Thursday, March 31, 2016

The Amazon Tax

Ben Thompson at Stratechery has an insightful post entitled The Amazon Tax on the 10th anniversary of the rollout of Amazon S3:
Until then Amazon Web Services had primarily been about providing developers with a way to tap into the Amazon retail store; S3, though, had nothing at all to do with retail,2 at least not directly.
Below the fold, some comments.

Thompson starts by providing the backstory to the famous but officially unpublished Jeff Bezos memo that I'm fond of referring to via this slide  from Krste Asanović's fascinating keynote at the 2014 FAST conference. The focus was:
transforming raw data center components into storage, computing, databases, etc. which could be used on an ad-hoc basis not only by Amazon’s internal teams but also outside developers:
The result, Amazon Web Services:
  • AWS has massive fixed costs but benefits tremendously from economies of scale
  • The cost to build AWS was justified because the first and best customer is Amazon’s e-commerce business
  • AWS’s focus on “primitives” meant it could be sold as-is to developers beyond Amazon, increasing the returns to scale and, by extension, deepening AWS’ moat
He quotes Chamath Palihapitiya who, when asked what company he would invest in if he could only choose one, responded on Quora:
AWS is a tax on the compute economy. So whether you care about mobile apps, consumer apps, IoT, SaaS etc, more companies than not will be using AWS vs building their own infrastructure. Ecommerce was AMZN’s way to dogfood AWS, and continue to do so so that it was mission grade. If you believe that over time the software industry is a multi, deca-trillion industry, then ask yourself how valuable a company would be who taxes the majority of that industry? 1%, 2%, 5% — it doesn’t matter because the numbers are so huge — the revenues, profits, profit margins etc. I don’t see any cleaner monopoly available to buy in the public markets right now.
Thompson asks:
what if the business model of Amazon’s e-commerce business has changed to “tax” collection?
The combination of increasing returns to scale and the idea of taxing the markets you're in is very powerful. Competitors operate at the scale of the single market they are in, you operate at the scale of all your markets combined.

Another way to look at the same idea is that if you are lucky enough to be the winner in your first market increasing returns to scale gives you a monopoly which allows you to extract rent. This rent can be used to invest in capturing a related market, one in which the economies of scale from the first market apply. Thus in this second market you are very likely to win, because even at the beginning your economies of scale are much greater than the competition's. And in the third market this effect is even stronger, and so on.

This effect must be worrying FedEx and UPS, for example:
It seems increasingly clear that Amazon intends to repeat the model when it comes to logistics: ... start with the fact that Amazon itself would be this logistics network’s first-and-best customer, just as was the case with AWS. This justifies the massive expenditure necessary to build out a logistics network that competes with UPS, Fedex, et al, ... I think it is a mistake to think that Amazon will stop there: just as they have with AWS and e-commerce distribution I expect the company to offer its logistics network to third parties, which will increase the returns to scale, and, by extension, deepen Amazon’s eventual moat.
A similar effect is behind the re-organization of Google into Alphabet, as described by Fast Company's Learning Larry Page's Alphabet, another insightful piece describing how Alphabet is a way to use the monopoly rents from Google's ad business ($73 billion in cash on hand) to move into markets where the scale of Google's technology can be applied. Apple needs to exploit these effects:
In fact, Apple captures a staggering 94% of the smartphone industry’s profits, according to research firm Canaccord Genuity.
W. Brian Arthur's Increasing Returns and Path Dependence in the Economy shows how random factors select one of a group of startups competing to enter a new market to be a winner because increasing returns to scale amplify the random factors. But this description applies to new markets where the competitors have about the same initial resources.

But that's not what is happening in markets targeted by these  monopoly rent extractors. The winner is pre-ordained because it starts out operating at much larger scale than the losers. Thompson writes:
nearly every startup of note to be founded in the last several years has started on AWS or one of its competitors.
Because Amazon's margins on AWS are formidable, Amazon could have dominated its customers in any of their markets that it chose to go after. Of course, it doesn't choose to go after many of them. It is possible to grow big enough on AWS to obtain substantial economies of scale and then leave Amazon, as Dropbox is doing.

The current issue of The Economist has a long piece entitled Too much of a good thing, arguing that the US economy has become insufficiently competitive, allowing companies that dominate their markets to accumulate cash:
Business theory holds that firms can at best enjoy only temporary periods of “competitive advantage” during which they can rake in cash. After that new companies, inspired by these rich pickings, will pile in to compete away those fat margins, bringing prices down and increasing both employment and investment.
That's obviously not what is happening:
An American firm that was very profitable in 2003 (one with post-tax returns on capital of 15-25%, excluding goodwill) had an 83% chance of still being very profitable in 2013; the same was true for firms with returns of over 25%, according to McKinsey, a consulting firm. In the previous decade the odds were about 50%. The obvious conclusion is that the American economy is too cosy for incumbents.
Because the pattern of market dominance by a small number of large firms is increasing:
Revenues in fragmented industries—those in which the biggest four firms together control less than a third of the market—dropped from 72% of the total in 1997 to 58% in 2012. Concentrated industries, in which the top four firms control between a third and two-thirds of the market, have seen their share of revenues rise from 24% to 33%. And just under a tenth of the activity takes place in industries in which the top four firms control two-thirds or more of sales.
And the result is rent extraction, exceptional profits being defined by The Economist as return on capital above the traditional 10%:
For S&P 500 firms these exceptional profits are currently running at about $300 billion a year, equivalent to a third of taxed operating profits, or 1.7% of GDP.
Which is a lot of money sucked from the economy into companies' cash stashes. The market believes that this money drain will continue, partly because a little of the excess cash is used to buy politicians:
Alphabet, Facebook and Amazon are not being valued by investors as if they are high risk, but as if their market shares are sustainable and their network effects and accumulation of data will eventually allow them to reap monopoly-style profits. (Alphabet is now among the biggest lobbyists of any firm, spending $17m last year.)
In the good old days anti-trust enforcement was effective, but that was so last century:
[Anti-trust authorities] cannot consider whether the length and security of patents is excessive in an age when intellectual property is so important. They may not dwell deeply on whether the business model of large technology platforms such as Google has a long-term dependence on the monopoly rents that could come from its vast and irreproducible stash of data. They can only touch upon whether outlandishly large institutional shareholders with positions in almost all firms can implicitly guide them not to compete head on; or on why small firms seem to be struggling.
See also the work of Izabella Kaminska at the FT's Alphaville blog. Johnathan Rothwell at Brookings points out that the same effect operates at the scale of individuals as it does at the scale of corporations.  The Economist suggests that in an ideal world there would be an alternative approach:
It would aim to unleash a burst of competition to shake up the comfortable incumbents of America Inc. It would involve a serious effort to remove the red tape and occupational-licensing schemes that strangle small businesses and deter new entrants. It would examine a loosening of the rules that give too much protection to some intellectual-property rights. It would involve more active, albeit cruder, antitrust actions. It would start a more serious conversation about whether it makes sense to have most of the country’s data in the hands of a few very large firms. It would revisit the entire issue of corporate lobbying, which has become a key mechanism by which incumbent firms protect themselves.
Of course, given the enormous sums incumbents have available to ensure that politicians prevent any such thing even being discussed, in the real world this isn't going to happen. Pushing back against network effects is really hard. Even Steve Randy Waldman's simple proposal to fix the tax code to dissipate Uber's network effects seems utopian.

This all makes Andy Grove's six-year-old How America Can Create Jobs prescient and deserving of the attention its been getting since his death:
Our fundamental economic beliefs, which we have elevated from a conviction based on observation to an unquestioned truism, is that the free market is the best of all economic systems—the freer the better. Our generation has seen the decisive victory of free-market principles over planned economies. So we stick with this belief, largely oblivious to emerging evidence that while free markets beat planned economies, there may be room for a modification that is even better.
For a more recent take on it, see for example Thomas Frank's The Blue State Model.

I should also note that the post Ben Thompson followed The Amazon Tax with, Andy Grove and the iPhone SE, is also a really insightful must-read.


David. said...

Noriel Roubini's Unconventional Monetary Policy on Stilts points out that the same two factors, short-termism and the resistance of economic theory to empirical refutation, apply to management of the economy and this reinforces both incumbent protection and the lack of good jobs:

"Unfortunately, the political economy of most structural reforms – with their front-loaded costs and back-loaded benefits – implies that they occur only slowly. At the same time, fiscal policy has been constrained in some countries by high deficits and debts (which jeopardize market access), and in others (the eurozone, the United Kingdom, and the United States, for example) by a political backlash against further fiscal stimulus, leading to austerity measures that undermine short-term growth."

David. said...

Jon Brodkin reports that according to Jeff Bezos letter to shareholders AWS has 1M users and $10B/yr sales:

"Measuring by total compute capacity in use, AWS is 10 times bigger than 14 other infrastructure-as-a-service providers combined, among those included in a Gartner report from May 2015. Those 14 include Microsoft, Google, CenturyLink, VMware, IBM, and Rackspace. The report did not include Alibaba's AliCloud service, which reported having more than 1.4 million customers as of June 2014."

David. said...

Some of Amazon's margins come from its reluctance to pay suppliers.

David. said...

Nick Wingfield at the New York Times reports on Amazon's second quarter:

"For most of its life, Amazon sacrificed profits if it could build another few warehouses to ship orders to customers more quickly or find some other investment to fuel its growth.

Now, it cannot avoid showing big profits thanks to the lucrative cloud computing business in which it has improbably become a leader.

On Thursday, Amazon reported net income of $857 million in its most recent quarter, the second quarter in a row in which it has shown a record profit. Its net income for those three months was also more than nine times the amount for the same period last year."


"The company reported operating income of $718 million from its Amazon Web Services business, up from $305 million a year ago. That is slightly more than the profit it showed from its North American retail business.

The North American retail business, though, brought in $17.67 billion in revenue compared with $2.89 billion for Amazon Web Services, an indication of the significantly higher profit margins in cloud computing."

David. said...

According to analysis by Canalys reported by Joe Fay at The Register, the $37.8B cloud market is increasingly dominated by the "big 4" vendors, Amazon (30.4%), Microsoft (16%), Google (7.4%) and IBM (6.7%). Amazon and Microsoft had more of the market than all vendors outside the big 4 (39.5%).

David. said...

Gavin Clarke at The Register reports that Amazon's dominance of the cloud continues to grow:

"AWS last week reported a $3.5bn revenue contribution to its parent's fourth quarter to December 31 – analysts had expected $3.6bn. A year ago, AWS made $2.4bn. ... AWS is now Amazon's single biggest growth engine. Net sales increased 47 per cent year-on-year, important in a quarter where commerce sales disappointed. ... Amazon dominated a fourth-quarter global cloud infrastructure service market worth $10.3bn with a 33.8 per cent share. ... Microsoft, widely touted as number-two to AWS with Azure, along with Google and IBM together accounted for 30.8 per cent."