Opendoor CEO: Building the Amazon for Homes

There are things you have to deal with in a public company that are amplified. In a private company, these problems are discussed with your VCs at a board table. In a public company, they’re discussed on Reddit, and they’re covered in The Wall Street Journal. So if you care a great deal about what’s said about you in The Wall Street Journal, running a public company is incredibly difficult. It’s just very difficult. Luckily, I just don’t care.

Kaz, welcome to the podcast.

Thanks for having me, man.

I feel like you’re the man of the moment. You recently took over as Opendoor CEO. It’s been a few weeks, a month.

It’s day 16, I think.

Day 16. I feel like you’re pioneering a new way of being a public company CEO. What was your mindset at first? What got you excited about the opportunity? And then, coming into it, what was your mindset about how you were going to be CEO of this company?

Look, I think Opendoor will become this generational company because people do a lot of this: when you go to business school, they’re like, “You shall have a business plan, and it should be 17 pages, and it should have Porter’s five forces on it, and that’s how it’s going to work.” That’s just generally not how great businesses are built—at least, not that many of them.

I think it’s important that great businesses start with a very simple statement that people can buy into or disagree with, right? That’s actually important: you’re saying something.

Yeah.

I think home ownership is good for the world. The more people that can own a home, the better off we are.

This is objectively a broken process, so we can fix it.

So I was just generally excited by the mission of the company.

Yeah.

I’ve just been a fan of it for a while, and looking from the outside in, I thought, “This is a ‘this feels good’ type of problem that I can help with.” So we’re 16 days into this journey. Let’s find out if that ends up being true.

Wow. Let’s trace a little bit of the history. Alex, you led our investment into the company. What was the thesis or vision that got you so excited about it?

I’m going to talk for a little bit about the background, and I’m going to start with Amazon and then get to Opendoor. I remember I met Eric Wu when he had first started Opendoor with Keith Rabois, and I think they had flipped maybe 70 or 80 homes right around Phoenix.

The reason why Phoenix is a very interesting market is there’s a term that I’m sure you’re intimately familiar with called cap rate, or capitalization rate. Think of it as the income of some asset divided by the price of the asset. The Bay Area has very low cap rates. You could buy a $20 million house in Pac Heights. If you wanted to rent it, it might rent for $10,000 a month or $20,000, which is a lot, but not as a percentage of the price of the asset. That is very, very low yield.

A place like Phoenix is different. All the homes are pretty similar—I’m exaggerating, so don’t come after me; I’m not trying to offend people who live in Phoenix. You might have a $200,000 house that rents for $20,000 a year, so that’s 10%. The net operating income divided by the asset—that’s your cap rate. At 10%, you can actually make a lot of things work because there’s always a default buyer, somebody who will say, “I can arbitrage this. I’m holding a house. I don’t have to sell it to another person who wants to buy a house. I can now rent it out.”

Because I can get a mortgage from a bank for 5%, I can make the whole math work. Eric had started flipping homes in Phoenix and made money on most of them. But I was like, “I don’t know. Flipping homes seems kind of challenging. It sounds great when all prices go up.”

What I was really drawn to was the vision that he laid out. When I finally invested, I thought he was pretty far along toward getting that vision to become a reality. This is what I mentioned: I’m going to talk about Amazon.

Amazon started off in the 1990s basically selling every book in the world—hence Amazon—in Jeff Bezos’s garage or basement, or the warehouse or something. By selling every book, by having an infinite supply of books, he got all of the demand. Then, because he got all of the demand, he could say, “All right, I want to now sell something that isn’t books. I’m going to sell CDs and DVDs.” That was next, but he still stocked those in the warehouse.

Eventually, it was like, “I’m going to sell TVs. I’m going to sell something else. I’m not going to stock them, but I already have all the demand.” I got the demand because I had all the supply for something else.

I start with this because it’s the chicken-and-egg problem. Do you start off with supply or do you start off with demand? Marketplaces, we know, are very, very valuable, but you have to start somewhere. Nobody wants to sell if nobody’s buying, and nobody wants to buy if nobody’s selling.

The Amazon model was: find something in one niche, use that to get all of the demand, and then, once you get all the demand, you can attract the supply in a non-principal-risk-taking way. Fast-forward to Opendoor: I remember when Eric was flipping all the homes in Phoenix. I was like, “That’s clever, but I don’t know.” I knew other people who flipped homes in Phoenix.

But I think it got to the point where, in one of the markets—I think it was Charlotte, if I remember correctly—almost 10% of the homes in Charlotte under a certain price, call it $600,000, were bought by Opendoor. A $600,000 home actually buys you a lot of houses in Charlotte.

Now imagine that I want to go buy a house in Charlotte. I can go to the multiple listing service, the MLS, which you see on Zillow, Redfin, and all these other sites. They kind of just mirror the MLS, but that’s only going to show me 90% of homes for sale. The other 10% are only on Opendoor.com.

You don’t have to get 100% of the homes in order to get 100% of the users. You get 10% of the homes. You get 5%. It’s kind of like the Laffer curve. I can’t exactly define what it is, but there’s some quantum—and I would call it 10%—where, if you get 10% of all the supply, you get 100% of the demand.

Once you have 100% of the demand, now you can break this horrible monopoly of real-estate-agent-dom. You can say, “You can list your house on Opendoor.com, and we’ll charge 1%.” If you pull that off, it’s the biggest market in the world.

That’s what made me so excited, because eBay is a marketplace for everything except homes, and it’s a $45 billion company. There aren’t actually big companies in residential real estate. It’s kind of strange. The biggest ones are CoStar and Zillow. There’s no $100 billion-plus company in residential real estate, which is kind of bonkers because it’s a bigger asset class than everything sold on eBay.

And so why is that?

Why is that? Because it’s so hard to aggregate the supply and the demand. What Zillow and others do is lead generation. It’s a terrible business model from a consumer perspective. It should have negative 100 NPS because you’re just getting called all day by real estate agents. You’re not actually improving the value proposition.

Opendoor—it’s not about flipping homes. Maybe it is; you run the company. It’s about how you build a marketplace, the biggest marketplace in the world.

To give you a sense of how crazy this is, there’s a company called Copart, a public company. You know what they do? It’s auctions for not used cars but total-loss cars. If you get into a car accident and State Farm is like, “Oh, that’s a total loss,” what do they do with the car? They sell it on Copart. Copart, I think, has a $70 billion market cap.

Who buys it?

Who buys it? People in Russia. Well, not Russia anymore, but all sorts of people. They buy it for parts, or they’re going to go fix it up. But Copart is a bigger company than Zillow, right? How is that possible? It’s just because nobody’s tackled this. It’s the biggest market in the world, but you have to build, hopefully, a capital-light marketplace.

I think Opendoor is deeply misunderstood as a company. You have to think about companies as categories. Category problems are real in public-market investing. I remember when I first joined Shopify in 2019, I think it was among the most shorted stocks on Wall Street because people made a category mistake about Shopify.

They were like, “Well, it’s another e-commerce company. What are the odds it will build all the warehouses?” Shopify does e-commerce; it’s not an e-commerce company. The very real leverage point for Shopify does not come from e-commerce. It obviously comes from software—the software.

I think that’s actually a problem Opendoor has had: externally, and admittedly for some time internally, the company thought of itself as essentially an investor in real estate as an asset class. That is not the job of the company. That’s not what the company does. It’s a software company designed to solve that problem.

I think if you attack it like that, you just fundamentally do different things, right? I mean, it’s a publicly traded company.

You actually look at the numbers. The company has repeatedly been buying fewer and fewer homes every year, right? Because they’re like, “Cool, I’m going to buy an asset class.” You only buy homes that are mispriced. But it turns out not that much in the world is mispriced for a very long time.

Yeah.

Right. And that just becomes a good business; it just won’t become a big business. And that’s, I think, a fundamental mistake the company has made in the past few years.

Yeah. Let’s understand this problem a bit deeper. Maybe, Alex, you could share a couple of points from the talk that you gave, or explain more about the monopoly real estate agents have and give some more context here.

So, real estate agents: for a very long time, there have been about 2 million registered real estate agents in the United States of America. The statistic that I quoted a couple of years ago—I assume it’s still the statistic, but I looked it up—is this: if you remember mean, median, and mode, mean is the average, median is the one in the middle, and mode is the most frequently occurring number. The mode number of transactions per agent per year is 0.

It’s kind of like, “Well, I’m an actor.” “Oh, well, what was the last film you were in?” “Well, I’m kind of a waiter right now. I’m waiting to get a film job.” That’s what a lot of real estate agents are. Even the really good ones don’t do that many transactions per year, and they’re fundamentally misaligned with their customers.

If I’m a buy-side agent, the more money you spend, the more commission I get. I normally get 2.5% to 3% as a buy-side agent. Normally, you have a 5% to 6% commission pool: half goes to the sell-side agent, and half goes to the buy-side agent.

There have been numerous lawsuits, some of which have gone to the Supreme Court, attacking this as a kind of evil—not oligopoly, but monopoly behavior. There’s an expression that I love by the playwright George Bernard Shaw: “Every profession is a conspiracy against the laity.” It’s very, very hard to deal with problems of concentrated benefit and diffuse harm.

There’s a concentrated benefit to the 2 million real estate agents—a small number of whom actually do active transactions—but they conspire to keep these commissions very high because you buy a house once every 10 years. You are absolutely harmed because, if you’re a buyer, the more you’re convinced to pay, the more the agent gets.

Why is it misaligned as a seller? The sell-side agent gets more if you sell for more because they just want to move on and get their check tomorrow. If I say, “Okay, you’re selling a $10 million house. That’s a lot of money,” and then there’s another buyer who will offer you $10 million and $10,000, the sell-side agent is like, “Well, I get 3% of that. I don’t really care.” But you get an extra $10,000. You really do care about that.

This is where the phrase “principal-agent problem” doesn’t come from, but it’s a personification—a real-life version—of the principal-agent problem. There were just so many things broken. Every other auction is fair. I can sell a used car on eBay, and they actually do a pretty good job of that. I see that there are 24 bids; here’s my reserve price. Real estate is just very opaque.

Because it’s a very infrequent transaction and you have concentrated benefit and diffuse harm, there have been so many attempts to violently disrupt this industry. One of them—and I should say that I’m on the board of Rocket Mortgage, which bought Redfin, and Glenn Kelman is an amazing guy who started Redfin and really wanted to disrupt real estate—was, “Okay, we’re only going to charge you 1%.”

Let me take a step back. The funny thing about real estate agents is that I’ve bought many houses, and I’ve never used a real estate agent. Actually, that’s not true. I did use a real estate agent one time to buy a house. I was trying to say, “I don’t want to use a real estate agent to buy a house because you’re going to charge me money.” He said, “No, no, no. You don’t pay me. The seller pays me.”

I said, “Where does the seller get the money from?” He said, “Well, from their bank account.” I said, “No, no, no. I send the money to the escrow agent, and then the seller gets the money from the escrow agent and pays you.” He said, “No, no, no. It’s different money.”

I remember talking to this real estate agent when I bought a house in Palo Alto. I told him, “With all due respect, of which I give you none, you must think I’m an idiot, or I think you’re an idiot. Those are the only two options.” It’s absurd to say that it’s free to be represented by a buy-side agent.

What Redfin started doing—and this did have a very positive impact on the industry—was asking, “Why don’t we bring down the 3% buy-side fee to 1% by rebating part of it back to the consumer?” But the state of Oregon bans that. Going back to “every profession is a conspiracy against the laity,” if I say, “Hey, use me, Erik, and not Kaz to buy your house, and I’m going to take that 3% that’s advertised as a commission and share it back with you,” nope, you’ll go to jail. You can’t do that. That’s absurd.

Even if you say, “I’m going to do buy-side representation for 1%,” the seller still has to pay 4% now, right? If it’s a 3% listing arrangement and then 1%, maybe it’s 6% all in, but of the 3%, 2% goes back to you. The whole thing is messed up.

The only way to really violently change this is to have your own marketplace. That’s my perspective.

I think the key part here is that transactions that happen incredibly infrequently usually end up being full of fraud. There’s a reason why carnies leave town. That’s a real thing, right? It’s like, “I got your money. I’m out of here.”

The odds are you’re going to buy maybe 2 homes in your lifetime, and the odds are you’re not going to use the same process to buy both of them. People intuitively understand this about used-car dealers. They’re like, “Hey, I’m not going to go to a used-car dealer that doesn’t offer certified pre-owned cars because I don’t trust a third party. I need someone else to certify this thing.” That’s a very real thing.

This is why, by the way, as people in software, we have a very hard time understanding this problem. In software, most of us get paid every month someone uses our product. It’s a long-term relationship: the more you use it, the more money I make. But software didn’t used to be subscription software. Back then, software also had this problem.

The best way to align counterparties is to take the transaction and stretch it out over time, to make sure that I’m interested in you liking me 10 years from now. If you do that, most of these problems tend to get solved. But in order to do that, you need a counterparty, or at least a third party, to certify that this thing is good and that has an interest outside that one transaction immediately.

That’s why Amazon works seemingly well. You buy things from Amazon, and if you don’t like them, you can return them. We launched this yesterday at Opendoor in Dallas, Texas. You buy a home from Opendoor, and if you don’t like it, you can return it. You can move in early.

Yeah. Try it out. Don’t like it? Return it.

Wow. No one would do that. This doesn’t ever happen regularly with a regular real estate transaction, which is why, if you’ve ever bought a house, your agent has said something along the lines of, “Oh, they just turned down an offer exactly like that,” or, “Oh, they have another offer coming on Tuesday.”

They never have another offer coming on Tuesday. They didn’t just turn down an offer like that. I think there’s a very real thing where you need, first, a counterparty who is interested in the long term, to start out a transaction, and, second, to remove the “I make all my money from this one transaction right now” dynamic.

Both of those things just tend to lead to bad outcomes in any market where there are agency problems that lead to terrible outcomes.

Yeah.

Well, there’s a corollary to that as well, which is on the financial-services side. Part of it is, if I only make a transaction happen once every 10 years, I don’t know it that well.

But if you’re only doing 2 of these a year, candidly, you don’t know it that well.

Yeah. And then you have this other complicated thing: how do you buy a house? How do you afford a house? To me, it’s kind of crazy that the financing stuff is totally divorced from the buying and selling, which is—part of how we became friendly is through Affirm, right? How do you divorce buying and selling from financing? They are one and the same.

I can’t buy a house until I sell my current house, and the real estate agent isn’t going to help me with that. They don’t have a big checkbook.

No, I think the agency problem that exists in real estate is actually multiplied along the chain because there’s usually an agent involved in the mortgage.

There's usually an agent involved in the insurance. There's usually an agent involved in every single thing you do, sometimes even escrow. So, let's say one principal-agent problem is a bad one in a real estate transaction; you usually have 5.

Yeah. So it's multiplied. And let's say one person making money off you once and never seeing you again is bad; in a real estate transaction, you usually have at least 5, sometimes a bigger number. Like, the guy who does the inspection for your house—you're never going to see that guy again.

And I think people can have very good people along this chain—the whole chain—and you end up with very terrible outcomes.

Right, because the system is basically designed not to lead to good outcomes.

Well, and the thing is, you have a lot of these subscale things that are actually pretty cool that have never been productized and rolled out.

So, if you are an English professor at Princeton or Stanford, you probably can't afford to buy a house, right?

The university helps you buy a house.

Or they will subsidize a mortgage.

One of my favorite examples is something called seller financing. Every now and then, you'll see a house sold by a very rich person who is like, “You know what? I don't need all the cash right now. You can pay me over time.” It's like BNPL. I mean, a mortgage is BNPL, but this is somebody saying, “I will be your financing option and not a bank.”

These are all really interesting ideas, but only Stanford and Princeton and other schools do that for their English professors, and only really rich people do this for homes. You could see, “Oh, here's a home that's listed for sale that has this special financing.”

That's done in the brick-and-mortar retail world every day. That's called every day that ends in Y. It's like, “Oh, we want to sell more stuff.”

Procter & Gamble—yeah, we want to sell more stuff. Here's a coupon.

Or Xbox wants to sell more things, or Lexus wants to sell more cars: 0% financing, 0% APR, Labor Day sale. Doing that for homes makes so much sense, but the real estate agents don't know what they're doing. There is one area in real estate where I think there are far fewer of these problems, which is when you buy a brand-new home from a builder.

Yes, when you buy a brand-new home from a builder, when you walk into a large community and buy a home from them, they've solved most of these problems. The production one—

I know exactly what they're like.

Yeah.

Everything is bundled. You have very few agency problems. You have some still, even then, if you shop with your own financing, but the odds are you shouldn't just take that financing because it's not optimized for you.

And this is, I think, the second very classical problem. Basically, it's a marketplace problem that you have to solve, and I agree with you: the way to solve it is you need to gather significant inventory and make it desirable for counterparties to come to you to buy it.

So you can actually start the flywheel.

Yeah. And you said a company that could do that would be one of the biggest in the world. Has a company like Amazon or other big companies tried to do something like this, or is it just so far afield that they would never?

Well, the other thing that's very unique about this industry is it's so fundamentally local. So if you say, “I want to beat the MLS,” there is no MLS, Inc.

It's not like I'm going to go beat those guys, and the commander's intent from the general is, like, “Take that hill.” It's every market is different. Just because you could dominate Charlotte, that does not make a single dent at all in Hawaii.

Actually, the one case where I used a real estate agent to buy my house was in Hawaii. I have a house in Hawaii, and they don't use the MLS there. It's just this captive system where they make sure that you go through them, and those agents do so well because it just shows the value of a marketplace. But you could have a company that decides, “I'm going to go run the table in XYZ place,” and it doesn't show up in a different geography.

What you do see, going back to these pockets of esoteric products, either financial or real estate, is executive moving. This is actually a big thing. Imagine that you're hired as—I actually have a friend who was hired as the CMO at Home Depot, and she and her husband lived in New York. Guess what? Home Depot is not in New York; they're in Atlanta. They're like, “Okay, we will buy your old house from you and we'll pay top dollar in cash. We'll help you buy a house over here. We'll put you up in an apartment for 2 months. We'll do all of these things.” Why? Because they want to get this person hired as the CMO of Home Depot. And they bundle in all these other things.

They're not trying to say, “I want to go disrupt the MLS.” I mean, they should. I would hope that eBay—and I don't know the history on this; I'm sure you probably have looked into it—has tried doing real estate. And there's another company called Auction.com that really got into the distressed commercial real estate space and some distressed residential real estate. If it's distressed, like you haven't paid your property taxes, the government seizes your home. The government doesn't want to hold on to a home and pay its own property taxes to itself, so it auctions it off, and it needs to close within 5 days. That's what something like Auction.com is very well situated to do.

But I find that you have some of these bespoke products, as an example, executive moves, and those were done in the very, very first-class way. Why can't—this is the cool thing about technology—would you rather be the richest person in the world in 1900 with no penicillin, no iPhone, and no Netflix, or a middle/lower-middle-class person in 2025 with penicillin, Netflix, and an iPhone? I'd much rather be the lower-middle-class person than the richest person in the world in 1900. What technology often does is help diffuse these amazing products down to everybody. Everybody gets the education of a billionaire. That's the, I think, motto of a16z, right? Let's do all of these things.

Yeah. I think it's a hard problem, which is why no one has solved it, and lots of smart people have worked on this problem, or at least some parts of it. But I think there have been 3 structural flaws that have basically killed every attempt to solve this.

The first one is, “Let me solve a part of it. Let me solve the tiniest, most profitable part of it first,” which objectively has failed every single time it's been tried. When Amazon started, they didn't say, “Hey, let me sell books.” Selling books is not that profitable, but it's a really good way to get inventory, right? You can get all the inventory of all books, and it has literally every book.

So, the first problem has been essentially narrowing yourself down to one thing. The second problem has been basically a channel problem, which is, “Hey, I'm going to solve this problem through the traditional channels,” and that objectively has just failed, I think miserably, every single time. And the third problem is that some people have gone relatively large. There are some relatively large businesses that do this in some pockets of the world, but they all essentially tip over and fail because they've solved it by throwing human beings at it and treating it like an operational problem.

Those are the next 3 buckets of problems explaining why no one has solved this. Before Carvana solved it for cars, this exact same problem—the exact same set of problems, in the same way—existed for cars. Carvana is, like, 3% of the U.S. car market. I don't know how much, probably somewhere around there.

So, I think there are very real parallels. There are things that probably could not have been solved in a way that was affordable and efficient 10 years ago. I generally think it's actually a real thing. And there were other problems that were solved the wrong way because, for a brief period of time, money was free.

Yeah.

Everyone made all the wrong decisions they possibly could. But I do think we're in this special time, this window, where basically all the tools you need to solve this problem kind of exist.

We just get deeper there for a second. How much is real estate like healthcare, where the regulatory landscape has distorted the market? Feel free to quibble with that framing of healthcare. Or is it just hard? Are the market dynamics you were describing simply emergent from how real estate works?

So, I think it's far more similar to automobiles than healthcare.

I agree.

It's just basically identical, honestly. If you look at how cars were sold basically until Tesla, everyone had a dealer network because once you had a dealer network, one was required. But Tesla's like, “Hold on a second. If I don't have 1 dealer, I don't have to have any dealers.” And the answer is yes, outside New Jersey, I think, right?

And I think in Texas, you're not allowed to buy a car unless it's sold at a car dealer. So, again, a couple of places.

So, for example, in North Carolina, Georgia, and Louisiana, I think those 3 states, you can't close a real estate transaction digitally.

You have to have a wet signature. That's kind of annoying, but that's a real thing. But it's not at all like the healthcare system; it's much more similar to how cars were sold. There are regulations, some good and some terrible, but they're much more similar in nature to how cars were sold.

How come we've solved this in cars?

Well, we hadn't solved this in cars. We had not solved this in cars until about 5 years ago. We solved this for new cars—Tesla solved this—and for used cars, Carvana solved this.

There are very real operational challenges that Carvana has that don't exist for homes, right? Carvana moves every single car they buy. Opendoor does not move homes; homes stay where they are. There are very real upsides to trying to solve this problem in homes, and cars obviously have downsides too.

But I think underwriting, for what it's worth, is similar-ish. The underwriting of a car is difficult; the underwriting of a home is difficult. They're similar-ish in difficulty—different problems, but similar-ish.

Well, this is the main thing: you can really subsegment the market. I could talk for 10 hours about why healthcare is messed up, but fundamentally, there are no prices.

Yeah.

How much does this cost? The doctor doesn't even know, and then, "Oh, well, we're going to bill your insurance company this much." They won't pay it, and because they won't pay it, the doctor charges more. The whole thing is messed up.

Here, you have price transparency, for sure. I know exactly how much everything is going to cost. You get a little bit misled by some of the agents who say, "Buy-side is free because the seller pays," blah blah blah. But you can chop this up into different markets, right?

This is what's very helpful to understand: selling $50 million luxury homes and how to price those. A $50 million house that's a spec home could sell for $25 million.

There's not really—but this is the cool thing about real estate: people need to live somewhere. They either rent or own, and there are different models in between. You can rent to own. You can get your friend to give you a free spot and not charge you anything, but you either rent or own.

The floor on valuation is the rental price. How does that compare—going back to cap rates—to the cost of capital? The vast majority of homes in America, number 1, qualify for what's called conforming mortgages. There already is this idea that for really expensive homes, you get a jumbo mortgage, and then there's everything else. The pricing of everything else is a lot more like cars.

It's actually transparent, too. It's far more transparent than cars.

With cars, there's a very real thing about repair. It's very opaque, and it's hard to know about structural damage.

Structural damage, I call it—or your fan belt. I don't know what a fan belt is, but there are things that are difficult. Then there's OEM versus non-OEM parts. There's a variety of things; the decision matrix is much bigger.

For homes, it's relatively transparent. And, by the way, to give Opendoor credit, this is something Opendoor is exceptionally good at: pricing renovations and changes. I have zero credit for this; they've dialed this in in a way that's actually surprisingly good.

Which makes sense, because who's going to get a better shipping rate with UPS—

Right?

Me, who ships 1 package a year, or Amazon, which eventually got such a good shipping rate that it couldn't get any better unless it started its own shipping network called Amazon Logistics?

You have economies of scale that never go to the agent, that never go to the customer, but should go to a company.

Yeah. I guess the real thing—and this is actually the real thing—is that I'm relatively ideological about this. I think you can make Opendoor an exceptionally good business, and the average person pays less for a house and sells a house for more. I think you can—it's actually, the math here is not that difficult.

How do you do that? There's so much friction out there. If Opendoor does nothing else than help you time the closing of your new house with selling your old house and closing on your new house, if that's all Opendoor does, you've now avoided, on average, about 3 mortgage payments.

Who was getting paid that money? Why? That's a real thing that everyone—everyone pays at least 1 extra mortgage payment or 1 extra rent payment, and it's frequently more. That's just 1 thing, and it's a service that Opendoor can deliver to most people even without buying their house. There's a very real thing there.

I think there are a bunch of these things that exist. The principal-agent problem is real. The very real thing on cost of capital is that when you aggregate lots of homes, your cost of capital tends to be lower. Therefore, that's a real thing.

The very real thing is the level of underwriting you can do on warranties. The reason Opendoor can offer a 7-day trial of a home is because we own lots of homes. You don't like this one? If we take it back, the odds are you're going to buy another house from us, right?

All these frictions exist because the system everywhere is subscale: there are so many principal-agent problems, and so many of these transactions are one-time transactions.

Okay, let's zoom back into the company for a second. Alex, you make the investment, and you've got this vision for the company and the opportunity. What have we learned about the feasibility of the opportunity based on the company's performance?

Another way of asking is: for people who haven't been following the company's trajectory, how have things gone? What have we learned? What have the ups and downs been, et cetera?

Well, there was a point in time where Opendoor was such a good idea. There's a saying that we use a lot in venture capital: you want to invest in a bad idea—something that looks like a bad idea but is actually a good idea—because if it looks like a good idea, it becomes a bad idea.

Opendoor was somewhere in between the 2 of these, because Zillow was like, "Oh my God." Rich Barton comes back to Zillow, Spencer leaves, and it's like, "I have to go do what Opendoor does." Everybody was getting into iBuying. Offerpad popped up, but Zillow was the big one.

They couldn't do it.

Well, here's why they couldn't do it. It's an interesting story. The other part of the story is that Ben Thompson wrote a post that almost summarizes, in a more eloquent form, what I was talking about: if you get all the supply—or just a small sliver of proprietary supply—you get all the demand. Once you have all the demand, supply comes to you, and that's how you finally build a marketplace.

I think rumor has it that Rich Barton reads Ben Thompson, as everybody should, because Ben Thompson is very, very smart. "Holy shit, we have to do this."

Yeah.

So they start doing this, and they're making infinite money. But this is where cohort math is so important—

Because the first homes to sell are the positive-selection homes. Imagine that it's October right now. Alex and I buy 1,000 homes today. The best homes will sell tomorrow.

The homes we're stuck with 2 years from now—there are ghosts that live in those homes, or termites, mega-termites. There are things that are wrong with them. Zillow thought, as a public company operating in the public limelight, "We're making so much money on iBuying," because it was holding everything that hadn't sold at NAV, net asset value, and yielding profits all along the way.

But you have to let the whole cohort cure, as the term goes. Then it's, "Uh-oh, we lost a lot of money on that cohort."

So what happened was Opendoor started getting outbid because this was such a good idea. Everybody thinks it's a good idea. It started becoming a little bit of a bad idea pre-marketplace. Zillow starts paying more. In fact, Zillow even had, "Hey, don't sell to Opendoor. Sell to us. We'll pay $1 more."

You always have to realize, if you talk to anybody who's a professional trader in the equity markets, they'll say, "You have to assume that it's an adversarial process." Anytime somebody hits your bid, there's something wrong. You should assume that you're about to get taken advantage of.

Zillow didn't really understand that, but they muddied the waters for everybody else. Eventually, Zillow pulled out of that business. Money started becoming not free.

If you're stuck with a huge amount of inventory, being a market maker—I mean, if you look at the most profitable companies in the world, Jane Street is a market maker, Virtu is a market maker, Citadel Securities is a market maker—they don't win all of the time, because if they did, then they would be doing something probably illegal.

They make money most of the time. They have a weighted coin, and they hold inventory and make money on this bid-ask spread from somebody who’s hopefully not taking advantage of them, which is why they want retail flow.

What happens, though, is that a lot of the market makers in the equities world don’t like to hold positions overnight because things could change. Opendoor is holding a lot of positions beyond just overnight, and then interest rates went from around 0% to around 4%. Four percent is not high in the history of the world, but going from zero to 4% in a few months is really high.

The pace at which interest rates increased—I mean, I think we can say in hindsight this was deeply irresponsible for the country.

Silicon Valley Bank went bankrupt, right?

Asset prices went down, yeah.

I think this is—look, I don’t think Opendoor made lots of mistakes regardless of interest rates, but there’s a very real thing that happened. Companies shouldn’t take credit or blame for macro, but what happened in the U.S. with interest rates had basically never happened before, and I think it will literally never happen again because it was so obviously stupid.

So basically, what happens is you’re left with all this inventory. The whole point is that if you’re just buying and selling, making money 51% of the time, and earning a spread between the top bid and the lowest ask, there’s this margin in between. If you just keep buying and selling, you can actually make this work.

This is why Jane Street makes a lot of money, or Citadel Securities, or any of these companies that you’ve heard of. They’re doing it very, very frequently. They’re typically not holding assets for a very long period of time. Some of them never hold assets overnight, just as a standing rule, and Opendoor has a lot of assets.

Then what happens is, when interest rates go up, what happens to asset prices? They tend to go down because now it’s like—think of it in terms of a mortgage payment. Before, my mortgage payment used to be $1,000. I only have $1,000 a month to switch from my rental payment to buying a house because I can make a down payment. Well, wait a minute: Now it costs me $2,000 a month, so therefore I don’t want to buy the house. What happens when aggregate demand goes down? Prices go down.

This didn’t happen as much in housing as would have been anticipated because we have a shortage of homes, and that’s a separate topic. But asset prices in general went down. SVB, Silicon Valley Bank, went bankrupt because interest rates went up, and if you’re stuck with a bunch of 75-basis-point, 10-year T-bills, they’re worth half as much as they were before. Now you have to go sell them. Uh-oh, you’re bankrupt.

The pace at which this happened was very, very high. People anticipated that rates were low, but they thought, “The Fed will raise by 25 basis points, then another 25.” It wasn’t like, boom. That was 5 sigma—or I’m making up the number of sigmas—but it was many, many standard deviations beyond the norm in terms of what you could anticipate.

All of the risk capital—it was kind of a double whammy. Risk capital, like venture capital, is risk capital. When interest rates are low, I don’t want to earn 75 basis points on T-bills. I’m going to invest in riskier things. I’m going to give it to venture capital. I’m going to give it to private equity. So risk capital pulls back. Asset prices go down. Demand for homes goes down because interest rates went up. It was kind of like everything went wrong at the same time.

The vision of the company—at least my vision of it, and I’m not the founder, but my vision as an investor—was that you guys had a chance at building a marketplace, but it was going to take time. You can’t just snap your fingers and, boom, you have a marketplace. How long did it take Amazon to build a marketplace? How many very, very sad, despondent shareholder letters did Jeff Bezos have to write before the marketplace appeared? A long, long time.

Opendoor, I think, was on the path and then just got hit with this triple, quadruple, or maybe quintuple whammy.

There’s also a real thing. Look, I think it’s very good when companies are blamed for their mistakes. It’s actually a good thing for the world. We should all admit our mistakes because that’s how we learn. The same thing happened to Amazon much earlier and to Carvana around the same time.

Yeah.

But Amazon and Carvana reacted differently to that thing happening. I’m like, “Cool. Was the original idea a good idea? All right, what was the mistake? Let’s shed the mistakes and go.”

I think what Opendoor publicly did was essentially abandon the original mission. It was, “Cool, we’re going to derisk this company a lot. We’re going to derisk the company across the entire segment.” I think that actually is a very hard spiral to recover from because you’ve now made a kind of mistake about who you are, which makes it really easy for everyone else to make the same mistake.

If Opendoor is an old-fashioned operational house flipper, that could be meaningful, but it’s not that big a business, right? And they lost faith in the original vision and the feasibility of it.

I mean, look, I think there’s a very real thing about being a public company. The things you have to deal with are amplified in a public company, but private companies basically all have the exact same problems. In a private company, these problems are discussed with your VCs at a board table. In a public company, they’re discussed on Reddit and described in The Wall Street Journal.

If you care a great deal about what’s said about you in The Wall Street Journal, running a public company is incredibly difficult. It’s just very difficult. Luckily, I just don’t care.

Right. So you’ve come back and you’ve joined and said, “Hey, let’s go back. The original vision was a good idea. Let’s bring it back”? Or how?

Well, look, I think the mission is a worthwhile one. I think Opendoor has made mistakes along the path, and we should learn from our mistakes. But the company is not made better by becoming weaker, right?

Yeah.

Will we make mistakes again? One hundred percent. We will for sure make mistakes again. We just launched Try Before You Buy, the 7-day trial of homes in Dallas, Texas. It’s day 1. I don’t know how it’s going to go. We’ll find out.

I was telling this to one of our product managers this morning. I got to Opendoor and I felt like, for 3 years, someone had stood by the company saying, “Hold, hold. Wait. Just wait for the macro to recover.”

Have you ever watched Braveheart? There’s a scene where Mel Gibson is standing in front of the Scottish army and the English are coming with weapons. Mel Gibson is standing there saying, “Hold, hold.” I’m like, “Don’t hold. Attack.”

You just don’t tend to build great software companies like that. I think you can build great hedge funds like that, but Opendoor is not a hedge fund. If it were a hedge fund, it needed to be 6 guys in New York with laptops. We’re not that. We’re a software company, and software companies need to be basically always on attack.

Yeah.

Always, always, always on attack. We’ll attack some long hills. We’ll accidentally lose some ships, but we’ll always be on attack.

Yeah. Alex, what advice have you given Kaz, or what would you be thinking about in terms of what strategic decisions he has to make, or what’s important if you were taking over as CEO?

Luckily, I’m not. So thank you for doing this service for mankind.

When I first heard that he was taking the job, I think I sent you my YouTube video. It’s actually my most-viewed video ever, largely by real estate agents who hate me and want to kill me. This was circulating, speaking of Reddit, and The Courage to Be Disliked. It’s a good book.

The Courage to Be Disliked is an excellent book. It’s among the best books you can read as a founder. You must read it. Excellent.

It’s this whole Socratic dialogue in this Adlerian psychology mode. It’s very, very interesting.

After giving that talk—oh my God—the amount of hate mail from real estate agents was incredible. I didn’t know there was such a thing as real estate agent hate mail because I was basically saying, “These people are leeches on society.” Lo and behold, they don’t like being called that.

They don’t like being called that.

And I know some real estate agents, but the whole process doesn’t make sense. There should not be a 6% tax, which, by the way, I make this point in the presentation. It’s unique to America. You can pick any other country in the world; it’s nowhere near that level of spread. You don’t have all this regulatory capture. It doesn’t make any sense.

If you have a marketplace for homes, it’s the biggest marketplace in the world. We know that the stock market is big, but the residential real estate market is bigger. So we know that—why are the NYSE and Nasdaq worth a lot? Why is eBay worth a lot? All of these things do everything but homes.

If you stick to that—which I think the company, and this is what Eric Wu was committed to doing—it just became an untenable situation where it’s hard to do it in the public eye.

They don’t like being called that.

For sure, because—

I think it’s harder, but I don’t think the company couldn’t have still done it in the public eye. The company had just learned a bunch of wrong lessons.

Yeah, and that was the point. When money is free, success is a terrible teacher. It’s like, “Oh, we’re making all this money flipping homes,” when the real magic is going to happen if Opendoor does not take principal risk on a house and it just means that it’s the lowest-cost way to sell.

It's the best way to buy, and that's what, by the way, buying stuff on Amazon is. That's what buying stuff—eBay has kind of lost its luster, but still, if I catch a home run at a Giants game, where do I go sell that? Who has the deepest liquidity? It's going to be eBay, and I'm going to sell it there. Everybody who wants to buy a baseball goes to eBay, and eBay does not have to take possession of the baseball. That's why they have an amazing business.

They probably have too many people who work there. It's a shadow of its former self. Apparently, it's still using the same .NET library from 1997. All sorts of issues there, but I'm excited about this vision of having a housing marketplace and also coming up with creative financing options. Finance—these are all one-offs, to your point. By the way, you can come up with better insurance options. If you control the entire thing and, to your point, have a lifelong relationship with that customer, there are so many things that you can do.

Underwriting a home for a mortgage, underwriting a home for insurance, and underwriting a home for buying it are fundamentally not that different as exercises. They really aren't. But all the players in this space come up with different answers to these questions, right? This should very clearly be done by one person. Very clearly, one company should do this. You and I are money nerds, so I literally wrote a book about that.

Yeah. I know you did the U.S.

So, I think there's a very—I'm generally excited about this. I do think, for what it's worth, that the right model actually is much closer to Amazon than to eBay.

Yeah.

Because I actually think Opendoor will have principal risk on quite a lot of its inventory for a very long time. Now, should it only have that? Obviously not. And it's also not a binary thing. It's not like zero risk or 100% risk; there's a gradient. You can live anywhere along the chain. You can deliver different values. We don't have to agree with a seller on the price of the home. We can say, “Cool. We agree it's at least this.”

Right. Right.

We'll give you at least that, and then let's sell it. You take the rest of the risk. We'll take the risk for the first bit. Yeah.

And there's a lot you can do here. It is not quite binary. That becomes a UX problem rather than an underwriting problem, but it's fixable.

Well, that's why Amazon is, last I checked, a $2 trillion company. eBay is a $40 billion company. So, objectively, I hope you become Amazon and not eBay.

That's the thing. It's like there's a market order and a limit order.

Yes. Yes.

And there's a lot of liquidity beneath the top bid and the lowest ask. There's always a lot of liquidity there, and those are the limit orders.

Gearing toward the future, what else can you hint at in terms of the biggest priorities, decisions, or strategic directions that you're focused on?

Okay. I think there isn't a 5-year Soviet plan here. There isn't one, because I think that's just not going to work out anyway. It just doesn't work out that well.

The thing I deeply admired about my time at Shopify is that, if you watch chess players go back in time and start chess, they'll teach you opening strategies: Open like this; this is what you should do. If you look at what modern chess players do, they do something called positional chess, right? Always put yourself in a better position. Play for the next position, the one after that, and just give yourself more options.

It's actually the right way to build a company transparently. You don't want to hold the strategy tightly—not the mission, the strategy. You want to hold the strategy very loosely. I don't have a secret bag of tricks, but I think there's a very real thing where Opendoor does too little for its sellers and does almost nothing for its buyers. We'll solve both of those problems.

I think there's a very real thing where, due to some bad advice, the company tried for a while to essentially only buy homes that were mispriced. “We will only buy homes that are mispriced by 20%.” It turns out, again, you can build a hedge fund doing that. You can't really build a software company doing that. Our goal is to buy and sell homes for a fair price. For a fair price, we will buy and sell homes.

I think those are 3 big things. Then what we'll do is just ask people to hold us to account against this mission. Hold us to account against this mission: What we are going to do out there is try to be the person you transact with quite frequently when buying or selling homes—not just 0.5% of the time, but a significant percentage of the time.

By the way, we've already—when I started at Opendoor, Opendoor was only available in 48 markets in the U.S. It is now available in every market in the U.S., because you can push pixels relatively easily.

Yeah.

So, there's a bunch there, a bunch of bad stuff that's—I think you should expect us to be a much more ambitious company.

Well, in terms of holding you to account, one thing that's so remarkable about the company, among other things, is how many people on the internet feel so passionate about the company—the Opendoor Army. How do we explain why Opendoor is a company that so many people feel so passionate about? And how does that impact the business, or how you think about it?

But I never—I never worked on Wall Street, and until I started this job, I owned one ticker. It was Shopify. At some point soon, I'll follow this ticker once I'm allowed to. I'm not a stock analyst. I'm just not one. I build parts for a living.

But I think there's a very real thing that happens where people feel like the natural intuitions of the average American are a very good indicator of what's true. William F. Buckley Jr. used to say he'd rather grab the phone book of Boston, Massachusetts, than go to a Harvard professor's book for advice.

There's a very real thing: the natural intuition of people is real. I think people look at real estate and how those transactions are done, and they say, “Well, this is stupid. This is not how this should be.” I think that's what it is. People are looking at it and saying, “Hey, this is not how this should work. There's a company out there that has told us they're going to fix this problem. We want them to go fix it.”

What I think is the most wonderful thing about this is that, if you engage with the average person, they tend to be deeply reasonable, ask really good questions, and have really good ideas. Whereas if you engage with supposed experts, they have preconceived biases and they're usually wrong. There's a whole pretense-of-knowledge thing that happens.

I think it's actually deeply helpful for the company to be a company where you know who I want advice from? The average person who owns the stock and is about to buy or sell a home. I want advice from that person, because the odds are they have a wider aperture of possibility.

This is a real question that someone asked me the other day: “Hey, why can I return what I bought on Amazon and not a home?” That's a really decent question. From the time someone asked that question to the time we launched the product was maybe 12 days.

There's something broken here. It is intuition. Intuition is crystallized knowledge, and intuition shared by lots of people is literally the knowledge of a significant number of people. I think it's important.

Yeah.

I think it's a good place to wrap.

Can I do the thing I'm supposed to do and plug for a second, please?

Yeah. So, please, please—any plugs?

Look, it is a bad thing for the world that people in this country can't afford to own their home. Kids that grow up in homes that are owned by their parents have better life outcomes. People who live in homes have better communities, lower crime, and better health outcomes. If you're so inclined to solve this problem with us, my DMs are open. Find me. We are going to build the most aggressive team in software.