Amazon 2.0

Reading this link made me think of a passage I recently read in a book called The Black Swan, by Nassim Nicholas Taleb. He wrote a section talking about how to accumulate wealth. He said it is obviously better to build something once and sell millions of copies, than having to build the same thing over and over again. His example is an author who writes a book one time and then sells millions of copies compared with say a plumber who learns his skills and then has to repeat his effort time and time again.

So Carvana has to build physical structure for the cars. Even if show rooms are few, the inventory has to be stored somewhere. If you live within 100 miles of a showroom, Carvana will deliver the car of your choice to your front door. If after an hour of driving you don’t like it they will take it back. If after six days of driving you don’t like it a truck will return and take the car back. Sounds like a lot of moving pieces. But they have the ability to purchase your trade-in cheaply. (and of course your car has to be made showroom ready.) Lots of moving pieces, people and capital investment.

Cars are becoming more and more similar. With quality going higher and higher, cars are becoming a commodity. two years ago I bought a car via e-bay. I shopped Carvanna and visited local dealerships to test drive the cars I was interested in. In the end, I saved thousands of dollars buying a car off e-bay. I thought the Carvana process terrific, but in the end, my wallet won. My car was a Hyundai Genesis and was then and still is under new car warranty. Now, for a car not under warranty, perhaps the results would have been different.

As I understand it, Carvanna makes money three ways: selling the car, making a loan, and buying inventory (think trade-ins) cheaply.

So, having said all this, I am still thinking build once and selling millions of copies, as in software, has to produce higher margins than buying and selling individual cars, regardless of projected revenue growth.



Interesting stuff about Zillow and Redfin. I tend to agree real estate is ripe for disruption, in some way. Just not sure how yet. However a friend of mine has been an agent with ExP World Holdings (ticker symbol EXPI) for a few years now and swears by it. They basically run their entire organization via Cloud based apps. No physical presence, etc. They are growing fast, which makes me want to consider them as a purchase. But from what I can tell their disruption is on the agent model itself and not necessarily on the customer-facing side (i.e. no reduction in fees). Not sold on this idea yet but intrigued by all three names here.

The important point here is that neither Zillow or Carvana have the same ability to scale as the SAAS companies that we are looking at here.

That’s wrong. Not only are Zillow and Carvana scaling their business, but they are doing so far more rapidly than SAAS companies.

A software company has a one-time big cost to create a sellable package, followed by much lower support costs. The scale leads to very high margins if the product sells well.

I see, you’re talking about scaling to profitability. You’re right. I absolutely agree that software companies have higher margins than retailers. That’s a no-brainer. I just disagree that retailers are bad investments. No, they’re not. If you’re buying the right retailer, they can be fantastic investments.

Here is a big negative when it comes to tech stocks: the market is limited. The more limited it is, the weaker the tech stock. I tried to invest in optical stocks for years and years, because I thought the technology was important. But a major problem with these stocks was that the market was so limited. How many potential customers did JDS Uniphase have? Basically they were limited to companies who were building out telephone infrastructure. Since the customers were so limited, all of those tech stocks were bad investments.

I’m not a techie, which means I’m really weak at analyzing how cool or important a technology is. I would suggest the more you have to explain or translate a technological innovation to people, the more dangerous that investment is. If it’s difficult to understand, that means the customer/decider pool is more and more limited.

Peter Lynch used to joke, “I don’t even know how electricity works.”

One of the powerful benefits of SaaS is that you don’t have to first convince the chief technology officer to buy your stuff. That’s what a “land and expand” strategy is. You can convince anybody in the business to try your service. And then as it catches on within the business, the uppity ups may decide to make a company-wide purchase.

Smart tech companies try to expand their market by making their technology easy for non-techies to understand. That’s why internet companies have been fantastic investments. They are tech stocks that are easily understood by many. I don’t know how Google works, but I understand Google. I don’t know how Carvana works, but I understand Carvana.

The potential market for a Google or a Carvana or a Zillow or an Amazon is far greater than the potential market for an Alteryx or a Twilio. Don’t kid yourself about TAM, it’s very important.

The margin on Zillow and Carvana isn’t going to increase substantially as they grow. They’ll continue to be heavily capital-intensive with inventory and carrying costs. And neither one has a unique product.

All this is true. Nonetheless, retailers can (and do) scale, and get very big. And internet retailers scale far more quickly than off-line retailers. And there are powerful networking effects at play. If Carvana has more car options than anybody else, that’s where you go to shop for a car. If Zillow is the company with all the housing data, that’s where you go to find your next house.

I feel like I was having these arguments 20 years ago!

1 Like

That’s wrong. Not only are Zillow and Carvana scaling their business, but they are doing so far more rapidly than SAAS companies.

Just look at the balance sheets and income statements of AYX and CVNA. We are seeing this play out in real time. CVNA now has $526 million in long term debt. $1.2 billion in total debt. AYX? $316 million in total debt. I don’t know why you’re so concerned over selling $2 billion, or $3 billion, and focusing on top line, when AYX is making more gross profit on a much smaller revenue number. So who is really scaling their business faster when AYX is making more gross profit? Who cares what sales are quite frankly! It’s not what matters.

Yes, CVNA has a large tam. By that definition you should also be buying Beyond Meat, because it too has a larger TAM than AYX. CVNA’s gross profit does not even cover their SG&A expense. They are a long way to profitability and they’re going to take a lot of debt to get there. Then that interest on the debt is going to be a bigger portion of the cost of selling a car.

AYX will be able to generate $1 worth of profit at a fraction of the cost of CVNA. All business is, is making the most profit possible with the smallest investment possible. The economics of AYX are much, much better than CVNA’s. I would never focus on just TAM but the economics of that TAM. Who cares if CVNA gets to $10 billion in sales when they can’t even cover basic operating expenses let alone show a profit. As I said before, Arrow Electronics, $25 billion year revenue business, worth $5 billion. Sales is not everything.

The reason JDSU is not a good business is because it’s a mature business catering to a few telco companies, and extremely competitive due to the concentrated customer base. It has been that way for years. CVNA, likewise is in a very competitive business selling cars, the same cars you can get at any other dealer. There is absolutely no differentiation in products. Other than “they have a bigger selection”, meaning a bunch of money tied up in inventory. That they make next to nothing on. There is no scaling up of buying used cars as a CVNA the same way WMT was able to buy in bulk from manufacturers.

It’s fine if you want to buy CVNA but in no way shape or form does it mean it’s better than AYX because it’s a bigger TAM market. Especially when one is growing dramatically and the other is not.


His example is an author who writes a book one time and then sells millions of copies compared with say a plumber who learns his skills and then has to repeat his effort time and time again.

This is an example of someone who actually tries business finds out. I found this out through two side businesses. 1) reselling vintage furniture, and 2) creating metal products with CNC equipment.

With reselling vintage furniture, finding the furniture at a price you could actually make money on is next to impossible. Sitting around at auction houses bidding against a bunch of other dealers not to mention people who wanted to buy it for themselves does is not an ideal method for acquiring product to sell. Nor is camping around in front of estate sales. Then the cost of selling it. And sitting on all the inventory until it sells. Only to repeat the process once you do sell it. All your money goes to inventory that sits there. That you possibly have to pay someone to refinish to actually sell it. This was always a hobby for my interest in furniture because I didn’t make a dime on it and I competed with other people doing the same or not making much.

With CNC manufacturing, getting inventory is just as easy as calling a steel supplier on the phone. Raw materials are only a small portion of the expense. I can buy sheet metal and let it sit there until it’s used with no problem, the raw materials are only 1/5 the cost of the sale. Cash flow is much better. Profit is much higher, and easier. There is differentiation between product so you can charge more. You can scale by actually making product and sell it at a price you could actually afford to employ people to make it for you. The only cost is the machinery and repairing it. Which can add up but it’s no worse than having all your money tied up in furniture inventory. Software companies have it even easier. They have no machinery they have to buy. Total cost is up front developing the software.

All business is is making the maximum profit with the smallest investment possible. Carvana is in a terribule industry for that to happen.

I just bought Taleb’s books again and reading Fooled by Randomness right now. Read two of them years ago but can’t even remember that example. Some things can just get glanced over without much thought at the time because they don’t mean anything to you.


I feel like I was having these arguments 20 years ago!

Ha. I’ll admit that somewhere in the Fool archives from 20 years ago, there’s a post where I said something like Amazon is just a bookstore, Walmart will eat their lunch!

The point is that for Carvana to sell a car, they need to go to an auction and buy it at a small discount to retail. To sell another car, they need to go out and buy another one at a small discount to retail. Merchandise isn’t going to cost substantially less with quantity. Same with Zillow.

Both companies have substantial competition that limits their asking prices. Carvana is competing against Carmax, AutoNation, Avis, and every new car/used car dealer lot in the area. Zillow’s pricing is limited by competitive values too.

A software company has roughly the same development cost whether they sell one copy or a hundred. The companies that we are looking at here are the ones that have already developed a marketable niche product that is somewhat differentiated from their competition. Profit margins will soar once enough copies are sold to cover costs. We’re buying (and selling) stocks here in the hope that they’ll reach that level quickly.

Does that make Carvana and Zillow bad investments? I hope not because I own some Zillow and Carmax, a Carvana competitor. I hope that both continue to give me good returns at relatively low risk, but I don’t consider either one part of my Saul’s Superstar portfolio.


Same with Zillow.

Except that Zillow doesn’t need to take title, so there is none of that inventory expense. And, if at all successful, they probably don’t have to work very hard at finding properties since sellers will bring the property to them.

1 Like

An interesting metric for carvana is their gross profit per car went from $2000 to $2400 in just one quarter. Not sure what caused that. If it’s scale we’re talking about

1 Like

I disagree, but maybe you’re right. Definitely a gorilla game going on. Zillow, Opendoor, Offerpad, Redfin, Keller Williams. It’s worth googling iBuyer and reading up on it.

Why would you think Zillow would be the winner? Any of these big RE agencies have the same data from the past 15+ years. Now they’re taking on buyer credit risk in the mortgage space as well. Most of us remember what it looked like for the non-bank and shadow-banking sector when they entered that space in 2005-07 it looked really great until it blew up right in their face.

It doesn’t make any conceivable sense to me that home buying by realtors/online lenders/data gatherers will be a gorilla game because you are going to sell to whichever of them offers you the most money. It is nothing like Amazon’s flywheel which improved results for buyers and seller by making it easier to shop there and had tech that was years ahead of their competition. RE sales are public information unlike Amazon’s internals.

I mean, Wal-mart, just passed Staples in online sales a couple years ago for crying out loud.

If I’m selling my home I’ll download the best 6-10 apps to find the iBuyer willing to pay the most in addition to my local B+M agent[s]. Nobody will care about Zillow unless they have the highest prices.

And if they always have the highest prices, that’s a recipe for disaster worse than any upcoming mortgage crash.

Not that Z can’t be successful and make money.

On a diff note, I do find a horrifying % number of Gen Zers don’t even have driver’s licenses both in the city and suburbs where I live. Ignore that trend to your detriment.


regarding re-selling homes. It seems to me that the company that will do will is the one that will let you represent your own home for 2 percent, but yet show it in multiple listing services. Selling my home to a Zillow or some such, does not make my wallet tingle. What does make me smile, is selling the house at full price and not paying the realtor 5 - 6 percent.

Who has sold every house he has owned FSBO

1 Like

I don’t think you actually understand Amazon’s business model. They are not a retailer. Let me clarify, yes, they sell a ton of stuff at retail prices to the general public off their web site, so they look a lot like a retailer, but that’s not the core of their business. Not to my way of seeing things anyway.

IMO what Amazon does, in fact, what they invented was to examine the most common activity on the planet (with the possible exception of sex), the exchange of goods for monetary consideration, and took note of all the primary frictions to this activity. Amazon created an infrastructure that reduced or eliminated as much friction as possible. They used the infrastructure to fuel their own business, but they also, in effect leased out excess capacity to anyone (with a few constraints) that wanted to take advantage of it.

They did exactly the same thing with AWS.

And BTW, you can buy a car on Amazon, Carvana so far as I can see has no moat.

And buying a house will never be as simple as buying a book. Aside from the fact that virtually every home sale requires financing, there’s all the stuff around title insurance (can’t get financing without it), recording, possible liens, tax status, etc. There’s just a whole lot of contingencies around real estate sales that add tons of friction (and expense) to the transaction. I don’t see this ever expanding to the scale of Amazon, but there is leverage in reducing the selling costs associated with brokerage fees. I’m not sure how/if Zillow is attacking that, but there’s an opening.