest runners in the world
competing on the same stage. That difference was
39 minutes in 1932. Now just take that
in for a second. This is the Olympic marathon. The guy has won the gold
medal, has taken a shower, is eating a sandwich,
and this other guy is still on the
course finishing up. That time is down to about five
minutes, seven minutes today. And I think we can
confidently say if we meet in the
future, that will be an even shorter
amount of time. So we see in swimming,
rowing, running, this convergence of times. And as you know, we need now
incredibly sensitive devices just to figure out
when the players leave and when they finish in
order to measure truly which one is the fastest
in that particular event. Let me shift gears a bit and
now talk about skill/luck. I’ve been depicting them with
these normal distributions, but of course,
that’s not the case. Skill– let me start with that. It’s actually quite
easy to describe. Skill tends to follow an arc. So we call this
the arc of skill. Again, let me start with
a physical endeavour. This is sports. So for example, if you’ve
ever played a sport when you start off
when you’re young, you don’t have a
lot of proficiency. But as you practice, you get
better, and better, and better. At some point you get
to peak performance as you’re getting stronger. And then you come
down the other side. So there’s this
very definable arc. By the way, for sports the
best way to predict this is slow twitch versus
fast twitch muscles. It turns out that in sports with
a lot of fast twitch muscles, players peak young. So sprinters– 22, 23. Basketball players–
their mid-20s. And then if there’s
more slow twitch, you peak in your
late 20s, typically. So baseball players– late 20s. Some football players–
late 20s, as well. By the way, if you’re a golfer,
there’s some good news for you. Golf tends to plateau
from roughly 30 to 35, and then it rolls
over after age 35. So here’s Tiger
Woods, who is now 38. So no matter what other
issues he’s dealing with now, Father Time is also
knocking on his door. Here’s the actual arc
of skill for tennis. This is quite interesting data. This is the winner of Men’s
Grand Slam tournaments in the Open era. So 1968 through 2013. So on top you can
see the distribution. And it turns out, the
mode is 24 years old. And it turns out in
the last 40 years or so, the mean is
actually 24 years old. So 24 years old is
basically the sweet spot for winning men’s
tennis tournaments. Now the bottom panel–
it’s a little hard to see from where you
are– but the bottom is actually the age of every
winner of every tournament since that time, with
a line at age 30. And what we know is in the
last 40 years– so 160 events– only four players have
been over 30 years old who’ve won a Grand Slam. Only four players. By the way, we had a
little bit of excitement about this just a couple
weeks ago with Roger Federer trying to win Wimbledon. He’s 32 years old. He’ll be 33 in the fall. It’s simply no one has won
Wimbledon over the age of 30 since Arthur Ash did it in 1975. So it was really– it just shows
you how amazing Federer is. But at his age, with
17 Slams, what happens is you physically slow
down just a tiny bit. And that difference
is the difference between you and a
younger generation. So we’ll see if in men’s tennis
this new, young generation starts to take over. This is a very, very
pronounced pattern that we see in every sport
and you know, bankable. Now your reaction to that is
probably something like, oh, that’s interesting. Maybe it helps you in your
fantasy sports league, or something. But what you care about is
cognitive performance, right? Psychologists study
cognitive performance. And it follows the
same arc, incidentally. And it turns out
psychologists often agree that cognitive
performance is a function of two
types of intelligence– so-called crystallized
and fluid intelligence. And you can see from
the picture this is sort of a bad news,
good news situation. Let’s start with the bad news,
and that’s fluid intelligence. So fluid intelligence
is effectively your ability to
deal with novelty. So if I present you with
a series you’ve never seen before, how good
are you at figuring out what’s going to happen next? Turns out we have
a couple people who are probably right at
their peaks in the room. Fluid intelligence tends
to peak in your early 20s and then pretty much goes
downhill throughout your life. So if you’re out of
college, basically, you’re probably on the
down swing in terms of your fluid intelligence. Crystallized
intelligence– which is the good news, of
course– is something that is exactly
what it sounds like. It’s your cumulative knowledge. It’s your wisdom, understanding
of facts, and so forth. That pretty much– barring
cognitive impairment– pretty much grows
throughout life. So you can see that. And it’s the combination that
defines this overall arc. So here are the
basic trade offs. Now I just want to
mentioned a couple numbers, and maybe I’ll ask you
guys to participate. I’m going to ask you a
couple numbers about peak age performance for certain tasks. So the first one
I’ll throw out there is there was a study done of
institutional money managers. So these are portfolio managers. And the question I’ll post
to you is, at what age do you believe institutional
money managers deliver the best excess returns? So what age do
institutional money managers deliver the best excess returns? Throw out some numbers. What do we hear? AUDIENCE: 72. MICHAEL MAUBOUSSIN: 70? God, I love that
answer, by the way. I’m holding on to that one. AUDIENCE: 55. MICHAEL MAUBOUSSIN: 55. AUDIENCE: 45. AUDIENCE: 40. MICHAEL MAUBOUSSIN: 40. 40 40. 45. Very specific. OK. So the answer– well,
the actual, exact answer is the range of 40 to 44. But let’s call it
42 as an average. So you guys, you know, some of
those guesses are really good. Let’s hold on to
that 70, though. I may modify that one. So institutional money
managers– early 40s. The other one that was even
more interesting for me– this was a paper done by David
Laibson and some colleagues at Harvard University. And this is where they used
massive data sets for this. They went out and studied– now
this is not people like you, but people in the
real world, out there. At what age do people make
the best household finance decisions? So this is at what
age do people get the best terms on
their auto loans? At what age do people
get the best interest rates on their mortgages? That type of stuff. So there were about a dozen
categories, tens of thousands of people participating. At what age do people make
the best household finance decisions? What number, what age
do you think that is? AUDIENCE: 30 to 35. MICHAEL MAUBOUSSIN: 30 to 35. AUDIENCE: 55 to 60. MICHAEL MAUBOUSSIN:
What was that? AUDIENCE: 55 to 60. MICHAEL MAUBOUSSIN: That’s good. I like that one. That’s good. So anybody else? Yeah? AUDIENCE: 40 to 45. MICHAEL MAUBOUSSIN: So 40 to 45. This is one where I’m thinking
the older– my thought was older is better, in part
because older people tend to have more assets
than younger people, on average, which is true. And really, there’s no
cognitive impairment issues even when you get into your
60s, typically, for most people. Turns out the age was 53. 53. And there’s a standard
deviation around that. But almost everyone in
the categories the mean tended to converge on age 53. So the question I was asking
myself is, why is it not 60, for instance. And it turns out
the cartoon version of that is that just
as when we get older it becomes more difficult for us
to react to things physically, as we get older we tend to get
cognitively lazier, as well. We tend to get
cognitively lazier. So I’m going to ask you
just to participate in this. Please don’t shout
out the answer. You guys just answer
it in your head or write it down
if you’re inclined. Here’s a very simple question
to try to illustrate this point. So here’s the set up. Jack is looking at Anne. Anne is looking at George. Jack is married. George is unmarried. Here’s the question
I’d like you to answer. Is a married person looking
at an unmarried person. All right? So Jack is married. He’s looking at Anne. Anne is looking at George. He’s unmarried. The question is,
is a married person looking at an unmarried person? So there are three
possible answers. A, yes. B, no. C, cannot be determined based
on the information you provided me. So let me ask this question. What is the first answer
that comes to your mind? AUDIENCE: C. MICHAEL MAUBOUSSIN: C, right? So by the way, how
many people C– first answer comes
to your mind, right? So of course, that’s
the wrong answer. But it is the first answer
that comes to your mind, right? Now let me see if I can
replicate the C mindset. See I have two
pairs to work with. There are no tricks. I have two pairs to work with. Jack and Anne– I know
Jack’s marital state. Right there, the light
bulb just went on. You know Jack’s marital state. You don’t know
Anne’s marital state. So I can’t tell
based on that pair. My second pair is
Anne and George. I know George’s marital
state, but I still don’t know Anne’s marital state. So I can’t tell. I’m basically stuck. Now the answer is A, yes. And the reason is Anne has only
two possible marital statuses. She’s either married
or she’s unmarried. And of course that’s exhaustive. So let’s assume for a
moment she’s unmarried. Jack is looking at Anne. Is a married person looking at–
then the answer would be yes. And now flip her status
and say she’s married. Anne is looking at George. Is a married person looking
at an unmarried person? The answer would be yes. Right. So no matter what I assume
about her marital status, the answer must be affirmative
for one of my pairs. Now what happens–
by the way, I rushed through that purposefully. Everyone answers C at first. And the only way to
get the correct answer is to check your work. Now if I had given
you all time, I assure you all would have
gotten that answer correct. But what happens
is, as you get older you’re less likely to check
your work– quite literally less likely to check your work. The first answer that
comes to your mind– And if you want to use
[INAUDIBLE] language, that question is meant to
evoke your system– one– rapid response. And unless you recruited
your system– two– to answer that
question properly, you’re going to come
to the wrong answer. That’s what happens. That’s the cartoon version of
what happens as we get older. We go with rules of
thumb and heuristics. Now let me turn to the topic
of luck, which is actually quite a bit richer than
the topic of skill, even. And when we talk
about luck there’s a very bright dividing
line between activities that are largely independent
activities and then path dependent activities–
where what happened before affects
what happens next. And what I’m going to
argue in these path dependent processes is
there is an inherent lack of predictability and
an inherent inequality. But let me start first
with a simpler case, and that is these
independent outcomes in luck. In the book we actually
used a baseball player named Adam Jones. His batting average in
2011– he’s an All Star now, by the way– his batting
average was .280. So we create a spinner
model– 28% of the time, hit. 72% of the time, out. We simulate 10,000 seasons. We compare it to the actual
Adam Jones statistics. And hitting is not
perfectly independent, but it’s close enough for all
intents and purposes, right? So that’s basically a close
to independent process– not perfect, but close enough. And there are other
things in business that we can model on
other things in life where that element
of luck is going to be pretty easy to
model and almost never perfect, but
typically pretty good. The more interesting case is
the case of path dependence. So this is my second quiz
question of the day and one that’s much easier
than the first one. Which is, what is the most
famous painting in the world? AUDIENCE: “Mona Lisa.” MICHAEL MAUBOUSSIN:
“Mona Lisa”, right? That’s a hint as everybody here
is looking at the painting. So there’s the “Mona Lisa.” By the way, how many
people have actually seen the “Mona
Lisa” in real life? Yes? And so what was your
impression of the “Mona Lisa”? AUDIENCE: Small. MICHAEL MAUBOUSSIN: Small. Anything else besides small? Any other descriptors? Inspiring? No, small. That’s usually what people say. Well, I have to say, my wife,
Michelle, is here today. And we last summer
took our whole family– we have five kids–
and mother-in-law. So all eight of
us went to Paris. And we, of course,
went to the Louvre. We got a guide
and we spent a day there so that our kids
could be disappointed as we were a generation before. So we’ve perpetuated
that tradition. Now you all know the basic
history of the “Mona Lisa.” It was painted by Leonardo
da Vinci, of course, in the early 1500s. It was in Italy for a number
of years in the early 1500s, but it’s been in France
continuously since about 1517. So let’s round it and say 500
years, basically, in France. Now the question is
an interesting one to pose is, why is the
“Mona Lisa”– by the way, “Mona Lisa” gets
85% market share when you ask that
question around the globe. 85% of people say– unprompted–
say the “Mona Lisa.” By the way, I was
curious about this because I was in
Singapore in Hong Kong, and I was giving a
version of the talk. And I asked people there–
just as I asked you– and the answer spilled out
just as rapidly and just as decisively. So that the truly
seems to be the case. So the question is,
why is the “Mona Lisa” the most famous
painting in the world? And if you were an art critic
you might say something like, well, there were only like 11
or 12 da Vincis in the world. But he depicted movement in
the background beautifully. He used oil paints. And you can always fall back on
the enigmatic smile of the Mona Lisa. But notice that
none of those things are truly unique to the
painting and many of them are actually quite
self-referential. So now what if I told you
that for most of “Mona Lisa’s” existence, it was not the most
famous painting in the world. And I want to give you two
bits of evidence on that. The first is in 1750. The painting was in Versailles–
about 20 kilometers away from Paris– at
the Royal Palace. And they said, we’re
going to take the top 110 pieces from Versailles
and bring them to Paris for an exhibition. So the top 100 pieces. “Mona Lisa” doesn’t
make the cut. This is 1750, so it’s about
a 250-year-old painting. It doesn’t make the cut. In 1797, it goes to the Louvre–
when it opened as a museum– where it remains
today, of course. And right around
1850– it’s 1849, specifically–
we’ll call it 1850. The curators of
the Louvre brought in experts to value
each of the paintings, in part for insurance purposes. And what you see is the “Mona
Lisa” was not even deemed to be the most valuable
da Vinci painting. And its value was
truly deemed to be a fraction of the most
famous Raphaels of the day. So what happened
between 1850 and 2014 where everybody says,
the “Mona Lisa” is the most famous
painting in the world? Now some of you may
have heard this story, but it’s a wonderful one. In the summer of 1911, an
Italian painter– patriotic, apparently– decides
that the “Mona Lisa” should be returned
back to Italy. So he sneaks into the
Louvre on a Sunday night– it was closed on Mondays at the
time– slept, woke up, and on Monday morning came in and
took the painting off its peg, put it in his jacket,
and walked out the door. Now when I asked you– for
those of you who saw the “Mona Lisa”– I said, what
was your description, what was your reaction? Most of you said small. And that was a huge asset
because it’s only 22 by 30. So it’s a fairly small painting. And that fit nicely
inside this guy’s coat as he sneaked out
with the painting. Now this guy– a lot of
theories about this guy. He was a little
bit of a crazy guy. But he basically went
back to his apartment, put it in a trunk, and
just slid it under his bed, and basically left it there. And no one heard anything of it. Now immediately–
by the way, this is a time when the Parisian
press was taking off, four million newspapers
in circulation– where’s our beloved “Mona
Lisa” becomes a national story in France and then ultimately
an international story. So this great da Vinci
painting is missing. Well, two years later, this
guy says, well, the gig’s up, decides to get rid
of this “Mona Lisa.” So he writes a letter to a
Florentine antique dealer. And he says, hey, my name–
I got the “Mona Lisa.” I want to bring
it back to Italy. Will you buy it off me? And the guy says, hey, if
it’s true, validate it, authenticate it, sure. So the guy goes
down to Florence. Sure enough, they
meet at a hotel, they authenticate
the “Mona Lisa.” They arrest the guy–
apparently shocked– as he’s dragged
off by the police. And they negotiate with
the French government. And then within
a couple of weeks it comes back to France
in early 1914– so 100 years almost,
exactly– to great fanfare. It gets its own room,
and 100,000 people come to see this. So it sort of kick starts the
popularity of this painting. And from there, of course, it
becomes part of pop culture. So on the left, you see the very
famous Marcel Duchamp parody where it’s got a little
mustache and a beard. And if you can read French,
read those letters quickly, it’s a bit of a scandalous
phrase about the “Mona Lisa.” And on the right, it comes
over to the US shores with Nat King Cole’s number
one song in the 1950s called “Mona Lisa,” wins an
Academy Award, and so forth. By the way, no one really knows
what the “Mona Lisa” is worth. The last time it came
to the United States was the early 1960s. It was brought over
by Jacqueline Kennedy. It was then put out for
appraisal for insurance. So there’s an interesting thing. There’s something called “The
Financial Analyst Journal” which writes about
different financial topics. And one of the things
they had in a recent issue was the art chain– so the art
prices over the last 100 years. So just for fun, I took
that number from 1962 and attached it
to the art chain. And you have to inflation
adjust and so forth. But if you work
out the math, sort of back of the
envelope, the best estimate for the value of the
“Mona Lisa” in 2014 dollars is 2.5 billion dollars. And that would be
roughly 10 times what’s paid for any painting. Now of course, it’s invaluable. Who knows what it’s worth
and certainly whether it would fetch that
in the real world. I have no idea. But just to give you
some sort of a sense of that exercise for fun. Now the problem with my
“Mona Lisa” story– hopefully it’s entertaining– but there’s
no way to prove it right and there’s no way
to prove it wrong. Right? What I’ve done is I’ve attached
a number of facts to a story– to a claim– that’s made
this painting so famous. But I want to share with you the
outcome of an experiment that lends credence to how
unpredictable hits really are. This is an experiment done by
three sociologists at Columbia University called Music Lab. The main researcher on this
is a guy named Duncan Watts. And what they did is ostensibly
about musical tastes. About 14,000 people
participated. So here’s the basic drill. You’re a college kids sitting
in your dorm, you get an email and it says, hey, we care about
what you think about music. Come into our site. You see 48 songs
by unknown bands. So they validate–
they made sure that no one had heard of
these songs or these bands. And they say, cruise
around, and listen to songs, and then rate them. Five stars, I love it. One star, I hate it. And if you really like it, you
can download it for your iPod. So that’s the basic set up–
love it, hate it, download it. Now unbeknownst
to the subjects is they came in–
20% came into what they call the
independent condition. The order of the
songs is randomized. Love it, hate it,
download it, but you can see what no one
else did before you. So effectively you’re in the
record store by yourself. The other 80%– of course,
you see that’s the control. The other 80% went into 10%
each into eight social worlds. You could think of
these, quite literally, as parallel universes. Identical initial set
up to the control, but now you could see what
other people did before you. You could see what
songs they downloaded and you could see what
songs they said they liked. In one extreme version
of the experiment they had a leader board–
so the most popular songs, the most downloaded
songs at the top. So the question is,
does the pattern of what people do before you
influence what you say you like and what you ultimately do? And the answer is fairly
unequivocally yes. Now I want to be really clear. The best songs in the
independent condition had a much better chance of
success in the social worlds. And the bad songs
really did well. So quality mattered. I want to be really
clear about that. But if you’re in the top
third, probably top half of the control independent
condition, pretty much anything can happen. Now there was one
song that I thought captured the
experiment brilliantly. It’s a song called “Lockdown”
by a band named 52metro. And it was number 26 in
the independent condition. So this is one out of 48. It’s number 26. It’s basically the
definition of average. Right in the middle. In one of the social worlds,
it was the number one hit. And in another one of the
social worlds, it was number 40. So the point is, if we roll
back time and replay time again, would we likely have the
same– would “Harry Potter” be “Harry Potter”? Would “Star Wars”
be “Star Wars”? Would YouTube be YouTube? The answer is highly
unlikely that the same things would succeed to the
same order of magnitude that they did succeed. It’s inherently very
difficult to predict winners. By the way, as a
side note– I’ll mention that I teach a course
at Columbia Business School. And this year– we always
bring in an executive guest. And this year we
brought in a guy who was an executive
at Time Warner. Now he runs Turner Broadcasting. And he must have
said to the students a half dozen times
in his session, we have no idea what’s
going to be a hit. You know, we obviously
work really hard at this. We try to think about formulas. We really have no idea
what’s going to work. So they’ve got a huge TV
studio, a huge film studio. They’re very incentivized
to figure it out, but it’s very difficult to do. Now the other thing I mentioned
is this inherent inequality. Economists call this
convexity, which says for a small
change in quality, there’s a huge change in payoff. The example– the
product– I want to use for this–
and by the way, you might think– there
are a lot of examples. So the x-axis here,
by the way, is just going to be skill, or quality,
whatever you want to say. The y-axis is some sort of
payoff– could be market share, could be profits,
whatever it is. One example– a trivial one–
is if you win the US Open tennis tournament, you make
twice as much money as the guy who
comes in second even though on any objective
scale of skill you’re sitting right
next to each other and vastly better than
the broader population. Now the product I want to
mention in this context is a product called
Stephen King. And you all know his
name, obviously– a remarkably successful
novelist for many, many years. His first very commercially
successful novel came out in 1973,
called “Carrie.” And from that point on,
he was off and running. He published about a book a year
with great commercial acclaim. But it turns out Stephen
King was actually writing more than
one novel a year. So he’s got these extra novels
stacking up on the side. So he turns to his
publishers and he says, hey, we’re doing great. How about we publish more
than one book a year? And the guy says,
Stephen, Stephen. One book per author per year. That’s how we do it in
the publishing world. So he says, you know what? I’m going to publish
anyway under a pseudonym. And the pseudonym
was Richard Bachman. Richard because there was a
Richard Stark novel sitting on his desk at that
time, and Bachman because Bachman-Turner Overdrive
was playing on the radio. Some of the older people
know that band and that song, “You Ain’t Seen Nothing Yet.” So Richard Bachman starts going
to the publishing businesses. And here’s what happens for
the next half dozen years. Stephen King– great
commercial success. Richard Bachman–
commercial flop. Stephen King– great
commercial success. Richard Bachman–
commercial flop. And this is going
on book after book. The last Richard Bachman book
came out in the early 1980s and it was called “Thinner.” There was a guy in a
Washington, DC bookstore who was reading this book. And he was like, this
guy Richard Bachman– he writes a lot
like Stephen King. He develops his characters
like Stephen King. You know, I think this
actually might be Stephen King. He so believes this he goes
down to the Library of Congress and looks up who’s got the
copyright to Richard Bachman. And it turns out– Stephen King. So he figures this out. And he calls him up
and says, Mr. King, I’ve realized that
you’re Richard Bachman. King comes clean
immediately, by the way, and says, hey, congratulations. You figured it out. I’ll give you the first
interview, and so forth. So the whole thing is over. Now what’s remarkable
and important for the context of the story
is that from the moment it was revealed that that book
was written not by Richard Bachman, but rather by Stephen
King, sales went up 10x. Same book. And by the way,
for those of you– this may have a
familiar ring to it. This actually
happened last summer. Anybody follow this whole
thing with JK Rowling? JK Rowling wrote a book under
the pseudonym Robert Galbraith. We actually have
the Amazon rankings. We kept track of them. It was actually very cool. So it’s puttering along. It’s doing fine. It’s puttering along but not
certainly at best seller. And then, this book is
actually written by JK Rowling, and it shoots to number one. So I can imagine
as a publisher that would be very hard to hold
back on that information because eventually you can
have a hit song, hit book. Now the last section I
want to finish up with and then we can have
a broader discussion is why we struggle
with understanding luck and then ultimately
what to do about that. Now if I tell you the
future has skill and luck, everybody here gets that. There’s no concern. You completely
understand that concept. But once an event
occurs, something that happens in all
of us– by the way, it happens effortlessly
and rapidly– which is your mind
creates a narrative to explain that
outcome and then you file that narrative
in your mind. And as you do that,
two things happen. One is called
hindsight bias, which is you start to believe that you
knew what was going to happen with a greater probability
than you actually did. And the second concept–
which is related– is called creeping
determinism, which is you start to believe
that what happened was the only thing that
could have happened. By the way, it’s very
natural because now you have the outcome and you have
all the facts that surround it. And your mind says, aha! Yes, it had to be that way. How could I have entertained
any other alternatives? Now you might say, why do
you have this weird picture? This guy on the left
is Michael Gazzaniga, who is a well-known
neuroscientist. He studied under Roger Sperry
and won the Nobel Prize. And he’s probably best known
for his work on so-called “split-brain patients.” Now these are people who
have debilitating epilepsy. They have failed all
their treatments. And as a last ditch
effort, they go in and they sever the
corpus callosum, the bundle of nerves between the
two hemispheres of the brain. By the way, the first thing
I should tell you is this is actually a very
beneficial surgery. People feel much
better after this. But the second
thing is, it sets up an incredibly interesting
experimental condition where experimenters can
feed information into the right
hemisphere and then ask the patient what’s going on. So the right hemisphere
has very little language. So for instance, they
might show a key to a door through the left eye to
the right hemisphere. And then they say
to the patient, point at the picture that
makes the most sense. And they point at the door,
no problem– easy to do. Then they ask them, why
are you pointing at a door? Now note, your left
hemisphere has got a problem because it can talk,
knows nothing of the key because it can’t
access the information in the right hemisphere. But it can see that
they’re pointing at a door. So what the left
hemisphere does– again, effortlessly and rapidly–
is it creates a gibberish story to explain what
you the person’s doing. And by the way, if you read this
split-brain patient research, it’s absolutely– it’s
all quite humorous because these people
basically go to no ends to explain what they’re doing. So pronounced is this
capability that neuroscientists call this the interpreter. The interpreter is part
of all of our brains, resides in our left hemisphere. If I give you an effect, you
will come up with a cause. And by the way,
it’s like an itch that demands to be scratched. I throw an effect at you,
you come up with a cause. Now here is the punchline
for our discussion. The interpreter knows nothing
of luck– never got the memo. So if it sees a
positive outcome, it assumes something
good has happened. If it sees a
negative outcome, it assumes something
bad has happened. And even check yourself. Even check yourself. If you’re looking at a
situation that’s probabilistic– and you know ahead of time the
range of possible outcomes– note that once the
outcome has been resolved, you automatically create a
story to explain why it happened and you somehow dismiss all
those other possibilities. So here’s the
thing I think is so fascinating about this
broader discussion is if the paradox of skill is
true– absolute skill never higher, relative
skill never narrower– means luck is determining
more outcomes. That’s colliding with a
mind that really struggles to understand the role of luck. So that’s one that hopefully is
something that can be helpful as you think through
these basic issues. Now what’s related to that
is our love of stories. There’s an interesting book
written a couple of years ago by Jonathan Gottschall
called “The Storytelling Animal.” One quote from this. It says, “A storytelling mind
is allergic to uncertainty, randomness, and coincidence. it is addicted to meaning. If the storytelling mind
cannot find meaningful patterns in the world, it will
try to impose them.” Now here’s the last line
that I think is the key. “In short, the storytelling mind
is a factory that churns out true stories when it can, but
will manufacture lies when it can’t.” And that, I think, is
the key concept for us. If we see outcomes that
are generated by luck, we will try to impose our own
stories to try to make sense, just as the split-brain
patients did before. Now you might say, well,
is that such a big deal? Let me give you one example from
the world of business that’s fascinating in this regard. There was a piece of research
done by some academics on all the books about
how to be a great company. “Good to Great,” “In Search of
Excellence,” “Built to Last.” You’ve heard of these. You may have a couple of
them on your bookshelf. In these books they tally
up about 700 companies that are mentioned
in these books. And the question these
researchers asked was, how many companies are
in these books by dint of luck and how many are there because
they’re truly skillful. So I’ll spare you
all the details. But what they basically do is
they have 50 years of data, they create a transition
matrix for return on capital, then they simulate
thousands of times. So it allows them
to create a template to sort so-called common cause
versus special cause variation. Basically, what does written
system tell me should happen? What’s special from the system? By the way, what they
found when they applied it to these companies is that
indeed, some companies truly do demonstrate skill, just
like in money management. Most of money
management performance can be explained by luck. But truly there
are some– you need differential skill to
explain actual outcomes. Same thing. But when they
applied this template to the 700 companies mentioned
in these books, what they found was that only 12%
of the companies could be confidently
coded as skillful. The other 80% they thought were
probably there because of luck. Now remember a while
ago I mentioned that when there’s a
lot of luck in activity you expect rapid
reversion to the mean? Go pick up your copy
of “Good to Great” and look at the Table of
Contents at the company’s mentioned there. And you’ll see that almost
from the moment the ink dried, the performance of
many of those companies rolled right over, just as
lots of luck would predict. So here’s an example of guys
selling tens of millions of copies of books purporting
to offer you an equation, or sets of rules, or attributes
that will allow you to succeed, when the foundation
of their research basically rests on
luck, for the most part. Now let me wrap up with
two final thoughts. How do we get better? How do we get better? So first let me talk
about the role of skill. If you’re on the far right
hand side of the continuum– the culture and skill side– the
answer is deliberate practice. By the way, I don’t know if
anybody saw the “New York Times” today– the science
section of the “New York Times.” There was an interesting
article by Benedict Carey about Zach Hambrick’s work
at Michigan State University on how much of true performance
is talent and how much of it is hard work and
deliberate practice. And by the way, I wrote
about this in my book. I thought the world
had tilted way too much toward this
Malcolm Gladwell cartoon version– 10,000 hours you
get great at anything– and too far away from
underlying talent. So I wrote a bit about this. But that’s the point
of the article. Hambrick’s work
shows quite clearly that there is clearly
differential talent and that talent plays a
major role in success. That said, if it’s a mostly
skill-based activity, deliberate practice is truly a
very key component to success. So this is 10,000 hours
right outside your ability where you’re getting
great feedback. Now here’s the key. In these types of activities,
the output of that participant is a very clear indicator
of his or her skill. If I want to know if
you’re a good tennis player or a good piano player–
and I know what I’m doing– I can listen to
you or watch you. And I can tell if
you’re good at it. And then I can give you
feedback to improve. As you slide over to the
luck side of the continuum, as you can imagine–
by the way, investing is a good example of this–
that connection between outcome and skill becomes broken, or
at least breaks down some. So for example, you
might go to Las Vegas and play blackjack and
play your cards foolishly and win for awhile or play
them intelligently and lose. Over the long haul that won’t
be true, but in the short run. So there’s no connection
between the quality of your playing and
your actual outcomes. As a consequence, as
you have more luck, you need to focus
more on process. And I won’t go on a
great deal on this. But the process–
I argue– should have three essential
components– one I’m going to call analytical,
which is finding edge. And then once you have
edge, figuring out how much to bet on that edge. Second I’m going to
called behavioral, which is understanding the
common biases that we all tend to fall for and trying to
weave into your process methods to manage or mitigate those. And the third I’m going
to call organizational. This most famously
goes as agency costs where agents and principles
might have different interests. But all of us work in
organizations, none of which are perfect. The question is,
is my organization helping or impeding the
quality of my decisions. So how you improve your skill–
I think– to some degree is a function of where you lie
on that luck/skill continuum and that dictates how
you think about that. Now whenever I tell
people I wrote a book about skill and luck
they go, oh, yeah luck. I know all about this. Yeah, luck is where
preparation meets opportunity. Or the harder I work,
the luckier I get. Yeah, see you guys said that. Hey, we don’t have
to go to this guy. I already know the whole story. It’s luck meets preparation. Now if you accept my
definition at the outset, none of those things
are actually luck. In other words, you
might think about it in a different way, which
is what is in your control and what is not in your control? If it’s in your
control, it’s going to be skill to some degree. Only if it’s out of your
control can it be luck. In some ways you can’t
improve your luck. All you can possibly do is
try to manage your luck. So let me give you two
very simple examples. On the left– it’d
be a simple case where if you’re in a
competitive interaction– let’s say a sports
match, for instance– and you’re the
stronger player, what you want to do if
you’re the stronger player is simplify the game. And by simplifying the game–
the dimensions of the game– that means your skill
almost assuredly will overwhelm that
of your competitor. If you are the underdog or
you’re the weaker player, what you want to do is complicate
the game, adding battlefields. Again, you’ll still
be the weaker player. But it dilutes the strength
of the stronger player. So some examples- well,
warfare is clear, right? If you’re the
weaker military, you don’t want to go toe to toe. You want to use
guerrilla tactics. In the world of
business, you don’t want to compete head to
head with an incumbent. You want to use disruptive
innovation– the Christensen type of stuff. In the world of sports,
you don’t want to– again, you use trick plays, and so
forth, different strategies. So those are the ways to try
to tilt the odds a little bit, especially if you’re
the weaker player, or tilt them in your favor if
you’re the stronger player. On the right is this
idea of little bets. This is something you
guys do a lot, probably, as an organization. Many years ago I
was an analyst that followed the food
companies– big packaged food companies– in
the United States. And they used to always
lament, we spent a lot of money on advertising and marketing. And we know that
half of it’s wasted, we just don’t know
which half it is. So we can’t stop doing it. And so what’s
happened, of course, is this concept of A/B testing. So now we can go along and
test two different things and figure out– for
example, an internet retailer might say here are two different
websites that people land on. Which one will sell more stuff? And so we constantly
test and improve. In a sense you’re not
really managing luck. What you’re doing is clearing
the clouds of uncertainty and focusing on causality
much more effectively. But I sort of threw
that in that bin as just dismissing
luck to some degree. With that I’ll stop. There are three
things I mime hope you’ll take away from this. The first is defining skill and
luck as being quite important. So just those definitions
are essential. But in the first section,
a particular thing I want you to walk
away with is this idea of the paradox of skill. That even as skill
improves absolutely, it’s often in many domains
getting narrower relatively, which is leaving more to luck. The second thing I want to leave
you with is the shape of luck, and especially these
path dependent processes. And in particular,
you really have to be very circumspect
about trying to predict the outcome of
these path dependent processes. And again– as you know–
also technology based. A lot of things do depend on
things like network facts, and so forth. And those are inherently
very difficult areas to predict winners. And then the third area
is what to do about this. And the main thing I
want to leave you with is this idea of the
interpreter, which is we have a part of our brain
that is constantly seeking causes for every
effect that it sees, and for you to be very
mindful and keep that in check as you’re considering
the outcomes around you. So with that let
me stop, and I’d be very happy to
entertain– [INAUDIBLE] is coming with a
white microphone. I’d be happy to entertain
any questions or comments. We have a few minutes? Yeah, we have a
few minutes, right? MALE SPEAKER: Yes, we’ll
open it up for questions. And thanks for the talk. MICHAEL MAUBOUSSIN: My pleasure. [APPLAUSE] Oh, thank you. [APPLAUSE] AUDIENCE: So just looking at
the financial community as kind of average Americans,
we don’t see a lot of certainty
or predictability. It looks to us like
there’s a lot of luck. So is there a lack
of skill there? Is the skill not growing? What’s going on? MICHAEL MAUBOUSSIN: Super
interesting question. I think the answer is the notion
that it’s all a huge amount luck– if I put it
back on the continuum, I would put investing way
over on the luck side. But again, I don’t think
it’s accurate to say it’s because of luck. So let’s go back to
the paradox of skill. I think that the
way to say this is that most professional
institutional investors are very skillful. They’ve gone to great schools. They’ve got incredible
information, and computing power, and information
access, and so forth. But the problem is they’re
all doing the same thing. So their skill is
extremely uniform. So I think that’s
a classic example of the paradox of skill. If I put you back in the 1960s
with the technology you have access to as a
financial money manager, you would run circles
around everybody. But now everybody’s
got the same thing. So I think it leaves
much more to luck. So that’s why it
appears to be luck– I think– for the most part. So that’s it. Now I will mention one
thing as a measure of that. I showed you the
marathon runners and how the difference
between the first and 20th guy has been declining over time. We actually did a very similar
thing for money managers. So what we do is we
take the standard– it’s basically the standard
deviation of excess returns– so it’s what does the
bell-shaped distribution of returns look like. And what’s happened
over the last 50 years is that it’s got skinnier. So the difference between
the very best and the average is less today than it was
a generation or two ago. So again, it is very
random, but it’s not because there is
not a lot of skill. It’s actually the opposite. The skill is very
high but very uniform, is the way to think about that. And by the way, that’s true
for athletics, too– sports. You might say that even the
World Cup– these matches are– you can see the
probabilities are not that high one way or the other. That certainly would
be no reflection of the skill of the players,
which is extraordinary. And certainly put any
one of those teams on the pitch versus
teams 20 years ago, and they’d run
circles around them. It’s that the skills
are evenly matched. And as a consequence, luck
determines the outcomes. Great question. Very important one. AUDIENCE: You mentioned
about the relationship between age and
cognitive ability. And you said that you picked
around 40, or 50, or something like that. Warren Buffett is over 80. Do you think he’s not
as good as he was? This is the first question. That is a fun question. The second one–
as you said, it’s more and more
difficult to find out who is a good value investor. But if you had to find one,
what would be a way to do it? MICHAEL MAUBOUSSIN: Awesome. Two excellent questions. Warren Buffett is,
I think– look, I think that he’s been pretty
much extraordinary at every age that he’s been. So that’s one thing. I think he’s doing less
traditional money management– as we would say in running
a portfolio of stocks. He’s obviously now
much more capital allocating for
Berkshire Hathaway. So I think that the game
has changed a little bit. But I do suspect–
I hate to say this– but I do suspect that he
might struggle competing with a 40-year-old version
of himself, I would guess. AUDIENCE: He also
has much more money. MICHAEL MAUBOUSSIN: And he
has much more money to deploy. Yeah, that’s right. That also impedes
his performance. The second part– your
second question was? AUDIENCE: How to find– MICHAEL MAUBOUSSIN: Oh, yeah. Value investors. So look, I think– I’ll mention
a couple things on this. First of all, I think
that the characteristic of many great value investors–
let me say it this way. There’s a quote which
I love by a guy named Seth Klarman at Baupost, who
is one of the great value investors. And he says, value
investing is at its core the marriage between
a contrarian streak and a calculator. So what does he mean? The contrarian streak, I
think, is the first element, which is the ability to go
against what everybody else is doing. So Buffett’s got
this great line. It is, be fearful when others
are greedy and greedy when others are fearful. Let me just say this–
investing is inherently a very social exercise. Very social. And it takes a
very unusual person who’s willing to do
the opposite of what everybody else is doing. So that’s the first element. The problem is
being a contrarian for the sake of being
a contrarian is not a very good idea because
sometimes the consensus is right. In other words, if the movie
house is on fire, by all means run out the door. Don’t– So the calculator is
the second component, which says as a consequence of
everybody taking one view, that means a gap between
price and value opens up. And so that becomes an
opportunity to invest. So the great value
investors I think are people with certain
characteristics– personality characteristics. In part, they are
people who tend not to care about what
other people think about what they think. And that’s very rare in
the regular population. Most people are very
sensitive to what other people think about them. And then often
the organization– this question about Buffett
is a great question. Berkshire Hathaway is set
up as an organization that’s very conducive to
quality decision making. He’s not influenced by a
lot of other pressures. Day-to-day business
pressures, for instance. So those would be
things I would look for. The main thing is this
ability to sort of operate independently of what
other people think. AUDIENCE: On the
[INAUDIBLE] a lot of people get the similar or the same
and they get a similar view. Instead of from one person
[INAUDIBLE] of luck. Or if we look from outside and
see this is a collective view. So would that actually the
collective view will actually be determining this
instead of luck? MICHAEL MAUBOUSSIN:
Yeah, exactly. Super interesting question. So let me make sure that
I’m on the same page as you. But I would just
say that we talk a lot about this in the
context of market efficiency. And it relates even
to this comment about being a contrarian. Well, we can use more formal
language like complex systems. But let’s just use a
more simple language like “The Wisdom of Crowds.” So when are crowds wise
and when are crowds mad? Crowds tend to be
[INAUDIBLE] when three conditions are in place. And Surowiecki wrote
about this in his book. One is diversity of
the underlying agents. So we need diversity
of points of view. The second is a properly
functioning aggregation mechanism. So you all have information
but I can bring it together in one place. And the third is incentives,
which is basically rewards for being right and
penalties for being wrong. So when those three
things are in place, I can demonstrate
that I get actually very efficient economic
results and similar to what the textbooks predict. But what happens is when one
or more of those conditions are violated, then I get
these inefficiencies. And I think that’s
why I mentioned those contrarian streaks. So when we all believe the
same thing, we lose diversity and we get the
[INAUDIBLE] of crowds flips over to the
madness of crowds. You can read the annals. I mean, there are famous– South
Sea Bubble, and Tulip Mania, I mean there are famous– the
internet now, maybe housing. I mean there are some– this
doesn’t happen all that much. But it does happen and it can
be very epic in its proportion. So that is a thing I
would say is, can we look for these so-called
diversity breakdowns. And for whatever
reason, everyone’s taking one side of the
trade over another. It could be mostly
psychological factors. But it could also be
technical factors. I own a portfolio that
has leverage against it. And now I’m getting
a margin call. So I have to sell things
because I have to. So it’s not because I
want to, but I have to. So those things can
contribute, as well. So that’s a very
excellent question. It’s a very rich question. But that’s the idea. So then you say, well, how do
skill and luck fit into that? I’m not so sure. That’s where the great
investors take advantage. And that’s where
it’s the contrarian streak plus the calculators
taking advantage of these diversity
breakdowns in effect. AUDIENCE: My name is Bruno. And my question is, I
think it’s safe to say that most of us in this room
are very– try to be outliers, try to be the best. And in that sense, it seems to
me that the take away from this is that I should try to be in
the group of the people who are the best. But like you said, that group
is getting bigger and bigger, and the standard deviation is
getting smaller and smaller. And after that I kind of
just have to hope for luck. And if that’s the case, how do
you deal with this powerless feeling of I did my best, but
it’s still not in my control? MICHAEL MAUBOUSSIN:
I mean, I don’t know if you’d say
it’s powerless. I think in some ways
it’s liberating. You do the best you can. So I would say it this way. I would go backwards
and say, if you look at the most successful
people in the world– however you want to measure
that– I think almost without fail you’d find
that they were incredibly lucky at some point. And by the way, in
chapter one of the book, I start with a story of–
you guys all know this. It’s a famous
story in technology of Bill Gates and Gary Kildall. Do you guys know the
story about Gary Kildall? Does everybody know this? You don’t know
the story of Gary. You know it, in the back, right? So if I get it wrong,
straighten me out on this. IBM launches a PC in
1980– started the project in 1980, ’81. And they need an
operating system. So it turns out
the chairman of IBM and Bill Gates’ mother are
on the same United Way board. So that’s a first start. She says, oh, my little Bill. He knows something
about those computers. You should go see him. The IBM teams flies to Microsoft
in Seattle to see Bill Gates. By the way, they were
building cards for Apple IIs. It had nothing to do
with operating systems. And Bill Gates– they say, do
you have an operating system? We’re building this PC. It’s all secret. Sign all these papers. And Bill Gates says,
no, I don’t do that. This guy Gary Kildall
down at California– he’s the guy that does this. Gary Kildall’s company had
80% market share for software for Intel chips. 80% market share in 1980. Dominant. And he’s considered the greatest
programmer of the 1970s. So they go down to Gary. This part of the story, no one
really knows what happened. But basically he
sort of blew off IBM. Like he either didn’t show up
or didn’t take them seriously. They kind of came to a
deal, but didn’t really. Anyway, the IBM guys
were very dissatisfied. And he refused to
sign all their papers. So they go back up to Microsoft. And they go, Bill, this
guy didn’t really help us. And Bill goes, I’ll hook you up. So he goes across town and
he buys basically a knockoff of Kildall’s product. Knock off for $50,000. And he called it MS
DOS, and he said, I’m going to sell
it to you, IBM. But I’m not going to sell it–
I’m going to license it to you. That was a genius. And from there– right now. So the PC is now
getting ready to launch. When you bought
PCs back in 1981, you actually bought
the IBM machine and then you bought
the OS separately. It wasn’t loaded on. So it turns out– so Kildall
gets wind of this– little Bill Gates is going to be selling
some operating system. So he calls up IBM and he
says, I thought we had a deal. They go, yeah, yeah,
OK, we have a deal. But it turns out, IBM
retained the rights to price the products. So when the first PC went
up for sale, MS DOS was $60. And Gary Kildall’s
product was $240. So which one do you
think they bought? And that’s the end of the story. So Gary Kildall ended
up– and the story, it’s an amazing story. By the way, Gary Kildall
was from Seattle. And it turns out, the story
ends that he went into a bar– a biker bar– in
Monterey, California. No one really knows
what happened. But he got drunk, got
into either a fight or just fell down, hit his head
on the bar, and died at age 53. And so it’s very–
you could say, could I have twisted that
story just a little bit and Gary Kildall
would be Bill Gates? Bill Gates would be still
super successful, I’m sure. But he wouldn’t be
Bill Gates, right? So those are the examples. That’s why I say, is
that liberating or not? I don’t know if
that is or isn’t. But my point is you
should do everything in your effort to succeed. But there’s no one that
is a massive outlier that isn’t lucky. It’s almost, by
definition, can’t be. If there’s luck in
what you’re doing, it’s almost– now
probably not so much true in things like athletics,
or music, or what have you, or if there’s any
quantifiable way to measure people’s performance. Less true. But when there are these
social processes kicking in. You know, JK Rowling. You guys know this story. Harry Potter was turned down
by nine or 10 publishers before someone grudgingly
willing to publish it. It’s the greatest selling
novels of all time, right? I feel liberated
by it, actually. I think it’s the other way. I feel the opposite,
which is if I do– and this is what I say
to my kids all the time. If you work as hard as
you can, the outcome doesn’t really matter. You’ve done all you can do. The fates. The fates of the gods. AUDIENCE: My name is Timmy,
and I had a quick question. We talked about
business and sports, but I just want to talk a
little bit about leadership. And as someone who’s high up in
a company like Credit Suisse, I’m sure when you guys
recruit talent and things like that you’re looking at
people who maybe have been successful at leadership
maybe through their own skill. But are there
times– where you’re talking about the
interpreter– that we always want to say it’s
skill, that you’re seeing something that
maybe isn’t pure skill and maybe is luck? And how do you go about
hiring them or thinking about are they going to– MICHAEL MAUBOUSSIN:
Well, I don’t think you want to hire
people who have been lucky, but I think that that’s
actually a huge business. This is a really
difficult question. And I have some thoughts
on leadership a little bit separate. But let me just
mention a couple things on being careful about this. There’s a guy at Harvard
Business named Boris Groysberg who wrote a book
called “Chasing Stars.” And what he actually
studied was this idea of hiring stars from
other organizations to join your organization. And it turns out
that this turns out to be a very poor,
poor practice. Most people don’t– their
skills don’t translate from one organization to another
very effectively. It’s true less in athletics,
but even true in athletic. It’s certainly true in
the world of business. So for example, they
followed 22 GE executives, who are obviously trained to
be the top in great leadership. And they found
that when they went to other organizations that
were very similar to GE, they tended to do well. But if they went to other
organizations that were not the same as GE, they tended
to flounder quite a bit. So we see this a lot in
the world of business. It’s hard for me–
what is leadership? To me the kinds of things
that are important in a leader is someone who– and there
are different ways to lead. But a lot of it is I do think
is actually quite intellectual. I think for me it’s thinking
about things strategically. It’s being fact-based,
data-based, and setting a tone or
direction for an organization that people are
enthusiastic to follow. So those are the
kinds of things. But it’s very hard. I think it’s very
hard to pin it. And the last thing I’ll
say is– and it goes back to the interpreter, which
you correctly drew out– By the way, there’s
a great book which I’d recommend to
everybody called “The Halo Effect” by a guy
named Phil Rosenzweig. It’s a very short
book– 175 pages. But an incredibly
important set of lessons. And what he says is basically,
when things are going well, the CEO walks on water. And when things are going
badly– the same CEO, by the way– may have no
idea what they’re doing. I think we spin stories
to fit the narrative, or to fit the facts, that
are not always well-placed. AUDIENCE: I found
it very interesting when you mentioned
that the talent gap between the person who is
the first and the person who is at 20th is narrowing. And in my head I was thinking,
what about the payoff gap between the two. If you compared the top guy
and the 20th guy 50 years ago versus now, now the gap is
huge, although the talent gap has narrowed down. It’s sort of a paradox. And if seemed frustrating. MICHAEL MAUBOUSSIN:
I’ve actually thought about
writing about this. I think it’s an absolutely
fascinating thing. And I think it’s very
true, by the way. So I think that there is a huge,
huge payoff– an increasing payoff– to cognitive surplus. And I think technology
has accelerated that to a large degree. So in the book I mention this. There’s a very famous paper that
was written more than 30 years ago by Sherwin Rosen called
“The Economics of Superstars.” And what he argued
in that paper was that increasingly people who
are just a little bit better than others are getting
disproportionate payoffs. And it’s more like
the tennis tournament. You’re getting
twice as much money. You guys all know
these data, right? If you look at people with
very advanced degrees– master’s versus college
degrees versus people with no degrees– the payoffs. But it’s even within
those subsectors there are still very, very
substantial differentials in pay. I think there’s just going
to be– and by the way, I don’t know how we stop this. I’ve been very influenced
by these books recently. “Average Is Over,” and
“The Second Machine Age” by a couple of– I
think they came here, actually– MIT economists. And those stories are– there
are positives that technology is helping everybody’s lives. I think in some ways
it’s really great. So we all have access to
education and entertainment we may not have had cost
effectively in times past. But it does feel like the
payoff to cognitive abilities is really getting very skewed. I don’t know what
stops or how it stops. Well, with that, thank
you so much for your time and attention. And have a great day. Appreciate it. [APPLAUSE]

Leave a Reply

Your email address will not be published. Required fields are marked *