The first ten years of their operating life were spent being famous for bleeding prodigious amounts of red ink. That's why the tag: Amazon.toast - it was assumed they would bleed to death without the easy money of the public market and a massive valuation.
Here's their net income statement going back to '97:
1997: -$32m; 1998: -$124m; 1999: -$719m; 2000: -$1.4b; 2001: -$567m; 2002: -$149m; 2003: $35m; 2004: $588m; 2005: $359m; 2006: $190m; 2007: $476m; 2008: $645m; 2009: $902m; 2010: $1.15b; 2011: $631m; 2012: -$39m; 2013: $274m
They've never been quite profitable. Only one year, 2004, did they show even a modest level of profitability (in part due to one-time gains). The rest of the positive years the margin was between 1% and 4% typically.
You'll forgive me if I don't trust in the judgment of a business plan built on losing money for two or three decades.
They can be a good company, grow to X hundred billion in sales, struggle to ever make money, and that's all perfectly fine - but it doesn't mean they deserve a hyper valuation. Target trades at a mere $38b valuation, and they only generated $8b in net income the past three years.
People say the enterprise cloud market is a Trillion dollar opportunity world wide in the next decade.
A Trillion dollars up for grabs, and Amazon has a healthy lead in the segment.
When you go to an AWS event the thing that stands out is the number of consulting and service companies that offer integration/application services to enterprises based wholly on AWS. I've been to Google Cloud events and its not even in the same ballpark as AWS. The Amazon lead is huge.
I think Amazon is still a good buy.
But they've only been unprofitable the last hand few of years. Before that they were quite profitable. Amazon's pursuit of businesses with high R&D costs is the reason for their current lack of profitability.