Latest / trading places / Nvidia $4.4T vs Cerebras $23B Val | Why Private Prices Are Wrong with Nick Fusco of PM Insights | Episode 27

Transcript

Nvidia $4.4T vs Cerebras $23B Val | Why Private Prices Are Wrong with Nick Fusco of PM Insights | Episode 27 · trading places

• 0:00
  Here's my hot take. Figure robotics is a little bit
• 0:03
  too focused on the human figure and not enough on the robotics.
• 0:07
  Talking about putting the data centers in space sounded science
• 0:10
  science fiction to me five years ago.
• 0:12
  But today I'm like, no, I'm all for it.
• 0:14
  To go to 19 billion in ARR. Now they added more ARR in one
• 0:18
  month than Databricks generates in a year.
• 0:21
  Sometimes maybe someone looked at the market and they said, oh,
• 0:24
  wow, this is great. There's no volatility over here.
• 0:26
  But what we can tell from the data is absolutely that's not
• 0:29
  terribly true. Good morning, and welcome to
• 0:52
  another episode of our really crappy, terrible, horrible DC
• 0:56
  podcast. What are we looking at here?
• 0:57
  We're looking at the Augment Power 20 again.
• 1:00
  Tell me, who's your Private Mag 7?
• 1:02
  Is this list right? Three, the top four I think are
• 1:06
  hard to argue with SpaceX, Anthropic and open AII think if
• 1:09
  you measure it by size, momentum, amount of money
• 1:13
  raised, like I think those are in there.
• 1:15
  I can understand why you put Andrew in there.
• 1:16
  Their market half of 60 billion. It's not 600 billion.
• 1:20
  But they're also, I guess they're especially with
• 1:23
  hostilities and the streets of Hormuz, there's a tension and
• 1:26
  they're doing the new rounds and maybe they maybe they get some.
• 1:29
  Fast riser. Fast riser.
• 1:31
  Hard to argue with data bricks and stripes so I think through
• 1:34
  the top 6 again. Size, momentum over 100 billion
• 1:36
  market cap. XAI is an odd one to include.
• 1:40
  That one's not there anymore. That's actually part of SpaceX
• 1:42
  now, right? Yes.
• 1:44
  But even as a stand alone, would you put XAI into the top 7
• 1:48
  before before they got bought by SpaceX?
• 1:52
  I thought that valuation was full of shit a long time ago.
• 1:55
  I still think it's full of shit as as a SpaceX shareholder I
• 1:58
  think I just got my pocket robbed.
• 2:00
  Attention and pickpocket. Attention never, said Jatrichi.
• 2:06
  Pickpocket. Think think Elon robbed Elon
• 2:10
  here Was right. Got diluted by 20%.
• 2:14
  He did XAI just for some marketing headlines and some AI
• 2:18
  sort of magics. Very dust, but.
• 2:21
  Do you think we'll get do you think in the long?
• 2:23
  Run. We'll get 20% the benefits.
• 2:26
  Will the benefits sex justify the illusion over?
• 2:29
  Time. I don't think so 'cause Elon's
• 2:31
  firing everybody in XAI and they're leaving like I think it
• 2:35
  was something like, you know, six of the eight people who were
• 2:38
  in the company have left already.
• 2:40
  And I think I think Elad feels like XAI is not measuring up to,
• 2:45
  you know, open AI and entropic. So we'll see.
• 2:48
  But but back to this list, Byte Dance is actually a #11 here
• 2:52
  when I really think that should be in the top seven.
• 2:56
  If I look at this list from CB Insights, that top group of
• 2:59
  four, I would say that's all definitely in the private mag 7
• 3:05
  and the next two data bricks and stripes.
• 3:06
  So by market cap, SpaceX is now #1 at like 1.2 open AI at 750 or
• 3:15
  800, something like that. Yeah, eight 3840 better than
• 3:19
  post money. They they raise 733.
• 3:21
  They're 840 post now. Byte Dance was 550 last week.
• 3:26
  We covered that. Yeah.
• 3:28
  Anthropic 380, I think and then Stripe maybe at 150, I think 160
• 3:35
  after last week and Databricks 134, I think that one's still
• 3:38
  there. So I think that's your top six
• 3:41
  of the top 7S. AI is now, you know, no longer
• 3:46
  solo. That's part of SpaceX.
• 3:47
  So then the question is who's who's 7?
• 3:51
  Is that Revolute, Canva, Androl? That's a good question.
• 3:55
  That one you can see Androl being at the 7th spot.
• 3:57
  I think Canvas got a really good case.
• 3:59
  They are we're billion in ARR. They're bigger than you know,
• 4:02
  they're they're data bricks in size, right in terms of revenues
• 4:05
  and they're also profitable growing.
• 4:07
  Yeah. Who?
• 4:08
  Shouldn't be in that list or who's maybe a little overhyped,
• 4:12
  overvalued. So we have covered Revolute
• 4:14
  before. I think they are aggressive at
• 4:17
  75 billion. A lot of that is because they're
• 4:19
  heading towards banking revenues and they're a they are AUK bank
• 4:22
  and banking revenues don't typically get the same
• 4:25
  valuations as tech. I'll go on record as saying I
• 4:28
  think an open the eye is starting to feel shaky.
• 4:30
  Even though they just raised the biggest round ever.
• 4:34
  I feel like Anthropic is catching up, Sam.
• 4:37
  Sam's waving his hands around and doing a lot of marketing and
• 4:40
  fundraising. But would you rather have
• 4:42
  opening eyes balance sheet and revenue growth rate or Anthropic
• 4:46
  balance sheet and revenue growth rate?
• 4:49
  Well, I think 2 weeks ago, we were both in the camp of, I
• 4:52
  think Anthropic looks a little, a little bit better $880 billion
• 4:56
  evaluation than Open AI at 8:30. That is the flash flow dynamics
• 4:59
  of Open AI. How much they're trying to raise
• 5:01
  and spend doesn't look great. Anthropic just added 5 billion
• 5:06
  in ARR in a single month in February, but it's only a 28 day
• 5:10
  month, so imagine what they could have done in a here's.
• 5:13
  Web for you, would you take a long short trade?
• 5:15
  Long on Anthropic and short on open AII.
• 5:18
  Think it'd rather be long anthropic.
• 5:19
  I don't know whether it's short open AI, but of of the two.
• 5:23
  Not brave, not that brave. What I think is really kind of
• 5:27
  crazy is figure robotics. And figure, I think recently
• 5:31
  raised a 30, nine, $40 billion evaluation.
• 5:34
  This is which is up from a $3 billion evaluation.
• 5:37
  Here's my hot take. I'm on figure robotics is a
• 5:40
  little bit too focused on the human figure and not enough on
• 5:44
  the robotics. The quote is form follows
• 5:47
  function. You might think that that means
• 5:50
  that form should follow the human form.
• 5:53
  But actually I think that's that's both.
• 5:56
  I think in this case the form should follow whatever the
• 5:59
  function is, which is not necessarily a human pushing
• 6:03
  things around, but like lifting and moving large objects and
• 6:07
  small objects and all kinds of shapes and sizes.
• 6:10
  And the world of robotics is much broader than just the shape
• 6:13
  of a human being. Interesting.
• 6:15
  Even though robotics is probably the hype that's going on right
• 6:18
  now, I would say that maybe we're a little too focused on
• 6:21
  humans and not enough focused on other things like what works.
• 6:27
  Well, I'm curious on your take jumps to your design principle.
• 6:29
  If you think about how we've imagined robots in the past, if
• 6:33
  you go back to some of the early imagination like the Jetsons,
• 6:36
  the robot did look like a person.
• 6:39
  And then I think during the then, then you get, you know,
• 6:42
  then you get popular imagination in movies where you got the
• 6:44
  assembly line and you get machinery and it doesn't look
• 6:46
  like humans, but it it works. You think that people will not
• 6:50
  want human, they'll not want humanoid robots around to be
• 6:53
  their personal assistant. You think it'll just be smaller
• 6:56
  robots and? First of all, when you mentioned
• 6:58
  Jetsons, I thought about, you know, the Jetson one, the flying
• 7:02
  car. Have you seen those videos on
• 7:03
  YouTube? Like, they're pretty amazing.
• 7:06
  I mean, I'm excited about robotics, and maybe that
• 7:08
  includes humans, but I think I'm probably more excited from an
• 7:11
  investment standpoint in the general field of robotics and
• 7:14
  maybe a little bit less specifically about the human
• 7:17
  figure. Well, the I can see there being
• 7:19
  a place for the Jetsons robot had Rosie and was kind of like
• 7:23
  this person, you Android housekeeper.
• 7:26
  So I can imagine there being like companionship being
• 7:28
  something that people will want. But I I kind of agree figure
• 7:31
  that valuation and based on such little revenue that doesn't seem
• 7:35
  to be justified by where they're at right now.
• 7:37
  It's a very that's a very ambitious valuation.
• 7:39
  Probably shouldn't be the top 10 or or even the top 20 honestly.
• 7:45
  Hello, anyone here I'd like to rent a maid?
• 7:49
  Yes, you have an appointment and your credentials.
• 7:53
  Credentials, Madam. You can't expect to get our maid
• 7:57
  without references. Let's jump to some macro news.
• 8:02
  I guess there's something been going on in the world lately,
• 8:04
  right Aman? You're talking about the the
• 8:07
  conflict in Iran. What do you think?
• 8:09
  You think that that's a major news event from the tech
• 8:13
  perspective or not? Yeah, I think it has just
• 8:16
  produced a lot of volatility. If I go back to the end of
• 8:19
  February when the initial strikes began with the US and
• 8:23
  Israel striking targets in Iran, the overall stock market is
• 8:27
  within about 3% of where it was on that day and and friend its
• 8:31
  overall highs. But the volatility on a
• 8:33
  day-to-day basis has has jumped and we've seen oil prices go as
• 8:37
  we're as we're filming this right now, it's about $100 for
• 8:40
  Brent crude. It's bounced around between 85
• 8:43
  and 115. And so a lot of the, I think the
• 8:45
  result has been won a lot of volatility which is put off a
• 8:47
  bunch of IP OS. Goldman Sachs just came out with
• 8:50
  their re forecast for the year and they re forecast PCE
• 8:53
  inflation. They were at 2.2% announcing
• 8:57
  2.9. So a little bit of a higher
• 8:59
  inflation rate for the year. Even for the core PCE, which
• 9:03
  excludes oil and gas prices, energy prices, they upped it
• 9:07
  from 2/1 to 2/4 for the year. So even if I just take out the,
• 9:11
  I guess direct impact of oil, just the knock on effects of
• 9:14
  trade being disrupted in the Straits of Hormuz.
• 9:17
  And then the fact oil goes into the as an input and so many
• 9:20
  other things like transportation inflation slightly higher.
• 9:23
  So they're not going to off their, their GDP growth rate by
• 9:25
  30 basis points as well. So it's having an, it's having
• 9:28
  an impact I think on, you know, slight impact in the markets,
• 9:32
  definitely impact on volatility and an outsized impact on IPO
• 9:35
  and tech. Yeah, I think that's really the
• 9:38
  issue for us. That's relevant, right?
• 9:40
  Like how does this have an impact on valuations in the
• 9:43
  market and IP OS? And that's that's largely driven
• 9:46
  by interest rates and inflation predictions and that's largely
• 9:50
  driven by the price of oil and how long the conflict is going
• 9:54
  to last. And yes, specifically, are the
• 9:57
  Straits, is the Strait of Hormuz, open for business or
• 10:00
  not? If we listen to our Department
• 10:03
  of War Secretary, Department of Events Secretary Pete Hexa, he
• 10:07
  says, hey, it's open for business, except that, you know,
• 10:10
  Arianne might fire on you. It's kind of like, yeah, the
• 10:15
  playground's open. As long as you don't mind the
• 10:17
  assassin sitting in the trees, they're ready to shoot you.
• 10:20
  So here's an interesting prediction market story about
• 10:27
  whether Iran, Israel, US conflict ends in various time
• 10:32
  frames. And right now it looks like the
• 10:34
  markets think that it ends somewhere around the end of
• 10:37
  April, maybe early May. That's when it rises above 50%
• 10:42
  prediction market. So that's, that's one sense of
• 10:45
  where the market thinks this, you know, war is going, which is
• 10:48
  another six weeks, 7 weeks, maybe 8 weeks, but not much more
• 10:52
  than that. But it depends what you mean by
• 10:54
  ends I guess. So this.
• 10:56
  This market resolves. I agree with you.
• 10:58
  I agree with you. When?
• 10:59
  When does the war end? Amman.
• 11:01
  Well, let me say it does is 1 answer.
• 11:04
  Another answer I guess is when hostilities led by US and
• 11:07
  Israeli forces seize something like that.
• 11:09
  Would that be? That could be another reason.
• 11:12
  There could be this tube, pull the troops back.
• 11:14
  Hey, now there's the reason, and I think that's actually the
• 11:18
  story, is when are the Straits of Hormuz open for business?
• 11:23
  And here's a polymarket bet on whether that traffic returns to
• 11:27
  normal by the end of April, and that's below 50%.
• 11:32
  So I think maybe there's a little bit of time between when
• 11:36
  the war quote ends and when traffic in the Strait of Formula
• 11:39
  returns to normal. But it seems like right now the
• 11:41
  market is saying that's not going to happen until at least
• 11:44
  May. So I guess if you, I don't know
• 11:50
  if there are other bets on whether this happens at the end
• 11:52
  of May or end of June, but it feels like it might take a few
• 11:54
  more months for that traffic to return normal.
• 11:58
  But I bet you the pressure on the Trump administration is got
• 12:01
  to be like rising, right? If the price of oil is up and
• 12:04
  inflation's up, that's not going to play well for US consumers.
• 12:09
  If it's up for other parties, maybe you might think that the
• 12:13
  Gulf states like the fact that the price of oil is high.
• 12:15
  But actually, I think Gulf states don't like the fact that
• 12:18
  Iran is now firing missiles in their capitals and driving
• 12:22
  business out of Dubai and Abu Dhabi and Riyadh.
• 12:25
  Yeah, absolutely. I got to think the pressure's
• 12:28
  building. I know there is political
• 12:29
  pressure building because the bitums are coming up in
• 12:32
  November. And but the Trump team had
• 12:34
  promised early on was this will be weeks, not months.
• 12:38
  And that's definitely the extent you've got an impact on prices
• 12:42
  on just Americans being involved.
• 12:43
  That's just, that's not something that you want to you
• 12:46
  want to be, you want, you don't want that to string out into the
• 12:47
  fall because the voters in the US right now, I'd say inside the
• 12:52
  Republican Party, it's probably 7030 in favor of continued
• 12:56
  strikes. But outside of the Republican
• 12:58
  Party, it's like, that's the. Public seat of the way.
• 13:00
  Yeah. Yeah, Well, the colors they're
• 13:02
  not happy about. This.
• 13:04
  Well, yeah. I can say though.
• 13:08
  Well, I was inside inside the Republican coalition, so it
• 13:10
  within among independents if opinion 7030 against you, that's
• 13:13
  not a good look for the midterms.
• 13:15
  And inside the Republican Party, there's an increasing voice as
• 13:18
  well for like, hey, let's let's wrap this up.
• 13:19
  This is this is not what we campaigned on in in 2024, right?
• 13:24
  This is this is the one thing we agreed on.
• 13:26
  Maybe. Maybe there were other issues.
• 13:27
  Where you campaigned on the fact that you said the Biden was
• 13:31
  gonna start to work. That's right.
• 13:34
  So another way to think about when when the conflict ends or
• 13:37
  when the market think thinks the conflict ends.
• 13:39
  I guess one way to do what is the the Poly market bet on what
• 13:42
  when is the ending. But the other question is how
• 13:44
  long do oil prices stay elevated?
• 13:46
  And this is the oil futures curve for Brent crude.
• 13:50
  So the price today is something like the 100 bucks in the spot
• 13:53
  markets, but the futures market is where do these rent crude
• 13:56
  barrels of oil trade one month out, three months out, 612
• 14:00
  months out. And you can see and at the end
• 14:03
  of June, which is sort of what you were saying was the Poly
• 14:05
  market prediction, it's still showing prices of oil changing
• 14:08
  hands at $80 a barrel by the end of the year.
• 14:12
  So end of 2026, we're down to low 70s, which is kind of where
• 14:16
  we were when we started the conflict.
• 14:18
  So it's really saying we're not going back to a normal oil price
• 14:22
  until probably the September to even December time frame.
• 14:26
  And so I can imagine the scenario where we, I guess as a
• 14:29
  country, we decide we're we're no longer operating American
• 14:32
  flights in Iran, but they continue to put pressure on the
• 14:35
  straights or even strike other allies in the least.
• 14:39
  And so the markets continue to be disrupted through the end of
• 14:41
  the year. I think that's probably more the
• 14:43
  market sentiment right now. We must be sure they go to
• 14:48
  happy, happy homes. But my mother told me.
• 14:54
  But good enough. Any reference from any mother is
• 14:57
  all we need. Well, on the job front, I guess
• 15:00
  we've been seeing some news of layoffs and I guess that started
• 15:04
  with Block announcing a 40% cut in the overall staff, but now
• 15:09
  we're also seeing it with companies like Oracle and Meta.
• 15:12
  Oracle announced thousands of job cuts.
• 15:15
  I think there's been expectations of as money as
• 15:18
  maybe 30 to 40,000 cuts in an overall company of 160,000.
• 15:22
  So that's that's probably around 20% of the potential, you know,
• 15:26
  jobs they're being cut. Oracle has made a big bet on AI
• 15:30
  is borrowed a good bit of money to make that bet.
• 15:33
  Probably there thinking about how can they make sure that they
• 15:36
  look more efficient to the market while they're sort of
• 15:38
  making such big bets, right? Yes, for sure.
• 15:42
  Well, it's, it's, there's a couple of things to think about
• 15:44
  there. On the one hand, they're
• 15:45
  announcing the layoffs. I think in in response to the
• 15:49
  stock price being down about 50%.
• 15:51
  There's some urgency I think around how they want to manage
• 15:54
  the business. This is a chart just showing the
• 15:57
  debt to market cap ratio of certain companies in the tech
• 15:59
  sector. So I'm just taking the total
• 16:02
  long term and short term debt of the company, dividing by the
• 16:05
  value of their equity. You can see Oracle in the far
• 16:08
  left. They're at about they're a,
• 16:09
  they're far more leveraged and far more indebted than anywhere
• 16:11
  else in the space. Just compare them to like Google
• 16:14
  or Meta on, you know, on the far right or NVIDIA, which have
• 16:17
  very, very little debt. And so there's definitely been
• 16:20
  pressure on the stock from the stock market on, on Oracle.
• 16:24
  They also announced a pretty significant commitment to CapEx
• 16:27
  and you know, part of the Stargate initiative earlier this
• 16:30
  year. So they've gone free cash flow
• 16:31
  negative, which is really not a, not a place where you want to be
• 16:34
  as a tech company. This is a chart to show in their
• 16:37
  free cash flow over the past, over the past two years.
• 16:40
  So they have gone from 9 billion and you know, free cash flow in
• 16:44
  2023 to 2024. They were up to over 11 billion
• 16:49
  in free cash flow. And then they announced
• 16:51
  Stargate, which is this massive. It's the probably the biggest AI
• 16:54
  infrastructure build out in history.
• 16:56
  The announced number was over $500 billion by 2029.
• 17:00
  It was supposed to be initially SoftBank and then a few other,
• 17:03
  you know, a few other open AI and SoftBank and a few others
• 17:05
  putting the capital in. But then Oracle came along and
• 17:07
  participated up to 50 billion annually.
• 17:09
  So they're now going free cash flow negative.
• 17:12
  And I think they are in a position because of their debts
• 17:14
  where they're pretty vulnerable to an AI downturn.
• 17:17
  And so I think they're trying to get ahead of this a little bit,
• 17:19
  which, you know, which makes sense.
• 17:21
  Another company that is discussing or at least thinking
• 17:24
  about layoffs is Meta. Sounded like 10s of thousands of
• 17:27
  job cuts, potentially up to 20% of the workforce.
• 17:31
  Yeah, 20% is what I heard as well.
• 17:33
  So we're talking about 16,000 of a company of 80,000.
• 17:38
  I think they've just been on the vanguard of this efficiency
• 17:42
  effort for the past couple of years.
• 17:44
  If you remember back in between 2015 and 2022, I would say they
• 17:47
  took almost no, despite being one of the most successful
• 17:51
  businesses of all time, they took almost no efforts to become
• 17:54
  more efficient. You can look at it, This is a
• 17:56
  chart that every CFO should love because they should be showing
• 17:59
  this now to the boards of directors at every quarterly
• 18:02
  meeting. This is revenue per employee.
• 18:04
  So if you look at Meta back in 2015, they were generating about
• 18:07
  1.4 million in revenue per employee.
• 18:10
  And then Fast forward seven years later, at the height of
• 18:12
  the pandemic in 2022, they've gotten a lot bigger, a lot more
• 18:16
  profitable. It was an era of innovation.
• 18:18
  It was an era of growth. And look at their revenue per
• 18:21
  employee, it's still at 1.5 million, so 1.4 million.
• 18:25
  So really kind of a pathetic effort in terms of operating
• 18:28
  efficiency. And in 2022, this is when
• 18:30
  Zuckerberg announced that 2022 would be the year of efficiency.
• 18:34
  He kind of, he kind of moved away from, from Meta and the
• 18:38
  metaverse and and ARVR and started focusing on operations
• 18:41
  and just being more capital efficient.
• 18:43
  And so this is where they're at. They've actually had a lot of
• 18:45
  progress now becoming more, more profitable, more efficient.
• 18:48
  And I think we're now best in class in terms of this revenue
• 18:51
  per employee metrics. So they're here, they are in the
• 18:53
  far left now, 2.5 million in revenue per employee.
• 18:57
  Compare that by the way, to Oracle, who we just said before
• 19:00
  was not doing great in this efficiency metric.
• 19:02
  And they're a point for revenue per employee.
• 19:04
  And they can see all the other companies between Meta best in
• 19:07
  class now and Oracle worst in class.
• 19:10
  So I just think Meta is on the cutting edge now of what you
• 19:13
  know, what can be done with technology and process
• 19:15
  improvement. I don't think they're going to
• 19:16
  be the last one. I think we'll be a lot more of
• 19:18
  these 1020% type refs coming in the valley now.
• 19:20
  And all the all the CF OS better start getting familiar with that
• 19:24
  chart and showing it to the boards because investors going
• 19:26
  to ask. So how big is your home?
• 19:31
  We have 3 bedrooms. A slum clearance project.
• 19:37
  So then you'll want to see our basic economy model from Great
• 19:40
  Britain. Well, there's been a number of
• 19:45
  announcements in the past couple of months about very, very large
• 19:49
  VC raising lots and lots of capital.
• 19:51
  Who's the VC? Mag 7?
• 19:54
  In the past week, we've had three huge announcements.
• 19:58
  General Catalyst is raising $10 billion.
• 20:00
  This is a company that already has.
• 20:03
  More than $40 billion under management and is going to
• 20:06
  increase that quite a bit. Founders Fund is raising $6
• 20:09
  billion a year this year. That's a pretty significant
• 20:13
  raise. Not too long after they raised
• 20:16
  4/4 and 1/2 billion. So kind of like $10 billion it's
• 20:19
  raised in the last year or two. And then Spark Capital is
• 20:22
  raising $3 billion. Spark was one of the early
• 20:25
  investors in Anthropic and I think has had over, I guess,
• 20:29
  maybe 100 X gain for their previous investments in AI and
• 20:33
  specifically Anthropic. So pretty big numbers for those
• 20:37
  3 firms. I think we were trying to kind
• 20:39
  of get a, a sense of who's big in BC.
• 20:43
  Here's a list of VC funds that are very large.
• 20:46
  The list that I've kind of highlighted there, the top five
• 20:49
  are all larger than 40 billion in a web.
• 20:51
  So that includes Andreessen Horowitz, Insights, Tiger,
• 20:55
  Sequoia Capital, General Catalyst kind of coming on their
• 20:58
  heels. Lightspeed raised $7 billion
• 21:01
  recently in Thrive, which has raised $10 billion recently,
• 21:04
  both of those kind of above 20 billion in size.
• 21:08
  Then we got a few others like Excel and Kotu, which have
• 21:11
  pretty large war chests, assets under management, maybe raised
• 21:14
  billion dollar funds recently. But then Founders Fund, Khosla
• 21:18
  and Spark all raising very large sums and Founders Fund really
• 21:22
  has raised 10 billion accounts or previous raises.
• 21:25
  And there's a few of those we could argue on this list.
• 21:27
  But overall, it seems like there's probably 10 to 15 VC
• 21:30
  funds that are larger than 10 billion assets under management
• 21:34
  recently raising, let's say upwards of three, four, $5
• 21:37
  billion or more for a single amount of fundraising.
• 21:41
  I guess the rich keep getting richer about.
• 21:43
  Is that Is that your view on this?
• 21:45
  Well, the bigger certainly getting bigger.
• 21:47
  I think what's happening is just take Founders Fund as an
• 21:49
  example. So there was a long period when
• 21:51
  they were focused on just more capital efficient deals.
• 21:54
  And now they're and they've deployed 3 billion in the past
• 21:58
  year and are back doing more. And they have done bets on like
• 22:00
  Polymarket and ramp and companies that you would say are
• 22:02
  capital efficient. But I think their biggest bets
• 22:05
  are they were in both open AI and Anthropic, which which is
• 22:09
  interesting said finally, it's one of the rules that VC, which
• 22:11
  is never invest in competitors and Darrell, they're they're
• 22:15
  putting on very capital intensive businesses now, right?
• 22:17
  AI and defense, it will produce great returns, but with great
• 22:21
  capital. And so the bigger funds and
• 22:23
  spark, you guess an example of this, some of these guys are
• 22:25
  getting big just to pursue the capital intensive businesses of,
• 22:28
  of AI and defense, if in general, capital is heading that
• 22:30
  direction as well. So they're moving off of their
• 22:33
  strategy of capital light and, and moving into some new
• 22:35
  territory, which which I guess necessitates the bigger funds.
• 22:38
  Plus, I guess if they can raise bigger funds, they, you know,
• 22:41
  they should, there's, there's reason for them to do that.
• 22:43
  But I do wonder if it'll hold up in the returns that they get in
• 22:45
  the future. Well, Aman, here's a question
• 22:48
  for you. With great capital comes great
• 22:51
  responsibility. Yes or no?
• 22:53
  Yes, absolutely. That's that is exactly that is
• 22:57
  the that is the dictum that the fiduciary duty requires of us,
• 23:00
  right. Here's a view of venture capital
• 23:03
  sort of segmented by whether we're talking about, you know,
• 23:05
  huge mega platforms, large VC, mid market, established and
• 23:09
  emerging managers and that that mega platform category maybe you
• 23:14
  know, occupied by 10/12/15 firms.
• 23:17
  They have over 300 billion under management and they've raised
• 23:20
  over 100 billion in the last couple years.
• 23:22
  The amount they've raised in the last three years a larger
• 23:25
  portion of the market than they currently represent by AOL.
• 23:28
  So they're piling up capital at greater than their market share.
• 23:33
  The next group of large institutional same story.
• 23:35
  They've, they've been 13% of total AOM, but they've raised,
• 23:39
  you know, 19% of the capital of the market in the last three
• 23:42
  years. So they're kind of kicking butt
• 23:45
  it seems like. I wonder if those next groups,
• 23:47
  the mid market and emerging managers in various groups are
• 23:51
  feeling a little pressure as a result of those, you know, top
• 23:55
  two groups really sucking up a lot of the capital in the
• 23:57
  market. Yeah, it's tough.
• 23:59
  I guess the mid markets doing OK, I guess they're still,
• 24:03
  they're still raising proportionately and not not as
• 24:06
  much as the other guys, but they're, they're hanging in the
• 24:08
  emerging market and the new ones, I think it's really tough
• 24:10
  and their track records are also suffering.
• 24:12
  The last five years hasn't been kind to that space as well loose
• 24:15
  in terms of the distributions, no turn.
• 24:17
  So I, I think it's a very tough cycle for emerging market
• 24:21
  managers. I hope it comes back because so
• 24:23
  much of the innovation that we need in the ecosystem comes from
• 24:25
  those emerging market managers making some of those seed stage
• 24:28
  bets. But the hype in the cycle right
• 24:29
  now is towards AI, capital intensive businesses, the rich
• 24:32
  getting richer and I, I guess the poor if not getting poorer
• 24:36
  than treading water, which is, you know, unfortunate.
• 24:39
  Some other news that's happening this week is Nvidia's big GPU
• 24:43
  tech conference is happening next week.
• 24:46
  NVIDIA GTC starts on Monday and lots of people in town, lots of
• 24:50
  events going on, huge focus on NVIDIA chips, AI, all kind of
• 24:56
  innovations that are happening there.
• 24:57
  Related to that, Entropic has really just been like, you know,
• 25:01
  hitting the bell one after another.
• 25:04
  Entropic has doubled their AR, tripled their AR in the last
• 25:08
  year. More they are.
• 25:10
  They've gone. They added 5 billion in February
• 25:14
  to to 19 billion reported now. There were about 2 billion a
• 25:18
  year ago. So they've actually gone 12X in
• 25:20
  the last year. If we look at the Sacra page for
• 25:24
  Anthropic, you can see kind of what's been happening with that
• 25:27
  growth rate. Just absolutely astonishing
• 25:30
  growth for a company that's already very, very large.
• 25:35
  Yes, we feel like this is supporting another raise for
• 25:38
  Anthropic at an even higher valuation in the future.
• 25:40
  It seems. Seems like they just finished
• 25:41
  raising, but then this news on revenue growth continuing.
• 25:45
  I wonder if there's more money being raised in the future now.
• 25:48
  Yeah, I think it's amazing just to put in context.
• 25:50
  So they put, they added 5 billion in ARR in in one month
• 25:53
  in the month of February. That was only a 28 day month.
• 25:56
  Imagine what they could have done in a 31 day month.
• 25:59
  But to go from to go to 19 billion in ARR.
• 26:01
  Now they added more ARR in one month than Databricks generates
• 26:06
  in a year, right. Databricks is one of the biggest
• 26:08
  stuff for companies out there. We set a top 7.
• 26:10
  They're doing 4 billion in ARR a year and these guys added five
• 26:13
  in a month. So they could be, if they're at
• 26:15
  19 now, they could be on their way to, you know, 50 by the end
• 26:19
  of the year. I think it's, they're just
• 26:21
  scratching the surface. I think what really did it is in
• 26:24
  February they had a couple of launches.
• 26:25
  Opus 4.6 was a, a big launch for them.
• 26:30
  It's bigot. It's we, we can see as soon as
• 26:32
  they launched that they were just destroying certain
• 26:34
  categories in public markets. Legal services took a 20 to 30%
• 26:37
  nosedive as soon as Claude began allowing legal companies and
• 26:41
  service providers to to do a lot of what LexisNexis and Thompson
• 26:46
  Roger does in terms of analyzing data, reviewing contracts.
• 26:50
  Cybersecurity took a bit of a hit when when that cloud code
• 26:52
  came out in middle of February. So they seem to be they
• 26:55
  definitely seem to be on a roll. I think cloud seems to be
• 26:57
  gaining traction getting share among businesses.
• 27:00
  I think businesses are using it to both cut cost and generate
• 27:03
  revenue and they seem to be catching up to, you know, open
• 27:07
  AI now in terms of their there are.
• 27:09
  So I, I think it's probably, I think it's, you know, it's all
• 27:12
  great momentum on top of the last raise.
• 27:14
  I'm sure the next raise will be in the next few months and we'll
• 27:17
  probably be at a premium to weather right now.
• 27:19
  Just that matter of revenue increase on both companies is
• 27:23
  kind of astonishing, but it seems like it's going asymptotic
• 27:27
  there. Yeah, I've never really seen
• 27:29
  this before in in any tech business and it seems like
• 27:33
  Anthropic has crossed a bit of a Moat here.
• 27:35
  I think they they seem to have picked up fraction now with the
• 27:37
  small businesses and enterprises that will will enable them to
• 27:41
  continue this growth at least for, you know, at least for the
• 27:43
  foreseeable future. 12X in the past year is amazing.
• 27:49
  Tea is served. You notice the very broad a This
• 27:54
  model was only used on weekends by an Old English teacher, Mr.
• 27:58
  Chips. I'm afraid not.
• 28:02
  Thank you, Agnes. Chip, Chip.
• 28:04
  Chin up, Cheerio. Chin up, Cheerio.
• 28:11
  OK. So last week we had Jason
• 28:13
  Saltzman from CB Insights on the program and he was showing off
• 28:16
  some numbers from the CB Insights State of Venture
• 28:19
  Capital 2025 report, which was kind of a look back on the year.
• 28:24
  Here is the TLDR that came from his slides.
• 28:28
  A lot of numbers on that slide, but the one I've gotten
• 28:30
  highlighted here is talking specifically about robotics
• 28:34
  being a rising category of venture investment, $40 billion
• 28:37
  last year spent on robotics. If we look at kind of like what
• 28:42
  that is in relationship to last year's category, that's up
• 28:45
  almost over 70% year over year in terms of robotics investment.
• 28:52
  And kind of the big news this past week was the return of the
• 28:57
  Messiah. And I am talking about not Paul
• 29:02
  Muadib, not Jesus Christ, but Uber ex CEO Travis Kalanick.
• 29:09
  Travis came out with a a one word Twitter post called Atoms
• 29:14
  and that is the name of his company I guess rebranded from
• 29:18
  cloud kitchens. And he wrote a relatively long
• 29:21
  thesis treatise on why atoms and robotics is very important.
• 29:27
  But interesting thing was he was talking about stuff that's all
• 29:30
  based on the physical world, robots in mining and transports.
• 29:35
  Some of this may be related to, you know, his previous company
• 29:38
  was called Cloud Kitchens. So some of it was delivery of
• 29:41
  food. But now I guess we're getting
• 29:43
  into mining and transport. And his quote was at Adams.
• 29:46
  We make gainfully employed robots, specialized robots with
• 29:50
  productive jobs that bring abundance to their owners and
• 29:53
  society at large. And this is where I'm kind of
• 29:55
  getting back to that comment we were making about figure
• 29:58
  robotics is robots don't have to look like humans to be useful.
• 30:02
  Yes, that's interesting. Well, his background was, is of
• 30:04
  course in transport. He was the former Uber CEO.
• 30:08
  So it sounds like he obviously he knows that space really well.
• 30:10
  I think he's his rebrand away from cloud kitchens.
• 30:14
  I guess is it just going to be there there they're broadening
• 30:16
  the remit to something that's going to be very, very
• 30:19
  interesting. Another robotics related
• 30:21
  announcement this week, Mind Robotics is raising $500 million
• 30:27
  at a valuation I believe of $2 billion to build AI powered
• 30:32
  robotics. And this is a spin out from
• 30:34
  Rivia. I don't know if the Rivian CEORJ
• 30:37
  Scaringe is running that company, but it seems like
• 30:40
  they've been able to successfully raise quite a bit
• 30:42
  of money. Yeah, 5. 100 million and from a
• 30:46
  plus investors from I guess Andreessen led the round and
• 30:50
  Excel and the company's valued at at 2 billion.
• 30:53
  But I know, you know, I know just from personal experience
• 30:55
  with both Andreessen and Excel who have worked with in the
• 30:59
  past, both are excellent investors.
• 31:00
  I would, I think both those guys would say they they really fund
• 31:03
  and look for teams that they wanted back.
• 31:05
  And so RJ is, you know, he's a pretty special guy from his
• 31:11
  background at at Rivian and AI and robot seems to be all the,
• 31:15
  you know, all the all the thing right now.
• 31:17
  Our last couple of items here are about Replit raising at a $9
• 31:22
  billion evaluation. That's only six months after
• 31:25
  raising it a $3 billion evaluation and Replit is notable
• 31:29
  in the category of sort of automating dev tools and AI
• 31:32
  driven coding. It's a big step up from the last
• 31:35
  round. They are, they said at the time
• 31:37
  of the September round that they were on track for 150 million in
• 31:41
  ARR. And remember you and I looked at
• 31:43
  that and said that doesn't sound right.
• 31:44
  We think they're, we think they're a lot north of that.
• 31:46
  I think they are now telling the the Forbes that they're going to
• 31:48
  be at a billion dollars in, in ARR by the end of this year.
• 31:52
  So that'd be like 9 times. Look, you kind of now, yeah, the
• 31:55
  9 billion valuation sort of nine times arr, which sounds, you
• 31:58
  know, much more. Doesn't sound that crazy.
• 32:01
  I just wonder whether AI coding assistance like, you know, like
• 32:04
  Replit agent is that it's just, I wonder if, if that's in the,
• 32:06
  if that's in the cloud product road map.
• 32:09
  And if, if so, you know, then, then then you got to worry.
• 32:13
  But I, I imagine the investors you know, have have thought that
• 32:16
  through and are making a bet that that Replit is either
• 32:19
  defensible or as you say can use actually use AI to accelerate
• 32:22
  their business like other SAS companies have.
• 32:24
  All right. We're going to close out our
• 32:25
  session today to just feature one of our own deals.
• 32:28
  And on the Practical Venture Capital website, if you go to
• 32:33
  the SPV deals section, we're featuring an SPV we're running
• 32:40
  called Exit Fiesta. This is actually our secondary
• 32:42
  play in the Latin American market.
• 32:45
  We're raising $10 million for that vehicle.
• 32:47
  We've already raised 7 so far and we're looking for additional
• 32:52
  investors at minimum of 100,000. This will be investing in a
• 32:56
  pooled vehicle of about 8 to 10 top late stage tech startups in
• 33:00
  the Latin American market, many of them coming out of Brazil and
• 33:03
  Mexico. A lot of emphasis on logistics
• 33:06
  and payments and we've had some great success investing in the
• 33:11
  Brazil market, led by our partner in Argentina, Car
• 33:14
  Teneboy. Today with us we have Nick
• 33:29
  Fusco, Founder, CEO of PM Insights.
• 33:32
  Nick is providing pricing data to all the vampires out there
• 33:36
  like me who want to buy your shares and LP positions at
• 33:39
  ridiculously big discounts to value, but he knows the true
• 33:43
  price of everything in the market.
• 33:44
  Is that right, Nick Bold? Statement Well, we're focusing
• 33:48
  on what is the market truth or at least a fair value in the
• 33:51
  marketplace. And you know, the truth, the way
• 33:54
  we the way we approach it is going to be finding where mid
• 33:58
  market should be based on willing buyers and sellers.
• 34:01
  In the absence of distress, we're going to drop a very
• 34:04
  specific definition on us as opposed to taking model of
• 34:08
  something distress. But what happens if there is
• 34:10
  distress? Well, before we jump into all
• 34:12
  this, Nick, welcome to the show. Wonder if you could give us a
• 34:15
  quick intro on yourself and PM insights?
• 34:17
  And then we'll we'll dive into all the juicy gossip about
• 34:20
  sellers both desperate and not. Thank you for the introduction
• 34:23
  and thanks for having me. I've been in the financial
• 34:26
  services industry for about 20 years now.
• 34:28
  For the majority of that, I I've really done three main things,
• 34:33
  providing reference data, understanding what it is that
• 34:35
  we're looking after, maybe what we're trying to price.
• 34:38
  From there we try to price it or put a valuation on it and if we
• 34:42
  know what it is and where it's going and what the performance
• 34:45
  is, then work on making inputs for benchmarks, index products
• 34:49
  and things of that nature. So I've done that from 2006
• 34:53
  until now on credit default swaps, asset backed securities,
• 34:56
  CDs written on asset backed securities.
• 34:58
  So I I was working on the ABX and CMBX indices.
• 35:01
  Is that if anyone saw the movie The Big Short?
• 35:03
  That was a lot of fun in my early 20s.
• 35:05
  I was just going to say yeah. There's quite a time, but then
• 35:10
  if you yadda, yadda, yadda, that idea of that, what I saw as the
• 35:13
  Holy Trinity between reference data, pricing, valuation and
• 35:17
  then index, having done that across many different asset
• 35:20
  classes has really been the focus.
• 35:22
  So always been in OTC markets, predominantly fixed income for
• 35:26
  many, many years until yeah, I found my my way into your neck
• 35:30
  of the woods back in 2021. Wonderful.
• 35:33
  I wonder if you could give us a little bit of thumbnail on who
• 35:36
  your customers are. I, I, from what I understand,
• 35:39
  you're not a buyer, you're not a broker.
• 35:41
  You, you are a market data provider, a Bloomberg of sorts,
• 35:45
  maybe for private markets, but who do you interact with and do
• 35:49
  they pay you for that or not? And how you interact with the
• 35:52
  rest of the market? Sure.
• 35:54
  So foundationally the the business was built on give get
• 35:57
  models with a consortion of broker dealers be that we're not
• 36:01
  a broker dealer, we are non threatening partner, you know
• 36:04
  it. So they are very willing to
• 36:06
  provide us with bids, offers and trades where they might not be
• 36:10
  so willing to do that with another broker directly or an
• 36:12
  inter dealer broker that immediately would say, OK, we
• 36:15
  want a piece of this fee. And that allows us to create
• 36:18
  more composite and faster, more timely, you know, transparency
• 36:24
  in the market. And in return we commit that
• 36:26
  we're going to give them those composites and a lot of that
• 36:28
  information back. You know, everything's going to
• 36:31
  be obfuscated. So you're not going to tell
• 36:32
  broker B where broker C was or anything of that nature.
• 36:36
  But it is really helpful when we're trying to add some
• 36:39
  transparency here that will bring more participants to the
• 36:43
  market. So that's that's kind of the top
• 36:45
  layer and that's where it has to start.
• 36:47
  It has to start with that broker dealer community.
• 36:49
  And then from there, we'll really work with a number of
• 36:53
  different parties on free trade. So price discovery is very, very
• 36:57
  clear, best execution, making sure we're not getting hurt too,
• 37:00
  too bad on on the the entry point.
• 37:03
  The majority of our time is spent on pricing as an input to
• 37:06
  valuations, whether it's going to be the valuation itself based
• 37:09
  on the market price or looking at revenue multiples, looking at
• 37:13
  performance, looking at comparables in the public and
• 37:16
  the private market, things of that nature.
• 37:18
  And then once we can get there, we want to build comfort very
• 37:22
  much beyond that initial valuation for the auditors, your
• 37:25
  risk controls or anything of that nature and so on.
• 37:28
  So when we look at all those aspects, who needs that?
• 37:32
  We've got kind of a medley of single family, multifamily, RAA,
• 37:37
  VCs, asset managers, hedge funds in in that category.
• 37:41
  That's a different beast looking more at signals and they don't
• 37:44
  even love our composites. Let's say.
• 37:46
  They want to look at all the activity and the relationships.
• 37:49
  And then we go to a number of external third parties like your
• 37:54
  S&P, Dow Jones, MSCIS, other third parties that would want to
• 37:58
  use these data sets as an input to anything that they're they're
• 38:01
  building. So it could be a derived work or
• 38:03
  a complement to their services. We really want to be looked at
• 38:07
  as an infrastructure layer for a lot of those bits and pieces
• 38:10
  across the market. It sounds like the customers and
• 38:14
  partners that you work with are probably handling, I would guess
• 38:17
  six and seven figure transactions, maybe even more
• 38:20
  than that. Is that kind of in the range of
• 38:22
  the folks that you deal with? So we're really exclusively
• 38:24
  focused on 2 buckets, A1 to 4.9 million bucket and a 5 million
• 38:29
  plus bucket. The reason why maybe that 5
• 38:33
  million could even be considered a, a low threshold to a certain
• 38:35
  extent is because we need to make sure that we're protecting
• 38:38
  against when there might be only one person out there that could
• 38:41
  do a $50 million transaction, it'll be blended in with other
• 38:44
  five and $10 million transactions.
• 38:47
  So that's that's the area of the market that we focus on.
• 38:50
  OK. So I was off by a zero there
• 38:51
  more like 7 and 8 figure transactions, correct.
• 38:54
  I guess I don't know if you deal with 99 figure transactions or
• 38:56
  not, but they're in there. There's just a better sense of
• 38:59
  of your market. OK, Now that you're in the world
• 39:01
  of private markets, what do you see as kind of the challenges
• 39:04
  that people are facing on a daily basis?
• 39:06
  Obviously, lack of access to data or lack of pricing is a big
• 39:10
  one. 1st, we, we don't really assist with access, right?
• 39:14
  That's going to be for you guys, OK?
• 39:16
  You're going to help your customers, your LP base access,
• 39:20
  access names or access funds or access, you know, the type of
• 39:23
  exposure that they want to have. But adding that level of trust
• 39:27
  and understanding of when something is getting sold to me,
• 39:31
  can I understand how it's going to be priced and what the
• 39:33
  volatility was historically? Because sometimes maybe someone
• 39:37
  looked at the market and they said, oh, wow, this is great.
• 39:39
  There's no volatility over here. But what we can tell from the
• 39:42
  data is, is absolutely, that's not terribly true.
• 39:45
  And when you look at the ways maybe AGP versus AA40 act fund
• 39:51
  would mark there, there will be some differences, right?
• 39:54
  I mean, you can, you can really accommodate your, your investors
• 39:58
  wishes by saying, oh, I use cost basis maybe or I use the last
• 40:02
  funding round. And and that always held for a
• 40:04
  long time. And we, we've started to see
• 40:07
  that dynamic change, particularly with what we call N
• 40:10
  port data that we we would pull from SEC filings from mutual
• 40:13
  funds and we could see that. It went from a step pyramid like
• 40:17
  an Aztec pyramid in the past where everyone in the market was
• 40:20
  really focused on, hey, what was the last round?
• 40:22
  We'll leave it there until we have new information in primary
• 40:25
  to taking into account the secondary information.
• 40:29
  And I think that's, that's kind of been a been a benefit to the
• 40:32
  market as a whole with all the money rushing in of the degree
• 40:35
  of comfort that that accredited investors and family offices and
• 40:39
  you know anyone that I'd previously mentioned had
• 40:42
  interest in. And then we have a fun runabout.
• 40:47
  We import from France a Blanche honey.
• 40:51
  Hello, Siri. Hello.
• 40:53
  This is Blanche card. Bonjour.
• 40:57
  Bonjour. Look at that suspension cheek,
• 41:02
  isn't she? Who do you see as competitors or
• 41:07
  other players? And we we've had a few other
• 41:09
  folks in the show, I guess Javier about this at Cap Light
• 41:13
  and Tyler Hendrickson that notice.
• 41:16
  Yeah. So the the main competition that
• 41:18
  that we tend to focus on is kind of that the mindset of how
• 41:21
  something would be priced. So if I look at the way the, the
• 41:28
  cap lights or notices of the world might price something if,
• 41:31
  if AI will get used or if their own crosses might get used.
• 41:36
  So they're, they're both doing broker dealer activity to my
• 41:39
  understanding. So we just don't want to be that
• 41:42
  we, we need to be arm's length. We want to be really able to
• 41:45
  accommodate an audit. So we don't want to be part of
• 41:47
  the economics of the trade. And also we don't charge our
• 41:50
  brokers, you know, there's, if, if we help the broker with
• 41:52
  something or just introduce something, we walk away.
• 41:55
  That's it. It's theirs to deal with and we
• 41:58
  never know the end counterparty. We don't, we don't want to be
• 42:01
  participatory to that, but the big competition is really like
• 42:05
  the traditional sense of how the big providers would be providing
• 42:08
  valuations based on let's say a discounted cash flow.
• 42:11
  Did public sector comparables point in time information the
• 42:15
  heavily model based or assumption based approach to how
• 42:19
  valuations would have been done historically?
• 42:21
  But can you share an example of a company where you see
• 42:23
  different market participants holding shares in the same
• 42:27
  company but at at different valuations and you know just
• 42:30
  that? Does that happen?
• 42:30
  And is there a dispersion? And like why would that be?
• 42:33
  Why would two different parties just be holding it at at
• 42:37
  different levels? What are some of the the causes
• 42:39
  of that? Here's an example of if you
• 42:41
  could see data bricks hopefully. And we look at the share price
• 42:44
  over time and all the triangles are the hollow triangles are
• 42:47
  going to be bids and offers 1 to 5 million in size.
• 42:50
  Everything that's filled in is going to be 5 million plus.
• 42:53
  The squares are going to be your exact transactions.
• 42:55
  And then I get rid of all that, right?
• 42:58
  So now we're focused on the composite.
• 42:59
  We're looking at secondary. So it's all the activity between
• 43:02
  these funding rounds. I'll show you these fun marks.
• 43:05
  So we'd never use a fun mark as an input to evaluation because
• 43:09
  you know, we, we find that we don't want the circular
• 43:11
  reference. If one of these funds is a
• 43:13
  client of ours and then they're taking our price and they're
• 43:15
  using our price, we're we're just going to get a circular
• 43:17
  reference. But that example calling on Bank
• 43:20
  of New York Mellon here, they're, they were pricing the
• 43:22
  common of data bricks at 102 per share.
• 43:24
  That's great. They had a certain assumption
• 43:27
  that was going to be a little bit different or or in this case
• 43:30
  quite a bit different than let me see who this is, this fund
• 43:35
  that was marking at $58 a share. But.
• 43:38
  Even here, yes, they're marking the preferred at almost half the
• 43:42
  price of the common. So anyone in market would say,
• 43:45
  OK, First off, that relationship should be flipped.
• 43:48
  And maybe, you know, that like might have a much more
• 43:51
  conservative approach. Maybe they have different
• 43:54
  information or, or perhaps really the IT it's that the
• 43:58
  assumptions were different. Some someone in the mix was
• 44:01
  having a very different assumption of where things
• 44:03
  should be. And you could see really when
• 44:04
  you're taking into the market activity.
• 44:06
  Let's just go to valuation. You should have a good
• 44:09
  indication of of how valuation will change over time and we can
• 44:14
  then do those comparisons. I know you didn't ask for this,
• 44:16
  but I'll, I'll throw up Snowflake.
• 44:18
  You can look at, you know, how does Databricks relate to a
• 44:21
  company like a Snowflake And if, if in that case, right, you
• 44:26
  know, one of those parties was using Snowflake as a heavy
• 44:29
  comparable for data bricks, you could see the the correlation is
• 44:33
  just falling apart. So that was now a bad assumption
• 44:36
  if you were saying that the secondary market is quite
• 44:38
  reliable or reflective of what a good valuation should be or
• 44:42
  maybe a predictor of the next funding round.
• 44:44
  Maybe for a little while there those two stocks were tracking
• 44:47
  closer. Looks like they started
• 44:50
  diverging September 24th. That's right.
• 44:53
  Do you provide speculation on why that occurred or you just
• 44:57
  like show that it did occur? So everything that we're going
• 44:59
  to provide is historical. We do provide research reports,
• 45:04
  bull bear cases, you know why, why we think something is, is
• 45:08
  the way it is. But a lot of the work is going
• 45:10
  to be based off comparables. So we can allow the customer to
• 45:14
  even load up revenue so they can look at things like revenue
• 45:18
  multiples. We've actually done this
• 45:19
  specific not too long ago, so this is very relevant to ours,
• 45:26
  our listeners. One thing that just you you take
• 45:28
  away from private markets and I guess public markets as well
• 45:30
  with your background is that different types of investors
• 45:33
  will set valuations differently. Just a like in the public
• 45:35
  markets you might think of smart money versus retail money.
• 45:38
  Like retail investors will value certain stocks in a certain way.
• 45:42
  The, the hedge funds or the sharps will do it in a different
• 45:44
  way, different methodology, different expectations, maybe
• 45:47
  different levels of information In private markets, you
• 45:49
  definitely want to be, you know, looking at pricing set by
• 45:52
  sophisticated venture capitalists, ideally on the
• 45:54
  board with information to sometimes these private
• 45:57
  companies turn into mean stocks or catch a retail bid.
• 46:00
  These separate out like you, you talked about Bank of New York
• 46:02
  and, and, and where they're holding it.
• 46:04
  And why can I just exclude Bank of New York or only include or
• 46:07
  exclude for certain of family officers to pick and choose
• 46:10
  who's setting the valuations? And so I get a, you know, maybe
• 46:13
  a better signal that I trust from people I trust versus noise
• 46:16
  in the market. Yeah, of course.
• 46:18
  So we'll provide all the underlying detail to the last
• 46:21
  thing I had pulled up was if you're not happy with with the
• 46:24
  way we've modelled something there was that 3D visualization
• 46:27
  that could even be time series to see the differences of of
• 46:31
  multiples over time versus growth rates versus valuation
• 46:33
  versus sector versus public versus private, anything you
• 46:36
  want that's your calculator. So I will say we, we always call
• 46:40
  ourselves a data company, but we're more of a data and
• 46:42
  analytics or model provider as well, right.
• 46:45
  So you could have that box and then when we do our composite,
• 46:49
  we never would have used Bank of New York or or Valic or, or
• 46:53
  anyone in that case as an input to anything.
• 46:55
  It's just exclusively going to be where the the bids, offers
• 46:59
  and trades are rolling through. But we do have very
• 47:01
  sophisticated customers that might say, you know what, I love
• 47:05
  what you guys do, but I would much prefer to heavily weight
• 47:08
  towards anything that's 5 million and above and exclude 1
• 47:11
  to 5,000,000 all together. And then they have the ability
• 47:14
  to do that on, on their side. So we, we will then we could
• 47:18
  always stand by the data. But if, if we were going to go
• 47:21
  to an audit, we'd suggest, you know what you've got that
• 47:23
  composite and that's, that's really a lot of our margin.
• 47:26
  How we get paid is when you come to us and you buy a price,
• 47:29
  great. We're selling that price to
• 47:31
  everybody. You know, everyone in the market
• 47:33
  has the same price. It's it, it, it's kind of a
• 47:35
  ticker. But then you'll, you'll
• 47:37
  obviously see the, the different funds marking different
• 47:40
  directions. And then also the levels of
• 47:42
  scrutiny will differ depending on who you are.
• 47:45
  And then airing a little bit away from last primary round,
• 47:49
  particularly when we saw the market slump in 2022-2023 and
• 47:53
  the distance between the funding rounds elongated so much,
• 47:56
  everyone was like, whoa, can we even use that?
• 47:58
  Well, Nick, can they even use that at this point four years
• 48:02
  later or or not? Because we're, we're still
• 48:04
  seeing those valuations on, on the books that we look at at
• 48:08
  least some of the time. And our, our questions were
• 48:11
  always like, hey, are the, aren't the auditors?
• 48:13
  Are they auditors doing their job?
• 48:15
  Are LP's, you know, holding their GPS accountable?
• 48:18
  And why are we still sealing seeing those valuations on the
• 48:22
  books when they're theoretically at least two years out of date,
• 48:25
  if not 3 or 4? Well, it's, it's, it's super
• 48:27
  interesting because you can easily jump to the three or four
• 48:31
  and, and I think you can jump to that because let's say you have
• 48:35
  an asset holder using last round.
• 48:37
  That's great. And then we find out through,
• 48:39
  you know, whether it's a FOIA request or any, any sort of way
• 48:43
  to, to ascertain what that holding was.
• 48:45
  You could say, you know, the Q3 marks just came out.
• 48:49
  They're always like quarterly delayed and then there's maybe a
• 48:51
  month delay on top. There are many service providers
• 48:53
  that can focus on that. The interesting part that I hope
• 48:57
  we do for this market or complement in this market is if
• 49:01
  we can have a really good sense of where things have gone since
• 49:05
  that publication. And since understanding how all,
• 49:08
  all of let's say your peer group or the big, the big asset asset
• 49:11
  owners, you're looking at pensions and balance and things
• 49:13
  like that where they're holding. When that's getting publicly
• 49:16
  disclosed, we can see what's happened since.
• 49:20
  So I think that that becomes a really cool part of the puzzle
• 49:26
  to see where the puck is headed. You mean that's all?
• 49:33
  Well, we do have. 1 old demonstrator model with a lot of
• 49:37
  mileage rouser. Coming, Sir.
• 49:43
  Here I am, Sir. But I think, you know, from our
• 49:48
  reference point, you know, both when we were primary fund
• 49:51
  managers, you know, the rules that we had were that you really
• 49:55
  can't hold a price for much more than two years, even 2 years
• 49:59
  might be pushing it. But that was probably the limit
• 50:02
  of an auditor's acceptance of the last primary round.
• 50:05
  And, and I guess if you could say, hey, we, we know something
• 50:07
  about the company and metrics are still heading in the same
• 50:11
  direction. Therefore, we're going to keep
• 50:12
  those numbers in the same position.
• 50:14
  But you know, we, we sort of know that the public market
• 50:17
  repriced within three months of events back in Q1.
• 50:22
  Q 222. And so these companies didn't
• 50:24
  have any other activity going on.
• 50:26
  You would think that at least by, I don't know, mid 24 people
• 50:31
  should have said, well, there's been no activity.
• 50:33
  We know the public market repriced.
• 50:35
  Shouldn't we be proactively marking these companies down
• 50:39
  sometime in 24, certainly, you know, by the end of 24, early
• 50:43
  25? Well, we're now a year past that
• 50:45
  time frame, if not a year and a half or two past that time
• 50:48
  frame. What's the justification for any
• 50:51
  GPS holding numbers, you know, prior to Q2?
• 50:55
  Q 122 valuations at this point in time.
• 50:58
  Just just to make it interesting.
• 51:00
  Well played devil's advocate right, though I wholeheartedly
• 51:05
  agree with you. I wholeheartedly agree with you
• 51:06
  that that you know you don't want to encourage using a stale
• 51:10
  mark for two years. It just it.
• 51:11
  It doesn't make any sense, but the only justification in my
• 51:14
  mind 2. Years might be the limit, but
• 51:16
  three or four years, yeah, 3-4 years, he'd be like, oh, that's
• 51:19
  not just stale, that's rotten fish.
• 51:24
  If you're comparing a company that might use comparables from
• 51:29
  the public market that are close enough, but there is no proper
• 51:32
  comparable, let's say Open AI, Anthropic or a SpaceX of the
• 51:36
  world, are you going to compare SpaceX to Raytheon or Boeing?
• 51:41
  And then when the revenue multiple just is just wildly
• 51:44
  divergent and you're just saying, you know, you know,
• 51:47
  Boeing's been pretty flat for a few years, Maybe we could just
• 51:50
  leave that SpaceX position flat. Nobody in their right mind would
• 51:53
  have really go in that direction.
• 51:55
  But maybe there's some sort of flip side of that coin where,
• 51:57
  you know, we can justify it because of it being so steady
• 52:02
  Eddie. But you know, you get that more
• 52:03
  maybe in the non venture market for privates.
• 52:06
  So yeah, I, I'm trying to be lenient here, but I would love
• 52:09
  everyone you're talking about or pointing fingers at to be a
• 52:12
  customer of ours because I think we'd be, I'm not pretty
• 52:15
  beneficial anyone. Nick, I'm not, I'm not pointing
• 52:18
  fingers at it, at it. Gosh darn soul.
• 52:20
  Well, where, where do you see kind of the the biggest
• 52:23
  mispricing going on? You know, we always leave that
• 52:26
  to our customers. We're again, we're not trading,
• 52:28
  we're not trying to facilitate trades or anything like that,
• 52:31
  but if someone could garner ideas from our data, have at it,
• 52:34
  you know, and I think some of the more interesting ideas are
• 52:37
  the ones that are, are heavily reactive.
• 52:40
  So if you really dig into the books of what some major public
• 52:44
  companies are holding, right? So if you're looking at public
• 52:48
  companies holding these private, private companies on balance
• 52:52
  sheet, that's pretty, pretty cool.
• 52:54
  And if we could all of a sudden realize that maybe the public
• 52:57
  company which is transacted on an exchange could be transacted
• 53:01
  and it has such a big position in some of these private
• 53:04
  companies that it actually starts to matter.
• 53:07
  And I'm not just talking like a Microsoft and Open AI.
• 53:11
  You know, if you look at what Salesforce owns, you look at all
• 53:15
  all these major companies and understanding if it could have a
• 53:18
  profound effect on the public stock.
• 53:20
  And it could be even be when a public stock gets beaten, the
• 53:24
  living daylights out of it in the market like what we're
• 53:26
  having right now. Maybe because the private market
• 53:29
  is not as correlated as everyone has thought.
• 53:32
  There's some trading opportunity there in the public market
• 53:36
  actually transacting a stock, not even a fund.
• 53:40
  And and that's getting you indirect access or, or direct.
• 53:43
  Yeah, indirect access to the private stocks.
• 53:45
  I just think that's one of the the more fascinating approaches
• 53:48
  here that maybe isn't, isn't getting picked up on or no one
• 53:52
  wants to call attention to because they they want to reap
• 53:55
  the reward. And then everyone finds out how
• 53:57
  smart they were a few years later.
• 53:59
  I guess one thing that we've seen from from our world is that
• 54:03
  there's substantial disadvantages for sellers of
• 54:07
  smaller transactions in the market, transactions where there
• 54:11
  hasn't been any primary around pricing or financial information
• 54:14
  disclosed recently possibly in geographies that aren't covered
• 54:18
  as often, you know, definitely outside the US And then
• 54:21
  complexity is another one. So you know, the discounts
• 54:23
  against fund interests are much more substantial than they are
• 54:27
  around single assets. Do you see that reflected in
• 54:29
  your data? Are there any, you know, sort of
• 54:31
  advantages for secondary buyers where because of size or
• 54:34
  geography or stage or complexity there, there's a mispricing
• 54:40
  going on in the market? Sure.
• 54:41
  The the one that's most tangible for us is that, that dynamic
• 54:45
  that you you hit on between institutional and retail.
• 54:48
  So you're not just seeing it in the price per SE, the the
• 54:52
  upfront price, you're actually in a lot of cases seeing it on
• 54:56
  the terms. So we do track when something is
• 54:59
  coming through, if it's a, a 00 SPV or two and 20 SPV, we track
• 55:04
  it if it's going to be a direct on cap table.
• 55:07
  And then we look at the size and invariably we've noticed.
• 55:10
  And part of the reason why we just don't use data South of the
• 55:13
  $1 million mark is because we, we found that the retail
• 55:18
  transactions that we we've observed or, or seen come
• 55:20
  through are, are just typically further, further away from where
• 55:26
  any sort of composite or truer institutional size takes place.
• 55:30
  And, and that's where you really want to, to have some like
• 55:33
  yourselves that will deal institutionally on behalf of
• 55:35
  many customers to be the fiduciary, really step in and
• 55:38
  understand where the deals are going to be had or just again,
• 55:43
  just making sure you're not getting your, your face ripped
• 55:45
  off on, on the transaction. For anyone doing retail, we do
• 55:49
  host an SPV calculator on our front end.
• 55:52
  It's free. You just go to the marketing
• 55:54
  website. But but I would, I would say,
• 55:56
  yeah, the, the delta between a retail and institutional is, is
• 56:01
  sometimes quite egregious. And it might not be clear
• 56:04
  because maybe the the composite mark that we're putting up,
• 56:07
  which is based on, you know, a director of 00 SPV, that's how
• 56:12
  we normalize it. It.
• 56:14
  If somebody's getting a three and 30 or something like that
• 56:16
  and it might have the exact same price you got to, they have to
• 56:19
  just understand they're not getting the same deal.
• 56:24
  I like her and I'll take. Oh, thank you, ma'am.
• 56:28
  Thank you. I think there's, there's one
• 56:34
  thing that we've been trying to work on as a strategy in the
• 56:37
  market, which we're calling, you know, Secret Stallions.
• 56:40
  And I guess I would say there's a, there's a big bifurcation in
• 56:43
  the market when you're talking about the Private Mag 7 and
• 56:46
  SpaceX Anthropic open AI are certainly in huge demand, tough
• 56:50
  to get access, tough to get there without, you know, paying
• 56:53
  a premium, maybe multiple layers of SPBS and, and probably big
• 56:57
  check that's even true, I would say for the next 3040 companies.
• 57:01
  But once you kind of get past that, that list, I think there's
• 57:04
  a couple of 100, maybe even up to 3 or 400 companies out there
• 57:07
  that are kind of below the radar for a lot of people.
• 57:09
  And that's typically where we go hunting.
• 57:11
  This is a a diagram we use to sort of talk about, you know,
• 57:14
  what's in the market. And these aren't like busted
• 57:17
  unicorns or horses that are too slow, too small to ever get to
• 57:20
  an exit. But they're kind of in that
• 57:22
  Goldilocks spot of what we think is big enough to go public or or
• 57:25
  at least in the near term, go public and get to an exit, but
• 57:28
  not on everybody's radar. And we found that there's just
• 57:32
  some really substantial mispricing employees who might
• 57:35
  be selling or early investors who might be selling
• 57:38
  particularly smaller transactions below 5 to 10
• 57:40
  million. It's a very inefficient market
• 57:42
  for those sellers. And and we can actually buy
• 57:44
  those positions at pretty decent discounts to where the market
• 57:48
  might have valued them in the last year or two.
• 57:51
  That's right. Do you see that from your data?
• 57:53
  Yeah, hopefully you saw my eyes light up here because the, the
• 57:57
  most recent thing that we've we've come out with is kind of
• 58:00
  what I, what I call maybe a dumb name, our time machine, it's
• 58:04
  that 3D rendering, right, So. So not a hot tub time machine, I
• 58:08
  hope 'cause I think that's already taken by somebody.
• 58:10
  Yeah, a little, a little different.
• 58:12
  So the the general sense is let's say that we were to turn
• 58:15
  back the clock on all the companies that are going game
• 58:19
  vesters and are now in the private Max 7 right.
• 58:22
  What did they look like when they were in the exact sleeve
• 58:25
  that you had? What what was the criteria, the
• 58:27
  investment criteria that could match or what was that dynamic?
• 58:30
  What did the, the executive team look like?
• 58:33
  What did the prior VCs look like?
• 58:35
  So that time machine, we, we, we encourage like we want people to
• 58:40
  be using that going in the Wayback machine, the time back
• 58:44
  machine, not the hot tub and utilizing that information to,
• 58:48
  to get a sense of, hey, the companies that I'm assessing,
• 58:52
  company X, company Y, company Z, do they look similar or have
• 58:56
  they been on that similar path? Have they been raising at
• 58:58
  similar multiples to what then became the next mag 7?
• 59:02
  So I love it the. Areas we tend to go after is
• 59:04
  like, hey, can we see a, a disconnect between the
• 59:06
  fundamental valuation of this company and where we can buy it.
• 59:09
  And when we look at, you know, those top five to 10 companies
• 59:13
  or, or even the top 30 companies, you know, the
• 59:16
  valuation is kind of way ahead of where you might think of
• 59:19
  fundamentals. Maybe you could just fight based
• 59:21
  on growth. But when we start talking about
• 59:24
  AI or robotics or space or defense, often times, you know,
• 59:29
  we're not really looking at revenue growth and profits
• 59:31
  anymore. We're, we're talking about, you
• 59:33
  know, science fiction and cool stuff, and people get really
• 59:37
  excited. Yvonne recently had with one of
• 59:40
  the Colson brothers over of Guinness and and talking about
• 59:43
  putting the data centers in space and how the the the
• 59:46
  efficiency there and the power that could be created there.
• 59:49
  I mean, sounded science, science fiction to me five years ago,
• 59:52
  but today I'm like, no, I'm all for it.
• 59:54
  That makes perfect sense. You're you're just so far ahead
• 59:57
  of me. Yeah, every.
• 59:58
  Everybody's all over that. I, I think the timing is the
• 1:00:01
  question, not necessarily the math or the physics there.
• 1:00:03
  It's we're like, is that achievable in three years or 10
• 1:00:06
  years? Well, if you if.
• 1:00:08
  You do an IPO that'll. Eli is actually bring some
• 1:00:10
  really light. He's just predicting the future
• 1:00:12
  really, really optimistically. Sorry, go ahead, I missed that.
• 1:00:17
  No, I appreciate it. I mean, if he if he's going to
• 1:00:19
  be able to pull capital from a new IPOI mean that that should
• 1:00:23
  speed things up too. Not that the man has a shortage
• 1:00:26
  of funds, but it it's certainly creating that hype cycle as
• 1:00:31
  well, right. So if you really think about
• 1:00:32
  why, maybe the interest has just exploded and you know, anybody
• 1:00:37
  talking to their financial advisor on a golf course is now
• 1:00:40
  talking about do they own SpaceX or not?
• 1:00:43
  It's, yes, it is, is very transparently in in everyone's
• 1:00:47
  face. He is a public figure as has
• 1:00:50
  been for a long time, but even more so.
• 1:00:51
  And then the talk of the IPO has brought it even more as almost
• 1:00:55
  feeling like it's tangible, you know, so, so I think that that
• 1:00:58
  momentum is really there for, for those purposes as as high as
• 1:01:03
  the valuation is. Yeah, Yeah.
• 1:01:06
  Well, that's good marketing. And it opens up the whole class
• 1:01:09
  and new investors, we do go public this now it's out of the
• 1:01:11
  it's into the public realm. It's not just limited to
• 1:01:14
  accredited investors and family offices.
• 1:01:16
  Now anyone who wants to piece of it, you know, can can get in.
• 1:01:19
  And that creates its own, its own hype, I guess, and its own
• 1:01:22
  demand, its own marketing around the project.
• 1:01:24
  And, and hopefully with, with Elon, the holders of SpaceX, you
• 1:01:28
  know, as they continue and as they, they take an IPO, maybe
• 1:01:31
  they have further interest in going back into the private
• 1:01:34
  market. And that's a, that's a benefit
• 1:01:35
  to our whole ecosystem in a, in a lot of cases, you know, the
• 1:01:39
  continued interest in private is is not going away to, to all of
• 1:01:42
  our benefits. You know, Nick, we've covered a
• 1:01:44
  lot of stuff today. I'm just wondering if there's
• 1:01:48
  any kind of insights that you feel like we've missed or any
• 1:01:50
  particular behaviors in your market that you think are
• 1:01:53
  interesting. Just a analysis is becoming
• 1:01:56
  easier, easier in this market with more transparency, more
• 1:01:59
  data breadth and depth. Really if you, if you thought
• 1:02:02
  the the access to VC was, was just too illiquid, I mean you
• 1:02:06
  secondary players like yourself solve for a lot of that, you
• 1:02:09
  know, begin being able to show that there is true liquidity and
• 1:02:12
  we could reflect that in data too.
• 1:02:14
  So getting comfort with with exposure and taking bigger
• 1:02:17
  exposure or or a decent portion of the portfolio to to generate
• 1:02:21
  alpha through VC with comfort, understanding where when you
• 1:02:25
  provide a valuation where that valuation has come from and that
• 1:02:28
  it's truly from market. And it's not as though we're
• 1:02:31
  going to have 100% write down from 1 cord to the next because
• 1:02:34
  we've been able to successfully track these companies.
• 1:02:37
  Again, it's almost like saying we're just trying to make things
• 1:02:39
  apples to apples across various different asset classes.
• 1:02:43
  I think I saw that the Robin Hood Private Markets venture
• 1:02:46
  fund just launched today. Do you think we're likely to see
• 1:02:49
  more public, private crossover, I guess virtualization,
• 1:02:52
  tokenization of stocks? Yes, simple answer.
• 1:02:57
  I, I think we, we, we already have it trending in that
• 1:03:00
  direction, right? You're, you're seeing more fun
• 1:03:03
  complexes dip into more private stocks.
• 1:03:07
  I, I think the, the, the general trend is you want to produce
• 1:03:12
  funds that are reflective of where, where the demand is and
• 1:03:15
  the demand is truly there. You know it.
• 1:03:17
  I don't think Robin Hood are doing it for fun.
• 1:03:20
  There's there's certainly a demand behind it.
• 1:03:22
  So would love to see, you know, great success across the board.
• 1:03:26
  Rising tide lifts all ships type type idea.
• 1:03:29
  I think there's a lot of people that will call into question
• 1:03:32
  different fund structures and what's better than the rest.
• 1:03:35
  And you know, if you're going to tie it back to the underlying
• 1:03:38
  asset value, we're here to help. We'll always do that.
• 1:03:42
  Well, Nick, thanks for spending some time with us and give us a
• 1:03:45
  thumbnail on kind of what you guys do over a PM insights.
• 1:03:48
  I I have a feel like I'm going to make a follow up, call myself
• 1:03:50
  here to learn more about the product and see what kind of
• 1:03:52
  insights we can get. The Time Machine.
• 1:03:54
  There we go. Thank you so much.
• 1:03:55
  Come. On let's go home, Rosie.
• 1:04:02
  Yes, ma'am. We can't guarantee how, you
• 1:04:06
  know. Welcome back to Another
• 1:04:19
  Evaluation Corner with Professor Aman.
• 1:04:21
  Virgie and Aman. We're covering Cerebrus this
• 1:04:24
  week. We are indeed fresh off of their
• 1:04:27
  latest announcement of a new fundraising.
• 1:04:29
  They are now on the that Augment top 20 list that we showed at
• 1:04:33
  the outset of our show. They got there with a $23
• 1:04:37
  billion evaluation. They just announced it last
• 1:04:40
  month with Tiger leading and then participation from
• 1:04:44
  Benchmark and Altimeter. So a a really good solid group
• 1:04:47
  of venture and financial investors.
• 1:04:50
  If you remember back in September 2025, they had
• 1:04:54
  announced a new round. They were trying to go public at
• 1:04:57
  an $8 billion evaluation. We knew that Fidelity was going
• 1:05:00
  to anchor the IPO, but then they pulled the IPO in October and
• 1:05:03
  said instead we're just going to do a private round with Fidelity
• 1:05:06
  in the league and then a few other strategics and and a few
• 1:05:09
  other investors at an $8 billion valuation, they're generating a
• 1:05:13
  couple 100 million in revenue or they were in 2024.
• 1:05:16
  So at the time it looked like a pretty aggressive valuation for
• 1:05:19
  an IPO and their timing wasn't great as they sort of hit the
• 1:05:23
  back half of the year. Some information came out in
• 1:05:26
  their S1. They just talked about customer
• 1:05:28
  concentration and some other risks, which we're, you know,
• 1:05:30
  we're going to go through. But the good news is they got
• 1:05:33
  that $8 billion evaluation done. They're now at a $23 billion
• 1:05:35
  evaluation with a new set of investors.
• 1:05:38
  So overall, a really interesting company to watch probably on the
• 1:05:41
  IPO path again for this year. What do they do?
• 1:05:44
  Well, they are competitor to NVIDIA.
• 1:05:46
  They are actually the designer of the world's fastest AI
• 1:05:49
  infrastructure and chips for training and inference.
• 1:05:52
  You might ask, how is it that they are so much more efficient
• 1:05:56
  and so much faster than NVIDIA? Well, their flagship technology
• 1:06:00
  and the wafer scale technology that they have is much larger
• 1:06:03
  and much more efficient than what NVIDIA does.
• 1:06:06
  It's apparently about 20 times faster.
• 1:06:08
  But the thing they sell, the wafer scale engine, is actually
• 1:06:11
  50 times larger than the Jeep and largest GPU sold by NVIDIA.
• 1:06:16
  So it is a interesting position versus NVIDIA.
• 1:06:20
  I think their IPO story is going to look something like this.
• 1:06:23
  They closed 2025 with around a billion dollars in revenue or
• 1:06:29
  revenue run rate. So that would that would be that
• 1:06:31
  they were traded at 23 times forward revenues.
• 1:06:35
  NVIDIA trades at about 23 times forward revenues now, AMD trades
• 1:06:39
  at 9 to 10 times revenues and Intel's at 4:00 to 5:00.
• 1:06:42
  But look at the growth rate for Cerebris.
• 1:06:44
  They're basically they're more than doubling their revenues
• 1:06:46
  year over year. In fact, if they generated 206
• 1:06:49
  million in revenue last year and they're projecting close to a
• 1:06:52
  billion this year, they might be up about five times in revenue
• 1:06:55
  growth. So they do have a great story in
• 1:06:58
  terms of their their acceleration of the top line.
• 1:07:01
  A couple of risks we just want to highlight, if you read
• 1:07:03
  through the S1 that they filed a year ago, you got to read into
• 1:07:07
  the fine print. So page 9, it's the beginning of
• 1:07:09
  the risk factors. Every company that goes public
• 1:07:13
  publishes about 50 pages of risk factors and it's it's easy to
• 1:07:17
  tune out 50 pages of what can go wrong.
• 1:07:20
  You don't have to read the risk factors of cerebrous I have so
• 1:07:24
  you don't have to. This is the biggest 1 you have
• 1:07:27
  to really pay attention to. They have a huge customer
• 1:07:29
  concentration risk. They they tell you right in
• 1:07:31
  there S1, but a significant portion of the revenue comes
• 1:07:34
  from a limited number of customers.
• 1:07:36
  And it turns out that the the big customer is G42.
• 1:07:40
  So that should mean like, hey, I want to learn.
• 1:07:41
  I want to learn more about that. Like keep going forward to page
• 1:07:43
  21 of the risk factors and now they tell you the specifics.
• 1:07:47
  Group 42 holdings accounts for between 83 and 87% of the total
• 1:07:52
  revenue for 2023 and then six months of 2024.
• 1:07:58
  So when 1 customer owns 83 to 87% of your revenue, a lot can
• 1:08:02
  go wrong. That customer could renege, that
• 1:08:04
  customer could renegotiate the contract, Maybe that customer
• 1:08:08
  switches to another provider like NVIDIA.
• 1:08:11
  It's also a bit of a red flag. I think that G42 is the biggest
• 1:08:14
  customer. So like who does NVIDIA sell to?
• 1:08:17
  It's Microsoft and Meta, it's Amazon, right?
• 1:08:20
  It's the hyper scalers. They're working with Open AI,
• 1:08:22
  they're working with Anthropic Cerebral System, working with
• 1:08:26
  any of those guys, they're working with the G42, and then a
• 1:08:29
  very small list of other hyper scalers and US based customers.
• 1:08:32
  So that's that customer risk. That customer concentration I
• 1:08:35
  think is a pretty major red flag.
• 1:08:37
  I think it's probably enough for them to, if I were to, if I were
• 1:08:41
  to redo the IPO story for you at NASA, Hey, let 2025 revenue is a
• 1:08:44
  billion dollars, which is great and the growth rate is 5X which
• 1:08:48
  is fantastic. But G42 is making up most of our
• 1:08:51
  revenue in the couple of quarters.
• 1:08:53
  And so there's a couple of, there's a couple of things we as
• 1:08:55
  a company have to do to mitigate that.
• 1:08:57
  One, you understand the concentration risk.
• 1:08:59
  They'd probably have to show a pipeline of revenue outside of
• 1:09:02
  the Middle East, outside of G42 to demonstrate that that revenue
• 1:09:06
  really is durable. You also have to wonder in this
• 1:09:09
  environment whether I know the UAE is an ally of the US right
• 1:09:12
  now, but if there's any kind of geopolitical risk in the Middle
• 1:09:15
  East, God forbid there are export controls put on the UAE
• 1:09:18
  for whatever reason by the current administration.
• 1:09:21
  That's just a existential risk to their business.
• 1:09:24
  If you keep looking at their S1, there's other stuff in there
• 1:09:27
  which you can also find. They've given a very sweet deal
• 1:09:31
  to G42 in the form of a preferred stock purchase
• 1:09:34
  agreement. The company G42 has the right to
• 1:09:37
  buy 23 million shares of their F2, their series F2 convertible
• 1:09:42
  preferred stock, at only $14.66 a share.
• 1:09:46
  So what does that mean for Cerebra shareholders?
• 1:09:49
  Well, in November of 2019, they raised at a $2.4 billion
• 1:09:54
  evaluation. The per share price at the time
• 1:09:56
  was 1832. They then raised again in 24,
• 1:10:00
  they raised a billion dollars at $8.1 billion evaluation and the
• 1:10:04
  per share price is 3623. So the fact they've given the
• 1:10:08
  right to G42 to buy 10% of their company at a strike price of
• 1:10:13
  only $14.00 a share just means that upon the IPO, if the share
• 1:10:17
  price does well, shareholders will be diluted because G42 will
• 1:10:21
  exercise the option. They'll buy at 1466.
• 1:10:24
  And then the higher the price goes upon the IPO, the worse
• 1:10:27
  they, the worse the dilution is just because the strike price is
• 1:10:30
  going to be more and more and more in the money.
• 1:10:32
  So my guess is Cerebrus will not be able to pull off the IPO that
• 1:10:35
  they want under those terms. I think what they will do in the
• 1:10:39
  next month or two is probably announce a broadening of their
• 1:10:42
  streets of partnerships. Maybe they're selling to Amazon,
• 1:10:45
  maybe they're selling to, you know, one of the other hyper
• 1:10:47
  scalers. The fact they can raise new
• 1:10:49
  money from new investors is a really good sign.
• 1:10:51
  But I think they have to work on that customer concentration
• 1:10:54
  story and demonstrate that they're not just a A1 trick pony
• 1:10:57
  with that one customer. So I'm going to I'm going to
• 1:11:00
  pend my buy or sell on Cerebrus on on just that customer story
• 1:11:06
  and we'll see what happens over the next three months.
• 1:11:07
  I think if they can broaden their footprint, they'll they'll
• 1:11:09
  do fine. The best time to go is right
• 1:11:11
  now. Get out there before Open AI or
• 1:11:13
  Entropic go public. Once they do, there are now
• 1:11:16
  multiple companies where institutional investors can go
• 1:11:19
  for an AI story. I think right now the only thing
• 1:11:21
  you can buy in the public market is an AI story.
• 1:11:24
  An AI per play was NVIDIA, right?
• 1:11:26
  There's no other company you can buy for Fidelity or if you're T
• 1:11:29
  Rowe Price or if you're a pension fund or a venture fund
• 1:11:31
  and you're being beaten on by your IC for like, where's my AI
• 1:11:34
  story? The only company out there is
• 1:11:36
  NVIDIA. Google's not purely an AI story.
• 1:11:38
  Met is not purely an AI story. I think Cerebrus is purely an AI
• 1:11:42
  story, and if they can get out and show some breath of customer
• 1:11:45
  footprint I think we'll probably do fine.
• 1:11:47
  If they can't though, I think it's a real concentration risk.
• 1:11:52
  Well, thanks for that summary, Aman, and I guess we'll see you
• 1:11:55
  next week on valuation Corner again.
• 1:11:57
  Sounds good. Thank you, Dave.
• 1:11:58
  See you next week. Sophisticated venture
• 1:12:07
  capitalist. That's that's us, right Aman?
• 1:12:09
  Yeah, I think so. The only ones, we're definitely
• 1:12:13
  sophisticated. I don't know what I.
• 1:12:15
  Hope so sometimes. When you said Messiah at first I
• 1:12:18
  thought you were talking about Paul at tradies, but that's I
• 1:12:20
  guess I'm just looking forward to dude.
• 1:12:23
  I think Travis is the modern day Paul at Tradies man, Silicon
• 1:12:30
  Valley nerds, Wall Street numbers.