Amazon Web Services: Is Something Wicked This Way Coming?

| About:, Inc. (AMZN)
This article is now exclusive for PRO subscribers.


Amazon shares keep making new all-time highs and did so again today.

The shares are now up by more than 40% since the end of March, which is significant outperformance, especially for a larger name in the tech world.

It is not surprising that investors are concerned about the outlook for one of the company's jewels, Amazon Web Services (AWS).

There are no signs that AWS is seeing growth slowing any more than might be expected because of its size. The competitive/pricing environment appears to have remained consistent.

Despite the strong outlook for AWS, investors may wish to trim positions in Amazon shares, as the valuation no longer reflects a close relationship to any likely derived DCF valuation.

What's the actual turf that this market leader needs to defend?

The title comes from Macbeth and his encounter with some witches. Well, the story is well known, I think, and certainly doesn't need any embellishment from me. Since the time in which Shakespeare's play was first written and performed, the quote has become a standard, and it has most recently been used by Ray Bradbury in a horror/fantasy novel which, since that time, became a film. I have never seen the modern versions; Hamlet works for me as a horror/fantasy read. I'll discuss the meaning of the title in the context of Amazon (NASDAQ:AMZN) and its shares in a few paragraphs.

I initially wrote about AWS at the end of March, and at that time, called out the growth potential and the profit opportunities that this underestimated piece of the company's empire could mean and was meaning to reported numbers. The shares have been on a tear since that point - although, to be fair, they were not doing so badly before that point. But anyway, since the published date of the initial recommendation, the shares have appreciated by just short of 40% - far greater than the tech index that I track which is up by 10% over that span, and far better as well than the performance of the SPDR S&P Retail ETF (NYSEARCA:XRT), which is down 5% over the same period.

There might be some observers, myself included, who would suggest that it is Amazon itself, which, while not in the retail index, has done the most to impede the operating performance and the share price performance of the sector.

In another article I wrote about Amazon, I observed that by now, the company's growth in retail was accounting for a noticeable and increasing proportion of the growth in US retail sales. It is hard for competitors to perform well if their life blood is being sucked up by a giant vacuum cleaner whose ambitions seemingly know few bounds.

I took a recent look at the company's retail headlines, and they continue to be staggering against the backdrop of an industry not known for much technological innovation. A service that allows parents to order 500,000 holiday toys via a voice command, the launch of a species of convenience food outlets, hiring 120,000 seasonal workers. One article talked about how Amazon had become the largest apparel retailer in the nation with the largest selection of items and how it was even encroaching on the luxury end of the market, where consumers had been thought to want to touch items before buying.

Not all of Amazon's initiatives are going to work - it never happens that way. Besides its retail initiatives, the company has announced a new music streaming service, and it is building a large wind farm in west Texas. And as the company becomes larger, it is going to face more macro issues that encompass things such as problems with shipping hazardous good, or issues with its Chinese suppliers. Shareholders simply must face the fact that Amazon is a giant company and that the results of giant companies, in whole or in part, are going to be influenced by the macro environment over which their influence is not pervasive.

And as mentioned earlier, the company's shares are up just short of 40% in little more than 6 months. Share price appreciation at that rate can obviously be worrisome, as it cannot continue indefinitely no matter how good the company's business is or how strongly it is growing. Almost all analysts have seemingly embraced Amazon and its future, and have concluded that Q3 earnings will be a beat. I have no reason to dispute that assertion, and I get it - but if everyone is already on board with that, the short-term upside is going to be limited when the earnings are actually released. The set-up just isn't helpful.

Because of its strong performance, I have had the occasion to trim my holdings in the shares just to keep it from becoming too large a fraction of my portfolio. (Sadly, my portfolio has not appreciated 40% since the beginning of April.) I have no idea as to what the company may be reporting at the end of this month, and prudently managing risk more or less insists that without a specific informational advantage, portfolios ought to remain balanced. (I suppose that statement can be set to match the platitudes that were the hallmark of Polonius in Macbeth.)

Price wars can be wicked. Will AWS face a price war?

In an excellent article that appeared on this site a few days ago, contributor Joseph Mwangi wrote about a series of price cuts that Microsoft (NASDAQ:MSFT) Azure had announced in some segments of its product line. I might, however, have turned the title around and asked if Microsoft can afford to have a price war with Amazon. And I would have concluded it cannot, and logic says it will not choose to go there.

Mr. Mwangi surmised that the price cuts were in response to AWS announcing that it was to open another of its regions in France a few weeks ago. I do not think the two items are particularly related. The entire business ethos of AWS is built on growing more rapidly than the market, and opening regions is one of the things it does. The company has announced plans to open multiple regions. A couple of months earlier, AWS opened its Mumbai, India, region. Other new regions coming soon include London, Montreal, Ohio and Ningxia, a large city in China. It is also the source of NingXia Red Wolfberry Juice, which... can be bought on Amazon. Can't say I have ever tried it.

In addition to the new regions, the company will open 13 new "availability zones" on top of the 35 it has been operating. Amazon's architecture is built on having multiple zones and regions which allow users to experience optimal availability, fault tolerance and scalability when compared to the performance possible from a single data center.

There are, to be sure, commentators on some of my earlier articles who have suggested that there are other ways to achieve the same results without building multiple data centers, but it is considered by most industry observers to be the appropriate strategy to follow in order to grow the availability of the network for an ever-increasing user population. I don't know, to be sure, but I doubt that Microsoft lowered some of its prices to respond to Amazon's planned region in Paris - I do think it is just part of the business as a whole. But the question of a price war in the space is not one to be dismissed casually, even if the opening of any particular data center is unlikely to be a precipitating factor.

The case against price wars

AWS has reduced prices 51 times since records have been kept. It is what it does. The last across-the-board reduction was earlier this year, but the company has been cutting prices for some offerings on an almost continuous basis. Amazon's pricing behavior is pretty much what one might expect to see in a highly competitive market with substantial price elasticity. As capacity increases, marginal costs decline, and the company passes along much of those savings to its customers, who respond by buying more services, which in turn leads to another round of marginal cost reductions. Many people describe this as a virtuous cycle.

It seems likely that Amazon has the lowest marginal costs of the major competitors in this space. Obviously, the company does not advertise its marginal cost for storage or availability or all of the other products that it offers. I might wonder if it even makes the calculation to determine the marginal cost of each of its services. At the moment, it is offering 8 compute services, 7 content delivery options, 5 data bases, 4 networking alternatives and 7 flavors of analytics. Amazon does offer tiered pricing, in which the price per unit declines quite significantly the greater the usage. And it offers what is described as reserved capacity. Most of the competitors in the space offer the same services priced in the same way - that is one of the advantages of being a market leader. Users can calculate their charges upfront, which means competitors can do so as well. Under the circumstances, with more or less perfect competitive information, falling marginal costs and very high pricing elasticity, the conditions that might create a price war in this space are not present - at least if competitors take rational steps to maximize their profitability. No one has the ability to make monopoly profits, and prices have declined so that at the margin, cost and price are in sync.

Again, I do not have the figures to hand, and I doubt there is some repository where they might live, but it seems fair to say that because of its scale, Amazon has the ability to offer more services that offer the lowest marginal costs to customers. There is something of another virtuous cycle in offering web services, which has been commented upon by many observers. The more a provider invests, the lower his costs and the more features and functions can be offered. Again, there have been commentators on this site who have suggested that it isn't that simple as just counting up the number of features - and I understand that objection - but so far, the data supports the contention that more features equals more user acceptance equals higher revenues equals more investment in more features. Here is a quote from one commentator: "On the whole, considering the number of service available, AWS in is in a league of its own, well ahead of GCE (Google Compute Engine)." This paper was published on a well-known industry web site. Her article also cites other industry professionals who have published on the subject, such as Jillian Mirandi, who has written that, "AWS has a more complete, enterprise-grade portfolio." (as compared to Google (NASDAQ:GOOG))

Amazon's larger competitors, such as Microsoft and Google are well aware that from a cost standpoint, they are unlikely to match AWS in marginal costs even though they both consume large amounts of their own service for other offerings such as search or Office 365. Again, there are some who say that the comparison is spurious, and to be sure, a case can be built about Amazon's costs. But using published data, the preponderance of the articles suggest that size is the major factor in determining cost and particularly marginal costs in offering web services.

Amazon's largest competitors also are aware of the same analysis that the author has made in quite some detail, at least at the moment, "Amazon boasts a wide array of services, many of which - such as the Simple Email Service and the CloudFront content delivery network - are not available in GCE." I am sure there are readers and commentators who will disagree with these citations and the others I could have chosen on this point. I actually do use Google cloud myself for some minor tasks, but it wasn't the result of any exhaustive competitive evaluation, which I would not be qualified to perform in any event.

Another observer has written something of importance in terms of the comparison. "Where AWS clearly has an edge over Google is in the marketplace of partners that run services on top of AWS's infrastructure. In the AWS marketplace users can spin up hundreds of types of services, like enterprise applications from SAP, or performance optimization tools from Riverbed and others." This source felt that Amazon and Google were more nearly matched in terms of price and performance, but ultimately gave the edge to Amazon based on the fact it has the most data centers.

Under the circumstances, launching a price war would be foolish, almost to the point of being suicidal on the part of Amazon's smaller competitors. All of the major competitors in the space are large companies, and their cloud operations are but tiny pieces of their current revenue. That is also true of Amazon, although it does get a very substantial percentage of its operating profit from AWS. At the moment, there are no signs that pricing is becoming more of an issue than it has been since the company first pioneered this space a decade ago, except...

The fly in the ointment

No fly, no story. I could finish this article and say that AWS is not currently seeing anything unusual in the way of price competition, and there is no sign of that changing anytime in the near future. But there is the issue of the black swan out there, or whatever you want to call Oracle (NYSE:ORCL) and the declaration by its chairman and CTO Larry Ellison that his company would be bringing to the market a service offering that had several times the performance at a remarkable price discount in the market. Oracle's offering is to be called Gen2, and one had best take claims of this kind with a grain of salt, especially given where they are coming from and the track record of the source regarding previous technology breakthroughs that turned out to be something else. Needless to say, Gen2 is not yet available, so measuring its claims is not really possible. That hasn't prevented some industry analysts from trying and doing a pretty fair job in assessing what it is that Oracle can actually offer.

An analyst, Marcel van den Berg of a consultancy called UP2V, did an analysis of what is known about Gen2 after the company's presentation at Oracle OpenWorld. Here is a quote taken from Gartner, an organization known for saying harsh words about companies that pay it millions of dollars a year: "I would characterize this early offering as minimum viable product; it is the foundation of a future competitive offering rather than a competitive offering today." Mr. van den Berg made reference to what Gartner analyst Lydia Leong (one of the co-authors of the "Gartner Magic Quadrant for IaaS" report) had to say on the matter, "For the moment, she said, Amazon has nothing to worry about but that challengers in the space may see serious competition from Oracle."

This is Oracle's second and probably last shot at getting IaaS right. It is a key part of the company's future, as without it, Amazon and Microsoft are going to keep eating its on-prem data base revenues - which, at one point or the other, will show up in declining maintenance revenues. Today, Oracle offers what the analysts call a best-effort cloud, which means the cloud platform is not designed to offer redundancy to the application. I won't go through the list of the deficiencies the authors of the different articles cite - for the purpose of this article, we only care about whether or not Oracle might be causing a price war that will hurt Amazon. One of the comments in Michael van den Berg's article is that "Oracle claiming the lead of Amazon is over is like a mouse attacking an elephant." I haven't seen Oracle heretofore described as a mouse. He goes on to suggest, "So Oracle might compete (with) AWS on niche solution but on general purpose IaaS Oracle will never beat AWS. (emphasis added)

At least in the short term, it would not appear that investors need concern themselves with Oracle initiating a price war in this space, as the company simply lacks competitive functionality which would be necessary for users to be able to buy from it.

In looking at AWS - and just AWS - I think the future cadence of growth will continue and that operating margins will continue to increase. Like most observers, I doubt it will continue to enjoy mid-50% growth for an indefinite period; it already has a very high market share, and it is probably resource-constrained in terms of sales and on-boarding capacity. Depending on definition, Amazon is said to have between 27% and 31% market share, with Microsoft having either half or one-third that amount, followed by IBM Corp. (NYSE:IBM) and Google. Gartner puts AWS almost in the very top right-hand corner of its August 2016 MQ survey - you can't get outside the box, but it looks as though it is trying to push Amazon to the equivalent of an asymptote.

Amazon has just signed a fairly significant deal to develop a closer relationship with VMware (NYSE:VMW) - a deal said to have been badly wanted by Google. (At some point, the opportunities that both companies might derive from their newly announced partnership should be better explored. The announcement came today while I was writing this article, and as it reinforces the conclusions, I didn't want to go back and rework the text. Suffice to say that working with VMW will enable AWS to better compete with MSFT for hybrid cloud opportunities, the main chink in the armor seen thus far for AWS.)

Microsoft has some competitive strengths compared to AWS, particularly in its hybrid cloud capabilities and in the ability it has to sell its own SaaS offerings as part of a complete cloud package. It neither has the incentive or requirement to launch a price war to remain very successful in the IaaS space. As of this writing, while Google has perhaps some incentive to use price as a lever, it simply doesn't have the critical mass to successfully launch a price war in the space in the hopes of turning market share upside down. I think AWS is likely to continue its dominance of the IaaS space for some years to come, and that it will continue to remain a strong component of the earnings growth that AMZN needs to sustain its valuation.

Disclosure: I am/we are long AMZN.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.