00:00:00
I pulled some data together, Nick, if
00:00:02
you could pull up the basics. I know we
00:00:04
talk about this a lot, but I thought it
00:00:05
would be good to show the difference
00:00:06
between kind of a normal distribution
00:00:07
and a power law distribution. like
00:00:09
venture returns or I don't like to call
00:00:11
it venture returns but I do think the
00:00:14
returns generally in free markets
00:00:18
create value creation in a power law
00:00:21
distribution which means that a few of
00:00:24
the many account for the vast majority
00:00:27
of the capital appreciation of the the
00:00:30
value creation and that's because of the
00:00:32
power of compounding. If you build a
00:00:35
better engine of technology, a better
00:00:37
business engine, you will compound and
00:00:39
that means that over time you will
00:00:41
accumulate more and more of the market
00:00:43
in an outsized way. As you accumulate
00:00:44
more of the market, you actually move
00:00:46
faster in accumulating more of the
00:00:47
market and eventually it becomes a
00:00:49
runaway flywheel. No one can catch up.
00:00:51
That's the key
00:00:53
return profile in free markets generally
00:00:56
speaking. And a manifestation of that is
00:00:58
that is also the return profile in
00:01:00
venture
00:01:01
in technology venture investing because
00:01:04
when you find that one or two great hit
00:01:06
the Uber that has a great network effect
00:01:08
or the Airbnb or the Google it has a
00:01:11
runaway effect in the market and it
00:01:12
accumulates all the capital and becomes
00:01:14
really valuable and you can see this in
00:01:15
the return. So if you just pull up this
00:01:17
is the latest Carta data. So this shows
00:01:20
what the irr by the vintage that the
00:01:23
fund was raised and then also by the
00:01:25
size of the fund and by how the fund has
00:01:28
performed on a percentile basis. Look
00:01:29
just at the 2017 row. So the top decile
00:01:34
funds are doing call it 30% versus you
00:01:37
know the median fund.
00:01:39
Hold on. They're not these are all
00:01:40
markups paper markup. These are not dist
00:01:42
these are all numbers.
00:01:44
Okay fair enough. But I'll make your
00:01:46
case in a second here. And these are not
00:01:47
DPIs. This is not distribution cash out.
00:01:50
This is just marked up value. And this
00:01:51
is 8 years old. This is the 2017
00:01:53
vintage. But just the difference
00:01:55
between, you know, call it a 30% IRRa
00:01:58
and a 10% IRRa that compounds over 10
00:02:00
years. So the multiple difference at the
00:02:02
end you should expect is the difference
00:02:04
of making 14 times your money or 2.6
00:02:08
times your money. The difference between
00:02:09
10% and 30%. And that's kind of what
00:02:12
this shows roughly is that the
00:02:13
difference between a 50th percentile and
00:02:15
a
00:02:15
problem with this data though is this is
00:02:18
Carter's product is used by a small
00:02:20
number of people for a small number of
00:02:22
years. That's why the other data we
00:02:23
showed from Cambridge is much better
00:02:25
because the 2018 forward are still in
00:02:28
the J curve. So they and to Chimat's
00:02:30
point they're paper so I knew I knew I
00:02:33
knew out there.
00:02:35
Okay. I knew we would go down this
00:02:36
rabbit hole so I shouldn't have shown
00:02:37
that but like
00:02:37
No, that's okay. There are two key
00:02:39
points I wanted to make and just pull it
00:02:40
up again, Nick. The first key point I
00:02:42
wanted to make is because of the power
00:02:44
law, the job of investing is to find the
00:02:47
power law winners. It is not to buy the
00:02:50
index. If you buy the index, you are
00:02:53
losing to the market. Even if these mark
00:02:54
even if these numbers are correct, which
00:02:56
they're not, as we've talked about,
00:02:57
you're losing to the index if you're
00:02:59
just buying the venture index. And it's
00:03:00
probably way worse than this. You have
00:03:03
to get the winners. And so what this
00:03:05
does show though is that the smaller the
00:03:07
fund, the less diversified it is, which
00:03:09
means the more they're likely to capture
00:03:11
the winners. And you can actually see
00:03:12
that the smaller funds generally have
00:03:14
better performance than the bigger
00:03:17
for sure they do because they're getting
00:03:18
in earlier as well, Freeberg, and
00:03:20
they're paying less value lower
00:03:22
valuation. So that's
00:03:23
but they're also not they're not getting
00:03:25
diluted away by the index. And the index
00:03:27
generally is going to return negative.
00:03:29
It's going to be a negative returner. I
00:03:30
think I told you guys I went to an LP
00:03:32
conference years ago. one of the biggest
00:03:34
venture funds and they showed that 45%
00:03:37
of their capital went into flat or down
00:03:39
rounds and they had a negative net
00:03:41
return on that allocation of capital
00:03:42
across 13 funds and if they had put that
00:03:45
if they had never invested in flat or
00:03:47
down rounds then they would have doubled
00:03:49
their IRRa overall as a as a fund really
00:03:52
important point that the the power law
00:03:54
matters which is your job is to just
00:03:56
find the winners and if you find the
00:03:58
winners they're going to compound so how
00:03:59
does that translate so so go to the next
00:04:00
slide Nick so this is the one that I
00:04:02
shared in 2023 three that Gokul Roger uh
00:04:05
published on Twitter and that's where I
00:04:06
first saw this which is if you buy just
00:04:09
the top 10 companies in the NASDAQ and
00:04:12
you just held it, you would have made a
00:04:14
24x multiple over a 24-year period
00:04:17
versus if you had just bought the the
00:04:19
NASDAQ. So just owning the best
00:04:21
companies in the NASDAQ is by far the
00:04:24
best way to drive multiple of return
00:04:27
over a 10-year period. you made a 9x
00:04:30
multiple 9x just by owning the top 10
00:04:33
companies in the NASDAQ. And this was
00:04:35
through roughly the end of 2023. So now
00:04:38
let's pull this up. I pulled this
00:04:39
analysis together this morning for our
00:04:41
conversation. And this shows that the
00:04:43
venture returns don't stop when you stop
00:04:46
being a private company. That the real
00:04:48
return and this highlights the point
00:04:50
that most of the value when you find
00:04:51
that power law winner, most of the value
00:04:54
is created when they are a public
00:04:56
company. So, Palunteer went public after
00:04:58
17 years and their market cap was 16
00:05:01
billion. So, they created 16 billion of
00:05:03
equity value over a 17-year period.
00:05:05
Since they've been public in 5 years,
00:05:07
they're now worth $436 billion. So,
00:05:10
they've created another $420 billion of
00:05:12
equity market value in just the last 5
00:05:15
years, probably eclipsing all venture
00:05:18
returns that have been made during that
00:05:20
same period of time. Airbnb went public
00:05:22
after 12 years, worth 47 billion. In
00:05:25
just 5 years, they added another, call
00:05:26
it roughly 30 billion. Uber went public
00:05:29
75 billion after 9 years as a private
00:05:32
company and now they're worth 190
00:05:34
billion. So they've added 120 billion of
00:05:36
value in just 6 years since being
00:05:38
public. And Spotify, another good
00:05:40
example, 27 billion market cap when they
00:05:42
went public after being private for 10
00:05:44
years. And then they added another 120
00:05:47
billion of value in the seven years
00:05:49
they've been public. And I've got the
00:05:50
whole list here. This is all these
00:05:51
companies that have 100red billion plus
00:05:53
market cap. And when you find the power
00:05:55
law winner, the value continues to
00:05:57
accrete. It continues to compound. And
00:06:00
you're better off just buying the public
00:06:02
stocks. And I think that this was my my
00:06:04
big kind of conclusion as I went through
00:06:06
all this data. Like
00:06:08
Chimath is exactly right. When you're in
00:06:09
the venture market, you learn a lot. And
00:06:12
ultimately, I think you probably want to
00:06:13
learn a lot to then allocate your
00:06:14
capital more wisely into what you've
00:06:16
identified are the power law winners
00:06:19
where most of the value continues to
00:06:20
accumulate whether they are private or
00:06:22
public. And having an IPO is just a
00:06:24
transitionary event for these companies.
00:06:25
Their their compounding engine will
00:06:27
continue as a public company if you've
00:06:29
identified them. Facebook such a great
00:06:30
example. Eight years as a private
00:06:32
company, 100 billion market cap at IPO
00:06:34
added another two plus trillion dollars
00:06:36
in the years since. Yeah, I think you
00:06:38
Freeberg you've cherrypicked all the
00:06:40
winners in hindsight, the biggest
00:06:41
winners for that list. What if you were
00:06:43
just to create a list of every
00:06:45
IPO
00:06:45
tech IPO and how it's fared afterwards,
00:06:48
right? No, it's a fair point. There be a
00:06:49
lot of stinkers in there that you
00:06:51
haven't
00:06:51
But my point, Sax, is just find the
00:06:53
power law winners and then that's where
00:06:55
all the the value
00:06:56
improves. Well, I mean that that's hard,
00:06:58
too.
00:06:58
Yeah. Here's the I think you have to
00:07:00
look at how we got here. I I started
00:07:02
doing this just, you know, in my second
00:07:04
decade
00:07:06
and companies decided to stay private
00:07:08
longer. That was the big trend. And then
00:07:10
there was a lot of reactions to that.
00:07:12
Secondary transactions started occurring
00:07:16
and then you just couldn't get DPI. you
00:07:18
had to hold it forever. And now the
00:07:21
industry is responding to this. So now
00:07:23
we have something called strip sales. We
00:07:25
have people creating continuation funds,
00:07:27
ways to take those early investors and
00:07:30
get them out into a new vehicle or to
00:07:32
create some liquidity. And Chimat's
00:07:35
point is really important because you do
00:07:37
get so much intelligence. And what I
00:07:39
learned having been an investor in Uber
00:07:42
and Robin Hood when they were $5 million
00:07:43
and $20 million companies. I was like
00:07:45
first investor into those companies is
00:07:48
when they crashed in the market. When
00:07:50
Uber hit 30, I bought up a bunch of it
00:07:52
as a public investor. Even though I had
00:07:53
a ton from when I was an angel investor
00:07:56
and I did the same for Robin Hood at $12
00:07:58
because I knew the management, I knew
00:07:59
Vlad, I knew Travis, I knew Dara and I
00:08:02
had that understanding of the market.
00:08:03
And the same thing I did with Facebook,
00:08:05
having listened to Chim talk about it
00:08:06
all these years and Brad Gersonner talk
00:08:08
about all these years, I backed into it
00:08:10
when it was $92. So what's happening now
00:08:12
is what RUF did with his continuation
00:08:16
fund and holding the private ones is in
00:08:18
the theme song of this podcast famously
00:08:21
when Zach said, "Let your winners ride."
00:08:22
That is the overarching thing here. The
00:08:25
industry, the venture industry is in in
00:08:27
massive transition. It's going to start
00:08:28
looking a little bit like private equity
00:08:30
and a little bit like Gavin or Brad
00:08:32
where they're private and public market
00:08:34
investors and you're seeing buyouts
00:08:36
occur. So people are saying, "Hey, this
00:08:38
SAS company's undervalued. We're just
00:08:41
going to buy the company. We'll start
00:08:42
operating it." So venture is in a
00:08:44
massive transition, but I think what
00:08:45
comes out on the other side is something
00:08:47
that looks a lot more like public
00:08:50
private investing like Ruof pioneered at
00:08:53
Sequoia and we're going to talk to him
00:08:54
about that at the summit. I'll just
00:08:55
respond to your last comment and Jal I
00:08:57
appreciate you saying that because yeah
00:08:58
Roloff is coming to the summit to talk
00:09:00
exactly about this which is going to be
00:09:01
great but that is the job of being an
00:09:04
investor to your point you're like hey
00:09:05
you know you cherry pick the winners but
00:09:07
I think that's the point the point is
00:09:08
that the returns acrew to these power
00:09:11
law winners and the job of being an
00:09:12
investor shouldn't be to index a market
00:09:15
anyone can index a market the job of
00:09:16
being a great investor is to find those
00:09:18
winners and
00:09:19
the question is how how hard that is or
00:09:22
easy it is and I think your just makes
00:09:25
it sound like, oh, all you have to do is
00:09:27
wait for these companies to go public.
00:09:30
Sorry, that wasn't my point. Yeah, that
00:09:32
wasn't my point. I kind of came across
00:09:33
that way, too. Yeah, sorry. I was just
00:09:35
saying that the power law winners
00:09:36
continue to acrew. In fact, not only do
00:09:38
they acrew as private companies and then
00:09:40
you if you get one of them, you you have
00:09:42
these 14x funds, but as a public
00:09:45
company, they continue to acrue. And so,
00:09:47
you know, this isn't just limited to
00:09:48
being a venture investor. pretty much
00:09:50
anyone has access to the public markets
00:09:53
and can make the decision that they've
00:09:54
made a bet on a power law winner. There
00:09:56
were a lot of people who were very vocal
00:09:58
supporters of Palanteer, very vocal
00:10:00
supporters of Uber, of Spotify when
00:10:01
these companies went public and a lot of
00:10:03
people said no, they're naysayers.
00:10:05
I think the problem that you would have
00:10:07
had trying to implement this strategy is
00:10:09
it would not have been clear what you
00:10:10
would have been underwriting. What would
00:10:11
you have been underwriting at Nvidia
00:10:14
for, you know, not the last five years,
00:10:17
but the many years before then? You
00:10:19
would have been underwriting video
00:10:20
games.
00:10:21
Video games. Yeah.
00:10:22
And you would have had this like very
00:10:24
odd data center business that didn't
00:10:26
make sense that, you know, Jensen took a
00:10:28
lot of heat for that he had to defend
00:10:31
and it it didn't actually make a lot of
00:10:33
financial sense. So, I I'm not sure.
00:10:35
I will give you a counter to that and
00:10:36
I'll I'll call him out and I'll make fun
00:10:38
of him. Now, my co-founder at Climate
00:10:40
Corp, his name's Siraj. is probably
00:10:41
listening right now. He bought Nvidia
00:10:43
years ago on the thesis that these GPU
00:10:46
chips would eventually be used for AI
00:10:47
models because he did all his work at
00:10:48
Stanford in distributed computing. He
00:10:50
did early work in AI. He was very close
00:10:52
with
00:10:52
that's different than what you were
00:10:53
saying before. I get it. He's a smart
00:10:55
guy who made a huge winner. But I'm
00:10:57
saying
00:10:57
but no, the pro the problem was he sold
00:10:58
Nvidia too soon because he's like, "Oh,
00:11:00
they're not actually."
00:11:01
Okay. But I'm just saying like there was
00:11:02
probably a person that was as smart as
00:11:04
Siraj and bought Intel. Then there was
00:11:05
another person that bought AMD. Like I'm
00:11:07
I'm not sure what the point is. Here's
00:11:08
the point about venture that I think is
00:11:10
worth noting. It's an exceptional
00:11:14
market that gives the practitioners
00:11:17
an incredible edge if you build a
00:11:20
business around the information
00:11:21
asymmetry that occurs.
00:11:23
Yes,
00:11:23
that you can monetize. After that, the
00:11:27
only business model is building an asset
00:11:30
gathering machine to generate fees
00:11:31
because even the consistency of being
00:11:35
able to generate these power law
00:11:37
distributions in successive funds
00:11:40
doesn't exist. And you can look at all
00:11:42
of the data. Cambridge has done this. A
00:11:44
lot of these guys have done this. All
00:11:45
the big LP fund of funds like Horses
00:11:47
Bridge have done this. Having a killer
00:11:50
fund
00:11:51
has zero correlation with your ability
00:11:53
to then have a next killer fund as an
00:11:55
LP. And I showed these returns on
00:11:57
screen, although we blanked out the
00:11:59
names, and I'm in all of them. It is
00:12:01
really hard to be consistent.
