00:00:01
I'd like to welcome Daniel Taylor to
00:00:04
knowledge at Wharton he's a professor of
00:00:06
accounting here and he's written a very
00:00:09
interesting paper on the connection
00:00:12
between political connections and and
00:00:15
insider trading during or just after the
00:00:19
financial crisis and during the period
00:00:22
where tarp was under consideration and
00:00:24
tarp is the Troubled Asset Relief
00:00:27
Program that's right and so I'm going to
00:00:31
let you give us a short summary first
00:00:35
and then we'll get into more specific
00:00:36
questions about what your paper is about
00:00:39
what it found okay so what we what we
00:00:41
did in the paper is we look at the
00:00:44
trading of corporate insiders so by
00:00:47
corporate insiders we mean officers and
00:00:50
directors of publicly traded
00:00:52
corporations all of these individuals
00:00:55
have to file what's known as form fours
00:00:58
with the SEC in the public record that
00:01:00
discloses their trades in their firms
00:01:05
shares or their firm stock so there are
00:01:07
allowed to trade in their firm shares
00:01:08
they file that information with the SEC
00:01:10
so what we did is we gathered all of the
00:01:14
information that's out there on these on
00:01:16
these form fours on their their trading
00:01:18
and we looked at the trading before the
00:01:21
financial crisis during the financial
00:01:22
crisis and after the financial crisis so
00:01:24
this is the trading of executives and
00:01:27
directors in their own firm and we look
00:01:30
specifically at financial institutions
00:01:32
and so one of the big questions that
00:01:34
comes out of the financial crisis that's
00:01:36
interested you know and Wall Street and
00:01:38
on Main Street is did anyone know you
00:01:41
know did anyone see the crisis coming
00:01:42
and so we initially started the project
00:01:44
by looking at okay can we see whether
00:01:48
insiders corporate insiders officers and
00:01:51
directors traded in advance of the
00:01:53
crisis so did banks that did poorly
00:01:57
during the crisis to their executive
00:01:59
sell shares before you know the crisis
00:02:01
hit and then we looked at the trading
00:02:03
during the crisis so around the bank
00:02:06
bailouts the tarp monies that you
00:02:08
alluded to and then sort of after the
00:02:11
crisis what what what did it stabilized
00:02:13
and we found that there was
00:02:15
there didn't seem to really be any
00:02:16
evidence of trading before the crisis
00:02:19
there was no evidence that insiders
00:02:21
traded in anticipation of the crisis but
00:02:23
we did seem to find some evidence that
00:02:25
insiders traded during the crisis in the
00:02:28
period in which the tarp funds were
00:02:30
dispersed and then we investigate that a
00:02:33
little bit more and what we find is
00:02:36
interestingly enough that it's only the
00:02:39
insiders that had political connections
00:02:42
and by political connections I mean
00:02:44
connected to a bank regulatory agency or
00:02:49
the current House or Senate at the time
00:02:52
so we looked at bank boards who had a
00:02:56
director or officer who had work
00:02:58
experience current or past at a bank
00:03:01
regulatory agency the senator the house
00:03:03
and we found that the boards of those
00:03:05
banks that had those political
00:03:07
connections traded more heavily during
00:03:10
the financial crisis their trades had
00:03:13
higher predictive ability of outcomes
00:03:16
during the financial crisis so they
00:03:18
predict for example the market reaction
00:03:20
and the amount of the bailout that the
00:03:21
bank would receive so we basically find
00:03:25
that during the crisis there is some
00:03:27
trading and anticipation of a bank of
00:03:29
bank bailouts and then after the crisis
00:03:32
of course we sort of things go back to
00:03:34
normal and we don't really find that the
00:03:36
day that those trades or those
00:03:38
individuals trades have any more
00:03:40
information than otherwise so this would
00:03:42
seem to be the definition of insider
00:03:44
trading on the face of it so I wanted to
00:03:48
just point out because I think it's very
00:03:50
interesting that this study had a very
00:03:52
large sample right looked at 7300
00:03:55
corporate officers is that right just
00:03:58
fit across 497 publicly traded tarp
00:04:02
eligible institution so this Greta says
00:04:05
this was a very wide net work that you
00:04:07
cast right I mean one way to think about
00:04:08
this paper is to think about it as using
00:04:12
big data and computer algorithms to sift
00:04:15
through public information on the trades
00:04:19
of officers and directors and to see
00:04:21
which trades of the officers and
00:04:24
directors are correlated with future
00:04:26
outcomes and so we would say it
00:04:29
trade is correlated with the future
00:04:31
outcome if it has a high predictive
00:04:32
ability it's more likely that that trade
00:04:35
might have been based on on on
00:04:39
information about that future outcome
00:04:41
but I do want to i do want to clarify
00:04:43
that you know in the popular press and
00:04:46
legal scholars they use the term insider
00:04:48
trading they're thinking of illegal
00:04:50
insider trading when we use the term
00:04:52
insider trading we're thinking of the
00:04:54
trading by officers and directors so we
00:04:57
say insider trading we mean trading by
00:04:59
corporate insiders and there's two types
00:05:01
the first type is the legal trading so
00:05:05
this is they can you know the the board
00:05:07
the CEO can trade in his shares you know
00:05:09
just like anyone else can with proper
00:05:11
disclosure with proper disclosure you
00:05:13
know as long as they do not have any
00:05:14
private information the illegal kind is
00:05:17
when they have private information that
00:05:20
they have not disclosed to shareholders
00:05:23
so this would this sort of relates back
00:05:26
to a set of rules called discloser
00:05:28
abstain so if you're a CEO or a manager
00:05:31
and you have private information it's
00:05:33
your duty to either disclose to
00:05:35
shareholders or abstain from trading and
00:05:38
so in this setting it's a little bit
00:05:40
more murky because when we're looking at
00:05:44
political connections and connections to
00:05:46
bank regulators you know it's not really
00:05:49
clear how the manager would disclose any
00:05:52
information that they would have gotten
00:05:54
from their connections so for example
00:05:56
because it's not definite information
00:05:58
right so you know that's something looks
00:06:02
someone is telling a bank insider it
00:06:05
looks likely let this is going to happen
00:06:06
so legally he can't he doesn't really
00:06:09
have an obligation to disclose something
00:06:11
that's not really certain is that right
00:06:12
that's correct that's correct and so I
00:06:13
think you know there's been a spate of
00:06:15
recent insider trading cases that have
00:06:17
gone on that the government is
00:06:19
prosecuted and you know out of those
00:06:21
insider trading cases the definition of
00:06:23
illegal insider trading has become more
00:06:26
murky slippery yes more slippery and so
00:06:29
I you know I don't think it would be
00:06:31
wrong to suggest that this paper has
00:06:32
sort of a smoking gun okay what we're
00:06:35
doing is we're reporting correlations
00:06:37
when as you mentioned a large sample and
00:06:39
casting sort of a suspicious I in the
00:06:42
direction
00:06:42
of insiders that have political
00:06:46
connections I mean if you if you go back
00:06:48
to all of the books and the popular
00:06:51
press during the period you know the
00:06:53
bailouts were effectively decided in you
00:06:55
know backroom meetings with sort of
00:06:58
insiders there's a great book you know
00:07:01
too big to fail yes by Sorkin it's sort
00:07:03
of details sort of all of these you know
00:07:05
his interviews with people and so a
00:07:07
natural question is is well you know if
00:07:09
if the bailouts being decided in private
00:07:11
and in consultation with you know
00:07:13
bankers or you know Treasury officials
00:07:16
or government officials how does that
00:07:17
information sort of leak out so this I
00:07:22
think that it's also interesting to note
00:07:25
that this was not the result you were
00:07:27
expecting when you started out on this
00:07:30
path could you just talk about that
00:07:32
briefly right this surprised you yes but
00:07:36
is all surprised yes absolutely and
00:07:37
frankly we're still kind of stunned at
00:07:39
the at the results and so when we
00:07:42
started the project several years ago
00:07:44
you know as I mentioned it was sort of
00:07:46
like a big data study you know we sat
00:07:49
down the co-authors and I and you know
00:07:50
talked about you know okay does this you
00:07:52
know this would be interesting does it
00:07:53
make sense because there's all of this
00:07:55
concern about did they anticipate the
00:07:57
crisis so we began the paper thinking
00:07:59
okay we would shed some evidence on
00:08:00
whether they anticipated the crisis or
00:08:03
not and so five years ago that would
00:08:05
have been sort of a very timely you know
00:08:08
finding they did anticipate the crisis
00:08:11
or they didn't anticipate the crisis but
00:08:14
it took us a little longer than we
00:08:15
expected to gather the data and to do
00:08:17
the analysis but the analysis was you
00:08:19
know a couple months of you know you
00:08:22
know coding and get gathering data and
00:08:25
and so when we started doing the
00:08:28
analysis and we realized okay we're
00:08:29
probably after the crisis let's add some
00:08:31
more data let's add the post-crisis the
00:08:33
bailout period data let's add you know
00:08:36
the data from you know 2010-2011 into it
00:08:39
and run the analysis we kind of we found
00:08:42
somewhat surprisingly that there was
00:08:45
this you know blip for lack of a better
00:08:47
term in the correlation between insider
00:08:51
trades and future performance during the
00:08:54
what we call the bailout period and
00:08:55
that's the
00:08:56
from when tarp began dispersed in
00:08:59
october through june and so when we saw
00:09:04
the blip the next question is okay where
00:09:07
is the blip coming from why is the blip
00:09:09
there and so then we found that the blip
00:09:10
only occurred in banks that actually
00:09:13
received tarpon money okay so then we
00:09:16
could you know think about like a you
00:09:18
know an investigative journalist dig
00:09:19
deeper okay so now the blip is only in
00:09:21
those banks that receives tarp money ok
00:09:24
so then dig deeper look at the
00:09:25
connections of the individuals or of the
00:09:28
bank okay well now the blip is not only
00:09:30
and banks that receives tarp money it is
00:09:32
also only within that set also those
00:09:35
that received tarp money and had ties to
00:09:38
former or current bank regulatory bank
00:09:43
regulatory agencies like the Fed the
00:09:45
Treasury the office of the comptroller
00:09:46
of the currency the FDIC and so that's
00:09:48
when we said you know like wait you'll
00:09:51
wait a second you know is this actually
00:09:52
you know what were is is this actually
00:09:55
where the blip is and so you know we're
00:09:59
my co-authors and I are actually pretty
00:10:01
suspicious and skeptical people by
00:10:03
nature so in terms of data analysis we
00:10:07
threw everything we could add it to make
00:10:09
it go away and so I I wouldn't be here
00:10:12
if it went away and so this isn't a case
00:10:15
where you know we're trying to you know
00:10:19
can some suspicion on something that may
00:10:21
or may not go away I mean this is a very
00:10:23
very robust result i also want to
00:10:26
mention that the data on political
00:10:28
connections it was not like we collected
00:10:32
this data or that we generated this data
00:10:34
everything in the paper is generated
00:10:37
from public available information so
00:10:40
it's perfectly conceivable that somebody
00:10:42
could go out there and you know
00:10:43
investigate it themselves the data on
00:10:46
political connections right where to get
00:10:48
data on political connections well in
00:10:51
the bios of directors and officers in
00:10:54
the annual reports they list what their
00:10:56
current and former work experience was
00:10:58
so everything that's that's in the paper
00:11:00
sort of out there in the public domain
00:11:02
so you aren't using any insider
00:11:04
information no we were not using it off
00:11:09
what are some of the practical
00:11:11
implications of these findings so I
00:11:12
think I mean it's really interesting you
00:11:14
pointed out that the alleged criminality
00:11:19
of all this is is questionable because
00:11:20
the information they had wasn't
00:11:23
necessarily actionable in the way that
00:11:25
they had to disclose it but if someone
00:11:27
was whispering in their ear and it was
00:11:30
good information you know it seems a
00:11:32
little bit it seems sleazy right I have
00:11:37
to put it mildly right so what what
00:11:41
practical implications could come from
00:11:44
this I mean for example I mean this
00:11:46
speaks to the damage that can be done
00:11:49
with the revolving door of government
00:11:50
officials you know having come from
00:11:52
industry and then going right back into
00:11:54
industry perhaps without a you know a
00:11:56
proper amount of gap in time but I'll
00:12:00
leave it to you what what are some of
00:12:02
the practical I mean I think that there
00:12:04
are probably I would say two main
00:12:05
practical practical implications the
00:12:07
first is we need to give some more
00:12:10
thought to what exactly constitutes
00:12:13
illegal insider trading I think right
00:12:16
now in light of recent court cases
00:12:18
there's a big gray area out there and I
00:12:21
think a lot of people would be surprised
00:12:22
at how gray the area actually is so you
00:12:27
need a smoking gun to be able to prove
00:12:30
an insider trading case and it's very
00:12:31
hard to get that smoking gun you have to
00:12:33
get the tipster the tipster has to
00:12:35
confess you know there has to be a
00:12:38
series of dots that you can that you can
00:12:41
collect a tipster who may have benefited
00:12:43
some way from this to begin with it's
00:12:45
it's it's difficult to do that using a
00:12:47
large sample with correlations but this
00:12:49
suggests areas where you know people may
00:12:53
want to shine a light so think about is
00:12:56
this this insider trading is a dark room
00:12:58
we're shining a light you know it's not
00:13:00
entirely illuminated but you know we've
00:13:02
got a good area where you know people
00:13:04
should look the second one is like you
00:13:10
mentioned the revolving door there are a
00:13:11
series of studies in the academic
00:13:14
literature on political connections and
00:13:17
specifically political connections of
00:13:19
financials
00:13:20
solutions so for example there's a
00:13:21
recent study in the journal of financial
00:13:23
economics and they found that this was
00:13:27
before Tim Geithner was was made
00:13:29
Treasury secretary that when he was
00:13:33
announced to be the Treasury secretary
00:13:35
the banks that had connections to him
00:13:37
either on the board seats or what not
00:13:39
experienced old massive increases in
00:13:42
shareholder value so what this is saying
00:13:45
is is that the market is aware that
00:13:46
there is this connection between bank
00:13:51
regulators and the banks at which they
00:13:53
are regulating so for example many
00:13:56
people are surprised when I when I say
00:13:58
go to the go to the Federal Reserve's
00:13:59
website you know you can even look
00:14:01
during the crisis Lehman Brothers CEO
00:14:03
was you know one of the board of
00:14:05
directors of the Federal Reserve you
00:14:08
know Jamie Dimon was on that was on that
00:14:10
board at all first calls himself correct
00:14:13
was heading that backroom deal right
00:14:16
that you alluded to earlier so you've
00:14:18
got this marriage is too strong of a
00:14:20
word but you have these very strong
00:14:22
connections between the regulators and
00:14:25
the people that they are regulating now
00:14:27
that's good and bad on the one hand you
00:14:29
want somebody knowledgeable about the
00:14:31
industry and how banks worked to
00:14:34
regulate banks but where it you know
00:14:38
where it becomes questionable is are
00:14:40
their conflicts of interest and sort of
00:14:43
our paper suggests that yes there are
00:14:45
conflicts of interest and these
00:14:47
conflicts of interest are potentially
00:14:49
pervasive and I don't think that's
00:14:51
really been given enough sort of you no
00:14:55
consideration out there is that you know
00:14:57
there's a good side to having a banker
00:15:00
regulate the banks and there's also uh
00:15:02
you know there's also a bad side did you
00:15:05
attempt to quantify any of this like how
00:15:08
much some folks may have benefited well
00:15:13
so the benefits are tricky because what
00:15:15
we're looking at is what we call
00:15:17
opportunity profits so when an
00:15:19
individual trades if their share price
00:15:22
goes up let's say six percent over the
00:15:25
next same month then that means that
00:15:27
they've earned a six percent return on
00:15:29
their on their money if they sell but we
00:15:33
don't we can't match buys and sells so
00:15:37
we can't actually compute a measure of
00:15:39
realized profit all we can say is well
00:15:42
you bought this month stock price went
00:15:44
up six percent right that's six percent
00:15:46
more than you would have made right if
00:15:48
you had waited a month usually better
00:15:50
when it goes up yes no that's right and
00:15:52
so in this case we look at the trades in
00:15:55
advance of the tarp infusion so we know
00:15:58
the date of the tarp infusions we know
00:16:00
the date of the trade and so we look at
00:16:02
the trades and surprisingly we find a
00:16:06
significant amount of trades both both
00:16:08
in volume and dollars clustered 30 days
00:16:11
before the bank actually receives a tarp
00:16:14
infusion now I have done some some work
00:16:18
on insider trading policies in my prior
00:16:21
research what was surprising to us is
00:16:22
that most companies have something known
00:16:24
as trading blackout windows in a trading
00:16:27
blackout window is imposed by the
00:16:28
company in the code of ethics on the
00:16:31
officers and directors and it says you
00:16:33
shall not trade n days typically 20 30
00:16:36
40 days in advance of a material event
00:16:39
so for example for an earnings
00:16:41
announcement that firms make on a
00:16:42
quarterly basis insiders are prevented
00:16:45
from trading you know several days in
00:16:48
advance of that earnings announcement in
00:16:50
this case if there was a blackout window
00:16:53
applied to the tarp infusion those
00:16:56
trades should not have existed so what
00:16:59
this suggests is is that either the
00:17:01
event the tarp infusion was
00:17:02
unanticipated so a blackout window could
00:17:04
not apply or that they sort of skirted
00:17:09
the their firms no blackout windows just
00:17:13
one final thing I just wanted to focus
00:17:15
on the no smoking gun comment that you
00:17:18
made at the outset so this is seems like
00:17:23
pretty clear evidence that there was
00:17:24
insider trading to me as I create it
00:17:27
very strong evidence and yet this isn't
00:17:32
pointing to any individuals or
00:17:33
necessarily I mean you were sort of
00:17:35
doing this globally and the information
00:17:37
that you have while interesting isn't
00:17:40
the kind of information that someone
00:17:42
takes the court it's the kind of
00:17:44
information that
00:17:45
together would use to say here are the
00:17:49
most interesting places to look if I
00:17:52
were going to try to find a wrongdoer
00:17:54
right so I guess I would think about
00:17:56
this as sort of as an academic paper on
00:18:00
the costs and benefits to political
00:18:02
connections so most of the academic
00:18:04
literature has focused on the benefits
00:18:06
that the shareholders in the firm
00:18:08
receives from having political
00:18:10
connections on the board right so it may
00:18:12
very well be that you know you like as a
00:18:15
shareholder in the firm you want your
00:18:17
bank to be politically connected because
00:18:18
you're more likely to get bailed out you
00:18:20
know you're more likely to get special
00:18:21
favors and so most of the academic
00:18:23
literature suggests this is a good thing
00:18:25
for shareholders you know not
00:18:27
necessarily for the government for the
00:18:30
public but for a narrow set of
00:18:32
shareholders what we're coming along and
00:18:35
saying in the you know sort of the a
00:18:36
Couture the academic debate is there is
00:18:39
also a cost what is the cost well the
00:18:42
cost is if you have these political
00:18:44
connections and during the crisis they
00:18:47
may engage in some flack riveter term
00:18:49
shady trading that shady trading comes
00:18:53
at a cost to shareholders so who loses
00:18:55
out from the you know the insider
00:18:58
trading if the executives traded at with
00:19:01
private information its other
00:19:02
shareholders who potentially are on the
00:19:05
other side of that trade and they lose
00:19:07
out so this is really about sort of in
00:19:10
if you think about academics is okay
00:19:13
political connections have a have a
00:19:14
benefit to the firm and we're
00:19:16
documenting one cost which is sort of
00:19:19
extracting rents will be say is rent
00:19:21
extraction and also the potential cost
00:19:24
or the risk of being exposed and what
00:19:27
that could do right to the bank's stock
00:19:29
right no so i don't i don't want there
00:19:32
to be you know to one shouldn't ascribe
00:19:35
any sort of policy agenda or anything
00:19:38
like that to the paper or two to the
00:19:40
findings we're just basically saying
00:19:43
look there's this cost of of political
00:19:45
connections and you know and by the way
00:19:47
this is this is not a trivial cost this
00:19:50
seems to be a big deal especially in
00:19:52
light of all of the debate about tarp
00:19:54
how the bailout monies were doled out
00:19:56
the decisions about how of those things
00:19:58
were
00:19:59
were given out and so hopefully
00:20:01
something like this doesn't you know
00:20:03
doesn't happen again but it does suggest
00:20:05
that you know these conflicts of
00:20:07
interests do have real costs for for the
00:20:11
bank shareholders even if the bank gets
00:20:13
bailed out you know the the managers or
00:20:16
the officers and directors may have
00:20:17
traded you know in private information
00:20:19
advance of that actual any idea of how
00:20:21
many people were involved in that
00:20:23
potential violation well its potential
00:20:27
violation it's difficult because it's
00:20:29
hard to say whether anything was
00:20:32
actually sort of whether you know I
00:20:34
can't go or it since everything's based
00:20:35
on correlations I can't go around and
00:20:37
say uuuuu this from that from this firm
00:20:40
all I can say is the set of firms and
00:20:44
the set of individuals where the
00:20:46
correlation is particularly high now
00:20:49
those trades may have been conducted
00:20:51
entirely legally you know within sort of
00:20:54
the per views of the existing and set of
00:20:55
trading rules and by chance there was a
00:20:59
very high correlation but the fact that
00:21:00
we find that you know these correlation
00:21:04
so pervasive as you mentioned and
00:21:07
specifically focused within those that
00:21:10
are politically connected suggest that
00:21:13
there is you know it's not just one or
00:21:16
two or three individuals they would not
00:21:18
be able to drive a result on 73,000
00:21:21
insiders or 7300 excuse me 7300 insiders
00:21:25
but you know it's so it's sufficiently
00:21:27
pervasive that you know it's able to
00:21:29
generate a large sample result thanks
00:21:32
for coming in today oh my did it my
00:21:33
pleasure
00:21:49
you
