Latest / The Punch List / Data center GCs should diversify, Housing Market Slowdown, CoreWeave's $50B debt

Transcript

Data center GCs should diversify, Housing Market Slowdown, CoreWeave's $50B debt · The Punch List

• 0:00
  We've stock dropped, so their equity went down and. Everyone got scared. There's a risk of them not
• 0:05
  being able to. Pay the debt I think is very low. Mechanical and electrical subcontractors are in
• 0:11
  really high demand for data centers. Mechanical, electrical and general contractor. All three are
• 0:17
  at the same level, contracting directly to the owner. Focus on diversification because you don't
• 0:22
  want to be concentrated in one markets prefab right now is it's not cheaper, faster for sure. On
• 0:29
  the legacy of that is most parking garages now are precast. That housing prices are down
• 0:34
  nationally. Government needs to do more to pump up housing construction. What we do is reduce housing
• 0:41
  cost prices. Welcome to the Punch List Podcast, the number one podcast in the world for architecture,
• 0:48
  engineering, construction. Probably the universe unclear, but we're here today with con off
• 0:55
  and TC. Hi guys, we got a lot to talk about today. Uh, some big news around data centers
• 1:02
  last week. We're going to talk about that. We're going to talk about what that means for you, the
• 1:06
  general contractor, uh, in terms of diversifying your pipeline. And then we're also going to talk
• 1:11
  about housing prices. So let's just jump in. Uh, last week, core, we've in the news among other of
• 1:18
  the Neo cloud, uh, data center developers. I'm going to frame this, um, based on what's in the news. And
• 1:24
  then we're going to talk about, um, kind of the bull and bear case for this. So core we've they
• 1:30
  build data centers essentially. Um, and they, uh, they don't just like, rent these things
• 1:37
  immediately. There's a long lead time. We talked about this in the pod of how long it takes to
• 1:41
  build a data center. So they got to take on a lot of construction debt, um, to get these things built.
• 1:47
  And they have to take on debt to buy the GPUs, etc.. And so to do that, they put what a standard
• 1:53
  in the industry. They put these data centers in a legal special purpose vehicle. They borrow against
• 1:59
  it, and they sign three year contracts with, like Microsoft and Google and all the big players, um,
• 2:06
  to fund this thing. So you got, you know, you're borrowing a lot of debt, your money going out way
• 2:11
  before your money's coming in. Typical. But that, you know, that spread of time is is some risk. So
• 2:17
  what what happened last week is basically the market started to put a premium on what they call
• 2:24
  unsecured debt. So, uh, not backed by assets. And um, this was really
• 2:30
  reflected in, if you guys remember, from The Big Short and the financial crisis, the credit default
• 2:35
  swap, the, uh, the lenders buy these things to insure against default when they loan lend money.
• 2:42
  And those spreads and the cost of those jumped up to reflect this kind of like market sentiment
• 2:48
  around the risk core. We've stock drop. So their equity went down. And so um, really it's like,
• 2:54
  okay, uh, all right. Debt costs more. Um. Lower value. You
• 3:01
  know, the cost of the debt going up lowers the value of this data center product. You know,
• 3:06
  ultimately, in the end, at least, the way it's modeled today, their stock went down. It's a lower
• 3:10
  equity value. So like this could be a spiral. You know, that makes it more difficult and challenging
• 3:16
  for these groups like core. We've uh, to be in business. And so that scared everybody. People
• 3:23
  started tweeting things. But really what this comes down to is like, are you an optimist? Do you
• 3:28
  think that in three years that Google and Meta and OpenAI and everyone else who's signing these
• 3:34
  three year contracts are going to pay up and use this and the demand is going to be there? Or do
• 3:39
  you think that, um, you know, this the fact that they have to keep using debt to build these
• 3:45
  things out and the cost of it's going up is going to make these things, uh, make this process and
• 3:51
  this business model unsustainable. Before all this data centers built out, that's what it really
• 3:54
  comes down to. So, um, some are saying core. We've I think we're, we're piled on the core. We've because
• 4:00
  they're in the news. But there's other people out there like the canary in the coal mine here, um,
• 4:04
  for what is like an unsustainable debt load for AI data center build out. I want to turn it over.
• 4:11
  And TC, what what are your thoughts here? I think before the thoughts, let's set the
• 4:18
  stage for the audience to kind of tell them how, uh, a data center
• 4:24
  project is rolled out in financed. Right. So that and there are different ways of doing so. Right. So
• 4:32
  I can either be a core wave, which is saying, hey, let me
• 4:39
  take on the risk of setting up these data centers. I'll find enterprise customers
• 4:46
  who want who will give me a long term contract commitment of using these data centers. So let's
• 4:53
  say, uh, a Google or Nvidia comes to code and says, hey, can you. Can you give me a
• 5:00
  100 megawatt commitment for next three years or five years? Seven years. And here's the fixed price,
• 5:06
  with some escalations pegged in the 10 million to $15 million per megawatt range, right
• 5:14
  on top of that contract being secured. Covid raises a debt. Uh, uses
• 5:21
  that money to fund the data center construction to fund the Nvidia GPU, purchasing, uh,
• 5:28
  to commission it. And then then then the money starts hitting, hitting the door month one when
• 5:33
  they when they start giving it out to those players. But that's every real estate development
• 5:39
  project, by the way. Right. That's straight forward. Yeah yeah yeah. The other way that happened
• 5:45
  accidentally, not realistically, is if you if you all saw the podcast from
• 5:51
  Zuckerberg, uh, when he went to the checkpoint part for the first time with with John John Callison
• 5:57
  of of of stripe. I mean, John was right. It was like
• 6:04
  you spend a hundred, you look $100 billion risk on building a new model. And you have no idea. You
• 6:10
  hadn't you didn't know it. Will it work or not? Right. It's like, yeah, that's true.
• 6:16
  But at least I could recover 30, 40 billion of that by leasing out my data center. If if our
• 6:23
  model did not work. So it's not a crazy investment. If I can at least lease it out.
• 6:30
  Uh, space. Just just that did that, right. They they over commissioned, they overbuilt and then
• 6:37
  what they needed and they when the demand when no one was giving an
• 6:45
  a similar trend. I'll show you. I'll tell you. Right. So what's his name? Uh, the
• 6:52
  Amazon chief was the AWS chief before, right? He said, uh, Jassy. Andy. Jassy.
• 6:59
  Andy. Jassy. Yeah. He's like, we have we have compute
• 7:05
  blocks till 2028. Like there's no compute left. If any new company needs a lot of
• 7:12
  compute, we don't have it to give it. It's all backlogged and reserved by big players. So this
• 7:19
  dearth of compute, and when there's no compute left, you go to you go to Elon and say, hey Elon, do
• 7:25
  you have some do you have some compute left with you? Like a lot. And say how much would you price
• 7:30
  it? I wouldn't price it 10 to $15 million or megawatt hour price uh 30,000,000
• 7:36
  megawatt because I can do so right. So that's that's how it is. And
• 7:43
  so if you go to Elon and say, hey, do you have do you have that, do you have that compute left? It's
• 7:48
  like, yeah, I'm going to I'm going to spot price it 2X3X for the demand is what the supply today is.
• 7:55
  Because I'm, I'm the only one left with that supply. Right. So these big players like meta and
• 8:00
  and and SpaceX also were building a lot of compute which on, on top of their own balance
• 8:06
  sheets, they didn't have to take much depth debt from the market. And they could they could price
• 8:13
  it higher as well. Right. So those are the two ways you can you can solve this demand right now
• 8:21
  with Covid, if everything was good, the only thing that changed in the last one, one
• 8:27
  month, I would say, and that got reported out last week. And that's why the stock price went down
• 8:34
  is the bond yield went up. The bond yield went up from 12% to 16%. They
• 8:41
  had $50 billion of debt on their books. And everyone got scared
• 8:48
  like will they be able to. Will they be able to, uh, pay their debt on time?
• 8:55
  Well, the on the other side, they have 104 billion of committed contracts.
• 9:02
  Uh, so. Sounds pretty sizable, but, yeah, it's it's gone to a state. It's a bit sensitive
• 9:09
  now. Uh, if if a downturn happens, if the ice consumption slows down,
• 9:17
  we can see a different world for curve. Uh, so the sensitivity, you know, flags have gone
• 9:23
  up in the market, and that's why you see that price going down. So that's that's the whole
• 9:28
  episode of what happened. Kind of. Thanks for explaining the overall architecture. Um, from a GC
• 9:34
  perspective, I'll know this. Uh, I'll say this when we are signing contracts, we
• 9:41
  if we can try to stay away from special purpose vehicles, because if they were to go down, you
• 9:46
  don't have a financial recourse. So if you do Tale of Two Cities, right. Two big new clouds. Ah, you got
• 9:51
  Crusoe. You got curve. Uh, Crusoe is not in the public market. Uh, they they raise
• 9:58
  money. You know, they have JVs with Blue Owl and all these other bigger firms. So the risk when
• 10:05
  you're doing a project as a general contractor is different. Covey versus Crusoe.
• 10:12
  Both. Both reliable. But because curve is relying on debt. If if they were to
• 10:19
  go bankrupt, then the risk of not getting paid is a lot higher with Covid.
• 10:26
  Um, coming back to the stock market reaction. I mean, I think fundamentally speaking, kind of
• 10:33
  you're right. I think the demand is there, the risk of them not being able to pay the debt, I think is
• 10:39
  very low, uh, because there is compute demand curve has a, have a has a good reputation of building
• 10:46
  good data centers. Um, I don't think we have had any issues with any of any one of their customers
• 10:52
  saying that curb has not delivered on time or to the capacity they promised. So I think those two
• 10:58
  bode well for them. Um, the downside is. Yeah. Market has decided that the the
• 11:05
  cost of debt needs to be higher. Interest rates have gone up to 15, 16%. Uh, for unsecured debt.
• 11:12
  They used to be around nine. What does that mean? They'll be paying more in interest payments every
• 11:19
  quarter. That means they're not going to be losing money, but they'll be making less money. So I think
• 11:24
  stock market is more in relation to. Well, why don't I put money in places which have
• 11:30
  higher profitability or higher margins compared to curvy where there are. This is going to cut
• 11:37
  into their margin. So it's not like they won't make money, they'll just make less money is what I
• 11:42
  think. So I mean that's the bottom line for them. I mean to my GC code. I'm gonna
• 11:49
  I keep always wanting this. Whenever you're signing a contract, especially the big guys, I mean,
• 11:54
  they already know this. Just know who you're going to sign a contract with. This is not just a code
• 11:59
  we thing. I think this is a more of a general thing as well. Even small real estate development
• 12:03
  projects. Just keep an eye out. Make sure you manage your risk whenever you're signing a
• 12:08
  contract with a special purpose vehicle versus a corporate financed project. I mean, I
• 12:14
  don't like. Feel like I have visibility into the full picture of this, that
• 12:21
  like, you know, people who are commenting on this, you know, in the news,
• 12:27
  but it does just feel like a risk repricing on debt. And it feels, you
• 12:34
  know, that's comes after this rate rise. Um,
• 12:41
  like it feels like a gyration in the overall arc versus like some kind of major
• 12:48
  flag. But, um, are we you know, there are people coming out and saying we're in a bubble. I think
• 12:55
  Ken Griffin from Citadel said that. I'm sure we are. But like, what does that mean? Uh, and
• 13:02
  what's going to be the end of that and who's going to go down and, but like right now it
• 13:07
  doesn't feel like it feels like they're going to recover from this. But you know, I don't know. Um.
• 13:14
  It's also a short term. Like, every state is trying to be.
• 13:22
  Uh, careful about the elections and not pushing the data centers forward. So
• 13:29
  there's a sentiment in the market about data centers being blocked by a lot of states. Once the
• 13:35
  elections are over, it's going to go back to thermal, and then it might just get better. Right. I
• 13:42
  mean, but I won't I'm Sorry. Go ahead. John. No, I was just saying, like, if we think. About the people who
• 13:47
  listen to this podcast. General contractors, architects, designers who are working on this, I
• 13:52
  don't think this means, like, your project you're working on right now is going to halt. Um, I think
• 13:58
  this is more like pipeline. Um, and I think, uh, I'll let
• 14:04
  TCU jump in. But I think that segues into our next topic about diversification of pipeline. You know,
• 14:10
  everyone's trying to feast on the data center thing, and you should you should go out there and
• 14:14
  pursue that business. But you should also be thinking about, uh, diversification in the event
• 14:20
  that this does, but, you know, crash out at some point. What were you going to say? No, I mean, listen,
• 14:25
  I won't minimize the political aspect of it. I think what what's happening with Stock Market's
• 14:32
  reaction to Cory? I don't think it's just limited to the debt going up. Um, the, the
• 14:38
  politically speaking data centers seem to be unpopular with the general public.
• 14:45
  That's just a fact. There's no way around it. And will will that lead to something
• 14:52
  after the elections or not? We don't know. But there are a lot of candidates running on the fact
• 14:56
  that, yeah, I'll make sure they can't build a data center in your backyard. So so that is another
• 15:02
  risk beyond the, the, the debt being higher that we have to consider. So if you're a
• 15:09
  general contractor, I'm sure you're thinking about that from a pipeline perspective, which like you
• 15:14
  said, John, kind of leads us into our next topic of diversification. Uh, I mean, bottom line is data
• 15:21
  center is still a very small part, uh, part of the overall construction market even now. Right?
• 15:29
  It's just a high, high profitability market. I think it has the highest margins right now of all
• 15:34
  the types of buildings we build. Yeah. Well, I think, uh, so our next topic, uh, there's an
• 15:40
  article in the press this week about, um, surprise, surprise. Mechanical and electrical
• 15:47
  subcontractors are in really high demand for data centers. Um, and by the way, like in big booms there
• 15:53
  always in those trades are always in high demand. And it's a great high skill trade to be a part of.
• 15:58
  Um, but, uh, TC was we were prepping this before this episode. You were talking a little
• 16:05
  bit about, um, diversification. And, um, why don't you just kind of expand upon your thoughts on
• 16:12
  that? Yeah. I mean, uh. You know, I start with this
• 16:18
  Aristotle quote of, like, where he says that sign of a mature mind is to be able to hold two
• 16:24
  competing ideas in your head at once. And, and I'm trying to do that here. The reason I say that is,
• 16:31
  okay, let's take the optimistic view. If you look at building, if you look at complex building
• 16:37
  systems like data centers, life science building. They are very mechanical and electrical. Heavy
• 16:42
  services rule the roost, especially with data centers. You know, 60% of the cost is in mechanical
• 16:49
  and electrical systems, and I'm not even including the cost of servers and all that. That's usually
• 16:53
  owner pays for that directly. I'm just talking generators, cooling systems distribution. So
• 17:01
  given that and they also lead. So what do I mean by lead. So when we do a commercial building
• 17:06
  architect leads, there is an esthetic component that drives what the building looks and feels. And
• 17:12
  and then all of us are in some ways subservient to it. I, you know, electrical system there is is
• 17:18
  like a waiter, right? It's serving the needs of what architect design decides room needs or
• 17:24
  equipment needs. And it just serves right. It's in a serving capacity. Whereas in a data center,
• 17:30
  servers are the king, right? You've built everything around supplying those things and they
• 17:35
  need a lot of power. They need a lot of cooling. So these entities come in a prime
• 17:41
  position to an extent that most data center builders do multi prime contracts, which is not
• 17:47
  typical in in in non data center settings. So what does a multi prime contract means. That means
• 17:54
  mechanical, electrical and general contractor. All three are in some ways at the same level
• 18:00
  contracting directly to the owner. GC has the responsibility to coordinate with other two, but
• 18:05
  they are not under the general contractors umbrella are for a commercial building or any
• 18:11
  other building type. They would almost always be under a general contractors umbrella. They won't
• 18:16
  do a multi prime contract data centers do because they understand the advantages of mechanical and
• 18:21
  electrical system. They want to buy equipment early. It's hard to get equipment right now. They
• 18:26
  want to block or block their um equipment in line of production and all that. So
• 18:32
  what that is leading to is it is leading to this conversations, especially on the mechanical
• 18:38
  contractors thinking, why don't we be the prime and general contractor comes under us like that?
• 18:43
  That's the level of conversation that's happening there. Which is which is very interesting. And, uh,
• 18:50
  could be possible. I, you know, I don't know. It'll be interesting to see how that goes. The other
• 18:56
  competing idea in my head is, okay, that's true for data centers, but that's not true for many other
• 19:02
  buildings. Like I can go on the other end of the spectrum and look at a parking garage where
• 19:06
  there's literally no mechanical system. Right? Or most commercial buildings where mechanical
• 19:11
  electrical pay play a lot lesser role, where GC has to be prime. So
• 19:18
  do you really want to, you know, pitch the GEC off by by being the prime right and
• 19:25
  putting your relationship kind of on a rocky road in other building types. So that's a, that's a
• 19:30
  question that comes to my mind. I'm sure they are debating that themselves too. But walk us through.
• 19:38
  I bet they do. For the money. Yeah, right. But. But walk us
• 19:45
  through. Um. I am a GC versus. I am a electrical sub. Okay, let's just
• 19:52
  have those two players for now. If I'm a GC, my
• 19:58
  backlog is 60%. Data centers for
• 20:05
  40%. Everything else. Uh, for the next two years
• 20:12
  now, I'm, I'm looking 3 to 4 years out. And I was like, hey, for my three year backlog, can I start
• 20:17
  diversifying more? Can I start, you know, doing 30% data centers,
• 20:23
  70% everything else? What kind of. Uh,
• 20:30
  diversification is do I play beyond a data center and kind of can a electrical or a mechanical sub
• 20:37
  do the same? Or are they stuck with this data center demand? Like walk us through like the the
• 20:43
  hedging strategy of diversifying and and you have you have been in this I think you talk to me
• 20:49
  about this. You have been in this situation at the 2008 crisis, uh, and boom. All right. So,
• 20:57
  so walk us how you get across a bust and boom cycles, uh, as a GC and diversify your backlog. If
• 21:03
  it were me, I would really focus on diversification, because in general, you don't want
• 21:08
  to be concentrated in one market, especially something as focused as data centers. More than 30
• 21:14
  to 40% max, 40%. The the tough part is, and I'm not in that role,
• 21:21
  but whoever is running a general contracting firm, if say, you're doing if 40% of your work is data
• 21:28
  centers, I bet that translates to 70% of your profits. It's a sugar rush.
• 21:35
  It's a sweet deal. It's. And. And in some ways, most DCS are
• 21:42
  engineering led. And data center is a very engineered building. Right? There's no there's no
• 21:47
  one complaining about esthetics or thermal comfort. You know, it's it's a building that is
• 21:53
  designed for servers. Like I said, there are prime candidates there, right? Like everyone, every other
• 21:58
  building type is designed for human beings primarily. And human beings tend to complain more
• 22:03
  and have a lot more ad hoc requests versus servers. You take care of the cooling for
• 22:10
  them, you get them power, they'll shut up and do their job. So. So it's it's it's a hard it's a
• 22:17
  hard decision to to to execute. But I would say hypothetically speaking, a courageous
• 22:24
  decision would be to diversify because I, I don't I don't know where I mean, as we just talked
• 22:31
  about that compute demand will keep going up. There is a lot of demand for it, but you look at
• 22:37
  the other side, it's still, you know, politically not very popular. Could you build them at some
• 22:44
  other location where it's cheaper to build it? You know, I know there's a whole fiber component to it
• 22:50
  as well, but but there are other risks. Right. Um, the other thing that's not helping. And like in
• 22:56
  the core, we've case as well, we didn't talk about the new generations of chips are coming back
• 23:01
  faster and faster and faster. So so the opex cost of these data centers, I mean, it depends on
• 23:08
  whether you call it opex or CapEx. But once a new generation of chips come in, you also either need
• 23:13
  to swap them out in your existing data center or the or a newer one becomes more attractive. So
• 23:21
  I don't know if it's a very reliable market. Um, you know, even though I think there is there is
• 23:27
  demand, but is there another way to address the demand? So I would say keep it below 40% if you
• 23:32
  can. That that that. That's how I would do it if I were doing it. When I was, I was
• 23:39
  reading somewhere that even data centers need a lot of warehouses these days to
• 23:46
  function, at least for temporary basis to keep the equipments in place. And
• 23:53
  warehousing in general is catching up. So I think that's a good diversification strategy to start
• 23:58
  doing warehouses as well. Uh, and it's it's also as bare
• 24:05
  shell as a data center is almost less cooling and services per
• 24:12
  se. I just wonder what it says about our society if data centers and warehouses become the top two
• 24:18
  markets. Well, hopefully infra and you know, I'm bullish on commercial making a
• 24:25
  comeback to I mean we we've talked about this before I feel like there is something
• 24:32
  we are finding out as we are building companies, that the the requirement for human connectedness
• 24:37
  by co locating and working together. I think that's coming back more and more as well. And,
• 24:44
  and the electrical guys can still diversify into energy distribution and and generation.
• 24:51
  Right. Because that's kind of a huge for our country actually. That's a good point. I didn't
• 24:55
  answer the mechanical electrical question. I think what I'm seeing, some of the mechanical and
• 25:00
  electrical, uh, uh, you know, trade partners do, which I think will be very valuable and would be a
• 25:06
  great legacy of data centers, is they are getting more and more into prefab because data center is
• 25:12
  so speed driven, um, that they are willing to pay the money because, see, the state of union
• 25:19
  on the prefab right now is it's not cheaper. It's conceptually and
• 25:26
  hypothetically feels cheaper. It's not it's faster for sure, but it's not necessarily cheaper.
• 25:32
  Data center for speed will pay for that, and once you do it enough, it will become cheaper because
• 25:38
  conceptually prefab should be cheaper. You're building it in a factory setting. It's a lot
• 25:42
  cheaper to build. You can build it where the labor is cheap, or are closer to where you have the most
• 25:48
  competent labor, and then ship it to a location where it needs to fit. I think data center can
• 25:54
  leave that legacy behind where a lot of these mechanical skills are. The whole electrical rooms
• 26:00
  can become prefab. I think that's coming. They can lead the you know, they can lead that revolution.
• 26:06
  And then that would really help the commercial building out. Just like liquid cooling is becoming
• 26:11
  cheaper and cheaper and more approachable. All right. Yes. Talking to some designers. And they were
• 26:15
  like they have to think differently to design data centers
• 26:23
  because they have to go out fast. And My my biggest question to them was, can
• 26:30
  you please document all these learnings and apply it to every building in the US because we need to
• 26:35
  ship things fast. So what are we are learning in building data centers? Fast need to reciprocate
• 26:41
  across all building types? Yes, absolutely. I
• 26:48
  mean, one example of that is right, the Apple spaceship building pushed precast concrete hard.
• 26:55
  That was such a big precast concrete job that I think later on, the legacy of that is most parking
• 27:01
  garages now are precast because we have fundamentally built that infrastructure to pull
• 27:06
  put a lot of precast out in Bay area. Right. And and
• 27:13
  so so it's just it's cheaper and faster to be able to do that which before that building we
• 27:19
  would usually gravitate towards cost in place garages. Yeah. We have the luxury in the Bay
• 27:26
  area, basically being able to do construction 365 without major weather interruption other
• 27:33
  than rain. And so that enables a lot of just in time
• 27:39
  stuff. Uh, it supports a lot of poured concrete, whereas like precast concrete is very
• 27:46
  popular in, um, you know, the northeast and the upper Midwest because you're really planning
• 27:53
  around weather. Um, and so, yeah, you're right. Having having that big flagship
• 28:00
  project being precast structural concrete was was good for the prefab industries, as was our project
• 28:06
  we did together. Um, so, yeah. Um, well, uh, very good.
• 28:13
  Last, um, topic for today. Uh, you know, other negative news last
• 28:20
  week was that it looks like the article that housing prices are, you know, down nationally a
• 28:26
  certain percentage point. Um, and, uh, wanted to get everybody's take on that
• 28:33
  kind of take. What what is what you read on. You know this now, what is a national headline? I would
• 28:39
  tell you. Housing prices are very much a market by market analysis. So sometimes these headlines are,
• 28:47
  you know, sensational. And you can look at it and but you know market by market that's not always
• 28:52
  true. Yeah. It's when the interest rates are low
• 29:01
  I think across the US it's almost the same story. When interest rates are
• 29:07
  high it becomes a very binary view, like places with huge demand are up
• 29:14
  and places with not good demand or way down. Right. So like we have the luxury of. Or
• 29:21
  marginally down relative to what they were. Almost. Yeah. I mean, not not crazy down. No one's crazy
• 29:26
  down in real estate. Like, real estate is one thing that cannot go crazy down. Unless the 2008 crisis
• 29:33
  that happened. You know, that was the one exception to it. But, yeah, I mean, we live in the luxury of
• 29:39
  Bay area right now where there's a lot of liquidity happening. So the prices are going up. I
• 29:43
  mean, I checked my my house prices. It's up pretty good in a good way.
• 29:51
  But yeah, because of the interest rates being up, being higher, I think uh,
• 29:59
  and uh, and the whole market confidence in general, not just like the
• 30:06
  interest rates, this is a whole the market has not been stable with all these geopolitical,
• 30:13
  uh, disruptions with, with, with, with the interest rate, with AI. Like, it's, it's a,
• 30:20
  it's for the first year The uncertainty is not just sitting in Bay area, it's sitting around the
• 30:26
  world. And if you combine those two things together, like people are not sure about
• 30:33
  buying houses, uh, or even like I was reading the news having babies for murder.
• 30:40
  Right? So, so that's so having babies, buying houses, like, have gone down this year. And once the
• 30:47
  certainty comes back, it's going to bump up again. But it's a very binary view like Miami
• 30:54
  or or Dallas or San Francisco are like going up because of the, the, the
• 31:00
  incoming demand. But everything else is like down. I mean, I think given where the market is, I think
• 31:08
  I mean, I hate to say this, but I think government needs to do more to pump up housing construction.
• 31:13
  Like we have talked about reduction in construction starts, interest rates being high is
• 31:19
  not going to help the housing market. your you're also losing labor to other
• 31:25
  like I say hi more profitable projects right data centers being one of them. So we need to
• 31:32
  find a way to make housing construction more attractive, because we all
• 31:39
  know that the percentage of home ownership for a developed nation in US is not very high. I mean,
• 31:46
  I can I'd have to look it up what it is, but I'm I remember looking at it. I mean, maybe one of us can
• 31:52
  look it up, but I think compared to other developed nations, it's not it's not very high. It
• 31:57
  needs to be high. When you own a house, you're more attached to a community. You tend to contribute
• 32:02
  more to your community. It's just it it unlocks different things for as a, as a, as a
• 32:09
  functioning member of a society. Um, and that really needs to be, you know, people need to be
• 32:15
  motivated more and it needs to be made more approachable because loans are not there. Mortgage
• 32:21
  is not coming down. Mortgage keeps going up. So cost of buying a house keeps going up.
• 32:27
  And not having enough supply. Doesn't help either. So you got low supply and high interest rate.
• 32:36
  Someone needs to do something about this. Yeah. You know, I think largely it is.
• 32:42
  Interest rates, you know, in the short term. But it points to a larger affordability
• 32:49
  issue in America. And kind of your point about people not wanting to have babies. I mean, that's
• 32:54
  the same thing. It's just too expensive. Or they don't feel confident enough to make that decision
• 33:01
  financially either to have the child, which requires expensive childcare and everything else,
• 33:07
  plus a house. So this is an affordability thing I think is a great thing right now. But, um,
• 33:14
  I'm, you know, I'm thinking that the squeeze on this, uh, is really going to drive, at least I
• 33:20
  hope, a more innovation in the home building space. Um, material cost processes.
• 33:27
  Obviously I'm doing offsite construction, but, um, all elements of this need to contribute. I mean,
• 33:34
  ultimately the way to solve this is to get lower cost housing like, and it's not through
• 33:40
  subsidy and everything else. It could be some of that. But we need to innovate the way we're
• 33:46
  delivering this product and the speed and the cost. That's how all technological
• 33:53
  technological revolution solve these things is reduction in cost and efficiency and
• 33:59
  innovation. So, John, if John, if, if I made you whatever you want to be
• 34:06
  made in the US, like what would you do to reduce the housing, housing cost
• 34:13
  basis. Yeah, I'd, you know, I'd have probably a a dramatic, uh,
• 34:20
  standardization of codes across states. Um, uh, I
• 34:27
  would, uh, you know, to have a dramatic reduction in regulation and process
• 34:34
  that we've layered upon home building. Um, and I think from a
• 34:41
  from a big, you know, country standpoint, the last time we had a massive
• 34:47
  homebuilding wave at like, massive, massive scale was post-World War two, largely from big wealth
• 34:54
  transfers from the government after World War two. And if we're going to have a generational, uh,
• 35:01
  refresh, we probably have to do all of that again. Uh, a generational refresh of the housing stock.
• 35:07
  Plus add more, and we need to do it with modern technology and modern efficiency. So it's not just,
• 35:14
  you know, really let the private sector drive it with, with, um, government, you know, some support. But
• 35:21
  that's that's what I would do. But I don't have a silver bullet. All I'm doing is just trying to
• 35:27
  build houses as fast and let you know, inexpensive as possible. Now, as we were talking, I looked
• 35:34
  up the data and actually, I'm I, I'm positively surprised that so U.S. homeownership rate is
• 35:40
  65%, which I think is pretty good. I thought it would be a lot lower. Um, but the key thing is it's
• 35:46
  better than UK, France, Germany, Switzerland. Like it's, you know, so so I think I think
• 35:53
  maybe, maybe why. Uh, so I'm thinking why it emotionally feels that it's lower. I think us has
• 35:59
  so much land. And I think that's the American dream, right. To own a house. So it feels like it
• 36:05
  should be higher. Um, the other key difference I see is in 2004, it was 69%. So it has gone down
• 36:12
  from 2004 to now. But that was also because of. Something. That was right at the kind of, you know,
• 36:17
  subprime, uh, bubble, which I think it was probably as high as modern percentage. And then we had a
• 36:24
  big drop, um, after financial. I mean, look, there's an equilibrium sum here, and it's a little bit
• 36:29
  controversial. I think we do have relatively higher homeownership rates than most of the world.
• 36:33
  And we kind of tinker at the edges here. But like, there's probably a natural equilibrium,
• 36:40
  uh, which is much more without massive government subsidized debt, uh, through the 30 year
• 36:47
  mortgage, that percentage would be much lower. And you can look at what those percentages were pre
• 36:54
  1940s, 1930s. New deal. Was 40%. Yeah 44%. So
• 37:01
  it is it is the it is the modern American dream. And I think it's a good one and a stabilizing one.
• 37:06
  And so I think we should keep supporting it. At a younger age. You want as a society you
• 37:13
  want people to be educated, people to own a house and people to invest in the equity market.
• 37:20
  Upside. Right. You want all those things to happen. I think the newer generation is
• 37:26
  is marginally not a lot like not a lot has changed, but marginally renting
• 37:33
  more and putting more money in the equity markets. Uh, so either you put money in real estate or
• 37:39
  equity or both, I think. But if you're doing both of them together, like you, you are in a good shape,
• 37:45
  like you are. You are. You are going up as the market is going up. Uh, right. And
• 37:52
  you feel contributing to the, to the economy and getting the benefit back as well. So
• 37:59
  as long as people are buying a house or, or investing in the stock market, uh,
• 38:06
  I think they are. They have a good upside and security, right? Yeah. I don't think there's
• 38:12
  anything wrong. There's a lot of, you know, stuff on the internet about how a house is not a good
• 38:16
  investment. And and that is true in some of these really high cost markets. I think sometimes, um, and
• 38:22
  in mid-tier markets where there's more fluctuation in the pricing, I think ultimately it
• 38:28
  does. I think a the house is like a stabilizing social function where anchors you to a
• 38:35
  place around school district and a community that you invest in, you know, um,
• 38:41
  and that is a good stabilizing social function. There's nothing wrong with renting. There's
• 38:48
  nothing wrong. A lot of people rent really high end stuff for a really long time, or live in the
• 38:54
  same place for a long time. That's a similarly stabilizing thing. Uh, they're also investing in
• 38:59
  the community. I think it's just more that house psychologically anchors you to a place.
• 39:06
  Um, it largely around child rearing, I think. Uh, but also just you invest
• 39:13
  you you care more about, uh, the local elections. You care more about local city services, whether
• 39:19
  you have kids or not. Your your you care about what the community looks like and you're going to
• 39:24
  pay more attention. And I think that's just and participate more. And that's good for society. No, I
• 39:30
  mean, specific. Example of that is I was uh, uh, we had few people visiting us and we were trying to
• 39:37
  rent a place for them, uh, you know, like, there were like, few different people coming in, and I found
• 39:44
  that, you know, it got me on this mathematical check of, like. So I live in San Jose.
• 39:50
  The rental rates are way lower than what you would pay for, like mortgage or ownership. So I'm
• 39:56
  like, why would anyone buy a like, you know, it doesn't even make financial sense. I mean, I know
• 40:00
  why I, you know, bought a house, but so I was talking to a guy who does data aggregation around
• 40:06
  it and he said, yeah, because like you said, John, people are willing to pay way higher than what
• 40:12
  they can financially get back if they were to rent the place out because they want to be
• 40:17
  psychologically owning something of their own, and they want to be tied to the community, they want
• 40:22
  to be part of the community. And that and we see that in many different markets where, um, you know,
• 40:28
  your people are paying a lot more for a house that they can ever get back if they were to rent
• 40:33
  it out. Yeah. Plays out with when I'm standing in my driveway with my kids and somebody drives a
• 40:39
  little too fast down the street, you know, I yell at them, I become that guy, I become. I can imagine
• 40:45
  John being one of those guys, you know, who yell at the cops. You know, I pay your salary. I'm. I'm
• 40:50
  yelling at. I'm I'm like, tell, you know, telling the UPS driver to slow down a little bit. Um. So
• 40:57
  crazy Uncle John in the neighborhood. I know it happens fast. It happens fast. Especially when you
• 41:03
  have little kids. Like, I get that, man. I mean, I have been whenever I invest in my house.
• 41:10
  I have both the reactions. Why am I doing that? And then at the same time, I was like, it makes sense
• 41:17
  from a family, you know, to just keep on doing this. It's like deeply satisfying to, you know, when I, we
• 41:23
  bought a house and I renovated it and like, my family moved in. You know, there was like a deep,
• 41:30
  like, psychological satisfaction. Like, I, you know, provide shelter for family, like caveman
• 41:37
  level kind of base evolutionary satisfaction and that and, um, look, I enjoyed it. I
• 41:44
  mean. But that being said, will not kill the surprise to early, but nothing next. Next time
• 41:51
  we're going to have. Special guest especially. You guys are talking about this at and someone who
• 41:58
  might have more power than us to to. Absolutely more power to us than to affect change. So
• 42:04
  stay tuned next week for our special guest.