I actually wouldn't be surprised if I'm misunderstanding something and you're right, but can I ask where you got these numbers?
Edit: On second read I realized the source I sent was misleading – the title talks about annual revenue, but hidden in the text there's a note saying that sub-headings actually show quarterly revenue. I'm still not sure about AWS-to-sales ratio, but my apologies for earlier confusion.
Yea no I'm also left scratching my head. I've never heard of a methodology to attribute market cap portions to business units; eager to hear some clarification too.
It is a well-established concept called https://www.investopedia.com/terms/s/sumofpartsvaluation.asp. A company's value is estimated on net present value (npv) of its projected, discounted future cash flows. If a company has a large, stable subsidiary that is growing slowly but is cash flow positive and a very fast growing but money losing subsidiary, it makes more sense to value parent company as if both subsidiaries were separate companies than the average of the two. A famous example was AWS within Amazon before it became noticeable in the bottom and top lines.
Further, it’s often applied backwards. Some financial engineering self proclaimed wizard will claim “x+y+z=a and the stock is trading at b, and a>>b, therefore the conglomerate discount is too large so we can earn risk free profit (aka arbitrage) by breaking this company up and publicly listing x,y, and z as independent units”. See: Xerox, for example.
Typically in mergers, you see accretion due to the removal of duplicate functions on the cost side (you don’t need two HR departments for example), but sometimes you get negative credit if you’re a too big / too opaque / too confusing.
Yes I understand NPV and DCF but not OPs point. The eCommerce arm makes a lot of money so I don’t understand how you could come to the conclusion that its present value is zero.
I see what you're saying. The logic to the OP's point is "If AWS were to be spun out, on its own it should be worth 12x revenue, or $750B. If the Amazon web business were to be spun out on its own, it should be worth 6x revenue or $250B. Since those add up to $1T and the market cap of AMZN is $950B, the market assigns a -50B valuation to the rest of the company." I agree it's flawed logic. The market may just disagree with someone's sum-of-parts. It maybe conglomerate discount. It may be that the market is undervaluing AMZN, too. I don't think that many people would disagree AMZN's ecommerce division is valuable.
Where are you getting this from? I cannot for the life of me find profits by business unit. I have glanced over the 2021 Annual Report but don't have to analyze it. If you have a source I would appreciate it.
This is done all the time when pieces of businesses are bought and sold. It gets a little complex when the businesses are related (like Amazon Advertising and e-commerce, as opposed to Amazon e-commerce and AWS which are more fully independent of each other).
I understand approximating present value by discounting cash flows but not how you would come to the conclusion that the eCommerce business is valued at zero because the sum of AWS and ads make the total of the market cap.
For AWS, 3rd quarter revenues alone were $20B, so $80B annual. I don't know where that other source of yours finds $17B, unless they are talking profit.
Like, on first page I've found in Google (https://www.insiderintelligence.com/insights/amazon-revenue/) it says: AWS – ~17B Ads – ~9B Online stores – ~66B
I actually wouldn't be surprised if I'm misunderstanding something and you're right, but can I ask where you got these numbers?
Edit: On second read I realized the source I sent was misleading – the title talks about annual revenue, but hidden in the text there's a note saying that sub-headings actually show quarterly revenue. I'm still not sure about AWS-to-sales ratio, but my apologies for earlier confusion.