Liberty

The main goal of these seems to be to learn from experimentation, and to sign people up for Prime memberships (which have all kinds of benefits for Amazon). It's basically a sales office that attract people in with books.

For those of you who listen to Patrick O'Shaughnessy's podcast, this is his father. Presentation notes (not mine, from the youtube page):

https://www.forbes.com/sites/chloesorvino/2017/05/23/james-leprino-exclusive-mozzarella-billionaire-cheese-pizza-hut-dominos-papa-johns/

Another interesting video by the same author, using the idea of inversion: "7 Ways to Maximize Misery"

You're welcome.

To add further to your point, while Liberties did not have GAAP earnings, they were *reasonably* valued on a FCF & EV/EBITDA basis. I don't think you can make the same case with Amazon... What TCI spent on m&a and debt interest, amazon is mostly spending organically. That's different optically, but it's all really growth expenses when you get to the bottom of it... If Amazon spent $1 billion to purchase a single billboard in Montana with the Amazon logo on it, it would likely grow revenue. Is this a "growth expense" that should be taken out to normalize earnings? If you gave me billions of dollars a year to piss away I could grow revenues really fast too. How do you determine that Amazon's "growth expenses" are good investments and not billboards in Montana? They may turn out to be great investments, time will tell. But it is very far from the TCI days where they were investing in monopoly cable systems where the unit economics were clear. There is only one thing I'm certain of when it comes to AMZN, that is that this (and the AMZN thread) is going to be an incredible thread to look through in 10 years. There will be some amazing lessons, one way or the other. Exactly, you have to judge if Amazon is getting good ROIC on its investments. That's what's hard about this situation. I didn't say that cable was the same as cloud computing and 1P/3P e-commerce, just that on one side TCI's FCF was apparent because it was used for M&A while Amazon's FCF isn't because it's mostly reinvested above that line. If Amazon was mostly growing through M&A, it's financial statements would look very different but that might not necessarily be better underlying economics. Not all sources of value have to be converted to FCF right now to be worth something. For example, Moody's or See's untapped pricing power had value even long before they actually use it fully. Edit: btw, one interesting thing that many don't understand about Amazon: The economics of third party sales on its platform are more attractive than 1st party sale. They basically just take a cut on what's sold through their website, like Ebay, and that's inherently higher margin than carrying inventory and operating warehouses, etc. 3P sales have been growing faster than 1P sales, but for those, only the actual amazon cut is recorded in revenues, not the whole dollar amount of the pricetag of what was sold. So accelerating 3P sales are actually a second-derivative headwind on overall revenue growth (ie. for the same amount of merchandise sold, the revenue will be lower for 3P than 1P but profits will be higher) but actually increase the velocity of value creation.

To add further to your point, while Liberties did not have GAAP earnings, they were *reasonably* valued on a FCF & EV/EBITDA basis. I don't think you can make the same case with Amazon... What TCI spent on m&a and debt interest, amazon is mostly spending organically. That's different optically, but it's all really growth expenses when you get to the bottom of it...

Internet History Podcast: 118. The Birth of Amazon’s 3rd Party Platform with John Rossman https://overcast.fm/+Bj7wXCyv8

Early Buffett wasn't as good as later Buffett, he was still learning. So logically we could expect later Buffett with early Buffett's AUM to do better. Greenblatt did 50% CAGR for a decade. I think it's not impossible that Buffett could do the same. Obviously this need to be adjusted for the context of when the comment was made; maybe with much lower interest rates you'd expect lower returns. But still, I think he could do very high numbers and he wasn't just boasting. Didn't he generate hundreds of millions in his PA while barely trying at all and not touching things BRK could want?

Well, not "always" unless your definition of "real" makes it tautological ;) I'm agnostic on growth. Growth sometimes destroys value. But highly profitable growth, especially if it's obfuscated, can certainly lead to interesting situations where value creation if under-appreciated.

MOSTLY MONEY with Preet Banerjee: 51: What is going on with the Canadian housing market? Ben Rabidoux weighs in. https://overcast.fm/+8XSGzLPc

No. But history suggests these are the situations that are prone to bubbles. The marginal buyer can justify any price with slight tweaks to their assumptions. Actually, this one looks pretty simple. Sell if it breaks 200 day MA. Otherwise, ignore fundamentals and hold tight. Agreed. I didn't mean to sound flippant. I just meant you have to look at the industry dynamics and try to understand the business model and make some assumptions about how much value is being created since it's not easy to separate maintenance and growth expenses and to know how many P&Ls are mature and how many are still in investment mode, etc. Not everybody might agree with your assumptions or take the time to dig in, but if it was obvious and everybody agreed, there'd be fewer opportunities out there. At just about any time since its IPO you could've invested in AMZN and done better than the market over a period of years, so so far, it has been undervalued. It reminds me of a piece I read a year or two ago about how for the first few decades of Walmart's existence you could've bought it at multiples that seemed quite outrageous yet done better than the market. We'll see what the future holds. Each should do what they're comfortable with, though. Some people prefer a lot of quantitative certainty, and there are certainly companies that offer that because of their models. Others might just need to know that the man is fat without knowing his exact weight, relying instead on other qualitative aspects to become comfortable. Both styles will lead to mistakes (including of omission), I don't students of one school should dismiss students of the other, especially since the master (Buffett) has been straddling the two for a while.

Josh Tarasoff (2013) comments on AMZN profitability: https://youtu.be/XAubRoZbI9U?t=23m15s There are also other comments about it before in the video, but I don't have a time stamp.

Josh Tarasoff (2013) comments on AMZN profitability: https://youtu.be/XAubRoZbI9U?t=23m15s

This is certainly true. But there is nothing that tells us much about the underlying economics and valuation of Amazon. Amazon's value is some function of revenue, margins, and ROIC at "steady state". None of those inputs can be estimated within +- 50% since they are likely 10 years in the future. Case A: Future Amazon is a low cost leader, like Walmart, with 2% net margins. If so, maybe it will trade for 15x earnings at "steady state". Case B: A capital light tech company, like Google, with 20% net margins. If so, maybe it trades at 25x earnings. Let's assume $1T in revenue. Case A: $300B market cap Case B: $5,000B market cap Both of those scenarios are plausible. But a simple change in assumptions increases valuation 16.7x. Yeah, nobody said investing was easy...

There's more than one way to reinvest into growth. He could do it by slashing his prices/margins. This way, he would be reinvesting the potential profits that he would otherwise be making into growth. He can also reinvest the profits into the business in other ways. For example, he could keep his profit margins normal, but take all that money and add more employees, more equipment, maybe rent a new locations, etc. He's still keeping the money inside of the business rather than taking it out. Both approaches look different from an accounting point of view, but the general idea is the same; use money that could go in your pocket and instead "spend" it on accelerating the creation of value. f.ex. When Amazon uses money from its more mature and profitable US retail business and uses it to pay for the startup of its operations in India, that's what it's doing. When it's using profits from AWS to invest in Prime Video content to attract more subscribers who then spend more on retail and are stickier sources of revenues, that's what it's doing. When it invests in CAPEX way over what it needs strictly for maintenance (more fulfillment centers, more data centers, more robots, entering more new markets, etc), that's what it's doing.

Almost everyone is short-term oriented. This only works for long-term oriented people, and for public companies, you have to spend a long time educating your shareholder base about what you're doing and they have to trust that you really mean it and can pull it off, otherwise they won't give you the latitude (also helps if you are a founder-majority owner)...

Correct. It's easy to beat your competitors on price and grow if you don't need to make a profit. For now Amazon is not a business. It's a non-profit. You're looking at it wrong. For sure I am. After all a company is not supposed to make money. A company is supposed to create value. Reinvesting all your earnings internally is one way to create value. TCI didn't have any earnings either. Update: A bit more here http://www.cornerofberkshireandfairfax.ca/forum/investment-ideas/amzn-amazon-com-inc/msg300226/#msg300226

Correct. It's easy to beat your competitors on price and grow if you don't need to make a profit. For now Amazon is not a business. It's a non-profit. You're looking at it wrong.

Exactly. Hard to draw useful conclusions from a faulty premise...

It's not gloom, just skepticism that this is quite what people thought it was even 5 years ago, and that outperformance is as likely as some think. This is a tough industry. Expectations for Sandstorm Metals & Energy were really high too. I wish shareholders that everything starts working and that your patience is finally rewarded. But even if it doubled overnight, it would still be lagging the sp500 for long term holders. Newer shareholders would be the ones getting a better deal...

That sounds great, but I think if you think about it some more, you can't make decisions like that rationally because you don't know how the business and management will evolve in the coming years. I suppose it could be an aspirational goal, but personally I'd keep an eye on things and revisit periodically and sell if I discover I'm wrong. I used to own Altius, and I used to look at the commodity sector and listen to a bunch of Rick Rule interviews. He's a very articulate guy, and he has great things to say about Altius (even though he never says the name). But ALS has been underperforming something like a SP500 ETF for over tens years, and has way underperformed even something pretty vanilla like MKL. At some point you have to decide if your capital is actually compounding faster than the market or not, and once you're behind by that much, what kind of returns would you need going forward to do decently, and is that realistic... Still thinking out loud here... ALS has a bunch of cheap call options on potential mines and royalties. But they had a bunch of cheap call options 10 years ago too, and I can't say that where they are today is exactly a huge success compared to that. 1% dividend yield, still debt to pay down for a while longer, then maybe the dividend will increase a little and finally help move the stock a bit.. But what CAGR are you getting after a decade? I think one of the biggest misconceptions about the ALS model back from when I was following it was the whole "royalties are great because you don't have to spend the capital to build the mine and take the risk yourself". That model takes you up to a point, but *someone* still has to spend all that capital and take all that operational risk, and if they can't make it, you don't get your royalty, or it's delayed potentially by years and years, as we've seen with Alderon (which many expected to be becoming a working mine by now). If your partners go bust because they're a mismanaged junior mining POS or if the commodity cycle whacks them (China slowdown, US slowdown, whatever), you still take a hit to your royalty. You also have political risk, always, mines have to be approved, are regulated (burden could go up, making some mines uneconomic) etc. Coal could be phased out faster in Alberta at the stroke of a pen (just like the pendulum in the US could swing hard the other way in a few years). All this to say that the model doesn't sound nearly as low-risk, predictable, and simple as it once did to me. But I could be wrong.

https://www.bloomberg.com/news/articles/2017-05-26/apple-said-to-plan-dedicated-chip-to-power-ai-on-devices

I don't want to re-litigate amazon, it's been done to death, but I don't think P/E tells us much about the underlying economics and valuation vs IV. It's a red herring. If they wanted to optimize for earnings, they could limit re-investment to what is required for, say, 10% growth (which might not be that much because of the secular e-commerce tailwind), they could serve most of that with their existing infrastructure for a long time, and margins and earnings would balloon. But they'd be creating IV slower than by reinvesting everything they make in new markets and product lines (the way to look at Amazon is as a collection of P&L, some mature and profitable generating cash to help pay for other P&Ls that are in startup mode and needing capital -- in aggregate they try to keep it balanced close to zero). And that's not just for retail. What if AWS tried to optimize for earnings? They could slow their capex a bit, slow their price cuts a bit, etc. They'd grow slower but generate a lot more earnings... So in the short term, they'd look better, pile up lots of cash on the balance sheet, etc, but over time they'd create less IV, because they are one of those businesses that happen to have big reinvestment opportunities and if they don't rush to grab them, eventually someone else will. They've been investing a ton in India, for example, and that could pay off big as that P&L matures, but for now it's depressing company-level earnings.

Or about $12k/share if you account for the stock splits around 1999...