WEBVTT

00:00.110 --> 00:03.790
PROFESSOR ROBERT SHILLER:
OK, good morning.

00:03.790 --> 00:08.950
So, I wanted to talk today about
Behavioral Finance or

00:08.950 --> 00:12.410
about Psychology and Finance.

00:12.410 --> 00:17.250
This is a longstanding
interest of mine.

00:17.250 --> 00:21.190
I've been involved with
it for over 20 years.

00:21.190 --> 00:27.080
It's not really emphasized in
your textbook, Fabozzi and his

00:27.080 --> 00:32.000
co-authors talk about a lot of
things in the financial world,

00:32.000 --> 00:34.750
but not about the underlying
human behavior.

00:37.280 --> 00:41.670
Behavioral Finance, or
Behavioral Economics more

00:41.670 --> 00:45.770
broadly, is a kind of revolution
that has occurred

00:45.770 --> 00:53.260
in finance and economics over
the last 20 or 30 years.

00:53.260 --> 00:56.740
And it remains somewhat
controversial.

00:56.740 --> 01:00.760
I don't quite fully understand
why it is that people polarize

01:00.760 --> 01:03.630
as much as they do, but some
people don't like this.

01:07.980 --> 01:11.100
We're coming along to be
the majority, I think.

01:11.100 --> 01:13.800
People are now regarding
Behavioral Finance as an

01:13.800 --> 01:15.350
important element of finance.

01:18.470 --> 01:24.240
But the real problem is that
people are complex and our

01:24.240 --> 01:28.750
financial institutions, as I've
emphasized, are designed

01:28.750 --> 01:32.490
for real people and their
functioning depends on the

01:32.490 --> 01:37.880
behavior of real people.

01:37.880 --> 01:40.520
And it's not as simple.

01:40.520 --> 01:45.630
You know, another revolution
that's occurring parallel, of

01:45.630 --> 01:50.430
bigger significance, is a
revolution in neuroscience

01:50.430 --> 01:52.420
about the human brain.

01:52.420 --> 01:57.760
And the human brain is a
very complicated organ.

02:02.790 --> 02:05.650
Economists have liked to
invoke the principle of

02:05.650 --> 02:10.700
rationality as an underlying
component of their theory, and

02:10.700 --> 02:15.900
that has been useful, but it's
of limited use, because people

02:15.900 --> 02:17.000
aren't rational.

02:17.000 --> 02:18.810
They are often rational,
they're

02:18.810 --> 02:21.240
not completely rational.

02:21.240 --> 02:24.210
And very often, people
behave stupidly.

02:24.210 --> 02:26.190
I'll put it that way.

02:26.190 --> 02:31.270
That includes everyone,
including me, because we're

02:31.270 --> 02:32.520
human and we have limits.

02:38.770 --> 02:42.800
One thing about the human
species is that we are aware

02:42.800 --> 02:45.560
of other people's weaknesses
and have an

02:45.560 --> 02:48.330
impulse to exploit them.

02:48.330 --> 02:52.090
So when you see other people
behaving stupidly, sometimes

02:52.090 --> 02:55.970
you think, maybe I can turn
this to my advantage.

02:55.970 --> 02:57.490
OK?

02:57.490 --> 03:03.490
And that becomes a problem.

03:03.490 --> 03:08.130
The history of humankind is a
history of exploitation of one

03:08.130 --> 03:11.380
person by another.

03:11.380 --> 03:13.220
Not entirely, but I'm saying
it has that as

03:13.220 --> 03:15.960
an important element.

03:15.960 --> 03:18.270
So, I'm going to talk about
these human failings.

03:18.270 --> 03:21.650
It's not to say that people are
stupid, I'm just saying

03:21.650 --> 03:24.130
they're people, and we're
all imperfect.

03:24.130 --> 03:28.120
We're smart in some dimensions
and we can be very smart, but

03:28.120 --> 03:32.340
we can also make important
mistakes.

03:32.340 --> 03:38.450
But before I start, I wanted
to try to put this into a

03:38.450 --> 03:40.150
perspective.

03:40.150 --> 03:45.800
Maybe, I'll return to this at
the end of the lecture, but I

03:45.800 --> 03:47.820
wanted to start out
on an upbeat note.

03:47.820 --> 03:50.640
I'm going to talk about all
kinds of human errors, but I

03:50.640 --> 03:57.310
wanted to start on an upbeat
note that the business world

03:57.310 --> 04:01.800
generally doesn't exploit
people terribly.

04:01.800 --> 04:11.880
I believe that very
characteristically successful

04:11.880 --> 04:18.000
businesses in finance and
elsewhere consider their

04:18.000 --> 04:23.020
long-term advantage and the
reputation they have. So,

04:23.020 --> 04:26.630
doing something that is
blatantly exploitative of

04:26.630 --> 04:30.240
human weaknesses will
work against

04:30.240 --> 04:31.660
their long-term advantage.

04:36.400 --> 04:39.570
You'll see a lot of human
failings, but we don't see

04:39.570 --> 04:45.210
people cashing in on them
as often as you'd think.

04:45.210 --> 04:48.900
And beyond that, I want to
emphasize also that another

04:48.900 --> 04:53.690
aspect of human behavior
is morality.

04:53.690 --> 04:58.550
Evolutionary biologists think
that this evolved along with

04:58.550 --> 05:02.840
our other traits, that we have
an impulse to be moral.

05:02.840 --> 05:06.920
And so, in the long run, you
might not really gain so much

05:06.920 --> 05:11.010
satisfaction from exploiting
other people's mistakes.

05:11.010 --> 05:17.200
And so, you don't necessarily
do that.

05:17.200 --> 05:21.130
So, that's why we have a lot of
weaknesses outlined and we

05:21.130 --> 05:25.960
won't see significant or serious
exploitation of them

05:25.960 --> 05:27.870
as characteristic.

05:27.870 --> 05:31.660
Now as you know, I have chapters
from my forthcoming

05:31.660 --> 05:34.550
book assigned for this course.

05:34.550 --> 05:39.170
And I looked back on what I put
up for you to read and I

05:39.170 --> 05:42.070
keep thinking, gee, this
really wasn't ready.

05:42.070 --> 05:46.820
So, I had a chapter for this
section of the reading list

05:46.820 --> 05:49.420
about Behavioral Finance.

05:49.420 --> 05:52.050
And I thought, I didn't
really get it right.

05:52.050 --> 05:54.870
I know what I was trying to
say, but maybe I should --

05:54.870 --> 05:58.950
what I start out in that chapter
is talking about Adam

05:58.950 --> 06:05.660
Smith and his book The Theory
of Moral Sentiments.

06:05.660 --> 06:12.390
Now just to remind you, Adam
Smith was a professor in

06:12.390 --> 06:17.530
Scotland in 1759.

06:17.530 --> 06:20.470
He was a professor of moral
philosophy, because there were

06:20.470 --> 06:24.800
no professors of economics
in those days.

06:24.800 --> 06:27.390
And he wrote in 1759 --

06:27.390 --> 06:29.400
Maybe I should write some
of this on the board. --

06:36.530 --> 06:39.870
He's probably the most important
figure in the

06:39.870 --> 06:43.060
history of economic thought.

06:43.060 --> 06:50.310
So, in 1759 he wrote his The
Theory of Moral Sentiments.

06:50.310 --> 06:56.480
And in 1776, he wrote the
more famous book,

06:56.480 --> 06:57.730
The Wealth of Nations.

07:01.760 --> 07:04.250
So, this is The Theory
of Moral Sentiment.

07:06.840 --> 07:14.610
The Wealth of Nations is
considered the first real

07:14.610 --> 07:17.500
treatise on economics and
it's a wonderful book.

07:17.500 --> 07:19.180
And it's still very
readable today.

07:22.280 --> 07:24.390
His The Theory of Moral
Sentiments is

07:24.390 --> 07:26.340
not so widely read.

07:26.340 --> 07:30.910
But it's not really economics,
it's a book about psychology

07:30.910 --> 07:33.330
and morality.

07:33.330 --> 07:39.130
I find it very good, even
250 years later.

07:39.130 --> 07:42.410
He went through many editions on
this book, because maybe he

07:42.410 --> 07:44.570
thought it was his most
important work.

07:44.570 --> 07:49.090
But the book starts out about
selfishness and altruism.

07:49.090 --> 07:53.450
And the real question, which he
thinks defines economics,

07:53.450 --> 07:57.000
is, are people really
completely selfish?

07:57.000 --> 08:00.990
Sometimes it seems that way,
that their presumed

08:00.990 --> 08:03.010
benevolence is just an artifice

08:03.010 --> 08:04.260
for their own benefit.

08:07.800 --> 08:10.410
But he wonders, how does an
economy work if everyone is

08:10.410 --> 08:11.540
totally selfish?

08:11.540 --> 08:14.470
And he ends up concluding
that they're not.

08:14.470 --> 08:15.590
I thought it was very
interesting,

08:15.590 --> 08:17.090
the way he put it.

08:17.085 --> 08:20.065
The thing he emphasized right
at the beginning of the book

08:20.070 --> 08:25.740
is that people inherently
love praise.

08:25.740 --> 08:29.840
We crave the approval
of other people.

08:29.840 --> 08:35.710
And so, praise is a fundamental
human desire.

08:35.710 --> 08:38.750
But then he reflected on it and
he said, do people really

08:38.750 --> 08:45.100
want praise itself or is it
something else that they want?

08:45.100 --> 08:45.880
Well, think of it this way.

08:45.880 --> 08:49.560
Suppose people made a big
mistake and thought that you

08:49.560 --> 08:52.650
had accomplished something,
but there was a mix up.

08:52.650 --> 08:54.220
You know, it was really
somebody else

08:54.220 --> 08:55.470
who did it, not you.

08:55.470 --> 08:57.800
And it's just a complete
mistake.

08:57.800 --> 09:00.660
You had nothing to do with it,
but you find lots of people

09:00.660 --> 09:02.150
praising you.

09:02.150 --> 09:05.640
Would that really
be pleasurable?

09:05.640 --> 09:08.730
And suppose you even know that
they'll never find out that I

09:08.730 --> 09:10.100
didn't do it.

09:10.100 --> 09:12.510
Well, Smith said, it
probably isn't.

09:12.510 --> 09:13.720
Think about it.

09:13.720 --> 09:16.990
You internally are thinking, I'm
getting all this praise,

09:16.990 --> 09:19.870
but I know I don't deserve it.

09:19.870 --> 09:22.060
So, I don't enjoy it.

09:22.060 --> 09:25.170
And then he went on to say that,
especially among people

09:25.170 --> 09:28.720
who are more mature, he says
more mature people --

09:28.720 --> 09:30.830
not everyone makes this step.

09:30.830 --> 09:35.550
But he says, adults, normal,
mature adults, make a

09:35.550 --> 09:39.960
transition from a desire for
praise to a desire for

09:39.960 --> 09:42.480
praise-worthiness.

09:42.480 --> 09:46.810
I want to know that I am the
kind of person who will be

09:46.810 --> 09:50.480
praised and I don't need
to get the praise.

09:50.480 --> 09:53.850
And he said, it's that tendency
ultimately, which

09:53.850 --> 09:56.860
makes an economy work,
that people don't

09:56.860 --> 10:00.080
care just about praise.

10:00.080 --> 10:02.650
He gives an example
of mathematicians.

10:02.650 --> 10:05.820
And he said he's known many
mathematicians in his life and

10:05.820 --> 10:08.390
he finds that they're
almost all obscure.

10:08.390 --> 10:10.760
The public doesn't know
about mathematicians.

10:10.760 --> 10:14.060
They couldn't explain to the
public what they do.

10:14.060 --> 10:18.730
And they don't seem to care at
all, because they know the

10:18.730 --> 10:21.310
public doesn't appreciate
mathematics.

10:21.310 --> 10:25.080
And so, there's a few
mathematician friends who

10:25.080 --> 10:28.520
understand what they do
and may praise them.

10:28.520 --> 10:31.190
But ultimately, a mathematician
can sometimes do

10:31.190 --> 10:35.730
the work completely unknown.

10:35.730 --> 10:43.850
And it's the praise-worthiness
that drives these people.

10:43.850 --> 10:46.850
You may think I'm being too
idealistic, when I say this,

10:46.850 --> 10:50.750
but I think that the finance
profession --

10:50.750 --> 10:53.860
this is what I was trying
to say in that chapter.

10:53.860 --> 10:58.940
That the finance profession,
like other mature professions,

10:58.940 --> 11:01.140
is really dominated --

11:01.140 --> 11:02.840
although there's a lot of funny
things that happen.

11:02.840 --> 11:07.610
It is really dominated by people
who have reached this

11:07.610 --> 11:10.040
desire for praise-worthiness.

11:10.040 --> 11:13.150
And so, you're not going to
exploit people extravagantly.

11:13.150 --> 11:15.100
Just because, why
would I do that?

11:15.100 --> 11:16.010
This is not a good thing.

11:16.010 --> 11:19.420
I wouldn't feel good about it.

11:19.420 --> 11:20.810
Well, some people will.

11:20.810 --> 11:25.470
Now, I wanted to also mention,
not everyone reaches this

11:25.470 --> 11:31.570
mature state that Adam
Smith describes.

11:31.570 --> 11:34.310
And that's one of the
complexities of human society.

11:34.310 --> 11:38.470
And I think that the finance
profession has a problem with

11:38.470 --> 11:39.910
other kinds of people.

11:39.910 --> 11:42.620
Now there's a whole branch of
psychology called Personality

11:42.620 --> 11:46.820
Psychology that categorizes
people by their personality.

11:46.820 --> 11:48.890
And we're not all the same.

11:48.890 --> 11:54.890
And in our society, we have
many different kinds of

11:54.890 --> 11:57.180
personalities.

11:57.180 --> 12:01.940
A successful society promotes
people up who have the

12:01.940 --> 12:03.930
praise-worthiness desire.

12:03.930 --> 12:06.440
We try to recognize them and
we try to put people of

12:06.440 --> 12:10.700
character into important
positions, with

12:10.700 --> 12:12.320
not complete success.

12:12.320 --> 12:14.380
But I wanted to just briefly
talk about --

12:14.380 --> 12:17.650
this is a lecture
on psychology.

12:17.650 --> 12:22.050
I wanted to talk about other
personality types.

12:22.050 --> 12:26.620
And I was going to use a book
called The Diagnostic and

12:26.620 --> 12:32.480
Statistical Manual Edition IV
published by the American

12:32.480 --> 12:34.620
Psychiatric Association.

12:34.620 --> 12:39.310
They're coming out with a
fifth edition in 2014.

12:39.310 --> 12:43.190
DSM-IV is kind of a household
word around my house, because

12:43.190 --> 12:48.830
my wife is a psychologist.
DSM-IV is actually

12:48.830 --> 12:51.970
controversial among
psychiatrists, because it's

12:51.970 --> 12:54.010
too cut and dry for
some of them.

12:54.010 --> 12:58.720
What it tries to do is, identify
mental illnesses and

12:58.720 --> 13:03.190
personality types in a
quantifiable, reproducible

13:03.190 --> 13:07.950
way, so that we can define who
has this mental illness or who

13:07.950 --> 13:11.940
has this personality type.

13:11.940 --> 13:15.240
And so, it gives you checklists
and it says, the

13:15.240 --> 13:20.390
patient must have exhibited at
least three of the following

13:20.390 --> 13:21.890
five behaviors.

13:21.890 --> 13:25.100
And then, there will be another
checklist. And so, you

13:25.100 --> 13:28.430
keep score and you can actually

13:28.430 --> 13:30.900
diagnosis personality disorders.

13:30.900 --> 13:32.100
I'm just going to mention
one of the

13:32.100 --> 13:35.240
many personality disorders.

13:35.240 --> 13:38.890
One of them is called
APD, called

13:38.890 --> 13:42.830
Antisocial Personality Disorder.

13:42.830 --> 13:44.310
And so they have checklists,
but just to

13:44.310 --> 13:45.860
give you a sense --

13:45.860 --> 13:52.070
oh, and the Antisocial
Personality Disorder is called

13:52.070 --> 13:58.080
psychopathy, or one kind
of APD is psychopathy.

13:58.080 --> 14:01.590
Another one is called
sociopathy and --

14:01.590 --> 14:02.190
I don't know.

14:02.190 --> 14:04.300
There's a huge literature
on these.

14:04.300 --> 14:11.980
But according to DSM-IV, 3% of
the male population in the

14:11.980 --> 14:20.230
world is APD, and 1% of
the female population.

14:20.230 --> 14:24.800
A simple definition
for APD is a jerk.

14:28.200 --> 14:32.380
There are more male jerks than
female jerks, apparently,

14:32.380 --> 14:34.480
according to their --

14:34.480 --> 14:36.540
this is all quantifiable
and done.

14:36.540 --> 14:41.740
But what is an Antisocial
Personality Disorder?

14:41.740 --> 14:44.370
It has the following
characteristics: lack of

14:44.370 --> 14:52.550
remorse, frequent lying, lack of
empathy, superficial charm,

14:52.550 --> 14:58.280
shallow emotions, distorted
sense of self, constant search

14:58.280 --> 14:59.420
for new sensations.

14:59.420 --> 15:01.020
Have you met someone
like that?

15:01.020 --> 15:07.430
You probably have, because
that's 3% of the population.

15:07.430 --> 15:09.660
I'm not anti-male, when I point
out there are three

15:09.660 --> 15:13.000
times as many jerks among
males as females.

15:13.000 --> 15:16.290
Females have characteristically
different

15:16.290 --> 15:17.620
personality disorders.

15:17.620 --> 15:21.640
And you can look down the list.
It's much more than 3%

15:21.640 --> 15:23.450
of the population that
would be diagnosed

15:23.450 --> 15:25.240
with one or the other.

15:25.240 --> 15:30.370
So, you know, an APD person
is manipulative, feigns

15:30.370 --> 15:34.590
affectionate or warm feelings,
but doesn't feel them, and is

15:34.590 --> 15:35.940
trying to deceive you.

15:39.010 --> 15:42.700
Once a student came to my office
and asked to sign up as

15:42.700 --> 15:44.640
my research assistant.

15:44.640 --> 15:46.510
I was talking and I thought,
well, maybe.

15:46.510 --> 15:48.000
I said come back
and talk to me.

15:48.000 --> 15:50.570
Later on, I read about him
in the Yale Daily News.

15:50.570 --> 15:52.970
He was an impostor student.

15:52.970 --> 15:55.180
He was not a Yale student.

15:55.180 --> 15:57.780
And he had been around to other
university campuses.

15:57.780 --> 16:00.580
He was an impostor at like
three different campuses.

16:00.580 --> 16:04.300
There was something wrong
with this person.

16:04.295 --> 16:06.815
Kind of made me feel -- then he
came later and asked me for

16:06.820 --> 16:07.970
a recommendation letter.

16:07.970 --> 16:12.450
I couldn't believe it after I
read about him in the Yale

16:12.450 --> 16:13.700
Daily News.

16:15.830 --> 16:17.540
This is extreme.

16:17.540 --> 16:24.660
And so, incidentally, someone
did a study of APD by going to

16:24.660 --> 16:28.970
a prison and categorizing
the inmates

16:28.970 --> 16:30.850
using DSM-IV standards.

16:30.850 --> 16:35.700
And they found that 40% of
the prisoners had APD.

16:35.700 --> 16:38.380
Also, neuroscience people have
found that there are

16:38.380 --> 16:42.140
differences in the prefrontal
cortex that are

16:42.140 --> 16:43.580
correlated with APD.

16:43.580 --> 16:45.650
So it seems to be --

16:45.650 --> 16:48.150
it's a problem we have in our
society that some people have

16:48.145 --> 16:50.685
a brain structure that's
a little different.

16:50.690 --> 16:54.310
And it may make it difficult
for them to

16:54.310 --> 16:55.780
behave in a good way.

17:01.089 --> 17:02.809
We're learning more and more
about neuroscience.

17:05.859 --> 17:08.529
It's interesting to me that Adam
Smith's book still rings

17:08.530 --> 17:09.830
true though after --

17:09.829 --> 17:12.149
There's some basic common
sense that we all learn.

17:12.150 --> 17:13.730
You have to judge people
and you have

17:13.730 --> 17:15.280
to learn their character.

17:15.280 --> 17:18.730
And you have to be a person of
character for, in the long

17:18.730 --> 17:21.230
run, that's what you want.

17:21.230 --> 17:23.180
It gives you what you
want in life.

17:36.320 --> 17:39.860
Oh, another thing I wanted
to say is that people are

17:39.860 --> 17:40.860
manipulable.

17:40.860 --> 17:43.270
Unfortunately, true.

17:43.270 --> 17:49.850
And unfortunately, we live in
a world where it's hard to

17:49.850 --> 17:55.990
avoid manipulating people at
all, because we have a free

17:55.990 --> 17:59.810
enterprise system that
encourages competition.

17:59.810 --> 18:03.450
And if the competition is
manipulating people, how do

18:03.450 --> 18:07.110
you completely stay
clear of that?

18:07.110 --> 18:10.210
I think, this is one of the
contradictions of our society.

18:10.210 --> 18:14.720
A very simple and obvious
manipulation is, they'll put a

18:14.720 --> 18:20.950
price on some item they're
selling, like $9.99.

18:20.950 --> 18:23.670
So you're like, well, why didn't
they just say $10.00?

18:23.670 --> 18:24.830
Well, you know why
they didn't.

18:24.825 --> 18:27.725
It's called ''pricing points''
in marketing.

18:27.730 --> 18:31.720
Because $9.99 sounds
a lot less,

18:31.720 --> 18:33.980
psychologically, than $10.00.

18:33.980 --> 18:38.620
So, everyone does it, almost
everyone does it.

18:38.620 --> 18:42.130
But is that bad?

18:42.130 --> 18:43.210
Well, it's bad in a way.

18:43.210 --> 18:44.950
It's manipulative, isn't it?

18:44.950 --> 18:46.430
I mean I'm annoyed by it.

18:46.430 --> 18:47.910
Maybe you are, too.

18:47.910 --> 18:52.730
But if you were in business,
would you do that too?

18:52.730 --> 18:56.140
You might feel that you have to
and it's a harmless sort of

18:56.140 --> 18:57.990
manipulation.

18:57.990 --> 19:00.080
It's not hurting
anyone really.

19:00.080 --> 19:04.190
They're maybe buying a little
bit more than they want.

19:04.190 --> 19:08.420
See, that's the kind of
thing that comes in.

19:08.420 --> 19:12.970
So, in looking at financial
institutions, they're often

19:12.970 --> 19:14.490
manipulative in that sense.

19:18.180 --> 19:19.540
It's similar to a politician.

19:19.540 --> 19:24.230
If you want to be a member of
Congress or whatever, you

19:24.230 --> 19:27.200
can't say what you
really believe.

19:27.200 --> 19:30.200
Because you won't get elected.

19:30.200 --> 19:34.360
You've got to kind of
doctor your opinions

19:34.360 --> 19:35.610
to the public opinion.

19:39.740 --> 19:42.080
But you might have a moral
purpose underlying it all,

19:42.080 --> 19:44.080
because you want to
get elected so

19:44.080 --> 19:45.330
you can do good things.

19:49.710 --> 19:51.320
So, you do end up
saying things.

19:51.320 --> 19:53.630
So, it's hard to judge
people, good or bad.

19:53.630 --> 19:57.410
It's an overall sense you get
of someone's character.

19:57.410 --> 19:59.520
That people are doing things
that appear somewhat

19:59.520 --> 20:03.930
manipulative and somewhat bad,
but you get an overall sense

20:03.930 --> 20:07.390
of the person through time.

20:07.390 --> 20:11.050
And ultimately, our society,
within limits, rewards people

20:11.050 --> 20:16.730
that show character through
all the confusing details.

20:16.730 --> 20:18.710
Now I wanted to move -- that
was my introduction.

20:18.710 --> 20:22.350
I wanted to move now to
discussing some particular

20:22.350 --> 20:25.800
aspects of Behavioral
Finance, or more

20:25.800 --> 20:28.300
broadly Behavioral Economics.

20:28.300 --> 20:37.450
And about human failings that
are exploitable by somebody

20:37.445 --> 20:45.515
and are somewhat exploited,
but remain.

20:45.520 --> 20:47.630
I wanted to start out with
what's probably the most

20:47.630 --> 20:51.460
famous element of Behavioral
Economics.

20:51.460 --> 21:01.820
It's Prospect Theory and
it was invented by two

21:01.820 --> 21:14.560
psychologists, Daniel Kahneman
and Amos Tversky, in the late

21:14.560 --> 21:16.380
20th century.

21:16.380 --> 21:20.680
They called it Prospect Theory
because it was a theory of how

21:20.680 --> 21:24.150
people form decisions
about prospects.

21:24.150 --> 21:26.330
And a prospect is a gamble.

21:26.330 --> 21:28.270
It's about people's decisions
under uncertainty.

21:31.020 --> 21:37.390
And in very simple terms, the
Prospect Theory says --

21:37.390 --> 21:39.520
now there's a huge literature on
this, so I'm trying to give

21:39.520 --> 21:47.200
you a very quick description
of it.

21:47.200 --> 21:50.920
That there's something called
a value function, which

21:50.920 --> 21:55.460
represents how people
value things.

21:58.450 --> 22:04.680
And there's a weighting
function, which shows how

22:04.680 --> 22:11.520
people infer or how they deal
with probabilities.

22:11.520 --> 22:15.610
And I just wrote simply what
Kahneman and Tversky say.

22:15.610 --> 22:18.680
I'll draw a picture of the
value function and the

22:18.680 --> 22:19.520
weighting function.

22:19.515 --> 22:24.645
And this will be very quick and
you'd have to read more,

22:24.650 --> 22:29.660
but the way people value
gains or losses --

22:29.660 --> 22:30.910
let me see.

22:33.334 --> 22:35.904
I better draw the line
in the middle.

22:39.280 --> 22:42.650
OK, and we're talking about
financial gains, so these are

22:42.650 --> 22:48.510
gains, and this is zero,
and this is losses.

22:48.506 --> 22:49.926
Well, negative.

22:49.930 --> 22:53.500
What I have on the horizontal
axis is wealth or money or

22:53.500 --> 22:55.140
something like that.

22:55.140 --> 22:57.810
Zero in the middle.

22:57.810 --> 23:01.560
And then, we have on this
axis value, which is

23:01.562 --> 23:04.512
something like utility.

23:04.510 --> 23:10.660
I'll erase my zero, so it
doesn't get in the way.

23:10.660 --> 23:12.900
What they find is,
that people's

23:12.900 --> 23:15.740
value has a funny shape.

23:15.740 --> 23:19.150
We don't weigh gains and
losses linearly.

23:22.890 --> 23:26.120
In the positive quadrant, when
we have positive gains,

23:26.122 --> 23:30.802
there's diminishing
value like that.

23:30.800 --> 23:32.570
It doesn't ever slope down.

23:32.570 --> 23:37.900
It's concave down like
diminishing marginal utility

23:37.900 --> 23:40.540
in economic theory.

23:40.540 --> 23:45.060
But for losses, it looks
something like this.

23:45.060 --> 23:47.550
It's concave up.

23:47.550 --> 23:50.710
I'm exaggerating a little bit,
but this is a diagram that

23:50.710 --> 23:58.150
Kahneman and Tversky wrote in
their famous Econometrica

23:58.150 --> 24:00.850
article 30 years ago.

24:00.850 --> 24:04.850
So, there's diminishing marginal
utility for gains,

24:04.850 --> 24:08.440
but there's the opposite --

24:08.440 --> 24:11.580
well, we have concave up.

24:11.580 --> 24:13.020
And there's a kink.

24:13.020 --> 24:16.620
Note that the value function
has a kink at the origin.

24:21.900 --> 24:25.030
So, what does this mean?

24:25.030 --> 24:30.010
OK, first of all, this
origin is a --

24:30.010 --> 24:34.500
from what point do I estimate
gains and losses?

24:34.500 --> 24:39.300
That's called the reference
point and it's psychological.

24:39.300 --> 24:41.570
And it's subject to
manipulation.

24:41.570 --> 24:43.740
The reference point
is the zero, from

24:43.740 --> 24:44.990
which I measure things.

24:48.280 --> 24:52.930
So, first of all, the reference
point is probably

24:52.930 --> 24:54.540
today's wealth.

24:54.540 --> 25:00.300
But it can be something else if
people are manipulated by

25:00.300 --> 25:03.860
the way something is
presented to them.

25:03.860 --> 25:09.300
Framing, according to Kahneman
and Tversky, is presentation.

25:09.300 --> 25:14.760
So, I can give the same prospect
to people, but word

25:14.760 --> 25:16.580
it in different ways
that suggests a

25:16.580 --> 25:18.840
different reference point.

25:18.840 --> 25:21.490
And that will change
people's behavior.

25:21.490 --> 25:25.620
So, you can manipulate people
by describing something in

25:25.620 --> 25:29.970
different terms, by suggesting
a different reference point.

25:29.970 --> 25:33.380
But typically, the reference
point is today's wealth.

25:33.380 --> 25:37.940
The kink means that people are
very conscious of little

25:37.940 --> 25:41.910
changes in their wealth and
they're spooked by them.

25:41.910 --> 25:45.540
I'm really afraid, because my
value drops very rapidly, even

25:45.540 --> 25:46.950
for a small loss.

25:46.950 --> 25:53.150
So, if you were to say,
lose $5 this morning.

25:53.150 --> 25:55.370
You had it in your pocket and
you lost it on the way to this

25:55.370 --> 25:59.810
lecture, you would feel
exaggeratedly bad about that.

25:59.810 --> 26:03.590
You should really regard $5 as
just nothing, because the

26:03.590 --> 26:05.740
present value of your lifetime
income is in the

26:05.740 --> 26:08.090
millions, so what's $5?

26:08.090 --> 26:10.090
But you don't think that way.

26:10.090 --> 26:16.120
So, you're spooked and deterred
by small losses, and

26:16.120 --> 26:19.940
less encouraged by
small gains.

26:19.940 --> 26:24.800
So, this kind of thing allows
businesspeople to exploit

26:24.800 --> 26:27.140
people if they want to.

26:27.140 --> 26:32.180
If people are so focused on
these little changes, then you

26:32.180 --> 26:37.340
are encouraged in business to
try to pick things out the

26:37.340 --> 26:40.990
people are paying attention
to, like small things, and

26:40.990 --> 26:43.800
sell insurance policies
on just those things.

26:48.620 --> 26:51.590
Insurance should be concentrated
on the really big

26:51.590 --> 26:53.550
things, like life insurance.

26:53.550 --> 26:56.920
You know the fact that one of
your parents could die and the

26:56.920 --> 26:59.650
children's family would be
out of money for the

26:59.650 --> 27:00.610
rest of their lives.

27:00.610 --> 27:02.070
That's a big thing.

27:02.070 --> 27:06.050
But it may not work to sell
that kind of insurance.

27:06.050 --> 27:11.730
You can do something that is
more focused on what people

27:11.730 --> 27:15.230
are watching and make it
something little so that it

27:15.230 --> 27:17.830
doesn't require them to
spend so much money.

27:17.830 --> 27:21.120
The classic example of that
is funeral insurance.

27:26.260 --> 27:29.320
You go around telling people
-- and for some reason this

27:29.320 --> 27:32.260
sales pitch works, and it's
worked for thousands of years.

27:32.260 --> 27:34.850
They sold this in
Ancient Rome.

27:34.850 --> 27:38.630
You tell people, if someone in
your family dies, you can have

27:38.630 --> 27:41.860
an expense of getting a proper
burial for this person.

27:41.855 --> 27:43.555
It costs money.

27:43.560 --> 27:47.720
And so, they would insure
that one thing.

27:47.720 --> 27:52.930
And it was a little thing, but
it works to sell that.

27:52.930 --> 27:57.100
Another example of it is airline
flight insurance.

27:57.100 --> 28:01.410
You're insured for this flight
on the airplane.

28:01.410 --> 28:04.720
I heard an ad for funeral
insurance recently.

28:04.720 --> 28:08.120
They're still doing this after
2,000 years, because it still

28:08.120 --> 28:10.110
works and it's manipulative.

28:10.110 --> 28:11.660
That's not what you should
do for people.

28:11.660 --> 28:14.040
You should not pick out
some little thing.

28:14.040 --> 28:16.150
Or they also have diamond
ring insurance.

28:16.150 --> 28:19.420
After an engagement, some women
will want to buy an

28:19.420 --> 28:21.560
insurance policy on the diamond,
because it can

28:21.560 --> 28:24.270
actually fall out of your
ring and you lose it.

28:24.270 --> 28:27.210
But that's like, what is it?
$5,000 or something?

28:27.210 --> 28:29.260
It's not big, it's
not essential.

28:29.260 --> 28:32.920
And if you're insuring that and
not other things, you're

28:32.920 --> 28:34.320
making a mistake.

28:34.320 --> 28:40.200
Fortunately, the insurance
industry is not too --

28:40.200 --> 28:42.800
it has come around to do
things that are more --

28:42.800 --> 28:45.020
they are doing things that
matter and are big.

28:45.020 --> 28:48.980
And that's because this Kahneman
and Tversky value

28:48.975 --> 28:51.775
function is --

28:51.780 --> 28:55.060
it represents an error that
people are prone to be making.

28:55.060 --> 28:59.180
But it's not total
and not complete.

28:59.180 --> 29:02.790
And so ultimately, people
don't go to insurance

29:02.790 --> 29:04.950
companies that manipulate
them.

29:04.950 --> 29:08.870
It gets around and eventually
people come around in wanting

29:08.870 --> 29:12.020
something better.

29:12.020 --> 29:16.660
So, while there is some
manipulation, it's limited.

29:16.660 --> 29:18.820
The other aspect of Kahneman
and Tversky is

29:18.820 --> 29:20.070
the weighting function.

29:25.470 --> 29:30.090
I'm going to draw again
a picture of it.

29:30.090 --> 29:34.730
This time it's how people
psychologically think about

29:34.730 --> 29:36.300
probabilities.

29:36.300 --> 29:40.230
This again is Kahneman
and Tversky.

29:40.228 --> 29:46.068
A probability is a number
between 0 and 1 or 0 and 100%.

29:46.065 --> 29:50.485
I'll put 0 here and 1 here.

29:50.490 --> 29:53.390
We can tell someone the
probability of something, but

29:53.390 --> 29:56.440
they can't accept it
psychologically.

29:56.440 --> 29:59.710
The errors that people
make are described by

29:59.710 --> 30:01.580
the weighting function.

30:01.580 --> 30:04.690
What the waiting function is,
it's the psychological impact.

30:04.690 --> 30:08.090
You are behaving as if you
just don't understand the

30:08.090 --> 30:09.330
probability.

30:09.330 --> 30:14.020
So, what Kahneman and Tversky
say is, that for very low

30:14.020 --> 30:18.820
probabilities people may
round them to zero.

30:18.820 --> 30:21.090
And for very high probabilities,
they may round

30:21.090 --> 30:22.890
them to one.

30:22.890 --> 30:27.080
But if they decide not to round
them to zero or one,

30:27.080 --> 30:30.360
they exaggerate the difference
between zero and one.

30:30.360 --> 30:34.330
You just can't think in terms of
a continuum of probability.

30:34.330 --> 30:36.070
So, this is what the weighting
function --

30:36.070 --> 30:39.400
this is the weight as a function
of the probability.

30:39.400 --> 30:42.470
For low probabilities
it's zero.

30:42.470 --> 30:44.250
I'm going to maybe exaggerate
it here.

30:44.250 --> 30:47.100
Then it jumps up.

30:47.100 --> 30:50.260
It's something like this, and
as it gets close to one, it

30:50.260 --> 30:53.230
jumps up to one again.

30:53.230 --> 30:56.730
So you see, it's like a
broken line segment.

30:56.730 --> 30:59.050
And there's been various
versions of this theory, but

30:59.050 --> 31:02.160
this is the simplest
version of it.

31:02.160 --> 31:07.360
So that means, if you're getting
on an airplane, you

31:07.360 --> 31:08.660
think, well, what's
the probability of

31:08.660 --> 31:10.660
this airplane crashing?

31:10.660 --> 31:13.150
Well it's probably something
like 1 in 10

31:13.150 --> 31:14.260
million or, what is it?

31:14.260 --> 31:16.640
Even less than that.

31:16.640 --> 31:19.030
So most of us, in our
mind, just say, it's

31:19.030 --> 31:20.580
zero and I'm done.

31:20.580 --> 31:21.660
I'm not going to think about.

31:21.660 --> 31:23.460
I'm not going to
worry about it.

31:23.460 --> 31:24.300
So, we're down here.

31:24.300 --> 31:26.540
We've rounded it to zero.

31:26.540 --> 31:29.870
But some people don't
round it to zero.

31:29.870 --> 31:32.730
And for them, they just blow
it out of all proportion in

31:32.730 --> 31:35.050
their minds, and it becomes
exaggerated.

31:35.050 --> 31:37.980
So, I have it here, so it
looks like it's a half.

31:37.980 --> 31:45.030
This would be one here and
this is about 0.4.

31:45.030 --> 31:47.680
And then ultimately, if the
probability gets really high,

31:47.680 --> 31:49.940
then I'm not even going to
think about it, it's one.

31:53.950 --> 31:57.090
In some of the most primitive
languages in the world,

31:57.090 --> 31:59.930
there's only a couple
of numbers.

31:59.930 --> 32:00.900
There's zero.

32:00.900 --> 32:02.190
There's one.

32:02.190 --> 32:03.840
Maybe there's two or
three, and then

32:03.835 --> 32:05.035
there's no more numbers.

32:05.040 --> 32:08.870
Well, our minds are still very
primitive in dealing with

32:08.870 --> 32:09.850
probability.

32:09.850 --> 32:12.530
So it's like there's only
three probabilities.

32:12.530 --> 32:17.610
Can't happen, may be,
and will happen.

32:17.610 --> 32:20.400
So, I think airline flight
insurance is an example of

32:20.400 --> 32:25.600
trying to manipulate this
personality characteristic.

32:25.600 --> 32:31.500
So, it means that they'll
catch all the people who

32:31.500 --> 32:33.340
exaggerate it.

32:33.340 --> 32:36.970
They used to have vending
machines outside of airlines.

32:36.970 --> 32:42.510
The vending machines would
encourage you to buy just for

32:42.510 --> 32:44.170
this flight.

32:44.170 --> 32:46.580
They put it right there when
you're getting on the flight.

32:46.580 --> 32:48.210
And so that's when you're
most nervous.

32:48.210 --> 32:50.800
If you're one of these
people who's up here.

32:50.800 --> 32:52.520
And of course, most people don't
buy it, but they don't

32:52.520 --> 32:53.610
have to sell it to everybody.

32:53.610 --> 32:55.220
They just sell it to these
people and they charge an

32:55.220 --> 32:56.700
outrageous price.

32:56.700 --> 32:59.130
But I haven't seen these vending
machines anymore.

32:59.130 --> 33:01.860
This is interesting.

33:01.860 --> 33:04.680
Economists wrote about them 30
or 40 years ago and they used

33:04.680 --> 33:05.750
to be everywhere.

33:05.750 --> 33:06.790
And they just kind
of disappeared --

33:06.790 --> 33:08.640
Do you ever see one of these?

33:08.640 --> 33:10.210
I think they're gone.

33:10.210 --> 33:11.030
Why is that?

33:11.030 --> 33:14.510
Well, somehow we get past
things like that.

33:14.510 --> 33:19.640
It's not like Kahneman and
Tversky are representing

33:19.640 --> 33:20.160
immutable errors.

33:20.160 --> 33:24.780
These are errors that
naturally happen.

33:24.780 --> 33:30.710
But you can get past it.

33:30.710 --> 33:34.630
And you end up wanting to deal
with people you trust. So, you

33:34.630 --> 33:36.990
see some vending machine at
the airport and you think,

33:36.990 --> 33:39.970
well, my insurance agent isn't
recommending I get this.

33:39.970 --> 33:41.260
I've got some kind
of insurance.

33:41.260 --> 33:43.310
So, you walk past it.

33:43.310 --> 33:52.370
There's a professor in Germany
at the Max Planck Institute in

33:52.370 --> 33:57.190
Berlin, Gerd Gigerenzer, who
has been taking on Kahneman

33:57.190 --> 34:02.930
and Tversky in saying that
they're right that people show

34:02.929 --> 34:05.429
these tendencies for errors.

34:05.429 --> 34:10.209
But I can train people out
of them with no problem.

34:10.210 --> 34:12.580
I just tell them this is an
error, and teach them then,

34:12.580 --> 34:13.830
and they don't do it anymore.

34:27.860 --> 34:32.790
Gary Gorton just did a seminar
here on errors that people

34:32.790 --> 34:35.380
make in financial --

34:35.380 --> 34:40.170
No, Nick Barberis here at SOM,
and he was using Caltech

34:40.170 --> 34:42.820
students and found that --

34:42.820 --> 34:48.100
and tested their ability
to prevent certain kind

34:48.100 --> 34:49.430
of errors like this.

34:49.430 --> 34:52.530
And he found that even the
Caltech students made these

34:52.530 --> 34:55.040
errors just horribly.

34:55.040 --> 34:57.350
We're wondering, aren't they
supposed to be bright?

34:57.350 --> 35:00.120
Those are young math geniuses.

35:00.120 --> 35:02.320
But about a third of them
got everything right.

35:02.320 --> 35:05.350
So I'm thinking, you know,
they're only undergraduates.

35:05.350 --> 35:07.700
By the time they get along, if
they go -- they'll eventually

35:07.700 --> 35:09.400
be trained out of
these errors.

35:09.400 --> 35:13.230
But right now, they're behaving
just like Behavioral

35:13.230 --> 35:14.930
Finance says they will.

35:19.150 --> 35:23.250
So anyway, I think that you will
find that Prospect Theory

35:23.250 --> 35:27.400
explains a lot of things that
go on in finance, but it

35:27.400 --> 35:30.520
doesn't explain everything.

35:30.520 --> 35:32.570
And let me move on.

35:35.300 --> 35:38.370
So, we want to talk about --

35:45.186 --> 35:46.436
let me see.

35:50.730 --> 35:53.860
I have got so many things
to tell you about here.

35:53.860 --> 35:55.850
And I'm thinking
about my time.

35:55.850 --> 35:58.380
It's a huge field, Behavioral
Finance.

35:58.380 --> 35:59.660
Let me mention a few
other things.

35:59.660 --> 36:03.550
Regret Theory is a
theory that --

36:03.550 --> 36:07.450
it's kind of related to
Kahneman and Tversky.

36:07.450 --> 36:11.180
It says that people fear
the pain of regret.

36:11.180 --> 36:12.620
There's an old expression.

36:12.620 --> 36:14.500
"I was kicking myself,"
because I

36:14.500 --> 36:16.750
made some bad decision.

36:16.750 --> 36:19.760
Well, that's a painful
experience when you did

36:19.760 --> 36:21.710
something wrong.

36:21.710 --> 36:25.680
This is represented somewhat
in the kink in the Prospect

36:25.680 --> 36:27.700
Theory value function.

36:27.700 --> 36:30.050
But Regret Theory says that
there's actually a painful

36:30.050 --> 36:32.810
emotion, that you're
wired not to like

36:32.810 --> 36:34.460
to have made a mistake.

36:34.460 --> 36:37.200
And so then, you end up
designing your life around

36:37.200 --> 36:39.880
that and trying to avoid
doing anything that you

36:39.880 --> 36:41.450
might regret later.

36:41.450 --> 36:46.480
And it can create problems. You
may make bad decisions,

36:46.480 --> 36:51.310
because you're overly worried
about regret.

36:51.310 --> 36:52.560
Gambling behavior.

36:55.180 --> 37:00.450
Anthropologists have reported
that gambling occurs in every

37:00.450 --> 37:01.700
human society.

37:03.780 --> 37:06.560
And so, it's one of the
human universals.

37:06.560 --> 37:10.460
Not that everyone does it, but
in every society you'll find

37:10.460 --> 37:13.870
people that do it.

37:13.870 --> 37:16.740
I have a 1974 study.

37:16.740 --> 37:22.260
It found that 61% of U.S. adults
actually gambled at

37:22.260 --> 37:25.020
least once for money
in that year.

37:25.020 --> 37:26.380
I bet, it has gone up.

37:26.380 --> 37:30.390
There's more opportunities for
gambling and it's gone up.

37:30.390 --> 37:37.590
1.1% of men are compulsive
gamblers and 0.5% of women.

37:37.590 --> 37:40.530
This is another male trait.

37:40.530 --> 37:43.250
Somehow men are more vulnerable
to compulsive

37:43.250 --> 37:44.640
gambling than women.

37:44.640 --> 37:47.420
But it's only a factor
of 2-to-1.

37:47.420 --> 37:50.890
But it's an addiction
that happens that

37:50.890 --> 37:52.740
distorts people's thinking.

37:52.740 --> 37:55.410
And it's such an addiction that
we have an organization

37:55.410 --> 37:59.980
called Gamblers Anonymous that
helps people with this.

37:59.980 --> 38:02.270
It ruins people's lives.

38:02.270 --> 38:04.980
People end up getting a divorce,
because you can't

38:04.980 --> 38:08.360
stay with someone, married to
someone, who is squandering

38:08.360 --> 38:09.790
the family money.

38:09.790 --> 38:10.230
They do it.

38:10.230 --> 38:13.580
They end up sneaking around to
gamble, like drinkers sneak

38:13.580 --> 38:14.850
around for the next drink.

38:17.490 --> 38:22.890
Gambling behavior, it seems to
be associated psychologically

38:22.890 --> 38:26.700
with a self-image, a sense of
who I am and why I'm an

38:26.700 --> 38:27.900
important and good person.

38:27.900 --> 38:29.860
A sense of competence.

38:29.860 --> 38:34.300
Most gamblers do things that
they think are revealing of

38:34.300 --> 38:35.550
their competence.

38:35.550 --> 38:39.850
And they tend to pick a certain
form of gambling that

38:39.850 --> 38:42.600
they become psychologically
identified with.

38:42.600 --> 38:48.020
And they avoid any other
form of gambling.

38:48.020 --> 38:51.170
Gambling behavior is part
of what goes on

38:51.170 --> 38:52.760
in the stock market.

38:52.760 --> 38:55.600
Certain people who have a
personality, which makes them

38:55.600 --> 39:00.480
particularly interested in
gambling, find that a life in

39:00.480 --> 39:05.810
finance can give them the
kind of stimulation.

39:05.810 --> 39:10.330
Gambling behavior, by the way,
is almost like a drug

39:10.330 --> 39:13.200
addiction in a sense.

39:13.200 --> 39:18.380
People who are depressed may go
to a gambling casino as a

39:18.380 --> 39:22.130
way of getting themselves
out of the depression.

39:22.130 --> 39:25.390
And they say that when they walk
into the casino, suddenly

39:25.390 --> 39:26.570
my troubles are gone.

39:26.570 --> 39:30.430
I feel invigorated and alive.

39:30.430 --> 39:33.610
It's almost like it creates a
hormonal difference that they

39:33.610 --> 39:36.360
seek, and it's almost like
injecting yourself with

39:36.360 --> 39:40.090
something, so it's a very
hard thing to conquer.

39:40.090 --> 39:44.030
I mentioned before, when the New
York state in 1811 created

39:44.030 --> 39:47.850
the first corporate law that
produced a lot of questionable

39:47.850 --> 39:51.690
companies, people then said,
this is just gambling.

39:51.690 --> 39:53.380
It's bad.

39:53.380 --> 39:55.980
But the other side of it is
that this same gambling

39:55.980 --> 39:59.410
behavior, it's not usually
a pathology.

39:59.410 --> 40:03.270
It's an aspect of human
sensation-seeking of various

40:03.270 --> 40:06.850
aspects of our psychology
that drive us.

40:06.850 --> 40:10.620
What the stock market is, in
some sense, is a way of

40:10.620 --> 40:13.960
channeling this kind of behavior
into something

40:13.960 --> 40:17.090
productive instead
of just a game.

40:17.090 --> 40:21.530
And so, they make it very clear
in the stock markets of

40:21.530 --> 40:24.160
the world, this is about
business and this is

40:24.160 --> 40:25.280
productive.

40:25.280 --> 40:29.670
The same emotional patterns that
created gambling behavior

40:29.670 --> 40:33.270
as a human universal
underlie some --

40:33.270 --> 40:36.500
this is not abnormal,
it's most people.

40:36.500 --> 40:42.600
Underlie traits that
work out well.

40:42.600 --> 40:44.830
OK, the next major thing I
wanted to talk about is

40:44.830 --> 40:46.080
overconfidence.

40:48.140 --> 40:53.440
And psychologists have found
that there's a human tendency

40:53.435 --> 40:57.385
to overestimate one's
own abilities.

40:57.390 --> 40:58.470
We all think --

40:58.470 --> 41:02.930
not all of us, most of us think
we're above average.

41:02.930 --> 41:06.100
Some of us think, we're
way above average.

41:06.100 --> 41:09.570
And this tendency has been
revealed in a number of

41:09.570 --> 41:10.750
experiments.

41:10.750 --> 41:12.310
I thought I would
try one on you.

41:12.310 --> 41:13.930
I don't know if it will work.

41:13.930 --> 41:19.310
I'll try it on this class, if
you will participate in this

41:19.310 --> 41:21.720
experiment by a show of hands.

41:21.720 --> 41:22.990
I'm going to ask you --

41:22.990 --> 41:26.100
I have three questions here.

41:26.100 --> 41:32.600
And I want to ask you
to write down a

41:32.600 --> 41:43.660
90% confidence interval.

41:43.660 --> 41:45.510
Do you have a pencil
to write this down?

41:45.510 --> 41:49.440
And then afterwards, I'm going
to tell you the answer and see

41:49.440 --> 41:51.340
if it fell in your confidence
interval.

41:51.340 --> 41:52.920
OK, so this is what it is.

41:52.920 --> 41:55.630
What is a 90% confidence
interval?

41:55.630 --> 42:00.030
It's a range of values, so that
you are 90% sure that the

42:00.030 --> 42:03.100
true value lies in this range.

42:03.100 --> 42:10.200
So, if I asked you, what is the
population of New Haven?

42:10.200 --> 42:14.990
You might say, 90% confidence
interval that's between 50,000

42:14.990 --> 42:16.670
and 150,000.

42:16.670 --> 42:21.030
That means, you're 90% sure that
the population falls in

42:21.030 --> 42:21.920
that range.

42:21.920 --> 42:25.730
And so, you should be right
90% of the time.

42:25.730 --> 42:28.930
If I ask you to give 90%
confidence intervals, you

42:28.930 --> 42:32.430
should be right 9
times out of 10.

42:32.430 --> 42:36.110
If I ask for a 99% confidence
interval, you have to widen

42:36.110 --> 42:38.550
the interval so that you
should be right 99

42:38.550 --> 42:40.450
times out of 100.

42:40.450 --> 42:43.120
So, what I'm going to ask you
to do, if you will cooperate

42:43.120 --> 42:47.480
with me, is give 90% confidence
intervals for --

42:47.480 --> 42:50.130
I have three questions I'm
going to give you.

42:50.130 --> 42:53.260
But I have to ask you to be
honest, otherwise this thing

42:53.260 --> 42:54.320
won't work.

42:54.320 --> 42:57.880
You could game me by just
giving excessively wide

42:57.880 --> 43:01.400
confidence intervals, right?

43:01.400 --> 43:03.310
From the ''zero-to-infinity''
type.

43:03.310 --> 43:05.330
Then you'll always
be right, but

43:05.330 --> 43:06.640
you're not playing honestly.

43:06.640 --> 43:11.240
So, I have to appeal to your
character to do this honestly.

43:11.240 --> 43:14.470
So, I have three
questions here.

43:14.470 --> 43:18.270
I just changed the questions.

43:18.270 --> 43:19.870
The first is --

43:19.870 --> 43:22.680
now what you have to write
down on your notepaper

43:22.680 --> 43:26.890
somewhere is your honest
estimate of a

43:26.890 --> 43:29.010
90% confidence interval.

43:29.010 --> 43:33.550
And so, the first question is,
the world population --

43:33.550 --> 43:39.820
how many people are alive
in the world?

43:39.820 --> 43:43.400
As of, I think it
was 8:00 a.m.

43:43.400 --> 43:45.250
this morning
[addition: February 21, 2011].

43:45.250 --> 43:49.420
The U.S. Census has something
called the World Population

43:49.420 --> 43:51.670
Clock and just go --

43:51.670 --> 43:53.670
don't cheat.

43:53.670 --> 43:55.230
I know, some of you
have laptops.

43:55.230 --> 43:56.340
Don't do it.

43:56.340 --> 44:00.530
But after this you can search
Google on World Population

44:00.530 --> 44:03.120
Clock and it shows you
minute-by-minute how many

44:03.120 --> 44:04.590
people there are in the world.

44:04.590 --> 44:05.970
Every birth and death.

44:05.970 --> 44:07.920
It's not actually recorded,
it's a fake.

44:07.920 --> 44:10.030
But I mean it's supposed
to be an estimate.

44:10.030 --> 44:12.000
So, I got the world
population.

44:12.000 --> 44:15.660
So, what I want you to do, can
you write down a lower bound

44:15.660 --> 44:20.110
and an upper bound for the world
population this morning

44:20.110 --> 44:22.910
as measured by the
U.S. Census?

44:22.910 --> 44:25.790
OK, can you write that down?

44:25.790 --> 44:28.380
But I don't want you to
make it too wide.

44:28.380 --> 44:31.550
Remember, I only want
you to be 90% sure.

44:31.550 --> 44:33.700
And I don't want you to make
it too narrow, because then

44:33.700 --> 44:36.360
you're more likely to fail.

44:36.360 --> 44:38.960
So, you've written that down,
the world population?

44:38.960 --> 44:43.030
OK, my next question is --

44:43.030 --> 44:43.320
2.

44:43.315 --> 44:46.925
The world, what does it weigh?

44:50.090 --> 44:52.900
Well, actually, it's the mass,
in kilograms, of the world.

44:52.900 --> 44:54.890
Can you write that down?

44:54.890 --> 44:56.140
This is astronomy.

45:00.350 --> 45:03.330
Let me say, I'm asking for it
in kilograms. But you can do

45:03.330 --> 45:06.100
it in metric tons.

45:06.100 --> 45:07.570
That just knocks off
three zeros.

45:07.570 --> 45:15.440
A metric ton is 1,000 kilograms.
And just so you'll

45:15.440 --> 45:20.510
know for sure, that's not the
same thing as a long ton,

45:20.510 --> 45:23.350
which is U.K. The United Kingdom
uses the long ton,

45:23.350 --> 45:25.910
which is 2,240 pounds.

45:25.910 --> 45:27.290
I just looked this up.

45:27.290 --> 45:30.770
And is 1.1 metric ton.

45:30.770 --> 45:33.850
And it's not exactly the same
as a short ton, which we use

45:33.850 --> 45:39.510
in the United States, which is
2,000 pounds, which is 0.98

45:39.510 --> 45:41.570
metric tons.

45:41.570 --> 45:43.230
Just tell me, how many tons.

45:43.230 --> 45:44.940
That's not going to
affect your 90%

45:44.940 --> 45:45.980
confidence interval, right?

45:45.975 --> 45:50.865
Just tell me how many tons
does the earth weigh.

45:50.870 --> 45:53.460
And then that's the second and
I have one more question.

45:58.390 --> 46:03.300
How many languages are
there in the world?

46:03.300 --> 46:07.750
Now, I know you might complain,
this is a matter of

46:07.750 --> 46:11.490
definition, because sometimes
two dialects might be

46:11.490 --> 46:14.900
considered a separate
language.

46:14.900 --> 46:17.280
Well, I'm asking you to
give me the number --

46:17.280 --> 46:25.460
the World Authority on languages
is an organization

46:25.460 --> 46:27.970
that has a website called
ethnologue.com.

46:33.180 --> 46:36.570
And if you go to that website,
they're always

46:36.570 --> 46:37.670
discovering new languages.

46:37.670 --> 46:38.710
They keep track of it.

46:38.710 --> 46:41.980
New languages keep getting
discovered, because some guy

46:41.980 --> 46:44.590
is hiking out in Siberia and
they go to this little

46:44.590 --> 46:48.910
village, and say, hey, these
people are speaking a language

46:48.910 --> 46:51.220
that has never been
documented before.

46:51.220 --> 46:54.870
So, it's this process of
learning languages.

46:54.870 --> 46:58.420
They also keep dying out,
because there's just elderly

46:58.420 --> 47:01.170
people in this village, and when
they die you know this

47:01.170 --> 47:03.080
language is going to
die with them.

47:03.080 --> 47:06.650
So anyway, the question is,
I want your 90% confidence

47:06.650 --> 47:11.030
interval for the number of
languages, as defined by

47:11.030 --> 47:13.310
ethnologue.com.

47:13.310 --> 47:15.360
You have to guess how they
define a language.

47:18.150 --> 47:20.890
But you have an idea more or
less what a language is.

47:20.890 --> 47:23.110
It's more than a dialect,
because they can still

47:23.110 --> 47:25.980
understand each other if they
speak different dialects.

47:25.980 --> 47:28.690
We're talking about really
different languages.

47:28.690 --> 47:32.240
OK, have you written down three
confidence intervals?

47:32.240 --> 47:34.410
OK, so I'm going to write
down the answers

47:34.410 --> 47:36.320
and I hope this works.

47:36.320 --> 47:39.160
I'm trusting to your honestly
in getting these things,

47:39.160 --> 47:42.170
because you good game me
and make this not work.

47:42.170 --> 47:43.560
But give me your honest count.

47:43.560 --> 47:46.250
So, I'm going to write down
the correct answer.

47:46.250 --> 47:50.490
So, the world population
as of 8:00 a.m.

47:50.490 --> 47:52.670
this morning
[addition: February 21, 2011]

47:52.670 --> 48:06.400
was 6,901,330,581.

48:06.400 --> 48:10.480
Now, can I get a show of hands,
how many peoples'

48:10.480 --> 48:13.260
confidence interval includes
that number?

48:16.320 --> 48:18.570
OK, can someone tell me
what percent that is?

48:18.570 --> 48:20.130
That is not 90%.

48:20.130 --> 48:22.240
You're doing pretty
well, though.

48:22.240 --> 48:23.980
What do you think, Oliver?

48:23.980 --> 48:24.850
STUDENT: About 80.

48:24.850 --> 48:26.290
PROFESSOR ROBERT SHILLER:
You think it's 80?

48:26.290 --> 48:29.820
Let me see them up again.

48:29.820 --> 48:30.870
Maybe it is 80, all right.

48:30.870 --> 48:34.050
You're doing really well.

48:34.050 --> 48:37.190
Some honest people here didn't
put their hands up.

48:37.190 --> 48:38.120
All right, well, that's one.

48:38.120 --> 48:40.430
So, we did 80 instead of 90.

48:40.430 --> 48:43.440
What about the weight
of the earth?

48:43.440 --> 48:55.970
In kilograms. Well, it's 5.974
times 10 to 24th power.

48:55.970 --> 49:00.220
And I'll give you
that in tons.

49:00.220 --> 49:08.000
It's 5,974 billion
billion tons.

49:08.000 --> 49:09.330
You got that?

49:09.330 --> 49:11.500
You might have to do
some calculation.

49:11.500 --> 49:15.910
It'd be 5.9734 times 10
to the 18th power.

49:15.907 --> 49:16.147
[correction: 5,974 times
10 to the 18th power is

49:16.150 --> 49:17.260
the weight in tons.]

49:17.260 --> 49:18.360
OK, can we do a show of hands?

49:18.360 --> 49:21.260
How many people had
a number in --

49:21.260 --> 49:24.480
how many people are in the
confidence interval there.

49:24.480 --> 49:26.510
All right, Oliver, what
do you think?

49:26.510 --> 49:27.790
What's the fraction?

49:27.792 --> 49:29.242
STUDENT: 5%, maybe 10.

49:29.240 --> 49:29.670
PROFESSOR ROBERT SHILLER: 10%.

49:29.670 --> 49:30.430
OK.

49:30.430 --> 49:34.180
So this one, we did really
well on world population.

49:34.180 --> 49:35.770
80, 10.

49:35.770 --> 49:39.370
The last one, how many languages
are in the world?

49:39.370 --> 49:44.450
Well, according to
ethnologue.com there are 6,909

49:44.450 --> 49:47.920
languages this morning
[addition: February 21, 2011].

49:47.920 --> 49:49.780
How many people got that within

49:49.780 --> 49:52.040
their confidence interval?

49:52.040 --> 49:54.620
OK, what do you think, Oliver?

49:54.620 --> 49:55.480
STUDENT: About the same, 10%.

49:55.480 --> 49:57.380
PROFESSOR ROBERT SHILLER:
10%, OK.

49:57.380 --> 49:59.450
Why did you do so much better
on world population?

50:02.250 --> 50:04.370
Well, thank you for being
honest with me.

50:04.370 --> 50:08.340
I think it worked once again.

50:08.340 --> 50:09.590
The overconfidence.

50:09.590 --> 50:13.080
So, why is it that people are
overconfident like this?

50:13.080 --> 50:21.770
And psychologists have tried to
describe, what it is that

50:21.770 --> 50:25.170
goes on in people's minds that
produces answers like this.

50:25.170 --> 50:28.620
One of them is, that people seem
to have a sense that they

50:28.620 --> 50:31.580
understand the world more
than they really do.

50:31.580 --> 50:33.140
It's an illusion.

50:33.140 --> 50:36.510
Actually, the world is just
infinitely complicated and

50:36.510 --> 50:37.760
there are so many surprises.

50:40.520 --> 50:42.260
When you think about a question
like this, there's

50:42.260 --> 50:45.750
many different perspectives
that you can take.

50:45.750 --> 50:49.790
And if you thought more about
it, your imagination might

50:49.790 --> 50:52.020
help you to widen your
confidence interval.

50:52.020 --> 50:55.110
But you can't think of all
the perspectives at once.

50:55.110 --> 50:59.600
And so, you tend to gravitate to
the first one that comes to

50:59.600 --> 51:04.670
mind and it gives you an
underestimate of the

51:04.670 --> 51:05.920
confidence.

51:10.080 --> 51:13.130
So, that is overconfidence.

51:13.130 --> 51:15.770
By the way, I think it's a
little bit higher in males

51:15.770 --> 51:17.250
than females.

51:17.250 --> 51:19.520
I didn't do a separate
male/female count.

51:19.520 --> 51:20.440
But females [correction: males]

51:20.440 --> 51:23.180
are definitely overconfident.

51:23.180 --> 51:26.140
That's the so-called macho
personality that's supremely

51:26.140 --> 51:30.140
overconfident, which is --
that's not in DSM-IV. I don't

51:30.140 --> 51:33.650
think it is, but it is more
common among males.

51:33.650 --> 51:37.110
But it's really, everyone
is overconfident.

51:37.110 --> 51:41.780
There's no important sex
difference here.

51:41.780 --> 51:44.540
Incidentally, I think that
overconfidence, and this is an

51:44.540 --> 51:47.610
important phenomenon, it
goes beyond yourself.

51:47.612 --> 51:49.462
It extends to your friends.

51:52.170 --> 51:55.330
And you exaggerate --

51:55.330 --> 51:57.500
there's a tendency for people
to think that I have very

51:57.500 --> 51:58.750
smart friends.

52:01.030 --> 52:02.290
I was reflecting
the other day.

52:02.290 --> 52:05.910
When I was an undergraduate at
University of Michigan, I was

52:05.910 --> 52:08.720
in the honors program and we
thought we were pretty smart.

52:08.720 --> 52:13.130
And I had a number of young
people that I just imagined

52:13.130 --> 52:17.840
were heading toward really
great careers.

52:17.840 --> 52:22.190
There was one student that we
called Young Jack Kennedy.

52:22.190 --> 52:23.770
You know, this was
some years ago.

52:23.770 --> 52:25.560
Jack Kennedy was president
of the United States.

52:25.560 --> 52:26.730
We thought he was a genius.

52:26.730 --> 52:28.940
That was probably wrong, too.

52:28.940 --> 52:32.050
None of these people
are geniuses.

52:32.050 --> 52:37.020
I was thinking that most of my
friends ended up in very good

52:37.020 --> 52:42.150
careers, but nothing that you
would think was spectacular.

52:42.150 --> 52:45.530
I had one friend as an
undergrad, who I thought was a

52:45.530 --> 52:50.400
genius, and his name was
Bruce Wasserstein.

52:50.400 --> 52:52.180
Anyone ever heard of him?

52:52.180 --> 52:53.700
Maybe not.

52:53.700 --> 52:54.790
Well see, that's it.

52:54.790 --> 52:57.430
But he founded his own
investment bank called

52:57.430 --> 53:02.620
Wasserstein Perella, became
really rich and then he bought

53:02.620 --> 53:08.100
interest in Lazard Freres, the
French investment bank.

53:08.100 --> 53:09.970
He was a real big shot
on Wall Street.

53:09.970 --> 53:14.860
I met him again about 10 years
ago and then he died.

53:14.860 --> 53:16.000
God.

53:16.000 --> 53:18.720
He had a heart attack
and died.

53:18.720 --> 53:19.860
So, I know his whole life.

53:19.860 --> 53:22.950
I saw him when he was 18 and
I've watched his whole life,

53:22.950 --> 53:23.910
and it's now history.

53:23.910 --> 53:25.900
It's kind of scary.

53:25.900 --> 53:29.480
But I remember thinking he was
a genius as an undergrad.

53:29.480 --> 53:31.410
I was wondering, what was
wrong with my judgment?

53:31.410 --> 53:33.060
Why did I see so many
other geniuses?

53:37.590 --> 53:41.620
Now that I think back on it,
he had a sort of real world

53:41.620 --> 53:44.140
common sense that amazed me.

53:47.040 --> 53:50.170
He just knew things that --
it wasn't fake knowledge.

53:50.170 --> 53:53.050
He seemed to know how
things worked.

53:53.050 --> 53:54.760
So, I guess I was right
about one of them.

53:54.760 --> 53:58.560
But not enough that none of you
have heard of him right?

53:58.560 --> 54:00.830
He has an investment bank
named after him.

54:00.830 --> 54:04.530
You should have heard of
him, but maybe not.

54:04.530 --> 54:07.110
But anyway, this thing affects
peoples' thinking, too.

54:07.110 --> 54:09.960
I think that we tend to think
that the head of state who is

54:09.960 --> 54:14.180
running, the head of our central
bank is a genius.

54:14.180 --> 54:17.920
And this really clouds
our thinking.

54:17.920 --> 54:21.380
It's like our ego extends to
the other people that we

54:21.380 --> 54:23.190
associate with ourselves.

54:23.190 --> 54:25.460
Now, the head of a central bank
in another country we

54:25.460 --> 54:26.910
have no respect for.

54:26.910 --> 54:29.590
It's only in our own country,
because it's part of our ego

54:29.590 --> 54:31.820
involvement that produces
this overconfidence.

54:37.710 --> 54:40.100
This tendency for overconfidence
produces a lot

54:40.100 --> 54:44.290
of anomalies and opportunities
for manipulation.

54:44.290 --> 54:49.720
So, for example, Rakesh Khurana,
who is a professor at

54:49.715 --> 54:57.305
the Harvard Business School,
has written a book called

54:57.310 --> 55:01.370
Search for the Charismatic
CEO.

55:01.370 --> 55:05.970
He claims that there's a
tendency for people to think

55:05.970 --> 55:08.410
that CEOs are geniuses.

55:08.410 --> 55:11.940
Or at least the one that
we found is a genius.

55:11.940 --> 55:16.970
And companies then seek out a
genius CEO to put in charge of

55:16.970 --> 55:19.770
the company as a kind
of manipulation

55:19.770 --> 55:21.250
of the stock market.

55:21.250 --> 55:28.670
They think if we get -- you
know, if we got some guy who's

55:28.670 --> 55:31.510
run other companies successfully
in the past, he

55:31.510 --> 55:32.370
must be a genius.

55:32.370 --> 55:35.410
Put him in our company, our
stock price will go up, then

55:35.410 --> 55:37.810
we can sell our executive --

55:37.810 --> 55:41.330
we can exercise our options
and make a lot of money by

55:41.330 --> 55:43.270
putting in this fake genius.

55:43.270 --> 55:48.880
And Khurana says, well, there
are some people who are maybe

55:48.880 --> 55:50.930
geniuses at management,
but most of the time

55:50.930 --> 55:52.300
they're just lucky.

55:52.300 --> 55:55.350
And we tend to develop
overconfidence.

55:55.350 --> 55:58.190
And then what happens, according
to Khurana, is, you

55:58.190 --> 56:00.760
put in some guy who turned
around some company

56:00.760 --> 56:03.370
spectacularly, supposedly.

56:03.370 --> 56:06.390
You bring him in to run a new
company and he doesn't know

56:06.390 --> 56:09.030
anything about this
new company.

56:09.030 --> 56:12.730
But he has to justify himself,
so he lays off a lot of people

56:12.730 --> 56:16.020
and shuffles things around, and
just destroys everything

56:16.020 --> 56:20.240
in the company, and
ruins things.

56:20.240 --> 56:25.480
This is related to another
author that I recommend.

56:25.480 --> 56:26.940
I've mentioned him
before, I think.

56:29.570 --> 56:33.320
Nassim Taleb wrote a book called
Fooled by Randomness.

56:37.990 --> 56:46.070
He's Lebanese, but now in the
U.S. Nassim Taleb, Fooled by

56:46.070 --> 56:52.660
Randomness, that says that most
of the things that happen

56:52.660 --> 56:55.430
in life are just chance.

56:55.430 --> 56:57.370
We tend to ascribe them.

56:57.370 --> 57:00.140
If they happen to us,
we conclude --

57:00.140 --> 57:01.800
we're very quick to conclude
that it's a

57:01.800 --> 57:04.170
sign of our own genius.

57:04.170 --> 57:07.230
And if it happens to someone
that is a friend of us, then

57:07.230 --> 57:11.570
you think, well, I have genius
friends, isn't that nice?

57:11.570 --> 57:13.410
And so, it leads to mistakes.

57:18.230 --> 57:21.000
OK, let me go to another --

57:21.000 --> 57:22.050
how much time do I have --

57:22.050 --> 57:25.510
cognitive dissonance.

57:25.510 --> 57:27.430
This is another psychological
principle.

57:31.152 --> 57:40.512
The term was coined by
sociologist Leon Festinger in

57:40.510 --> 57:42.590
the 1950s, I believe.

57:42.590 --> 57:44.660
I actually met this guy.

57:44.660 --> 57:46.620
That's the nice thing about
being in academia, you meet

57:46.620 --> 57:49.130
all these great names if
you're in long enough,

57:49.130 --> 57:51.290
eventually.

57:51.290 --> 57:54.290
But what is cognitive
dissonance?

57:54.290 --> 57:59.540
It's a judgmental bias that
people tend to make, because

57:59.540 --> 58:01.600
they don't want to admit
they're wrong.

58:01.600 --> 58:03.550
Maybe I'm oversimplifying
this mistake.

58:03.550 --> 58:07.470
It's painful to think, that I
believe something and it was

58:07.470 --> 58:13.710
wrong, so people will cling to
old beliefs and try to find

58:13.710 --> 58:16.050
evidence that supports their
beliefs, because they have an

58:16.050 --> 58:19.380
ego involvement with
the belief.

58:19.380 --> 58:20.760
And so, I will be biased.

58:20.760 --> 58:23.800
The famous experiment
indicating cognitive

58:23.800 --> 58:26.170
dissonance, done by some
psychologist, had the

58:26.170 --> 58:27.720
following form.

58:27.720 --> 58:30.960
They got a list of people who
had just bought a car and they

58:30.960 --> 58:32.420
knew what make of car.

58:32.420 --> 58:35.360
They got the list from car
dealers, so they knew exactly

58:35.360 --> 58:37.630
what car they had just bought.

58:37.630 --> 58:39.450
And they called these people
up and asked them to

58:39.450 --> 58:41.350
participate in a psych
experiment.

58:41.350 --> 58:43.750
Or I think they said a
marketing experiment.

58:43.750 --> 58:46.240
They didn't let them know that
they knew what car they had

58:46.240 --> 58:47.360
just bought.

58:47.360 --> 58:49.700
And then, the experiment
was the following.

58:49.700 --> 58:53.050
Let's go through a number of --
what magazines do you read?

58:53.050 --> 58:54.430
And they said, let's
get these out.

58:54.430 --> 58:57.030
They got all the magazines that
were on the newsstand.

58:57.030 --> 58:59.560
And they said, let's look
through page by page and tell

58:59.560 --> 59:02.770
us which ads you remember
reading.

59:02.770 --> 59:06.250
What they found is, that people
read the ads for the

59:06.250 --> 59:08.990
car they just bought.

59:08.990 --> 59:14.190
And they avoided especially the
car that they thought they

59:14.190 --> 59:16.900
might buy, but decided
not to buy.

59:16.900 --> 59:20.220
So, after you buy a car,
you want to confirm

59:20.220 --> 59:21.130
your belief in it.

59:21.130 --> 59:24.040
So, you selectively get
information that

59:24.040 --> 59:26.450
confirms your belief.

59:26.450 --> 59:31.470
And so, this cognitive
dissonance is another factor.

59:31.470 --> 59:32.530
It's been demonstrated.

59:32.530 --> 59:33.920
It's an error that
people make.

59:33.920 --> 59:35.380
It doesn't mean that people --

59:35.380 --> 59:38.290
again, none of these
errors is unviable.

59:38.290 --> 59:41.030
People will make the error and
then they'll learn from their

59:41.030 --> 59:42.590
mistakes and they'll correct.

59:42.590 --> 59:46.060
They're not totally cognitive
dissonant, but it's just a

59:46.060 --> 59:49.250
kind of error that
keeps coming up.

59:49.250 --> 59:52.700
So, I give you a couple of
examples of cognitive

59:52.700 --> 59:55.500
dissonance and its effects
on finance.

59:55.500 --> 1:00:00.620
So, Will Goetzmann, who is a
professor here at the Yale

1:00:00.620 --> 1:00:05.390
School of Management, and a
couple of his co-authors found

1:00:05.390 --> 1:00:09.040
that mutual fund investors --

1:00:09.040 --> 1:00:12.140
when a mutual fund does very
badly in its investment

1:00:12.140 --> 1:00:16.960
performance, many or most
investors sell the stock and

1:00:16.960 --> 1:00:17.660
get out of it.

1:00:17.660 --> 1:00:21.060
But some of them hang on.

1:00:21.060 --> 1:00:23.750
And they thought that that was
due, perhaps, to cognitive

1:00:23.750 --> 1:00:24.290
dissonance.

1:00:24.290 --> 1:00:31.470
Because I bought this fund,
I don't want to sell

1:00:31.470 --> 1:00:35.430
because I was right.

1:00:35.430 --> 1:00:38.190
So, what they did is, they
interviewed these people and

1:00:38.190 --> 1:00:42.080
they found that these people
didn't even know how badly the

1:00:42.080 --> 1:00:42.800
fund has done.

1:00:42.795 --> 1:00:46.235
They had blocked it out and
they had an exaggerated

1:00:46.240 --> 1:00:47.490
impression.

1:00:52.135 --> 1:00:55.075
An exaggerated impression of
this, which is characteristic

1:00:55.080 --> 1:00:56.380
of cognitive dissonance.

1:00:56.380 --> 1:01:00.930
You just forget the evidence
that's contrary to your theory

1:01:00.930 --> 1:01:02.890
and you keep assembling
evidence that

1:01:02.890 --> 1:01:05.150
supports your theory.

1:01:05.150 --> 1:01:09.970
I have another example of
cognitive dissonance and this

1:01:09.970 --> 1:01:17.290
one was produced by Professor
Sendhil Mullainathan at the

1:01:17.290 --> 1:01:20.940
Harvard Economics Department.

1:01:20.940 --> 1:01:26.120
And Mullainathan looked at
financial advisors --

1:01:26.120 --> 1:01:27.370
and his co-authors.

1:01:29.410 --> 1:01:32.010
What they did in this study --

1:01:32.010 --> 1:01:33.590
that's a whole big profession.

1:01:33.590 --> 1:01:35.750
I remember at the beginning I
pointed out how many hundreds

1:01:35.750 --> 1:01:38.380
of thousands of financial
advisors there are.

1:01:38.380 --> 1:01:40.640
What they did, it's an
interesting experiment.

1:01:40.640 --> 1:01:45.290
They hired actors to go to
financial advisors and ask for

1:01:45.290 --> 1:01:46.780
their help.

1:01:46.780 --> 1:01:50.370
And the experiment was
the following.

1:01:50.370 --> 1:01:55.650
They would say the same thing to
each financial advisor, but

1:01:55.650 --> 1:01:58.600
the different actors would
present their existing

1:01:58.600 --> 1:02:00.370
portfolio differently.

1:02:00.370 --> 1:02:03.640
In other words, you'd go to the
advisor and you'd say, I

1:02:03.640 --> 1:02:08.470
have a portfolio of investments
and I'm almost

1:02:08.470 --> 1:02:11.050
entirely in money
market funds.

1:02:11.050 --> 1:02:11.450
That's all.

1:02:11.450 --> 1:02:12.130
You'd just say that.

1:02:12.130 --> 1:02:14.620
You wouldn't express
any opinion at all.

1:02:14.620 --> 1:02:17.960
Another actor would go and
say, I've got all of my

1:02:17.960 --> 1:02:21.460
portfolio in tech stocks.

1:02:21.460 --> 1:02:24.720
Or I've got all of my portfolio
in options.

1:02:24.720 --> 1:02:27.190
Now, what should advisors
tell people?

1:02:27.190 --> 1:02:30.700
Well, if they were acting really
professionally, they

1:02:30.700 --> 1:02:33.910
should question the assumptions
that made the

1:02:33.910 --> 1:02:36.950
actor supposedly put all their
money in one kind of

1:02:36.950 --> 1:02:39.300
investment.

1:02:39.300 --> 1:02:42.300
And many of the financial
advisors did, but

1:02:42.300 --> 1:02:44.920
usually they didn't.

1:02:44.920 --> 1:02:49.650
They didn't question
the actor.

1:02:49.650 --> 1:02:54.660
They assumed that the actor, who
had put, supposedly, all

1:02:54.660 --> 1:02:58.800
of the investments in money
market funds, was someone who

1:02:58.800 --> 1:03:04.640
was very risk averse, or thought
that was the right

1:03:04.640 --> 1:03:05.410
thing to do.

1:03:05.410 --> 1:03:08.220
And they didn't want to
challenge them, so they would

1:03:08.220 --> 1:03:10.740
walk out of there with maybe
a slightly different mix of

1:03:10.740 --> 1:03:12.170
money market funds.

1:03:12.170 --> 1:03:15.140
And somebody else who was in a
very risky portfolio, they

1:03:15.140 --> 1:03:17.340
didn't challenge them.

1:03:17.340 --> 1:03:22.140
And they even sent actors in
with almost all of their

1:03:22.140 --> 1:03:25.680
portfolio in their own
company's stock.

1:03:25.680 --> 1:03:28.350
Now if you work for Ford
Motor Company --

1:03:28.350 --> 1:03:30.100
I noticed my uncle --

1:03:30.100 --> 1:03:32.200
I had conversations with
him about this --

1:03:32.200 --> 1:03:35.420
who worked for Ford Motor
Company and put all of his

1:03:35.420 --> 1:03:38.250
life savings in Ford
Motor Company.

1:03:38.250 --> 1:03:41.740
I said, Uncle Ralph, you
shouldn't do that.

1:03:41.740 --> 1:03:44.930
Because it's your job and all
of your life savings.

1:03:44.930 --> 1:03:47.180
What if something happened
to Ford Motor Company?

1:03:47.180 --> 1:03:49.730
Fortunately, he didn't
work for GM, which

1:03:49.730 --> 1:03:51.680
became worthless recently.

1:03:51.680 --> 1:03:52.660
But it can happen.

1:03:52.660 --> 1:03:55.150
You know, Ford could be
completely wiped out.

1:03:55.150 --> 1:03:56.890
That's your life savings.

1:03:56.890 --> 1:03:59.500
And you know what
he said to me?

1:03:59.500 --> 1:04:02.690
He said, you know, I've worked
at Ford all my life.

1:04:02.690 --> 1:04:04.830
They treat me well, I
believe in them, I'm

1:04:04.830 --> 1:04:07.600
not going to sell.

1:04:07.600 --> 1:04:09.130
So, there's lots of
people like that.

1:04:09.130 --> 1:04:11.780
So, when they show up at a
financial advisor, the first

1:04:11.780 --> 1:04:15.100
thing that they should do, the
financial advisor should tell

1:04:15.100 --> 1:04:18.580
them, get out of your
Ford stock.

1:04:18.580 --> 1:04:20.250
That's just too risky.

1:04:20.250 --> 1:04:24.700
But only 40% of the financial
advisors did that.

1:04:24.695 --> 1:04:29.715
The 60% left them mostly in
their own company stock.

1:04:29.720 --> 1:04:31.000
Why did they do that?

1:04:31.000 --> 1:04:33.890
Well, Mullainathan thought, it's
because the advisors know

1:04:33.890 --> 1:04:37.220
there's cognitive dissonance,
and they're afraid to drive

1:04:37.220 --> 1:04:39.070
away a new client.

1:04:39.070 --> 1:04:42.120
Maybe they'll gradually do it
over a while, but you just

1:04:42.120 --> 1:04:44.640
don't challenge their
deep beliefs,

1:04:44.640 --> 1:04:46.580
whatever they say is true.

1:04:46.580 --> 1:04:48.430
And so, they're kind
of yes-men.

1:04:48.430 --> 1:04:52.000
Not all of them, and maybe
they'll come around.

1:04:52.000 --> 1:04:54.950
It relates then, again,
to a moral dilemma.

1:04:54.950 --> 1:04:59.520
If you are a financial advisor
working in private practice,

1:04:59.520 --> 1:05:03.940
what do you do with people who
come to you, if you know from

1:05:03.940 --> 1:05:07.330
experience that challenging
their deep-seated beliefs will

1:05:07.330 --> 1:05:08.990
drive them away?

1:05:08.990 --> 1:05:11.320
So, in the real world,
this is again --

1:05:11.320 --> 1:05:13.630
I'm not sure that these
financial advisors are doing

1:05:13.630 --> 1:05:15.630
the wrong thing.

1:05:15.630 --> 1:05:19.430
If they would eventually tilt
them toward a more responsible

1:05:19.430 --> 1:05:21.920
portfolio, they can't
drive them away.

1:05:32.640 --> 1:05:34.330
I have a lot of -- so many.

1:05:34.330 --> 1:05:37.900
Let me list some of the
others and move on.

1:05:37.900 --> 1:05:39.710
What else should I talk about?

1:05:39.710 --> 1:05:40.960
Anchoring.

1:05:50.850 --> 1:05:59.000
Anchoring refers to a tendency
to anchor your opinions on

1:05:59.000 --> 1:06:02.330
something that captures
your attention.

1:06:02.330 --> 1:06:05.730
The famous anchoring experiment
by, it was again,

1:06:05.730 --> 1:06:06.980
Kahneman and Tversky --

1:06:09.730 --> 1:06:12.020
I could almost do this
experiment here in class with

1:06:12.015 --> 1:06:13.045
you if I had a wheel
of fortune.

1:06:13.045 --> 1:06:15.795
A wheel of fortune is
like on a game show.

1:06:15.800 --> 1:06:19.360
You spin the wheel and it comes
up with a number between

1:06:19.360 --> 1:06:23.540
zero and a hundred in this
particular wheel of fortune.

1:06:23.540 --> 1:06:25.140
So, this is the experiment.

1:06:25.140 --> 1:06:28.720
They asked their subjects,
how many people --

1:06:28.720 --> 1:06:30.580
it had to be something
that had an answer

1:06:30.580 --> 1:06:31.650
from zero to a hundred.

1:06:31.650 --> 1:06:36.470
One of their questions was, how
many nations in the world

1:06:36.470 --> 1:06:36.650
[addition: in percent]

1:06:36.650 --> 1:06:39.680
belong to the United Nations?

1:06:39.680 --> 1:06:40.610
So, they asked the question.

1:06:40.605 --> 1:06:42.415
They said, don't answer
me just yet.

1:06:42.420 --> 1:06:44.010
Think about that question.

1:06:44.010 --> 1:06:47.080
What percent of nations
in the world belong

1:06:47.080 --> 1:06:48.080
to the United Nations.

1:06:48.080 --> 1:06:50.750
Then, they spun the wheel, and
it came up and it showed a

1:06:50.750 --> 1:06:52.840
random number.

1:06:52.840 --> 1:06:54.830
And then, they asked people
for the answer.

1:06:54.830 --> 1:06:57.320
Well, it turns out that people
tended to give an answer close

1:06:57.320 --> 1:07:00.170
to the number that just
came up on the wheel.

1:07:00.170 --> 1:07:03.560
This is totally irrational,
right?

1:07:03.560 --> 1:07:05.360
That wheel has nothing to
do with the answer.

1:07:05.360 --> 1:07:07.620
And yet people were
influenced by it.

1:07:07.620 --> 1:07:10.500
So then, they would follow-up
and ask them, hey, that number

1:07:10.500 --> 1:07:12.400
you gave is the same as the one
that just came up on the

1:07:12.401 --> 1:07:14.271
wheel, or it's close to it.

1:07:14.270 --> 1:07:15.470
Why did you do that?

1:07:15.470 --> 1:07:17.360
The guy would say,
just coincidence.

1:07:17.360 --> 1:07:18.630
I wasn't influenced
by the wheel.

1:07:18.630 --> 1:07:19.960
Of course not.

1:07:19.960 --> 1:07:22.540
But you know they were, because
statistically they

1:07:22.540 --> 1:07:24.110
proved that they were.

1:07:24.110 --> 1:07:29.960
So, anchoring means that people
are attracted to --

1:07:29.960 --> 1:07:36.630
they're affected by subconscious
things.

1:07:36.630 --> 1:07:38.300
I shouldn't say subconscious.

1:07:38.300 --> 1:07:40.590
They didn't make a logical
connection.

1:07:40.590 --> 1:07:43.380
When you face real ambiguity and
you don't know the answer

1:07:43.375 --> 1:07:46.335
and you've got to come up with
a decision, you are swayed by

1:07:46.340 --> 1:07:49.320
the most silly and
random things.

1:07:49.320 --> 1:07:50.950
There's a representativeness
heuristic.

1:07:58.580 --> 1:07:59.930
This is also Kahneman
and Tversky.

1:08:02.440 --> 1:08:05.960
And that is, that people
overemphasize certain patterns

1:08:05.960 --> 1:08:08.720
that they think are
representative of what they've

1:08:08.720 --> 1:08:10.380
seen before.

1:08:10.380 --> 1:08:14.250
So for example, certain patterns
in the stock market

1:08:14.250 --> 1:08:18.280
that may be very rare
and unusual.

1:08:18.279 --> 1:08:20.749
If they remember it, if it
somehow attracted their

1:08:20.750 --> 1:08:23.080
attention, they begin
to look for that

1:08:23.080 --> 1:08:25.710
pattern again and again.

1:08:25.710 --> 1:08:29.090
And they see it too often.

1:08:29.090 --> 1:08:31.290
So for example, Head
and Shoulders, we

1:08:31.290 --> 1:08:32.650
talked about that.

1:08:32.649 --> 1:08:37.339
McGee, the technical analyst, he
saw the Head and Shoulders

1:08:37.340 --> 1:08:38.780
pattern in the stock market.

1:08:38.779 --> 1:08:40.939
And he saw that it crashed
after that.

1:08:40.939 --> 1:08:42.889
But actually, it's pretty
hard to find those,

1:08:42.890 --> 1:08:44.050
they're kind of rare.

1:08:44.050 --> 1:08:47.250
And it's not the right way to
process data, to be looking

1:08:47.250 --> 1:08:49.020
for patterns that are
representative.

1:08:49.020 --> 1:08:51.140
And it invites manipulation.

1:08:51.140 --> 1:08:53.230
So, I'll give you some
other bad behavior.

1:08:53.229 --> 1:08:56.239
If people believe in the Head
and Shoulders, if they believe

1:08:56.240 --> 1:08:59.150
that the Head and Shoulders
pattern of stock price

1:08:59.149 --> 1:09:03.079
movements predicts a decline,
here's what I can do.

1:09:03.080 --> 1:09:05.340
I'll take some thinly
traded stock.

1:09:05.340 --> 1:09:06.670
I'll get a friend.

1:09:06.670 --> 1:09:09.840
We'll trade back and forth and
we'll influence the price to

1:09:09.840 --> 1:09:11.490
create -- we'll deliberately
create a Head

1:09:11.490 --> 1:09:14.000
and Shoulders pattern.

1:09:14.000 --> 1:09:17.350
And then we'll short the stock
massively, right at the time

1:09:17.350 --> 1:09:19.060
when the head and shoulders
pattern

1:09:19.060 --> 1:09:21.760
would give a sell signal.

1:09:21.760 --> 1:09:24.710
We can make tons of
money doing that.

1:09:24.710 --> 1:09:26.840
So, why don't we do that?

1:09:26.840 --> 1:09:29.330
Well, we don't because
it's a manipulation.

1:09:29.330 --> 1:09:42.690
[SIDE CONVERSATION]

1:09:42.689 --> 1:09:44.309
PROFESSOR ROBERT SHILLER: I
want to then just conclude

1:09:44.310 --> 1:09:48.090
with social contagion, because
it's so important.

1:09:48.090 --> 1:09:49.340
This is my last --

1:09:51.470 --> 1:09:57.630
and this is really social
psychology.

1:09:57.625 --> 1:09:59.925
I'm running out of room here.

1:09:59.930 --> 1:10:01.920
That's social contagion.

1:10:01.920 --> 1:10:06.470
Social psychology reflects on
the fact that people are

1:10:06.470 --> 1:10:09.980
interdependent, and what I
think is affected by what

1:10:09.980 --> 1:10:11.230
others think.

1:10:15.340 --> 1:10:18.360
There's something called
herd behavior.

1:10:18.360 --> 1:10:19.870
That's a popular term.

1:10:19.870 --> 1:10:25.750
It refers to the tendency for
people to move with the herd,

1:10:25.750 --> 1:10:26.720
not consciously.

1:10:26.720 --> 1:10:29.340
They don't think that they're
moving with the herd.

1:10:29.340 --> 1:10:31.990
I might bring up a little
sociology here

1:10:31.990 --> 1:10:34.210
and I'll use a term.

1:10:34.210 --> 1:10:37.490
Everything has been psychology,
but the great

1:10:37.490 --> 1:10:40.790
sociologist, one of the founders
of the discipline of

1:10:40.790 --> 1:10:50.990
sociology, was the French
scholar Emile Durkheim at the

1:10:50.990 --> 1:10:53.140
late 19th, early 20th century.

1:10:53.140 --> 1:11:03.820
And he used the word
''collective consciousness.''

1:11:03.820 --> 1:11:07.450
And that is, that our opinions
about what's happening are

1:11:07.450 --> 1:11:12.180
formed by a collective
understanding of

1:11:12.180 --> 1:11:14.940
what's going on.

1:11:14.940 --> 1:11:20.980
We have a tendency to think of
ourselves as rational and

1:11:20.980 --> 1:11:21.860
common sense --

1:11:21.860 --> 1:11:24.500
all of our views come
from common sense

1:11:24.500 --> 1:11:27.620
processing of facts.

1:11:27.620 --> 1:11:31.140
I have a sense of belief about
what goes on in the world, but

1:11:31.140 --> 1:11:36.700
I underappreciate the extent to
which my views are a little

1:11:36.700 --> 1:11:40.790
bit arbitrary and shared by
millions of other people.

1:11:40.790 --> 1:11:44.710
You live in a certain point in
time in history, and there are

1:11:44.710 --> 1:11:48.450
certain kinds of facts and ideas
and anecdotes that are

1:11:48.450 --> 1:11:50.700
circulating.

1:11:50.700 --> 1:11:56.680
There's another term, this is
a zeitgeist. That's German,

1:11:56.680 --> 1:11:58.500
but it's now English.

1:11:58.500 --> 1:12:01.890
That means "spirit
of the times."

1:12:01.890 --> 1:12:06.020
So, what Durkheim and other
sociologists allowed us to

1:12:06.020 --> 1:12:10.510
understand is, that the
zeitgeist is determined by a

1:12:10.510 --> 1:12:13.310
collective memory, a collective
set of facts that

1:12:13.310 --> 1:12:14.700
we operate on.

1:12:14.700 --> 1:12:19.380
This herd behavior creates big
swings in the stock market and

1:12:19.380 --> 1:12:21.590
other things.

1:12:21.590 --> 1:12:25.050
So, it has a huge financial
impact.

1:12:25.050 --> 1:12:29.190
But anyway, I've listed a number
of Behavioral Finance

1:12:29.190 --> 1:12:31.800
principles that really come
from psychology, and I've

1:12:31.800 --> 1:12:36.370
talked about some tests or
examples of their proof of

1:12:36.370 --> 1:12:40.450
their importance in finance.

1:12:40.450 --> 1:12:42.770
But what do we conclude
from this?

1:12:42.770 --> 1:12:51.930
I think that my conclusion is,
that we are evolving toward

1:12:51.930 --> 1:12:56.620
better and better financial
institutions.

1:12:56.620 --> 1:13:01.150
There is a lot of manipulation
and exploitation, but we as a

1:13:01.150 --> 1:13:03.670
society have outlawed
a lot of it.

1:13:03.670 --> 1:13:08.340
For example, I mentioned doing
a stock market manipulation

1:13:08.340 --> 1:13:11.140
trick to create a Head and
Shoulders pattern.

1:13:11.140 --> 1:13:12.090
That is an offense.

1:13:12.085 --> 1:13:14.585
It will get you in jail
for doing that.

1:13:14.590 --> 1:13:17.100
And we prosecute that now.

1:13:17.100 --> 1:13:19.550
So, you can't do that.

1:13:19.550 --> 1:13:22.160
I'm going to talk more about
this in the next lecture about

1:13:22.160 --> 1:13:23.380
regulation.

1:13:23.380 --> 1:13:26.710
But it's also people's moral
judgments that the people who

1:13:26.710 --> 1:13:30.270
evolve to become important in
finance are people who have an

1:13:30.270 --> 1:13:35.710
internal compass, a desire for
praise-worthiness that

1:13:35.710 --> 1:13:37.510
eliminates --

1:13:37.510 --> 1:13:42.400
I'm going to give just two
examples of some recent

1:13:42.400 --> 1:13:45.250
articles about this.

1:13:45.250 --> 1:13:51.240
In the current issue of the
Harvard Business Review that,

1:13:51.240 --> 1:13:55.170
I assume, is still
on newsstands --

1:13:55.170 --> 1:13:58.210
This is Harvard Business
Review.

1:13:58.210 --> 1:14:05.530
There's an article by Michael
Porter and co-author Mark

1:14:05.530 --> 1:14:11.510
Kramer, Porter is a well-known
professor at Harvard, in which

1:14:11.510 --> 1:14:17.000
he argues that we're coming to
realize more and more about a

1:14:17.000 --> 1:14:21.080
principle called -- he calls it
''shared value.'' Or they

1:14:21.080 --> 1:14:29.390
call it ''shared value.'' And
that is that the manager of a

1:14:29.390 --> 1:14:38.660
company shouldn't be
underestimating the importance

1:14:38.660 --> 1:14:45.520
of shared value creation with
society, with other people.

1:14:45.520 --> 1:14:50.430
That is, we're all in this
together, and if we're mature,

1:14:50.430 --> 1:14:53.760
we recognize, for example,
that I don't want to be

1:14:53.760 --> 1:14:55.160
exploitative.

1:14:55.160 --> 1:14:59.890
I don't want to make the local
people in my town upset with

1:14:59.890 --> 1:15:00.800
our company.

1:15:00.800 --> 1:15:05.040
I don't want our labor force
to become disenchanted.

1:15:05.040 --> 1:15:08.820
Now, what he's saying it's not
really about morals exactly.

1:15:08.820 --> 1:15:11.110
It's more about long-term
value.

1:15:11.110 --> 1:15:15.410
But I guess morals somehow
creeps into the same judgment.

1:15:15.410 --> 1:15:20.670
That mature businesspeople see
shared value and that there

1:15:20.670 --> 1:15:24.610
was maybe not enough
emphasis on that.

1:15:24.610 --> 1:15:28.130
Financial theory was leading
us too much toward thinking

1:15:28.130 --> 1:15:32.930
that a manager should be
selfishly pursuing a narrow

1:15:32.930 --> 1:15:36.030
focus, like maximizing
the short-run

1:15:36.030 --> 1:15:38.610
value of their shares.

1:15:38.610 --> 1:15:43.130
Anyway, the other example I
have, which is also recent,

1:15:43.130 --> 1:15:46.890
not quite as recent as that, is
a book that came out last

1:15:46.890 --> 1:15:52.250
year by Anna Bernasek,
who is actually a

1:15:52.250 --> 1:15:56.920
journalist, not an academic.

1:15:56.920 --> 1:16:01.000
But it's called Economics
of Integrity.

1:16:01.000 --> 1:16:02.250
Is that the title exactly?

1:16:05.270 --> 1:16:06.830
Yes.

1:16:06.830 --> 1:16:08.120
Economics of Integrity.

1:16:12.880 --> 1:16:19.200
And her point in that book is,
that a sense of personal

1:16:19.200 --> 1:16:23.670
integrity has dominated what
people do in the business

1:16:23.670 --> 1:16:28.460
world much more than we
thought recently.

1:16:28.460 --> 1:16:33.290
There has been too much
disregard of the fact that

1:16:33.290 --> 1:16:36.140
people do things because
they're right.

1:16:36.140 --> 1:16:39.050
She gives an example in the book
-- and I'll close with

1:16:39.050 --> 1:16:40.210
this concept.

1:16:40.210 --> 1:16:44.650
She said, let's consider milk.

1:16:44.650 --> 1:16:46.700
Now, you drink milk
regularly, I hope.

1:16:46.695 --> 1:16:48.175
It's good for you.

1:16:48.180 --> 1:16:51.020
But it could poison you.

1:16:51.020 --> 1:16:54.890
People used to get sick from
drinking contaminated milk.

1:16:54.890 --> 1:16:56.930
And you don't ever hear of
anyone getting sick.

1:16:56.930 --> 1:16:58.850
So she said, why is that?

1:16:58.850 --> 1:17:02.590
Well, we have government
regulation of milk production

1:17:02.590 --> 1:17:06.850
and there are laws about
purity of milk.

1:17:06.850 --> 1:17:09.500
But she looked into it
and found that --

1:17:09.500 --> 1:17:13.430
actually, she didn't think it
was mostly the regulation.

1:17:13.430 --> 1:17:18.280
She thought that you are safe
drinking milk because of the

1:17:18.280 --> 1:17:21.330
integrity of our
people, mostly.

1:17:21.330 --> 1:17:24.510
That if you go out to some milk
company and talk to the

1:17:24.510 --> 1:17:28.130
employees, they might not
generally even know about the

1:17:28.130 --> 1:17:29.340
regulations.

1:17:29.340 --> 1:17:31.740
But if you ask them,
they'll tell you --

1:17:31.740 --> 1:17:34.410
I mean, are you careful to
keep this milk clean?

1:17:34.410 --> 1:17:37.050
They'll tell you, well,
someone's going to drink this,

1:17:37.050 --> 1:17:40.550
so obviously it's common sense
I'll keep it clean.

1:17:40.550 --> 1:17:43.890
And what she says, it's not
primarily the regulation, it's

1:17:43.890 --> 1:17:47.820
the integrity of the people that
makes the economic system

1:17:47.820 --> 1:17:49.210
work as well as it is.

1:17:49.210 --> 1:17:51.280
So, anyway, I've emphasized
both sides.

1:17:51.280 --> 1:17:54.170
I've talked about human failings
and about people

1:17:54.170 --> 1:17:57.180
exploiting these failings, about
people with antisocial

1:17:57.180 --> 1:17:58.860
personalities.

1:17:58.860 --> 1:18:02.070
But we have a system that
somehow eliminates this from

1:18:02.070 --> 1:18:04.930
being the major factor
in our markets.