Liberty

If these are numbers you're comfortable with, Cameron, that's fine. I'll note that you make it sound like you're being generous but are using almost year old numbers (in a fast-growing business) and giving retail operations a margin that is lower than the margin that WMT has had in many of the past years when I think the AMZN model is quite a bit more efficient and Amazon's gross margins are a lot closer to WMT than COST (higher, actually, but you can make adjustments for AWS). I'll also note that giving AWS 40x isn't being generous. A dominant low-cost operator, high-margin business with recurring revenue that is successfully going up the value chain while growing at 40-50% CAGR while only having addressed a very low % of its potential market usually can't be bought for a 2.5% net yield. If you find one elsewhere, please let me know via private message.

I haven't done the math lately but I don't thing you get a valuation anywhere near $500B if you assume AMZN is going to have WMT or COST margins. You need to assume much higher margins. That's not what I said. I said their own retail operations would probably have that or better. Then the 3P stuff significantly higher, then AWS higher. But the margins aren't the whole story, the turnover, growth, and runway over time are also important factors, as well as options in other fields (they're building a Netflix competitor, they have a new fast-growing business with Alexa (in their own devices and licensed to others), they seem to want to slowly build a Fedex/UPS competitor, go into groceries, go against pharmacies, auto parts, etc)...

A few randoms thoughts: With some businesses, everything's pretty clear, and that's always nice. With visa you can look at the earnings and FCF and ROIC and ROE and do your valuation. With some others, things aren't as clear. You can either put them in the too hard pile, which I did with Amazon for a really long time (esp. considering I've been a customer since the late 1990s), or you can try to think things through and get a portrait that might be good enough to have an opinion. I think that's sometimes worth doing because there's usually little edge to be had by just looking at the same income statement and balance sheet that everybody's looking at and hoping to have a unique insight. So here are some things that I think often get overlooked: What are the chances that Amazon's mature retail businesses don't get margins comparable to WMT or COST? Is anything about the amazon model inherently less efficient or profitable? Their prices aren't that different. Seems to me like it should be even better, since all they need are huge increasingly automated distribution centers/warehouses with fewer employees, fewer niceties (customers don't see them), and a lot more throughput (rather than holding inventories on the shelves for people to see), and secular tailwind from the shift to e-commerce. And it seems to me like Amazon is at least getting as good a deal from suppliers as WMT and COST. Over time as you have more and more FCs close to big population centers, you can start to offer 2-day and 1-day shipping (very hard to match for most competitors) and cut into your shipping costs by leasing your own planes and ships and create your own last mile network (you first cherry-pick the highest-return routes, leaving UPS and Fedex with the rest). And if the mature retail P&Ls within amazon are getting margins at least as good as WMT and COST for equivalent merchandise (maybe with higher turnover, so better ROIC), then what are the chances that the loss-making startup P&Ls are going to never grow into profitable mature P&Ls later on? Then you have to consider third-party sales. This is a very large part of their sales (especially since in revenues they only recognize their cut and not the whole item price), and this is more of an eBay model. You connect to our APIs, list your product on our site. When you sell it we take a cut, ka-ching the money ends up in our bank account and we haven't touched anything. This seems to me a much better business than traditional retail. And on top of it Amazon is now offering FBA, so that people can pay even more just to hold inventory in amazon's logistics network. Amazon takes no inventory risk and gets even better utilization out of its fixed assets, and everything still takes place on their site, so they still own the customer relationship. Thanks to the breadth of inventory that 3P permits, they now have so many things that to many people, Amazon is the main search engine for stuff. There's surveys down this thread showing how people increasingly don't use google to search for products but go directly to Amazon. How much is it worth to be the google of products and have people not even consider competitors most of the time? And with half the country and increasingly other countries paying for Amazon Prime, people have a sunk cost that pushes them to use Amazon even more (same as a costco membership). Once you're the Google of products, you can start selling targeted ads in search results, which they're increasingly doing. Seems like another great business to be in. Then there's AWS. On-premise IT is a trillion dollar business, and over time most of it will move to the cloud because it's simply better in almost all aspects. Utilization is better, scalability is much easier, security is better unless it's a core competency of your company, etc. So I think there's a huge secular tailwind there, and scale matters a ton. There's probably going to be 3-4 big winners, and Amazon has a big lead. They keep cutting prices because they're riding Moore's Law and hard drives and ram getting cheaper, and the "economics of scale shared" model creates a nice flywheel. If you share the benefits of scale with your customers, you grow faster and are much harder to compete with, making you bigger, leading to more scale benefits which you can then use to keep growing, etc. Hard to catch up with the giant in an industry where scale matters a ton. Then there's the fact that most people seem to think AWS is just IaaS, a bunch of dumb servers, but they've actually been going up the stack pretty quickly. They sell a bunch of software as a service: Databases, all kinds of ML modules and lambdas, serverless stuff, etc. This is up the value chain and also a very good business, especially once other companies build their backend on your platform. They tend to stick around for that stuff. With a lot of these things, it's a flywheel where the bigger you get, the more you can do the things that makes you even bigger and keeps smaller competitors from being able to keep up with you. So you become an ever better value for customers, so they're even more sticky (Amazon Videos might make it even more likely that you'll renew your Amazon Prime, which keeps you buying more on Amazon because you don't worry about shipping, etc). There's a bunch more stuff, but this is some food for thought when you compare Amazon to Walmart or whatever. How much is that worth? Well, there's not one good answer. If there was, someone would've given it and there'd be no more debate about it. But I think too many people don't think things through enough to have an opinion. Looking at the numbers in filings isn't the same as understanding the business.

FB talks about ad load a lot. They increased it last year, not this year. The ads you're seeing are not necessarily the ads someone else is seeing. The whole point is these are targeted. The ad pricing is determined by automated auction. So when supply goes down, or grows slower than demand, pricing tends to go up (and vice versa). Pricing going up as supply growth slowed down is a good sign that they have pricing power, ie, advertisers are still getting good ROIs and aren't switching to other ad networks when prices go up.

One thing to keep in mind is that we're just talking about Facebook users. How's the MAUs been going for Instagram, Whatsapp, and Messenger, and how monetized are these things?

I guess I wasn't clear. What I'm saying is that this aspect of Graham's thinking has been fleshed out and explained in more detail by these other people. Of course it's true that the value is from that aspect. I'm saying that to understand it you can use these other frameworks.

I think it's more about adding to Graham's framework some of the learnings of Buffett, Munger, and Fisher to take into account more of the business quality, industry dynamic, qualitative aspects (culture, management, etc), and understanding how real value is created even if GAAP doesn't always capture it well because it was designed for different kinds of businesses. Other frameworks are also useful, like Ben Thompson's "aggregator" framework, which explains why some of the big internet businesses are very different from legacy businesses. Buffett has actually praised Bezos quite a bit in recent years, so he doesn't seem to think it's all a mirage of smoke and mirrors and no profits.

It's all in the presentation here: https://s21.q4cdn.com/399680738/files/doc_financials/2017/Q3/Q3-'17-Earnings-Presentation.pdf ARPUs in Europe are about 1/4 of NA, and ARPUs in Asia-Pacific are about 1/10th. Seems like an opportunity more than a problem.

Feel free to short it if that's your analysis of the situation ¯\_(ツ)_/¯

That's correct. Coal is still on the way out and EVs are still massively more efficient and cleaner compared to gasoline well-to-wheels. I wasn't talking about CO2, I was talking about air pollution as in smog. Having millions of gasoline and diesel vehicles in a city doesn't help. Obviously coal is terrible for smog too, but it's going away and tends not to be next to cities at least. Oh yeah, coal is the worst. No argument from me. We need to transition away from coal regardless of EVs or not, and since we have to do that, EVs are better than gasoline car in any scenario. Doesn't matter, over the lifetime of a vehicle as the grid gets cleaner, the vehicle will get cleaner, and the number of EVs will take a while to ramp up. So they might be a bit worse at first, but by the time there's a ton of EV, we'll likely have a much cleaner grid and much saner environmental policies. Waiting is not an option, we've waited way too long to address this problem already. I once saw some math about how the shift to big screen LCD TVs in the past decades had a bigger impact on the grid than transitioning to millions and millions of EVs would (suddenly everybody has a freaking 50-inch turned on for hours each day during peak time...). Trump's all hot air. The economics of coal don't make sense and will make less sense with each passing year, and the pendulum will swing the other way after Trump, there might be a price on carbon. Fracking will keep natural gas inexpensive, and wind and solar and storage keep getting cheaper year after year after year. There are also all kind of interesting nuclear techs being developed that might lead to a renaissance someday (my fave would be liquid-fluoride thorium, but there are all kinds of smaller breeders and such) I saw some LCA studies a few years ago that showed well-to-wheel for gas ICE vs EV on various sources, and they pretty much all came ahead up to 100% coal depending on the coal source and coal plant technology. I wish I had kept all those sources. People don't realize how inefficient ICEs are, and they don't take into account the whole life cycle of producing, refining, and distributing gasoline (it doesn't just fall from the sky). You also have to look at the longer-term rather than just a moment in time. Every year new cleaner generation capacity is going up on the grid. Over the 15+ years of use a EV gets, it'll get progressively cleaner, while an ICE vehicle will get progressively dirtier as it wears out.

The model has been explained often and in many places, I'm sure you can find them. They run a lot of things through the income statement that others might capitalize, and lots of growth expenses go into things like headcount. They basically target break even in the aggregate, but it's run as a conglomerate with mature profitable P&Ls funding startup P&Ls. What if they targeted 10% growth a year? Do you think profits would stay the same? Do you really think they need all this headcount growth and R&D and growth in leases and massive investments in places like India just to serve their existing customer base growing at 10%? Do you really think their retail operations are less efficient than Walmart and Costco (with the secular tailwind of e-commerce at its back), and that their cloud is a worse business than the cloud business of MSFT and GOOG (despite much bigger scale)? It's clear that the aggregate numbers hide all this, and that's why the market has underestimated Amazon for many years (if it hadn't, you wouldn't have gotten this much outperformance out of it, right?). Personally, I tend to think the people who are betting that Bezos is stupid and doesn't know how to make money are mistaken. Yeah, that's not surprising. Malone hadn't taught the world how to optimize growth vs tax efficiency back then. Still, I bet they'd have paid a lot more taxes if they had opened half as many stores each year, right?

That remains to be seen. At some point finding reinvestment opportunities for so much cashflow will be impossible and/or Bezos will want to start taking a profit, possibly to fund other ventures like Blue Origin or to start his own philanthropy (which will very likely be quite innovative and results-based). What about them? What about all of Amazon's customers who feel they're getting more value from Amazon than from legacy stores? You have to win people's business, you're not entitled to it. Otherwise, all the stores with 40% gross margins that existed before the discounters (Target, Walmart, etc) would still be around. What happens if any company blows up? Good question. That's not how Amazon operates. It's basically a conglomerate with dozens of P&L, the profitable and mature ones fund the faster growing but loss-making startups. This has been explained many times in many places. We only see the aggregate numbers, so it only looks like the whole think is breaking-even. Oh yeah? You mean they're constantly raising lots of equity? Or lots of debt? Or that by slowing investment a bit they couldn't pay employees with cash rather than with stock?

What do you mean that it remains to be seen if FB penetrates other countries? MAUs on FB for US & Canada are 239m. Europe is 364m. Asia-Pacific is 794m and Rest of World is 675m. I think that' pretty well penetrated in "other countries", depending on "other" compared to what. But if it's compared to N-A, then 88.5% of their users are outside. Even compared to N-A + Europe it's 70.9% outside of those "western" countries. On the other hand, Tencent is pretty a Chinese story (where they dominate). It might be a more compelling investment case, I don't know, but I don't think there's much question that FB is well penetrated around the world, with China a notable exception.

I don't think they're getting the message... http://nypost.com/2017/11/01/softbank-definitely-in-talks-with-charter-about-merger/

Changed to what? Taxing revenues?

Breaking news: Taxes are paid on profits. If you don't make much profits, you don't pay much taxes! Now for the weather... btw, it would be interesting to see if WMT paid much taxes during it's hyper-growth phase, when they were constantly investing their cashflow in opening lots of new stores. The government doesn't really lose, it's just deferred. If you have good reinvestment opportunities that help you grow much faster, you might not pay as much tax now, but you'll be much bigger later, so when you do ease off the gas and turn on the profit tap, the government will likely get much bigger absolute numbers of tax dollars than if you had shown profits much earlier and foregone growth opportunities. This makes sense as long as the growth gives good ROIC. Otherwise, it'll be customers and vendors capturing most of the value, not the corporation or the government...

Q3: https://investor.fb.com/investor-news/press-release-details/2017/Facebook-Reports-Third-Quarter-2017-Results/default.aspx Revenue up 47%, EPS 77%. :P

This looks so much better than TouchID:

Q3: http://s2.q4cdn.com/242125233/files/doc_financials/2017/Q3/3Q17-MA-Earnings-Release.pdf

This was interesting a little after the 6 minute mark: This was part of the keynote reveal of FaceID. It's safer than TouchID, 1 in a million rather than 1 in 50,000. I've been reading lots of reviews of the X, and watching lots of videos, listening to podcasts. So far people seem to like FaceID at least as much as TouchID or better. A conclusion that I've heard a few times is that after a while, it feels like you are using your phone without authenticating anymore (like in the early days, back before people put PINs on their phones), even if in fact you are secure. It reduces friction even more than what existed before -- you just look at your alarm in the morning and it lower volume because it knows it has your attention, you look at your lock-screen notifications and the content of the messages pop up once it IDs you, every thing that worked with TouchID automatically works with FaceID because the APIs were built for biometric ID without being dependent on one method, etc.

skanjete, where are you located? The rules for these things might be different where you are. Here, the TWC-CHTR-BHN merger didn't trigger anything taxable (in Canada, and same in the US afaik).

Testing FaceID with hats, glasses, in the dark, etc:

Good podcast interview with Brad Katsuyama of IEX: http://investorfieldguide.com/katsuyama/ IEX market share keeps going up steadily:

http://www.csisoftware.com/2017/10/constellation-softwares-jonas-operating-group-completes-acquisition-of-mcr-enterprise-solutions-and-its-wholly-owned-subsidiary-mcr-systems-limited/

https://www.imore.com/iphone-x-review https://techcrunch.com/2017/10/31/review-the-iphone-x-goes-to-disneyland/