Popular in Professional

Business

The Allegory of the Favela: The Multifaceted Effects of Socioeconomic Mobility

University of Pennsylvania Leandro S. Pongeluppe

Architecture

Connecting Cities and Nature

University of Virginia Timothy Beatley

Human Rights Law

Human Rights and Corruption: Problems and Potential of Individualising a Systemic Problem

Max Planck Institute Anne Peters

Finance

Chinese Exchange Rate Policy: Lessons for Global Investors

University of California San Diego Michael Melvin

Law

A statistical approach to law school citation rankings

University of Virginia Joshua Fischman

Education

Firm Performance Pay as Insurance against Promotion Risk

Stockholm School of Economics Alvin Chen

On the Other Side of Hedge Fund Equity Trades

View more insights in Professional - Finance

View more figures from Lancaster University

View the page of Olga Kolokolova

Read the Study

Next in Finance

Carbon Emissions Trading and Environmental Protection

Jennie Bai | Georgetown University

The Risk and Return of Impact Investing Funds

Jessica Jeffers | HEC Paris

Olga Kolokolova discusses that, trading against highly-diversified low-turnover institutional investors (quasi-indexers), hedge funds generate positive ex-post abnormal returns and avoid negative abnormal returns on their equities portfolios.

Image courtesy of interviewee. January 24, 2024

Log-in or Sign-up to Faculti
Currently viewing this subject insight as a guest. You have insight(s) remaining for this month. Login to view 8000+ figures on the platform.

Generating Workspace...

Export as PDF

The paper examines hedge funds, financial institutions known for exploiting market inefficiencies to generate profits for their clients while enhancing market performance. It notes that while hedge funds earn abnormal returns (alpha), someone on the other side of the trade must incur losses. The authors aim to investigate the counterparties that provide these alpha-generating opportunities for hedge funds and the possible implications of those trades for other investors.

Key findings indicate that hedge funds display distinct trading behaviors, including a preference for illiquid stocks and a tendency to engage in frequent trading, often operating as short-term investors compared to other institutional players. Additionally, they find a specific relationship between hedge funds and quasi-indexes (long-term, diversified institutions), referred to as an "alpha for beta swap." This video video suggests that hedge funds earn positive alpha when buying stocks that quasi-indexes simultaneously sell, while quasi-indexes generate negative alpha from purchasing stocks that hedge funds sell.

The paper highlights that quasi-indexes may lose money because of their risk-averse strategies, often preferring high-beta stocks that do not consistently provide high alpha. Furthermore, hedge funds utilize market anomalies—predictors of future returns—effectively to identify profitable trading opportunities while many other institutions trade in the opposite direction. This video video divergence creates sustained profitable opportunities for hedge funds.

The findings lead to two primary takeaways:
1. The market can facilitate beneficial trades between investors with differing utility functions, even if it appears detrimental for some.
2. The existence of large assets under management by non-alpha-maximizing institutions like quasi-indexes means that there will always be exploitable market anomalies that hedge funds can capitalize on for profit.

Language:

Export as PDF

X-TIMESTAMP-MAP:LOCAL=00:00:00.000,MPEGTS=900000

00:00.065 --> 00:01.885
In this paper, we look at hedge funds.

00:02.345 --> 00:04.125
So hedge funds, it's pretty well known.

00:04.125 --> 00:05.805
These are financial institutions

00:06.265 --> 00:08.685
and they are seen as being arbitrages.

00:09.025 --> 00:10.525
So what they quite, what they do,

00:10.635 --> 00:12.445
they exploit market efficiencies.

00:12.825 --> 00:15.605
And by doing so, they not only earn money for their clients,

00:15.865 --> 00:18.045
but they also help markets to perform better.

00:18.385 --> 00:19.885
So they help price discovery

00:20.305 --> 00:22.485
and they make our markets more efficient.

00:23.185 --> 00:26.605
So these are kind of the key, well part

00:26.605 --> 00:28.165
where we start about the hedge funds.

00:28.515 --> 00:31.165
Well, however, if you think about hedge funds,

00:31.305 --> 00:34.565
if they continuously earn, well alphas abnormal returns,

00:34.905 --> 00:38.445
so they, it means that there is somebody on the other side

00:38.585 --> 00:40.005
who actually provides 'em the offer.

00:40.225 --> 00:42.925
So if they earn money, somebody must be losing money.

00:43.825 --> 00:45.805
And even though we know a lot about the skills

00:45.805 --> 00:47.765
of hedge funds already from the previous literature,

00:48.025 --> 00:49.845
we thought that it would be really great

00:49.865 --> 00:52.285
to look at the other side of hedge fund epi traits

00:52.285 --> 00:54.605
and actually look at what are the,

00:54.605 --> 00:57.125
who are the counterparties who provide them

00:57.125 --> 00:59.125
with these opportunities to earn alpha

00:59.545 --> 01:01.725
and then try to uncover why they do so.

01:01.905 --> 01:04.125
And if they do so does it mean that there is another group

01:04.125 --> 01:07.005
of investors who actually lose money for the investors?

01:07.345 --> 01:09.245
So this was the starting point of our analysis.

01:09.785 --> 01:12.245
The, as I mentioned, we know really a lot already about

01:12.595 --> 01:14.125
what hash funds, do,

01:15.065 --> 01:16.525
or even though I shouldn't admit

01:16.525 --> 01:17.965
that they're quite secretive and we don't

01:17.965 --> 01:19.045
observe all the trades.

01:19.385 --> 01:22.085
But, at least some indirect evidence suggests, first

01:22.085 --> 01:24.845
of all, that hedge funds have, stock peaking

01:24.845 --> 01:25.845
and market bilities.

01:26.595 --> 01:29.085
This means that among all the stocks, they're able to,

01:29.345 --> 01:31.245
to pick up stocks which perform the best

01:32.145 --> 01:34.165
and also to turn them correctly.

01:34.345 --> 01:37.965
So buy them and exit the market, in the correct time.

01:38.385 --> 01:40.325
So we also know that hedge funds tend

01:40.325 --> 01:42.165
to trade more illiquid stocks

01:42.545 --> 01:45.045
and that they trade much more frequently than any

01:45.045 --> 01:46.245
other institutional investors.

01:46.825 --> 01:48.645
And they as an aggregate tend

01:48.645 --> 01:51.005
to be much more short-term investors than

01:51.145 --> 01:52.365
say, a pension funds.

01:53.235 --> 01:55.375
But we also know that sometimes hedge funds

01:55.915 --> 01:58.055
do front run mutual funds, for instance,

01:58.875 --> 02:01.335
and if mutual funds are in distress, so

02:01.335 --> 02:03.735
that hedge funds sometimes start selling stocks

02:03.735 --> 02:05.135
before mutual funds selling them.

02:05.755 --> 02:09.255
So that kind of early money, the kind of short positions.

02:09.755 --> 02:12.775
So we also know that hedge funds tend

02:12.775 --> 02:14.215
to trade on market anos.

02:14.595 --> 02:16.935
So let me just mention quickly what market anos are.

02:17.325 --> 02:18.895
What we usually think about the markets

02:19.005 --> 02:21.975
that they're information efficient, meaning that as soon

02:21.975 --> 02:23.495
as information becomes publicly available,

02:24.035 --> 02:26.055
the prices should adjust pretty

02:26.055 --> 02:27.135
much immediately or instantly.

02:27.515 --> 02:29.455
And then let's say there is no possibility

02:29.615 --> 02:31.375
to earn any abnormal returns

02:31.595 --> 02:33.375
or excessive profits, thereafter.

02:33.995 --> 02:37.415
But what we observed that in fact there are lots

02:37.415 --> 02:39.655
of predictors of future returns,

02:40.345 --> 02:43.015
quite often these predictors are based on accounting data,

02:43.075 --> 02:44.175
for instance, for profitability.

02:44.715 --> 02:46.215
And then months

02:46.225 --> 02:48.095
after this information has been released,

02:48.525 --> 02:50.015
most profitable firms keep

02:50.085 --> 02:51.655
outperforming in the stock market.

02:52.035 --> 02:53.975
So what we know in hedge funds that they tend

02:53.995 --> 02:55.135
to use this information

02:55.275 --> 02:57.135
and they tend to trade profitably on this

02:57.135 --> 02:58.775
information. So again,

02:58.775 --> 03:03.505
This, as I said, we know a lot of what do hedge funds,

03:03.525 --> 03:04.945
and how do they generate alpha,

03:05.485 --> 03:09.305
but it is still not really clear who the counterparts

03:09.305 --> 03:10.825
of hedge funds are.

03:11.005 --> 03:13.985
So, and that's where we try to like add to the literature.

03:14.445 --> 03:16.105
We proceed in several steps.

03:16.155 --> 03:18.905
First of all, and since we're talking about different groups

03:18.965 --> 03:22.225
of institutions, it's important to identify who's who.

03:22.725 --> 03:25.185
So we started from hedge funds

03:25.325 --> 03:26.905
and for these, we have a union

03:26.925 --> 03:28.865
of several hedge fund databases.

03:29.165 --> 03:32.385
So we have U Hatch, Barclay Hedge Morning Start and Task.

03:33.045 --> 03:34.225
We merged these databases,

03:34.225 --> 03:35.585
we clean them from the duplicates.

03:36.245 --> 03:40.345
And then we looked at the filings over 30 F form

03:40.405 --> 03:41.825
to security and exchange commission.

03:42.365 --> 03:45.785
So this is the filings which all institutions which manage

03:45.785 --> 03:48.465
more than 100 million have to file

03:48.465 --> 03:49.825
with security exchange commission,

03:49.825 --> 03:52.185
where they report their kind large equity positions.

03:53.085 --> 03:55.265
And what we did, we then matched the names

03:55.365 --> 03:58.065
of the hedge funds from our list to the names

03:58.565 --> 04:01.185
of institutions, which filed this third F form.

04:02.325 --> 04:06.105
And once we have done this, we then, take our time to go

04:06.105 --> 04:09.385
through Bloomberg and to check that all this match pairs,

04:09.385 --> 04:11.505
which we find are actually really hedge funds.

04:11.605 --> 04:13.425
And they don't have any other business on the side.

04:13.425 --> 04:14.865
So it's not a part of the mutual fund.

04:14.865 --> 04:15.905
They don't have insurance company.

04:16.085 --> 04:18.745
So that we identify really pure hedge fund companies.

04:18.885 --> 04:20.065
So it's our first step.

04:20.485 --> 04:24.625
And the second step is who are these other institutions,

04:24.675 --> 04:26.745
which we're looking at and this

04:26.745 --> 04:27.905
big question, how do you classify them?

04:27.905 --> 04:31.985
Yeah. And then one could go by their nominal labels.

04:32.095 --> 04:34.585
Okay, we have pension funds, we have mutual funds,

04:34.605 --> 04:38.345
we have ETF, yeah, but then we also have banks.

04:38.375 --> 04:40.305
Yeah. But then the question

04:40.305 --> 04:42.465
of course is if you have a mutual fund,

04:42.615 --> 04:44.345
what exactly does this fund do?

04:44.845 --> 04:47.065
Is it probably, there are active mutual funds,

04:47.065 --> 04:48.425
there are very passive mutual funds.

04:48.965 --> 04:50.785
So, and that's what we did in this stage.

04:50.925 --> 04:53.345
We adopted the classification of Brian Boucher,

04:53.805 --> 04:57.305
who suggested looking not at the nominal labels,

04:57.365 --> 04:59.345
but action what institutions do.

05:00.285 --> 05:02.985
So in particular, how frequently they rebalanced their

05:02.985 --> 05:05.345
portfolios and how diversified their portfolios are.

05:06.365 --> 05:08.865
And then he provided this very nice classification

05:08.925 --> 05:11.985
and I should meet, we use his database to look at it,

05:12.125 --> 05:14.785
and divide it in the institutions to several big groups

05:14.925 --> 05:17.385
and the largest, quasi indexes.

05:18.005 --> 05:19.745
So these institutions which have,

05:19.745 --> 05:23.865
very diversified portfolios who follow, long only

05:24.165 --> 05:25.625
and buy whole strategies largely

05:25.925 --> 05:27.705
and quite often trace the indices.

05:28.865 --> 05:31.005
So it's one group and another group which is much more

05:31.005 --> 05:32.565
active institutions transient.

05:32.905 --> 05:33.925
So these institutions,

05:34.015 --> 05:36.405
which also hold diversified portfolios,

05:36.865 --> 05:38.605
but they rebalance them much more frequently,

05:38.705 --> 05:40.525
so much more short-term institutions.

05:41.105 --> 05:43.365
So again, coming back to our certain F filings,

05:43.365 --> 05:47.285
we have the list of all institutions who file their holdings

05:47.285 --> 05:50.565
with the SSC, we first identify hedge funds,

05:51.105 --> 05:53.925
and the rest we classify using the,

05:53.925 --> 05:57.405
brands database on transcend and quasi indices.

05:58.875 --> 06:01.055
So once this is done, once we know the institu,

06:01.055 --> 06:04.295
and then I should admit that we cover around 63 to 65%

06:04.595 --> 06:06.135
of all institutions by doing.

06:06.135 --> 06:07.375
So. Of course there is a chunk

06:07.375 --> 06:09.215
of institutions which we cannot classify,

06:09.475 --> 06:11.335
and this would be, others.

06:11.435 --> 06:12.495
We also have them in the paper.

06:12.495 --> 06:14.495
Yeah, but that's kind of unclassifiable.

06:14.915 --> 06:18.855
So, we look at these institutions

06:18.855 --> 06:21.095
and with the next step we look at how they trade.

06:21.995 --> 06:25.455
So we estimate change in their holdings in each

06:25.465 --> 06:26.735
stock from quarter to quarter.

06:27.705 --> 06:30.005
So, and we look at it as the kind aggregate holdings

06:30.005 --> 06:31.765
as a group, for instance, for each stock,

06:31.855 --> 06:34.285
let's say which ones they priorit and Apple, Microsoft.

06:34.745 --> 06:37.205
Uh, so we'll take the holdings

06:37.345 --> 06:41.285
of a given class institutions say all hedge funds at the end

06:41.305 --> 06:42.765
of second quarter of year

06:43.665 --> 06:45.325
as a fraction of the firm which they held.

06:45.575 --> 06:48.085
We'll look how much of this firm the hedge funds held

06:48.085 --> 06:49.445
during at the end of the previous quarter.

06:49.625 --> 06:51.565
And we take a difference. This kind

06:51.565 --> 06:53.845
of gives us the rough idea of

06:54.105 --> 06:57.565
how much on the aggregate given types of institutions bought

06:57.785 --> 06:59.485
or sold in each stock.

07:01.015 --> 07:02.675
So the, these are kind of the first step.

07:02.935 --> 07:06.075
Second, we try to identify which stocks are hot

07:06.215 --> 07:08.395
or cold in for each of the types of institutions.

07:09.015 --> 07:11.875
And we looked at these changes in holdings

07:12.015 --> 07:14.755
and we classify them in large sales

07:15.255 --> 07:17.275
or large buys.

07:17.855 --> 07:21.315
So if this change in holdings is above 20 percentile,

07:22.745 --> 07:23.945
actually about 80 percentile,

07:23.945 --> 07:27.585
20 top changes in holdings would be, in,

07:27.605 --> 07:31.465
stocks which these group of institutions intensively buys

07:31.925 --> 07:33.105
during this quarter.

07:34.155 --> 07:36.295
And we, if you look on the negative changes in the bottom

07:36.295 --> 07:39.815
20%, the off what we'll see, we'll see that this is the,

07:40.035 --> 07:43.095
stocks which were intently sold by given group prostitution.

07:44.795 --> 07:47.055
And last step is the intersection of this

07:47.055 --> 07:49.335
because again, we're looking at this counterparts

07:49.435 --> 07:50.455
and we look want

07:50.455 --> 07:53.095
to identify the stocks which are traded in one direction

07:53.235 --> 07:56.175
by hedge funds and in the opposite direction

07:56.315 --> 07:57.775
by other groups of institutions.

07:58.805 --> 08:00.255
Yeah, so makes sense. Yeah.

08:00.255 --> 08:02.415
So we look on the, the largest, for instance,

08:02.415 --> 08:03.935
large purchases by hedge funds

08:04.555 --> 08:07.815
and also large sales by was the indexes.

08:08.435 --> 08:10.015
And this would be the stocks which are in

08:10.015 --> 08:11.135
our small portfolios.

08:12.515 --> 08:14.935
And the last step we evaluate performance of the stocks.

08:14.935 --> 08:18.495
They're to see whether they do go good or not so good.

08:18.835 --> 08:21.135
Uh, depending on whether hedge funds buy other solutions

08:21.165 --> 08:22.175
sell or vice versa.

08:22.405 --> 08:25.015
Once we, did them, as we evaluate performance

08:25.015 --> 08:28.415
of these portfolios and what we find out that first of all,

08:28.515 --> 08:31.655
we can identify the group of investors, which seems

08:31.715 --> 08:33.455
to really provide this extra opportunities

08:33.515 --> 08:34.615
to earn alpha, alpha hedge funds.

08:35.355 --> 08:36.855
And it looks like we understand at

08:36.855 --> 08:37.935
least partly the reason for this.

08:38.515 --> 08:41.455
And what we find, we find that between hedge funds

08:41.475 --> 08:43.255
and quasi indexes, this kind

08:43.255 --> 08:45.335
of long-term diversified index follow institutions,

08:45.725 --> 08:48.215
this seem to be an alpha for beta swap.

08:49.675 --> 08:53.815
So the hedge funds earn positive alpha on those stocks,

08:53.865 --> 08:56.485
which they buy while quasi sell.

08:57.925 --> 09:01.305
And on the opposite, so the alpha of stocks,

09:01.305 --> 09:02.345
which quasi indexes buy

09:02.345 --> 09:04.785
and hedge fundal is negative over the next

09:04.785 --> 09:05.985
quarter after the trade.

09:06.015 --> 09:09.305
Yeah. So as we look at in the future, so basically indeed,

09:09.365 --> 09:12.545
so it looks like quasi indexes lose abnormal returns

09:12.545 --> 09:14.225
where they trade against hedge funds

09:14.805 --> 09:16.145
or they provide the abnormal returns

09:16.165 --> 09:17.305
if they sell stocks to them.

09:18.205 --> 09:20.705
And what's happening in the parallel is

09:21.135 --> 09:23.385
that we have the opposite effects of the beater.

09:23.845 --> 09:25.585
So it looks like once that indexes,

09:25.585 --> 09:28.025
they purchase from hedge funds, high beater stocks,

09:28.125 --> 09:29.985
the market beater, and they sell

09:30.315 --> 09:31.585
lower beater stocks to them.

09:32.085 --> 09:34.065
So we call it alpha beta swap.

09:34.875 --> 09:38.695
So, just a side remark, if you compare hedge fund rates

09:38.695 --> 09:40.255
with other institutions like transient,

09:40.395 --> 09:41.815
we don't see this pattern whatsoever.

09:42.315 --> 09:45.055
So this is really this alpha beta swap, which pronounced

09:45.055 --> 09:47.495
between, quasi indexes and hedge funds.

09:48.075 --> 09:49.215
It is also very stable.

09:49.235 --> 09:50.375
We look at it over time

09:50.915 --> 09:53.535
and we find that we observe pretty much the same pattern,

09:53.915 --> 09:58.375
throughout, our sample, which is from 19, 94

09:58.395 --> 09:59.815
to 2017.

10:00.275 --> 10:02.935
So, and now the question is why is it happening?

10:03.595 --> 10:05.255
And so why, why would,

10:06.125 --> 10:09.335
well quasi indexes basically lose kind of money.

10:09.595 --> 10:12.495
And our deportation of that is that the reason is

10:12.495 --> 10:15.415
that they have not maximizing strategies.

10:15.515 --> 10:18.735
The optimization functions very different in particular

10:18.765 --> 10:22.415
because they cannot short sell, they cannot,

10:22.415 --> 10:24.495
quite often they are, they they benchmarked.

10:24.755 --> 10:26.735
So the only way for them to earn a bit

10:26.735 --> 10:29.375
of more money is take a bit of a more market risk.

10:29.675 --> 10:32.575
So that's why they tend to prefer high beta stocks,

10:32.605 --> 10:34.615
basically maximizing their own utility.

10:35.395 --> 10:38.375
But high beta stocks, they tend to also generate

10:38.935 --> 10:39.975
a smaller alpha.

10:40.795 --> 10:43.215
And we try to dig into this in more detail

10:43.235 --> 10:47.375
to see whether indeed this is just this, this,

10:47.375 --> 10:49.775
that they could just usually lose money

10:49.775 --> 10:51.295
because they buy high be of stocks.

10:51.715 --> 10:53.935
And we had a counterfactual example

10:54.105 --> 10:57.655
where we looked at the trades of quasi indexes where they

10:58.435 --> 10:59.655
say buy a lot of stocks,

10:59.995 --> 11:02.255
but hash funds do not sell these stocks to them.

11:03.035 --> 11:04.855
So this kind of counterfactual.

11:05.315 --> 11:08.975
And what we found is that we still see this beta preference.

11:09.035 --> 11:11.455
So quasi indexes prefer stocks with high beta,

11:12.035 --> 11:14.655
but there is no losses in alpha.

11:15.635 --> 11:18.095
So which tells us that's actually among all the

11:18.195 --> 11:20.215
so quasi indexes, they prefer high beta stocks

11:20.635 --> 11:21.935
and they want to buy high beta stocks.

11:21.935 --> 11:25.975
They want to sell low beta stocks while hedge funds are able

11:25.975 --> 11:28.655
to identify out of the pool of all the stocks,

11:28.875 --> 11:30.255
really profitable trades.

11:30.755 --> 11:34.335
And that's when they act as counterparts for wza in indexes.

11:35.635 --> 11:39.055
And one of these factors which we found is the,

11:39.055 --> 11:42.115
lottery demand factor of al.

11:42.175 --> 11:43.595
So what is the lottery demand?

11:43.595 --> 11:46.235
This, this is, so the lottery type stocks,

11:46.235 --> 11:49.355
this is the stocks which pretty much work like lottery.

11:49.935 --> 11:51.955
So from time to time with small probability,

11:52.025 --> 11:53.885
they may give you very high return.

11:54.915 --> 11:57.695
And these stocks on average tend to have high beater

11:58.035 --> 11:59.735
and actually lower alpha.

12:00.835 --> 12:04.335
And we found out that indeed stocks which has funds sell

12:04.355 --> 12:05.495
to quasi indexes,

12:05.665 --> 12:08.775
these are the stocks which exhibit this high exposure

12:08.795 --> 12:09.975
to the lottery demand factor.

12:11.385 --> 12:14.125
And this exposure to this lottery demand factor,

12:14.365 --> 12:17.525
explains the negative alpha of stocks, which,

12:17.725 --> 12:19.765
quasi indexes buy from hedge funds.

12:21.125 --> 12:24.375
Well, however, so we still have a part of stocks,

12:24.375 --> 12:27.135
which hedge funds, hedge funds buy from quasi indexes,

12:27.515 --> 12:29.815
and their positive alpha cannot be explained

12:29.815 --> 12:30.855
by the lottery demand.

12:31.195 --> 12:34.495
So there still is something else in hedge funds skills,

12:34.495 --> 12:35.975
really, there's a opportunity

12:35.975 --> 12:39.255
to identify really profitable traits out of those stocks,

12:39.305 --> 12:42.095
which quasi indexes, which are wishing to sell.

12:42.675 --> 12:44.415
We did lots of digging trying

12:44.415 --> 12:45.775
to figure out what is in there.

12:45.995 --> 12:47.735
So for instance, we looked at liquidity.

12:48.475 --> 12:51.175
Is it true that hedge funds just earn liquidity premium?

12:51.995 --> 12:55.455
So yes, it's true. Uh, they do buy a lot of illiquid stocks.

12:55.605 --> 12:59.535
However, if we exclude illiquid stocks from our trades,

12:59.595 --> 13:01.735
we still see this alphabet swap.

13:01.765 --> 13:04.855
It's still there. So then we looked at other things,

13:04.855 --> 13:08.175
for instance, trading of corporate events like m and a.

13:08.195 --> 13:10.295
We also know that hedge funds quite often traded on m

13:10.295 --> 13:13.375
and a, so we exclude from our sample all firms,

13:13.425 --> 13:16.255
which were involved in emerging acquisitions like two

13:16.255 --> 13:18.135
towards before the quarters after our trade.

13:18.515 --> 13:20.615
And we still find in this cleaner sample,

13:20.685 --> 13:21.775
this alpha beta swap.

13:22.595 --> 13:27.415
We looked also at contrarian versus, momentum trading

13:27.415 --> 13:28.535
because it's also known

13:28.535 --> 13:31.815
that hedge funds go into contrarian trades while like QA

13:31.815 --> 13:34.775
indexes, mutual funds quite often go into momentum trades.

13:35.355 --> 13:38.335
So that we looked at past performance

13:38.595 --> 13:41.735
and we exclude stocks with, really bad past performance

13:41.735 --> 13:43.655
or really good past performance kind of to control

13:43.655 --> 13:46.535
for this extreme contrarian versus momentum trades.

13:46.835 --> 13:49.295
And we still see this alpha for beta swap.

13:50.545 --> 13:52.125
The only thing which also add to this,

13:52.505 --> 13:55.445
but again, well largely on the side

13:55.445 --> 13:56.845
where hedge fund sell stocks

13:56.845 --> 14:00.045
to quasi indexes are trading on market omas.

14:01.025 --> 14:02.925
So because if we think about the market anos,

14:02.925 --> 14:05.685
so these are the factors which predict future returns

14:06.065 --> 14:08.085
and they kind of prescribe trading.

14:08.225 --> 14:10.805
So if you have a, a firm with high profitability,

14:10.985 --> 14:12.365
you need to buy this firm.

14:12.665 --> 14:13.685
If you have firms with low

14:13.685 --> 14:15.005
profitability, you need to sell the firms.

14:15.745 --> 14:17.885
And, it was kindly shown in the past

14:17.885 --> 14:22.405
that in general institutions as the whole tr tends

14:22.425 --> 14:25.565
to trade in the opposite direction of what, this kind

14:25.565 --> 14:27.445
of prescriptions or recipe would be

14:27.445 --> 14:28.645
according to market omas.

14:29.185 --> 14:32.165
And what we find is indeed, if we look into detail

14:32.225 --> 14:35.045
to the swap stocks according to long

14:35.145 --> 14:39.445
or short, lag of this market, anoma trades, we see

14:39.445 --> 14:40.645
that within the short lag.

14:40.665 --> 14:43.405
So these are the kind of stocks which are predicted

14:43.405 --> 14:46.085
to perform poorly by the market anomalies.

14:46.855 --> 14:49.565
There. Uh, quasi indexes tend

14:49.565 --> 14:51.885
to buy stocks while hedge funds sell to them.

14:52.025 --> 14:54.485
And this is the other part which generates again,

14:54.485 --> 14:56.645
the negative alpha for,

14:56.725 --> 14:58.165
quasi indexes when they buy stocks.

14:58.545 --> 15:00.325
Key takeaways are kind of twofold.

15:00.465 --> 15:04.285
The one probably more philosophical is that,

15:04.465 --> 15:07.845
if you observe trading in the market, even if you think

15:07.955 --> 15:11.165
that it is, bad for some institutions,

15:11.185 --> 15:13.245
in fact it may be still a win-win situation.

15:13.785 --> 15:15.285
If we have different types of investors

15:15.515 --> 15:18.645
with their not perfectly aligned utility functions,

15:18.985 --> 15:21.245
you end up having trades which benefit both sides.

15:22.315 --> 15:24.005
It's one takeaway. And the second one, if,

15:24.005 --> 15:27.125
if you look at the size of the institutions,

15:27.705 --> 15:31.445
the quasi indexes, they, they manage way larger portfolios

15:31.505 --> 15:33.525
and way more money than hedge funds or,

15:33.625 --> 15:35.285
or transit institutions together.

15:36.065 --> 15:40.405
So, and it means that while we have such large,

15:40.745 --> 15:43.245
amount of money invested with institutions which do not have

15:44.085 --> 15:46.005
explicit alpha maximizing strategies,

15:47.535 --> 15:50.045
there always will be some profitable opportunities

15:50.705 --> 15:53.605
on the table, based on market anomalies,

15:53.605 --> 15:56.485
based on some other factors which hedge funds these

15:56.485 --> 15:58.765
productive arbitrators can exploit

15:58.945 --> 16:00.725
and then earn money on this for their investors.

Hedge Funds: Market Efficiency and Counterparties

00:00:00 - In this paper, we look at hedge funds. So hedge funds, it's pretty well known. These are financial institutions and they are seen as being arbitrages. So what they quite, what they do, they exploit market efficiencies. And by doing so, they not only earn money for their clients, but they also help markets to perform better. So they help price discovery and they make our markets more efficient. So these are kind of the key, well part where we start about the hedge funds. Well, however, if you think about hedge funds,
00:00:34 - so they, it means that there is somebody on the other side who actually provides 'em the offer. So if they earn money, somebody must be losing money. And even though we know a lot about the skills of hedge funds already from the previous literature, we thought that it would be really great to look at the other side of hedge fund epi traits and actually look at what are the, who are the counterparties who provide them with these opportunities to earn alpha and then try to uncover why they do so.
00:01:04 - of investors who actually lose money for the investors? So this was the starting point of our analysis. The, as I mentioned, we know really a lot already about what hash funds, do, or even though I shouldn't admit that they're quite secretive and we don't observe all the trades. But, at least some indirect evidence suggests, first of all, that hedge funds have, stock peaking and market bilities. This means that among all the stocks, they're able to, to pick up stocks which perform the best
00:01:34 - So buy them and exit the market, in the correct time. So we also know that hedge funds tend to trade more illiquid stocks and that they trade much more frequently than any other institutional investors. And they as an aggregate tend to be much more short-term investors than say, a pension funds. But we also know that sometimes hedge funds do front run mutual funds, for instance, and if mutual funds are in distress, so that hedge funds sometimes start selling stocks
00:02:05 - So that kind of early money, the kind of short positions. So we also know that hedge funds tend to trade on market anos. So let me just mention quickly what market anos are. What we usually think about the markets that they're information efficient, meaning that as soon as information becomes publicly available, the prices should adjust pretty much immediately or instantly. And then let's say there is no possibility to earn any abnormal returns or excessive profits, thereafter.
00:02:37 - of predictors of future returns, quite often these predictors are based on accounting data, for instance, for profitability. And then months after this information has been released, most profitable firms keep outperforming in the stock market. So what we know in hedge funds that they tend to use this information and they tend to trade profitably on this information. So again, This, as I said, we know a lot of what do hedge funds, and how do they generate alpha,

Identifying and Classifying Hedge Fund Institutions

00:03:09 - of hedge funds are. So, and that's where we try to like add to the literature. We proceed in several steps. First of all, and since we're talking about different groups of institutions, it's important to identify who's who. So we started from hedge funds and for these, we have a union of several hedge fund databases. So we have U Hatch, Barclay Hedge Morning Start and Task. We merged these databases, we clean them from the duplicates.
00:03:40 - to security and exchange commission. So this is the filings which all institutions which manage more than 100 million have to file with security exchange commission, where they report their kind large equity positions. And what we did, we then matched the names of the hedge funds from our list to the names of institutions, which filed this third F form. And once we have done this, we then, take our time to go through Bloomberg and to check that all this match pairs,
00:04:11 - And they don't have any other business on the side. So it's not a part of the mutual fund. They don't have insurance company. So that we identify really pure hedge fund companies. So it's our first step. And the second step is who are these other institutions, which we're looking at and this big question, how do you classify them? Yeah. And then one could go by their nominal labels. Okay, we have pension funds, we have mutual funds, we have ETF, yeah, but then we also have banks. Yeah. But then the question
00:04:42 - what exactly does this fund do? Is it probably, there are active mutual funds, there are very passive mutual funds. So, and that's what we did in this stage. We adopted the classification of Brian Boucher, who suggested looking not at the nominal labels, but action what institutions do. So in particular, how frequently they rebalanced their portfolios and how diversified their portfolios are. And then he provided this very nice classification and I should meet, we use his database to look at it,
00:05:14 - and the largest, quasi indexes. So these institutions which have, very diversified portfolios who follow, long only and buy whole strategies largely and quite often trace the indices. So it's one group and another group which is much more active institutions transient. So these institutions, which also hold diversified portfolios, but they rebalance them much more frequently, so much more short-term institutions. So again, coming back to our certain F filings,
00:05:47 - with the SSC, we first identify hedge funds, and the rest we classify using the, brands database on transcend and quasi indices. So once this is done, once we know the institu, and then I should admit that we cover around 63 to 65% of all institutions by doing. So. Of course there is a chunk of institutions which we cannot classify, and this would be, others. We also have them in the paper. Yeah, but that's kind of unclassifiable.

Analyzing Hedge Fund Trading Strategies and Performance

00:06:18 - and with the next step we look at how they trade. So we estimate change in their holdings in each stock from quarter to quarter. So, and we look at it as the kind aggregate holdings as a group, for instance, for each stock, let's say which ones they priorit and Apple, Microsoft. so we'll take the holdings of a given class institutions say all hedge funds at the end of second quarter of year as a fraction of the firm which they held.
00:06:48 - during at the end of the previous quarter. And we take a difference. This kind of gives us the rough idea of how much on the aggregate given types of institutions bought or sold in each stock. So the, these are kind of the first step. Second, we try to identify which stocks are hot or cold in for each of the types of institutions. And we looked at these changes in holdings and we classify them in large sales or large buys.
00:07:22 - actually about 80 percentile, 20 top changes in holdings would be, in, stocks which these group of institutions intensively buys during this quarter. And we, if you look on the negative changes in the bottom 20%, the off what we'll see, we'll see that this is the, stocks which were intently sold by given group prostitution. And last step is the intersection of this because again, we're looking at this counterparts
00:07:50 - to identify the stocks which are traded in one direction by hedge funds and in the opposite direction by other groups of institutions. Yeah, so makes sense. Yeah. So we look on the, the largest, for instance, large purchases by hedge funds and also large sales by was the indexes. And this would be the stocks which are in our small portfolios. And the last step we evaluate performance of the stocks. They're to see whether they do go good or not so good. depending on whether hedge funds buy other solutions
00:08:22 - Once we, did them, as we evaluate performance of these portfolios and what we find out that first of all, we can identify the group of investors, which seems to really provide this extra opportunities to earn alpha, alpha hedge funds. And it looks like we understand at least partly the reason for this. And what we find, we find that between hedge funds and quasi indexes, this kind of long-term diversified index follow institutions, this seem to be an alpha for beta swap. So the hedge funds earn positive alpha on those stocks,
00:08:57 - And on the opposite, so the alpha of stocks, which quasi indexes buy and hedge fundal is negative over the next quarter after the trade. Yeah. So as we look at in the future, so basically indeed, so it looks like quasi indexes lose abnormal returns where they trade against hedge funds or they provide the abnormal returns if they sell stocks to them. And what's happening in the parallel is that we have the opposite effects of the beater. So it looks like once that indexes,

Understanding Alpha-Beta Dynamics in Investments

00:09:28 - the market beater, and they sell lower beater stocks to them. So we call it alpha beta swap. So, just a side remark, if you compare hedge fund rates with other institutions like transient, we don't see this pattern whatsoever. So this is really this alpha beta swap, which pronounced between, quasi indexes and hedge funds. It is also very stable. We look at it over time and we find that we observe pretty much the same pattern, throughout, our sample, which is from 19, 94
00:10:00 - So, and now the question is why is it happening? And so why, why would, well quasi indexes basically lose kind of money. And our deportation of that is that the reason is that they have not maximizing strategies. The optimization functions very different in particular because they cannot short sell, they cannot, quite often they are, they they benchmarked. So the only way for them to earn a bit of more money is take a bit of a more market risk.
00:10:32 - basically maximizing their own utility. But high beta stocks, they tend to also generate a smaller alpha. And we try to dig into this in more detail to see whether indeed this is just this, this, that they could just usually lose money because they buy high be of stocks. And we had a counterfactual example where we looked at the trades of quasi indexes where they say buy a lot of stocks,
00:11:03 - So this kind of counterfactual. And what we found is that we still see this beta preference. So quasi indexes prefer stocks with high beta, but there is no losses in alpha. So which tells us that's actually among all the so quasi indexes, they prefer high beta stocks and they want to buy high beta stocks. They want to sell low beta stocks while hedge funds are able to identify out of the pool of all the stocks, really profitable trades.
00:11:35 - And one of these factors which we found is the, lottery demand factor of al. So what is the lottery demand? This, this is, so the lottery type stocks, this is the stocks which pretty much work like lottery. So from time to time with small probability, they may give you very high return. And these stocks on average tend to have high beater and actually lower alpha.
00:12:04 - to quasi indexes, these are the stocks which exhibit this high exposure to the lottery demand factor. And this exposure to this lottery demand factor, explains the negative alpha of stocks, which, quasi indexes buy from hedge funds. Well, however, so we still have a part of stocks, which hedge funds, hedge funds buy from quasi indexes, and their positive alpha cannot be explained by the lottery demand.

Understanding Hedge Funds and Quasi Indexes

00:12:34 - really, there's a opportunity to identify really profitable traits out of those stocks, which quasi indexes, which are wishing to sell. We did lots of digging trying to figure out what is in there. So for instance, we looked at liquidity. Is it true that hedge funds just earn liquidity premium? So yes, it's true. they do buy a lot of illiquid stocks. However, if we exclude illiquid stocks from our trades, we still see this alphabet swap.
00:13:04 - for instance, trading of corporate events like m and a. We also know that hedge funds quite often traded on m and a, so we exclude from our sample all firms, which were involved in emerging acquisitions like two towards before the quarters after our trade. And we still find in this cleaner sample, this alpha beta swap. We looked also at contrarian versus, momentum trading because it's also known that hedge funds go into contrarian trades while like QA
00:13:35 - So that we looked at past performance and we exclude stocks with, really bad past performance or really good past performance kind of to control for this extreme contrarian versus momentum trades. And we still see this alpha for beta swap. The only thing which also add to this, but again, well largely on the side where hedge fund sell stocks to quasi indexes are trading on market omas. So because if we think about the market anos,
00:14:06 - and they kind of prescribe trading. So if you have a, a firm with high profitability, you need to buy this firm. If you have firms with low profitability, you need to sell the firms. And, it was kindly shown in the past that in general institutions as the whole tr tends to trade in the opposite direction of what, this kind of prescriptions or recipe would be according to market omas. And what we find is indeed, if we look into detail to the swap stocks according to long
00:14:39 - that within the short lag. So these are the kind of stocks which are predicted to perform poorly by the market anomalies. There. quasi indexes tend to buy stocks while hedge funds sell to them. And this is the other part which generates again, the negative alpha for, quasi indexes when they buy stocks. Key takeaways are kind of twofold. The one probably more philosophical is that, if you observe trading in the market, even if you think
00:15:11 - in fact it may be still a win-win situation. If we have different types of investors with their not perfectly aligned utility functions, you end up having trades which benefit both sides. It's one takeaway. And the second one, if, if you look at the size of the institutions, the quasi indexes, they, they manage way larger portfolios and way more money than hedge funds or, or transit institutions together. So, and it means that while we have such large,

Exploiting Market Anomalies for Profit

00:15:58 - explicit alpha maximizing strategies, there always will be some profitable opportunities on the table, based on market anomalies, based on some other factors which hedge funds these productive arbitrators can exploit and then earn money on this for their investors.

Faculti AI can make mistakes

Research

Policy

Analysis