Amazon won't spin-off Amazon Web Services

  1. WeWrite-down
  2. No deal Brexit is not a hedge fund conspiracy
  3. Europe’s digital infrastructure issue
  4. Let’s give a helping hand to Andrew Yang
  5. Anatomy of a malware scam
  6. ARK Invest’s Tesla model gathers dust
  7. A delirious defence of Uber
  8. WeLiquid: Adam Neumann pockets $700m
  9. Yesterday, in efficient markets
  10. The warm fuzzy feeling of indirectly owning Tencent
  11. The best of Morgan Stanley's Adam Jonas
  12. Apple/Tesla: M&A and heartbreak
  13. Did Beyonce make $300m from Uber's IPO?
  14. Bitcoin is the 10-year Treasury of our time
  15. High resolution music is a solution looking for a problem
  16. Amazon is furious about this negative review
  17. Missing: $500bn of American savings
  18. Blockchain for Brexit: a wonderfully terrible idea
  19. The Bank of Hodlers [sic] (sigh)
  20. Behind the curtain at China Ding Yi Feng
  21. An answer to Mark Cuban's question
  22. Crumbs! It's CRYPTO: the movie!
  23. National Beverage Corp loses its fizz, and its mind
  24. Mensch! Dan McCrum is innocent, ok?
  25. Europe's $1 trillion tax gap
  26. Why online propaganda mobs are an investment red flag
  27. Davos has produced an amazing new guide on precisely how not to think about risk
  28. When the public relations industry does PR for itself
  29. Who wants to be crippled by student debt?
  30. The bitcoin price is wrong
  31. The warm fuzzy feeling of Goldman debt
  32. “Cryptoassets” are crashing again. Is it time to start calling them cryptoliabilities instead?
  33. Puff the tragic cryptowagon smokes out the Mumsnet demographic
  34. Don't write off the public sector
  35. Initiative Q: an elementary pyramid scheme with grandiose ideas [Update]
  36. Moral investments aren't outperforming
  37. No one is killing it in crypto (not even Woz)
  38. Too smooth: the red flag at Patisserie Valerie which was missed
  39. No, the housing crisis will not be solved by building more homes
  40. Sorry Civil, 'crypto-economics' and 'constitutions' won't save journalism
  41. 'Short-termism' isn't a thing, say Fed economists
  42. Coinbase wants to be “too big to fail”, lol
  43. Regulation and innovation don't have to be enemies
  44. Retailers get so lonely around the holidays
  45. Folli Follie: $1bn of fake sales, and what to learn from the debacle
  46. The new green evangelism
  47. Tilray, how low can it go?
  48. The ICO behind the tragic Everest stunt is now “airdropping” tokens from rockets
  49. Beware the Hindenburg Omen?
  50. The broken conversation about financial regulation
  51. The improbably profitable, loss-making Blue Prism
  52. The EM rout is not made in America
  53. Wages and growth and honestly we just give up
  54. Britain's first blockchain-enabled co-working space isn't blockchain-enabled
  55. There is a FIRE that never goes out
  56. The WeWork Garden of Eden
  57. IQE: lumpy 'Apple' sauce at the pricey Cardiff chip shop
  58. There's only so much a central bank can do alone
  59. Eight questions every first-time buyer should ask
  60. MiFID II: not all doom and gloom
  61. Tesla: getting to Q3 profitability
  62. Turkey contagion fears are overblown [Update]
  63. The chance of an inflation shock may be higher than you think
  64. Sorry Tim, the humanity is not being drained out of music
  65. Digital crop circles
  66. What could go wrong here?
  67. Sirius Minerals: money for a hole in the ground
  68. The Bank of England has a strange idea of what QE achieved
  69. One for the ladies...
  70. 'Of course, many ridiculous papers appeared'
  71. Is a change goin' to come?
  72. The capacity's not there yet (and probably never will be)
  73. Musk and Tesla are not inseparable
  74. Libraries, from Carnegie to Bezos
  75. Crypto & government: from anarchy to amity in the USA
  76. 'I'm sorry Dave, I'm afraid I cannot sanction this Series B round'
  77. RBC, through the FANG barrier
  78. Self-help to buy
  79. CFA: Chartered crypto analysts -- updated
  80. The Netflix dilemma -- updated
  81. Fujitsu's new blockchain offering: really cheap or really expensive?
  82. Nothing But the Shirt on Your Back
  83. Universities of Britain: cosying up to crypto is a bad look
  84. How to make a living in the cult of meritocracy
  85. Spotify: Drake-oil salesmen
  86. Oh, the digital humanity
  87. Sports are not markets, predictions ain't investment
  88. Spot the difference, Steinhoff edition
  89. Larry Robbins, a cautionary tale
  90. The node to serfdom
  91. Carney is down with the crypto kids
  92. Samsonite: inventory, excess baggage, and unresolved questions
  93. It might be a long wait for “the equivalent alternative to ICOs”
  94. Don't blame it on the sunshine
  95. In corporate America, brands develop you
  96. One in ten dollars of US housing were anonymous
  97. Should AT&T worry more about its debt?
  98. Who cares if Elon is incinerating capital?
  99. Let’s not try make 'crypto chicks' a thing
  100. Tokens all the way down
  101. Eight-dimensional chess with Elon Musk
  102. A lopsided trade is a good trade, Italian inflation edition
  103. How to buy Italian fire insurance
  104. Atlas bugged
  105. Inflating inflation
  106. Crypto's most devout believers are suffering a crisis of faith
  107. Plus500: past performance is no guide to the future
  108. Noble rot in a shrinking Harbour
  109. In defence of ticket touts
  110. Please don't tell individual investors to buy leveraged loans
  111. RIB Software: the unicorn rainy-day fund
  112. Retail is not dead
  113. Did Soros really give Tesla a “vote of confidence”?
  114. At a crypto conference in New York, it feels like 2017 all over again
  115. Egregious expectations - Intelsat edition
  116. Bitcoin cash is expanding into the void
  117. Stop getting The Flintstones wrong
  118. Bond investors do not care if Argentina is solvent in 100 years
  119. Ubiquiti Networks: of cash and borrowed time
  120. “We're very disappointed in you, Spotify”
  121. 'Sex redistribution' and the means of reproduction
  122. Tesla probably needs to raise capital this year
  123. No entitlement crisis in America
  124. Free cash flow to whom?
  125. Hey crypto bros! Journalism ≠ advertising
  126. Human capital and the jobs guarantee
  127. This is a tech bubble, when's the crash?
  128. The magic of adjustments: ebitla-dee-da
  129. FUD, inglorious FUD
  130. A complex analysis reaches same conclusion as simple one: hedge funds suck
  131. The jobs guarantee and human-capital “nationalisation”
  132. These hedge fund numbers can't be right
  133. The Vomiting Camel has escaped from Bitcoin zoo
  134. Lies, damn lies, and charticles
  135. The world doesn't need more Elon Musks
  136. No, Facebook should not become a nonprofit
  137. Sell all crypto and abandon all blockchain
  138. Immutable ledgers meet European data protection
  139. Amazon is not a bubble
  140. Japan's economic miracle
  141. Have you ever meta crypto joke you didn't like?
  142. Delaware should change its rules to let the light in
  143. Who needs the labels anyway?
  144. Baby Boomers want your family to finance a larger share of their retirement
  145. No, America would not benefit from authoritarian central planning
  146. No one needs to buy Tesla
  147. How to win a debate in the cult of meritocracy
  148. Steinhoff International and the case of Pepkor Global Sourcing
  149. Sorry Jack, Bitcoin will not become the global currency
  150. The “academic’s cryptocurrency” is an elegant waste of time
  151. Cigarettes are the vice America needs
  152. Well that’s one reason to buy yen…
  153. Musicians, don't just blame the labels for your lack of dough
  154. Giving stock away to staff doesn't absolve share buybacks
  155. A penny for Macpherson’s thoughts on the nominal anchor
  156. Monopoly and its discontents
  157. A State of Mind
  158. America is not the least protectionist country in the world
  159. This is nuts, when does Netflix crash?
  160. No Bloomberg, the world's richest people did not lose $114bn...
  161. Someone is wrong on the internet, government employee pensions and passive investing edition
  162. Someone is wrong on the internet, possibly fragile
  163. Someone is wrong on the internet, consumer financial regulation edition
  164. Someone is wrong on the internet: tontine tokens [Update]
  165. Someone is wrong on the internet, road economics edition
  166. Someone is wrong on the internet, wages and the stock market edition

We like Scott Galloway on Alphaville. The colourful professor of marketing at NYU Stern School of Business holds big, acerbic opinions on tech, media and finance. And he isn’t afraid to share them. He also once graced Alphachat, so there’s that too.

Every year Galloway posts his top predictions for the year, which verge from the ridiculous to the curious. They’re also often right. For instance, in 2017, he called Amazon buying Whole Foods.

But reading through his 2019 list, one of his seemingly reasonable predictions felt a bit off to us: Galloway thinks Amazon will spin-off Amazon Web Services (AWS), its cloud computing business.

Galloway’s argument, to be fair, is pretty good. He thinks anti-trust against big tech is coming, and chief executive Jeff Bezos wants to get ahead of the Department of Justice. By spinning off Amazon’s extremely profitable cloud services company, not only will Bezos appease the powers-that-be, but also end up creating one of the ten most valuable businesses in the world. Particularly as there are so few ways to invest in a pure cloud company following IBM’s $34bn acquisition of Red Hat last October.

It’s an elegant argument, granted. And Galloway spins it with such enthusiasm that its hard not to feel like he’s right. But it has one big problem. It marginalises the financial symbiosis between Amazon’s retail businesses and cloud arm that has propelled it, and the stock, to giddy heights since the financial crisis. So it’s going to be hard to let go.

When it was a pure play online retailer, Amazon had the same problem most retail businesses do: profit margins are slim. From its IPO in 1997 to the launch of Amazon Web Services in 2006, its operating margins only once crossed six per cent, according to data from S&P Capital IQ.

Small margins limited Amazon's ability to expand its research and development, and labour force, faster than revenues, no matter how fast they were growing. A drop into the red would not have been welcome on Wall Street, even if absorbing these costs may have meant faster growth in the future.

But Amazon did have one secret weapon: its cash flow. Thanks to its leverage over its suppliers, and its efficency in collecting payments from customers, Amazon generates a lot of cash. From 1997 to 2007, it chalked up $3.6bn of operating cash flow, versus just $877m in operating profits., according to S&P Capital IQ.

This cash had to be reinvested, or passed back to shareholders. Amazon, still in growth mode, opted for the former, ploughing dollars back into capital expenditure to help build its distribution network and website functionality. Crucially, only the depreciation and amortisation from these investments had to pass through the profit and loss statement, protecting margins.

However this mix of low profits, and high cash flow, didn’t deal with Amazon's evident existential threat: the cyclicality of its business. Its exposure to the consumer promised pain in the next downturn due to the high fixed costs of sourcing and delivering goods. It needed to diversify.

AWS, which launched in 2006, was the product Bezos had been waiting for. The business, which provides the web plumbing for a range of digital businesses such as Netflix and Monzo was not only less-detached from the vagaries of consumer spending, but also an accounting shadow to Amazon’s core retail business. In other words, it was high margin, but likely (we must stress) cash flow negative.

Let us expand on this point. AWS was first broken out in Amazon’s financial statements in the first quarter of 2015, and according to the filings, its operating margins have averaged 23 per cent since 2013. Amazon’s North American and International businesses, which covers its various retail offerings, including Prime, and hardware sales, managed just 1.5 per cent operating margins over the same six year period.

Assuming these margins held fairly steady before AWS's big financial reveal, this extra profit wiggle room has helped Amazon to increase its research and development costs at a much faster rate than before. For example, from 2001 to 2009, Amazon’s R&D spend grew at an annual average of 19 per cent, according to S&P Capital IQ. From 2010 to last year, the figure was 42 per cent. It is not unfair to speculate that without AWS, bets on products such as the Fire phone, Amazon Echo, Alexa and Prime Video would not have been as large, or even possible.

Yet, AWS also requires hard cash to fuel its expansion. Data centers, with their large fixed costs for cooling, hardware and construction, don't come cheap. Amazon’s latest 10-K suggests as much, disclosing:

[Capital expenditures] primarily reflect additional capacity to support our fulfillment operations and additional investments in support of continued business growth in technology infrastructure (the majority of which is to support AWS)

While we don't know the capital expenditure mix between retail infrastructure and AWS, it is not outrageous to suggest that Amazon's $13.4bn capex spend in 2018 was driven by AWS. Indeed, the acceleration in capital expenditure since 2010 is remarkable. In the nine financial years from 2001 to 2009, Amazon's capex grew at an annual average of 25 per cent. In the past nine years, this growth more than doubled to 57 per cent, according to S&P Capital IQ.

However, thanks to the cash generative nature of its retail business (and a heady dose of stock compensation), Amazon has been able to support AWS's dollar-hungry growth without excessive support from the capital markets.

In turn, AWS has expanded the headroom available for research and development spending, fuelling Amazon's exploration of hardware, voice and other novel retail concepts. As in, helping to set the business up for future growth.

Galloway himself has somewhat acknowledged this delicate dance, arguing AWS would be more valuable without it subsiding Amazon.

But he has also cited Amazon's low cost of capital as one of its main competitive advantages. Without this unique source of dollars, it is fair to say AWS might not be quite as competitive, and therefore enticing to investors, as Galloway suggests. Companies which rely on the kindness of the capital markets often command a lower valuation also (although in this cycle, not so much). To boot, Amazon's retail businesses may also not appear quite so strong if it's left saddled with the accelerated R&D costs subsidised by AWS's profits.

Of course, Alphaville may be wrong about this. Galloway has form, and Bezos has shown he's not afraid to play politics to his advantage.

The question is though: what's worse for Amazon - a potentially long period of anti-trust scrutiny, buffered by a crack team of lawyers and lobbyists? Or wrenching one of its most important assets out of its commercial ecosystem?

For Bezos, we bet it's the latter.

Related Links:

"Don Draper has been drawn and quartered"
- Alphachat
Amazon is not a bubble
- FT Alphaville
Free cash flow to whom?
- FT Alphaville
Amazon learns a little about itself and America, leaves New York
- FT Alphaville
Pivot podcast (w/ Scott Galloway and Kara Swisher) -- Recode Media

  1. WeWrite-down
  2. No deal Brexit is not a hedge fund conspiracy
  3. Europe’s digital infrastructure issue
  4. Let’s give a helping hand to Andrew Yang
  5. Anatomy of a malware scam
  6. ARK Invest’s Tesla model gathers dust
  7. A delirious defence of Uber
  8. WeLiquid: Adam Neumann pockets $700m
  9. Yesterday, in efficient markets
  10. The warm fuzzy feeling of indirectly owning Tencent
  11. The best of Morgan Stanley's Adam Jonas
  12. Apple/Tesla: M&A and heartbreak
  13. Did Beyonce make $300m from Uber's IPO?
  14. Bitcoin is the 10-year Treasury of our time
  15. High resolution music is a solution looking for a problem
  16. Amazon is furious about this negative review
  17. Missing: $500bn of American savings
  18. Blockchain for Brexit: a wonderfully terrible idea
  19. The Bank of Hodlers [sic] (sigh)
  20. Behind the curtain at China Ding Yi Feng
  21. An answer to Mark Cuban's question
  22. Crumbs! It's CRYPTO: the movie!
  23. National Beverage Corp loses its fizz, and its mind
  24. Mensch! Dan McCrum is innocent, ok?
  25. Europe's $1 trillion tax gap
  26. Why online propaganda mobs are an investment red flag
  27. Davos has produced an amazing new guide on precisely how not to think about risk
  28. When the public relations industry does PR for itself
  29. Who wants to be crippled by student debt?
  30. The bitcoin price is wrong
  31. The warm fuzzy feeling of Goldman debt
  32. “Cryptoassets” are crashing again. Is it time to start calling them cryptoliabilities instead?
  33. Puff the tragic cryptowagon smokes out the Mumsnet demographic
  34. Don't write off the public sector
  35. Initiative Q: an elementary pyramid scheme with grandiose ideas [Update]
  36. Moral investments aren't outperforming
  37. No one is killing it in crypto (not even Woz)
  38. Too smooth: the red flag at Patisserie Valerie which was missed
  39. No, the housing crisis will not be solved by building more homes
  40. Sorry Civil, 'crypto-economics' and 'constitutions' won't save journalism
  41. 'Short-termism' isn't a thing, say Fed economists
  42. Coinbase wants to be “too big to fail”, lol
  43. Regulation and innovation don't have to be enemies
  44. Retailers get so lonely around the holidays
  45. Folli Follie: $1bn of fake sales, and what to learn from the debacle
  46. The new green evangelism
  47. Tilray, how low can it go?
  48. The ICO behind the tragic Everest stunt is now “airdropping” tokens from rockets
  49. Beware the Hindenburg Omen?
  50. The broken conversation about financial regulation
  51. The improbably profitable, loss-making Blue Prism
  52. The EM rout is not made in America
  53. Wages and growth and honestly we just give up
  54. Britain's first blockchain-enabled co-working space isn't blockchain-enabled
  55. There is a FIRE that never goes out
  56. The WeWork Garden of Eden
  57. IQE: lumpy 'Apple' sauce at the pricey Cardiff chip shop
  58. There's only so much a central bank can do alone
  59. Eight questions every first-time buyer should ask
  60. MiFID II: not all doom and gloom
  61. Tesla: getting to Q3 profitability
  62. Turkey contagion fears are overblown [Update]
  63. The chance of an inflation shock may be higher than you think
  64. Sorry Tim, the humanity is not being drained out of music
  65. Digital crop circles
  66. What could go wrong here?
  67. Sirius Minerals: money for a hole in the ground
  68. The Bank of England has a strange idea of what QE achieved
  69. One for the ladies...
  70. 'Of course, many ridiculous papers appeared'
  71. Is a change goin' to come?
  72. The capacity's not there yet (and probably never will be)
  73. Musk and Tesla are not inseparable
  74. Libraries, from Carnegie to Bezos
  75. Crypto & government: from anarchy to amity in the USA
  76. 'I'm sorry Dave, I'm afraid I cannot sanction this Series B round'
  77. RBC, through the FANG barrier
  78. Self-help to buy
  79. CFA: Chartered crypto analysts -- updated
  80. The Netflix dilemma -- updated
  81. Fujitsu's new blockchain offering: really cheap or really expensive?
  82. Nothing But the Shirt on Your Back
  83. Universities of Britain: cosying up to crypto is a bad look
  84. How to make a living in the cult of meritocracy
  85. Spotify: Drake-oil salesmen
  86. Oh, the digital humanity
  87. Sports are not markets, predictions ain't investment
  88. Spot the difference, Steinhoff edition
  89. Larry Robbins, a cautionary tale
  90. The node to serfdom
  91. Carney is down with the crypto kids
  92. Samsonite: inventory, excess baggage, and unresolved questions
  93. It might be a long wait for “the equivalent alternative to ICOs”
  94. Don't blame it on the sunshine
  95. In corporate America, brands develop you
  96. One in ten dollars of US housing were anonymous
  97. Should AT&T worry more about its debt?
  98. Who cares if Elon is incinerating capital?
  99. Let’s not try make 'crypto chicks' a thing
  100. Tokens all the way down
  101. Eight-dimensional chess with Elon Musk
  102. A lopsided trade is a good trade, Italian inflation edition
  103. How to buy Italian fire insurance
  104. Atlas bugged
  105. Inflating inflation
  106. Crypto's most devout believers are suffering a crisis of faith
  107. Plus500: past performance is no guide to the future
  108. Noble rot in a shrinking Harbour
  109. In defence of ticket touts
  110. Please don't tell individual investors to buy leveraged loans
  111. RIB Software: the unicorn rainy-day fund
  112. Retail is not dead
  113. Did Soros really give Tesla a “vote of confidence”?
  114. At a crypto conference in New York, it feels like 2017 all over again
  115. Egregious expectations - Intelsat edition
  116. Bitcoin cash is expanding into the void
  117. Stop getting The Flintstones wrong
  118. Bond investors do not care if Argentina is solvent in 100 years
  119. Ubiquiti Networks: of cash and borrowed time
  120. “We're very disappointed in you, Spotify”
  121. 'Sex redistribution' and the means of reproduction
  122. Tesla probably needs to raise capital this year
  123. No entitlement crisis in America
  124. Free cash flow to whom?
  125. Hey crypto bros! Journalism ≠ advertising
  126. Human capital and the jobs guarantee
  127. This is a tech bubble, when's the crash?
  128. The magic of adjustments: ebitla-dee-da
  129. FUD, inglorious FUD
  130. A complex analysis reaches same conclusion as simple one: hedge funds suck
  131. The jobs guarantee and human-capital “nationalisation”
  132. These hedge fund numbers can't be right
  133. The Vomiting Camel has escaped from Bitcoin zoo
  134. Lies, damn lies, and charticles
  135. The world doesn't need more Elon Musks
  136. No, Facebook should not become a nonprofit
  137. Sell all crypto and abandon all blockchain
  138. Immutable ledgers meet European data protection
  139. Amazon is not a bubble
  140. Japan's economic miracle
  141. Have you ever meta crypto joke you didn't like?
  142. Delaware should change its rules to let the light in
  143. Who needs the labels anyway?
  144. Baby Boomers want your family to finance a larger share of their retirement
  145. No, America would not benefit from authoritarian central planning
  146. No one needs to buy Tesla
  147. How to win a debate in the cult of meritocracy
  148. Steinhoff International and the case of Pepkor Global Sourcing
  149. Sorry Jack, Bitcoin will not become the global currency
  150. The “academic’s cryptocurrency” is an elegant waste of time
  151. Cigarettes are the vice America needs
  152. Well that’s one reason to buy yen…
  153. Musicians, don't just blame the labels for your lack of dough
  154. Giving stock away to staff doesn't absolve share buybacks
  155. A penny for Macpherson’s thoughts on the nominal anchor
  156. Monopoly and its discontents
  157. A State of Mind
  158. America is not the least protectionist country in the world
  159. This is nuts, when does Netflix crash?
  160. No Bloomberg, the world's richest people did not lose $114bn...
  161. Someone is wrong on the internet, government employee pensions and passive investing edition
  162. Someone is wrong on the internet, possibly fragile
  163. Someone is wrong on the internet, consumer financial regulation edition
  164. Someone is wrong on the internet: tontine tokens [Update]
  165. Someone is wrong on the internet, road economics edition
  166. Someone is wrong on the internet, wages and the stock market edition
Copyright The Financial Times Limited 2019. All rights reserved. You may share using our article tools. Please don't cut articles from FT.com and redistribute by email or post to the web.

Read next:

Read next:

Lookout, there’s a dollar crunch!

FT Alpha Tweets