00:00:05
our paper is about scale and skill and
00:00:07
active management we started with a
00:00:09
simple idea which is that a mutual
00:00:12
fund's performance depends not just on
00:00:14
skill but also potentially on scale so
00:00:18
in order for us to say anything about
00:00:19
scale we need to make sure we first
00:00:21
understand scale so the goal of this
00:00:24
project is is to understand the nature
00:00:26
of returns to scale an active mutual
00:00:29
fund management
00:00:30
we test two different ideas the first is
00:00:33
decreasing returns to scale at the fund
00:00:35
level there the idea is as a fund gets
00:00:38
larger that causes the fund's
00:00:41
performance to to drop for example
00:00:43
because larger funds make larger trades
00:00:46
that push prices more and that hurts
00:00:49
performance that's the first idea the
00:00:51
second idea is decreasing returns to
00:00:53
scale at the industry level the idea
00:00:56
here is as the overall mutual fund
00:00:58
industry gets larger that makes all
00:01:00
funds performance decrease and here the
00:01:03
logic is as there are more mutual fund
00:01:06
dollars out there competing and chasing
00:01:08
after mispriced stocks that makes prices
00:01:11
move and it makes harder for any mutual
00:01:13
fund to find a a good trading
00:01:16
opportunity our main result is strong
00:01:18
evidence of decreasing returns at the
00:01:21
industry level in contrast we find mixed
00:01:24
evidence of decreasing returns at the
00:01:26
fund level so more simply we find that
00:01:29
it's the size of the mutual fund
00:01:30
industry that's very important for
00:01:32
performance the size of an individual
00:01:35
fund is less important these results
00:01:38
that we're finding on scale have some
00:01:40
interesting implications for how we
00:01:42
think about skill we offer a new way of
00:01:45
measuring skill that adjusts for scale
00:01:48
with this measure we show for example
00:01:50
that uh skill has been trending upwards
00:01:52
uh in the mutual fund industry over the
00:01:54
last 30
00:01:58
years well an active mutual fund is U
00:02:03
distinguished from a passive mutual fund
00:02:05
in the sense that a passive mutual fund
00:02:07
would just buy a basket of stocks that
00:02:10
represent an overall investment in the
00:02:12
stock market or segment of the stock
00:02:14
market what we typically call an index
00:02:16
fund uh an active fund deviates from
00:02:19
that from that portfolio by
00:02:22
overweighting stocks that it believes
00:02:24
are underpriced and offer Superior
00:02:27
returns and uh down weights or sells off
00:02:30
the stocks that it believes are
00:02:31
overpriced or or offer worse returns uh
00:02:35
so an active fund will uh trade more
00:02:38
than uh a passive fund and its return
00:02:41
will have what we call tracking error
00:02:44
relative to the passive uh uh Benchmark
00:02:47
against which it's
00:02:52
judged well the Surge and popularity of
00:02:54
active mutual funds really coincides
00:02:56
with the surge in mutual funds overall
00:02:59
in in fact uh index funds or passive
00:03:02
funds have actually had a somewhat
00:03:04
higher rate of growth than active mutual
00:03:06
funds but the overall mutual fund
00:03:08
industry has grown so much uh over the
00:03:11
decades that active funds too have
00:03:14
become uh much bigger in the sense that
00:03:16
they control a much bigger segment of
00:03:19
the uh US Stock
00:03:21
Market uh we believe that uh some the
00:03:25
overall trend in mutual fund growth just
00:03:27
coincides in general with the uh
00:03:31
increasingly smaller fraction of the US
00:03:33
Equity Market that's owned directly by
00:03:36
individuals uh the growth in retirement
00:03:38
plans such as uh you know defined
00:03:41
contribution of 401K plans is is is one
00:03:44
uh Big Driver of that growth and there
00:03:46
other institutional factors that have
00:03:48
been identified as leading to this surge
00:03:51
in in mutual fund uh
00:03:57
size as there more Mutual fund dollars
00:04:01
out there uh competing with each other
00:04:03
and looking for mispriced
00:04:06
stocks then there's more trading on
00:04:08
those ideas that pushes prices and
00:04:11
that's going to make it harder for any
00:04:12
mutual fund to find a mispriced stock so
00:04:15
the analogy we like to use is that if
00:04:18
there are more people fishing on the
00:04:19
same Pond well that's going to make it
00:04:21
harder for any individual person to
00:04:24
catch fish on that
00:04:28
pond well as Luke mentioned earlier in
00:04:31
order to understand skill we have to
00:04:33
understand what the role of scale or
00:04:34
size of of industry in the fund is in in
00:04:37
generating performance so we Define
00:04:40
skill as the ability of a manager to
00:04:43
perform relative to a benchmark uh
00:04:46
abstracting from the effects of Industry
00:04:48
size and his own fund size U so one way
00:04:52
of thinking about that is we try to ask
00:04:54
uh how much outperformance would this
00:04:57
manager have if he were the only only
00:04:59
one operating in the stock market and
00:05:01
didn't have competition from all the
00:05:04
other mutual funds and in fact didn't
00:05:06
have to worry about how big his own fund
00:05:08
got in some sense it's his ability to
00:05:10
generate Returns on the first dollar
00:05:12
you'd invest with that with that fund
00:05:18
manager well we don't really know for
00:05:21
sure but uh we do know that uh certainly
00:05:25
the education of uh those entering the
00:05:28
financial services industry has
00:05:29
increased over time uh so the training
00:05:32
has gone up um also uh over time new
00:05:36
strategies get discovered uh we have uh
00:05:39
for example quantitative approaches to
00:05:42
Investment Management that exploit
00:05:44
technology and Computing capabilities uh
00:05:47
to much greater degrees than we could
00:05:49
decades ago uh so we believe things like
00:05:52
training and
00:05:53
Technology uh lead to uh higher degrees
00:05:57
of skill as well as uh the usual ual
00:05:59
sort of learning on the job effects uh
00:06:03
uh we believe that skill is uh increased
00:06:06
for those reasons and possibly others we
00:06:08
don't really know why for sure we show
00:06:10
that skill has gone up over the last 30
00:06:12
Years yet performance has not how is
00:06:15
that possible our explanation combines
00:06:18
two elements uh the first is the mutual
00:06:21
fund industry has been growing steadily
00:06:22
over the last 30 years the second
00:06:25
element is decreasing returns to scale
00:06:27
at the industry level and in other words
00:06:30
funds are becoming more skilled but the
00:06:32
industry is also growing and those two
00:06:35
effects tend to offset each other in
00:06:36
other words yes fund managers are more
00:06:39
skilled today but they have to be more
00:06:41
skilled just to uh keep up with the rest
00:06:43
of the
00:06:47
pack so as a fund gets larger it starts
00:06:51
to make larger trades those larger
00:06:53
trades push prices more against the fund
00:06:56
which hurts the fund's profits yeah for
00:06:59
examp example if you have a fund that uh
00:07:01
thinks a given stock is underpriced and
00:07:04
wants to buy 5,000 shares of that stock
00:07:07
uh it has to go and uh find someone
00:07:10
willing to sell that stock in the market
00:07:11
if if another fund identifies a stock is
00:07:14
under pric and wants to buy 50,000
00:07:15
shares of that stock it's got to find
00:07:18
more potential sellers of that stock to
00:07:21
do so it either has to offer a bigger
00:07:23
concession willing to pay more for that
00:07:25
stock to get it quickly or it has to try
00:07:28
to parcel out it's its purchases over
00:07:31
time and take longer and perhaps thereby
00:07:34
miss out on some of the underpricing
00:07:36
that would get corrected in the
00:07:41
meantime I think we were surprised uh
00:07:45
because some some of the uh the leading
00:07:48
uh theory about mutual funds these days
00:07:50
assigns a a central role to the notion
00:07:53
that as a given fund gets bigger uh it
00:07:56
should become become harder for it to
00:07:57
perform um so I think we were somewhat
00:08:00
surprised that even though our estimates
00:08:02
uh pointed in the direction we were not
00:08:04
able to get strong statistical Precision
00:08:08
associated with those estimates uh so I
00:08:10
think that was one aspect of our study
00:08:12
we found somewhat
00:08:14
surprising the logic for for why a fund
00:08:17
Size Matters is compelling but there is
00:08:20
a a different logic pointing in the
00:08:22
other direction for example imagine two
00:08:25
mutual funds that follow exactly the
00:08:27
same strategy they make the same trades
00:08:30
at the same times well you can see that
00:08:33
their individual sizes may not matter
00:08:36
instead what's going to matter is their
00:08:37
combined
00:08:42
size well again here's one of these
00:08:44
findings we don't have a a a definitive
00:08:46
answer for um but one one potential
00:08:50
reason is that uh younger funds could
00:08:53
also come with younger managers on
00:08:55
average uh so we get back to this issue
00:08:58
of uh increases in skill and and uh sort
00:09:01
of technological prowess perhaps being
00:09:04
associated uh with with age um the other
00:09:08
thing is the younger funds could be
00:09:09
exploiting newer strategies that the
00:09:11
market has not sort of sort of caught on
00:09:14
to and exploited to as significant a
00:09:16
degree again this is one of those things
00:09:18
like uh um Rising skill we don't we
00:09:22
can't really say for sure why it's
00:09:24
happened but but uh we think it's not
00:09:26
implausible that uh that these factors
00:09:28
could contribute to Superior performance
00:09:31
for younger funds we find that there's
00:09:33
also a second way in which fund age
00:09:35
matters and uh that's a relation over
00:09:39
time between performance and age we find
00:09:41
that a fund's performance typically
00:09:43
deteriorates as the fund ages uh if you
00:09:47
take all these results together they're
00:09:49
consistent with a simple story the story
00:09:51
is that new funds entering the industry
00:09:54
have more skill than the existing funds
00:09:57
possibly because of better education or
00:09:59
better grasp on technology because of
00:10:01
this Superior skill the young funds
00:10:03
outperform initially they outperform
00:10:06
their benchmarks and they also
00:10:07
outperformed the older funds but
00:10:10
performance tends to deteriorate as a
00:10:12
fund ages and the reason is that as a
00:10:15
fund ages the mutual fund industry is
00:10:17
getting larger and larger and that hurts
00:10:19
everyone's
00:10:23
performance investors should prefer
00:10:25
newer active funds over older active FS
00:10:29
funds uh each month we form a portfolio
00:10:33
uh that contains all of the youngest
00:10:34
funds say funds that are of age between
00:10:37
Zer and three years we also form a
00:10:39
portfolio of funds that are 10 years old
00:10:41
and older and we find that the portfolio
00:10:44
of very young funds significantly
00:10:46
outperforms the portfolio of of older
00:10:53
funds well here's sort of the bad news
00:10:55
of this of this uh finding is that even
00:10:57
though we find younger funds
00:10:59
significantly outperform older funds in
00:11:01
the active fund Universe uh active funds
00:11:04
in general be they young or old uh on
00:11:08
average have
00:11:09
underperformed uh passive index funds so
00:11:14
uh our message would be to someone who
00:11:16
has decided to invest at least part of
00:11:19
their money in active funds uh younger
00:11:21
funds do seem to on average offer spr
00:11:23
performance but uh the overall averages
00:11:26
do favor index funds uh
00:11:29
and potentially because of this finding
00:11:32
that the uh industry Size Matters so uh
00:11:36
you know as long as the industry is as
00:11:38
big as it is the active fund industry is
00:11:40
as big as it is uh index funds may offer
00:11:43
a a better return uh you know were that
00:11:46
industry to be smaller were it to shrink
00:11:48
and indeed index funds are growing more
00:11:51
rapidly uh there could come a time when
00:11:54
active funds uh are able to keep Pace uh
00:11:58
and perhaps even offer some prep
00:11:59
performance but uh the evidence for now
00:12:02
anyway would seem to be that uh index
00:12:05
funds uh uh offer offer Superior average
00:12:10
returns based on historical
00:12:15
evidence well we are following up
00:12:17
actually in fact we have a a uh a recent
00:12:19
working paper that also explores the
00:12:22
role of fund size in a somewhat
00:12:24
different way we're looking at the uh
00:12:26
the turnover of of funds how how much
00:12:29
they trade and to what extent that
00:12:31
turnover seems to be related to
00:12:33
Performance do funds earn more after
00:12:36
they after they trade more does a given
00:12:38
fund when it trades more heavily produce
00:12:40
Superior returns as a result and uh
00:12:43
there we do find that that fund size
00:12:45
does seem to play a role in particular
00:12:47
small funds do seem to exhibit a
00:12:49
stronger relation between their turnover
00:12:51
and their performance suggesting they
00:12:53
can more readily exploit these time
00:12:56
varying opportunities to identify Mis
00:12:58
pric
00:13:21
see
