Amazon – The Good, The Bad And The Ugly – Seeking Alpha
Amazon – Has anyone been getting into their wallet
A bit more than 10 years ago, I went to India for the first time and took the drive from Delhi to Agra to see the Taj Mahal. In those years, the trip was incredibly tiresome, although a fascinating representation of life in India. The distance on the road that existed at that point was 220 kilometers and it was called a highway – but it was one that was shared with elephants and bullock carts and sometimes troops of monkeys. The monkeys were bold – while the nominal toll on the road was minimal, the monkeys knew they could get their cut by begging for bananas. (If you are thinking of going to the Taj any time in the future – fear not – a few years ago, the Yamuna Expressway opened – it is shorter and almost empty and the trip takes two hours or less – and no monkeys, elephants or bullock carts). Of course, as soon as I saw the Taj, the misery of the trip took second place, but when I returned to New Delhi and asked how it was possible to drive consistently in India, I was told that one needed great brakes, an excellent horn and lots of luck.
Trying to cover Amazon’s (NASDAQ:AMZN) stock is a bit like that. This is a company that exists on many planes of development with operations spanning the globe, some of which do well and others less well and some do incredibly. There are observers who want to evaluate this company just on the numbers. There are others who want to analyze a piece of the results or to focus on a single product. For those of you who read my articles, it ought to be obvious that I would be happier if all I had to comment on was AWS – but none of us can buy AWS. I do not stream music or video, although I am an enthusiastic user of Prime. Maybe one day AWS will be spun-off, but I really doubt it. Companies rarely sell their crown jewels.
Amazon reported the results of its Q4 and 2016 year Thursday. The results were not an unalloyed pleasure to consider and the shares have pulled back a bit (3.5% by Friday’s close). One contributor on this site called the quarter a disappointment. Another contributor felt before the earnings that it was time to exit the shares and thought his thesis was confirmed by the results. Yet another observer was concerned about a red flag on the balance sheet in terms of Amazon’s payables balances.
So just how well did Amazon do, how concerned should investors get about some of the self-evident weak spots and how should investors look at guidance. Is the story over, is it time to exit the name or is the story just beginning?
There is a famous tale about Ben Franklin and the Constitutional Convention. He remarked at the end of convention that he had looked every day at a chair in which the convention president, George Washington, had sat and remarked “that for weeks… in the course of the session… looked at that sun behind the President without being able to tell whether it was rising or setting. But now at length I have the happiness to know it is a rising and not a setting sun.” As I will discuss in this article, there is not a better metaphor for the quarter than Amazon just released. If there is anything in the IT space that is more a rising sun, I really can’t imagine what it might be.
I do not think anyone sitting in their office or home today can really predict either the cadence or magnitude of Amazon’s profitability in its many different businesses. Whether readers or observers use some of Amazon’s streaming services or try to buy everything they can from Amazon Prime is not really dispositive of that question.
Recently, various industry analysts published data showing Amazon with outsize level of the online business to consumer market. I do think that using the growth in reported non-AWS revenues of a quarter or two as a proxy for the long-term performance of Amazon retail is more than a bit short-sighted. I think determining that Amazon will never achieve significant levels of profitability because it is investing in fulfillment centers at a prodigious rate is backward looking thinking. The number of businesses that need to expand their fulfillment capabilities by 30% in a year and need to do so for several years is more of a good thing than a bad thing. At some point, it is important to take a look at a bit more than the specific published quarterly numbers in order to draw reasonable conclusions. I think that AWS alone is one of the jewels in the IT firmament. Overall, I find it hard not to expect that Amazon will continue to confound skeptics for many years to come.
The anatomy of a disappointment
As the saying goes, the point of view depends on the point of view. But no denying that Amazon’s results were below what most observers, myself included, had thought most probable. Just for the record the headlines included revenues for the period were $44 billion, up 22% as reported and 24% in constant currency. The percentage growth rate declined in Q4 from the full year percentage growth rate of 28% in constant currency.
Published expectations had been that Amazon revenues would be about $1 billion higher. The company saw an impact of about $558 million from currency fluctuations and much of that was not in the prior analyst consensus.
Looked at traditionally, free cash flow reached just less than $10 billion for the year. A significant component of the increase in free cash flow was driven by the increase in deferred revenues. Amazon reports cash flow several ways to take account of capital lease commitments, but most analysts look at the traditional free cash flow metric.
As pointed out elsewhere, the increase in unit shipments came in at 24% compared to a historical trend of 28%. It is one of those statistics that entrances some but there are many sub-issues that make the headline less than useful in forecasting future growth. On the other hand, not quite so broadly commented was the relatively sharp acceleration in the growth of unearned income which grew by 53% this past quarter vs. 37% in the year earlier period, a function of substantially greater bookings than usage for Amazon Web Services (AWS) and a sharp increase in Prime subscriptions.
Stock-based comp, the bete noir of some, continued to increase and was $887 million in the quarter, up 46% year on year and was 8% of CFFO. That said, the dilution from stock-based comp has been moderate, overall with 497 million shares outstanding at the end of 2016 compared to 490 million shares outstanding a year earlier.
The accounts payable contribution to CFFO, which had been flagged as a danger sign by one contributor, increased by almost 16% year on year and increased by 17% for the 12 months’ period ending December 31st as well. The increases in payables were a significant component of last quarter’s CFFO. On the other hand, Amazon has been using its size to force suppliers to finance inventories for many years now without any noticeable problems. See the link shown for a review of the history of the practice. Receivable balances also increased in the quarter which again is a seasonal norm for the retailing business. Gross margins (GAAP) grew from 32% to 33.6%. Given all of the components that make up gross margins, I do not think this is a necessarily salient trend.
Fulfillment costs, a closely watched metric, increased by 30 basis points from 12.7% to 13%. Fulfillment cost increases are a function of the sky-high costs of delivery in India coupled with a very back-end loaded cadence of the opening of fulfillment centers in the US and in Europe. Every time Amazon opens a new fulfillment center, it is planned to run at sub-optimal levels for several quarters. When lots of fulfillment centers open simultaneously as was the case in the 2nd half of 2016, the results are predictable. The other significant operating expenses also showed noticeable increases in expense ratios. Are these signs of future trends or are they an indication that the cadence of the company’s growth is strong and management is investing to accommodate even stronger growth. I think the latter argument has the preponderance of the evidence.
Results in North America showed somewhat more favorable trends than did results in the rest of the world, with operating income rising by 28% year on year, while international saw a growth in operating losses. Revenues in North America were up by 22% compared to growth of 18% internationally. I suspect that increasing losses internationally are attributable to the continued substantial investments the company is making in India particularly and to other less developed geos and thus are likely to continue for some period.
AWS growth declined to 47% (constant currency). Full year 2016 growth was reported at 55%. GAAP operating margins for AWS reached 26% this quarter which compares to 25% for the year. GAAP operating margins were 27% last quarter. That growth seems even more impressive in the wake of noticeable price cuts for AWS which took effect in December. As I wrote about in an earlier note, the demand elasticity from the price cuts is likely to have had minimal impacts on revenue growth in the December quarter, but may have cut operating margins by a couple of hundred basis points.
As mentioned above, AWS saw an increasing level of multi-year commitments in which users buy a given quantity of services on a prospective basis. Amazon does not report AWS bookings and probably will not do so in the foreseeable future, but it seems reasonable to suggest given the strong growth in deferred revenues, particularly on a comparative basis, that growth in proxy bookings showed very favorable trends in the period.
Guidance – What it implies and what it doesn’t
Probably the strongest cavil on the part of some investors had to do with the company’s Q1 guidance which was in a very broad range and did not show the kind of growth many were hoping for. Company guidance for Q1 was forecast to be $33.3-$35.7 billion compared to the prior revenue consensus of $36 billion. Much of the weaker guidance is a function of FX which is estimated to be a headwind of $730 million in the quarter and the balance of the estimate difference relates to the fact that 2017 is not a leap year. Overall, adjusting for those two factors, Amazon guidance would have been 23% at a mid-point, probably in line with most investor expectations.
Amazon is also forecasting that operating income will decline to between $250 million and $900 million in the current quarter, a significant fall from $1.1 billion of operating income the company posted in Q1 2016. Some of the fall relates to the loss of a selling day in Q1; whereas fixed costs continue unabated revenues do not in a given quarter. That alone is most of the difference. Further, the company is continuing to open fulfillment centers at rates far greater than the rates it opened such facilities in Q1 2016.
Amazon announced the opening of its own air cargo facility in northern Kentucky which will be another major expense. While some of the expense will be capex, inevitably hiring thousands of new staff and starting up this kind of an operation will put pressure on cost ratios in the first quarters of development.
In retail, Amazon can and has been vicious in competing when it needs to do so. In India, Amazon’s price competition has simply upended Flipkart and smaller rivals. But in the US, the CFO said on the call that promotional activities have been consistent throughout 2016 and that he saw no change in the level of promotional activities. Guidance, he said, was not cut because of price competition.
It would not be feasible to attempt to look at all of the revenue growth drivers at Amazon in a single article and provide value add for readers. Some investors and observers may want to focus on Amazon streaming video which saw 100% growth in hours utilized last year. Others may want to look at the statistics relating to Fulfillment by Amazon (FBA) which showed a noticeable increase in its ratio within total deliveries. Other analysts are particularly interested in India and China these days. Given the number of moving parts that need to be forecast to provide a specific quarterly estimates, it isn’t terribly surprising that the results usually turn out to be quite a bit different than the forecasts. It would make little sense to try to use quarterly results as a marker indicating longer term trends, and if investors or commentators wish to do so, they have to look behind the numbers in some detail to understand what might really be happening.
There are many ways that investors look at growth proxies as opposed to just looking at the revenue forecast. One of those is the growth in square footage of the company’s fulfillment centers. Last year the company increased its growth in square footage by 30% up from 20% in the prior year. That growth in square footage went to service the 40% growth in Amazon fulfilled units. Company CEO Brian Olsavsky refused to forecast specific square footage growth this year other than to say it would be substantial. There is very little to suggest that Amazon’s retail business is slowing down in terms of its growth and operationally, the company continues to make investments to augment efficiency and profitability.
Overall, I believe that when looked at excluding AWS, and considering the FX headwind and the impact of the lost selling day, the growth the company is projecting is 21% for Q1 at its mid-point. Given the latest share statistics regarding e-commerce, Amazon continues to gain ground compared to the rest of e-commerce sellers. The link above not only shows Amazon’s dominance in some typical e-commerce categories but also the number of adjacent markets that are still targets of opportunity for Amazon. That suggests that the forecast is conservative and designed to be exceeded Being #1 and gaining share is not really a bad place to be.
It is clear that Amazon continues to invest heavily in order to show spectacular growth. There are going to be some Amazon shareholders who would like management to trade off growth for higher short term profitability. Amazon’s operating margin guidance for Q1 reflects a pattern of substantial investment that shareholders have seen before and should expect to see again. It is the behavior expected from a well run company investing for its future.
I am perfectly happy to see Amazon reinvest the profits it is earning from more mature markets to enter new and different business lines. Will automated grocery stores lead to a new growth cycle for the company? Will the company’s entry ion auto parts business have a visible impact on merchandise revenues? Amazon as a company continues to enter new lines of business, and while some of the initiatives will not work, enough have in recent years to sustain growth.
I think the fact that operating margins are forecast to be at levels below prior expectations in Q1 by some amount has much to do with some mythical deterioration in the company’s market position or some different, lower profit paradigm for the company’s retail business. It does suggest that Amazon is making appropriate investments to reduce the cost of fulfillment and to improve user satisfaction. If stakeholders think that building fulfillment centers and opening and air traffic facility are wrongheaded investments, then clearly they are likely to be unhappy with Amazon’s capital allocation decisions for many years to come. If these investments are not considered by investors to be optimal use of the company’s profits, then my suggestion is that investors shouldn’t be looking at large cap growth companies.
Some further thoughts on AWS, still the Jewel in Crown
Last quarter, AWS operating income was 74% of the company’s total operating income and that compares to 52% of operating income in Q4 2015. For the year, AWS contributed 74% of operating income compared to 67% for all of 2015. As mentioned earlier, AWS growth was 47% year over year this quarter in terms of reported revenue compared to growth of 55% last quarter and 69% in the year earlier quarter. This apparently is a cause for alarm to some commentators. Now in the eyes of some readers, a business that has reached a revenue run rate $14 billion/year and has GAAP operating margins of 26% and which is still growing at 47%/year is probably not a huge problem. Is it a sign of overvaluation? What is a reasonable valuation for this asset?
Before proceeding there are a couple of misconceptions that I have seen expressed that need to be addressed. There appears to be some belief that the Web Services space is becoming more competitive which has led to the slowdown in percentage growth at AWS and thence to the price cuts that Amazon implemented in December. The thesis sounds great until one looks at the details.
In the interest of brevity – well, in the interest of not bloviating – I will make this short. Competitors are leaving the web services space. They have and they will continue to do so. Why? The space is too capital intensive for most potential entrants. While capital intensity is declining for Amazon at scale, the opposite is true for most of the other players. Even companies the size of Workday (NYSE:WDAY) and Salesforce (NYSE:CRM) are partnering with Amazon to avoid having to compete aggressively for IaaS dollars.
It is relatively easy to see that Amazon has major cost advantages compared to its largest competitors in the space, Microsoft (NASDAQ:MSFT), Google (NASDAQ:GOOGL) (NASDAQ:GOOG) and IBM (NYSE:IBM). Amazon simply has lower average and marginal costs when compared to competitors both current and potential and that is something that is well recognized.
In December, Amazon implemented a series of price cuts. The cuts were no secret – they had been announced the prior month at the Amazon convention and I had written an article detailing the cuts and their likely impact. Anyone paying attention would have known that the cuts were at hand and would impact the results of the quarter. It is a travesty when analysts use a well known event that should have been incorporated in their models to reach a spurious conclusion.
But a couple of things to note. Amazon always cuts prices. Not once, not twice but multiple times every year. The price cuts in December were not the first of the year by any means and the company will continue to cut prices. As pointed out at the time by SA contributor Alex Cho, the cuts in December were the smallest annual cuts since 2008. It is no doubt, fun to write about price wars and increasing competition. It is also, I imagine, fun to propagate false news and makes authors feel good. The problem is that is just ain’t so. I have to say that the lengths to which some analysts have gone to defend their short thesis knows few ends. I could cite a few offenders – but heck why spoil their party.
So why does Amazon cut prices. It is pretty basic and it is called price elasticity. I will spare readers my own anecdotes on the subject, but the fact is that cutting prices on web services is an attempt to exploit the strongest demand driver there is in the space – price. Users migrate to the cloud not because they can use pretty pictures in their offices of blue skies dotted with puffy white objects, but because the cloud is much cheaper than on-prem. The cheaper it is, the faster the migration. How fast does price elasticity work – probably not in the time scale with which financial publications need to produce stories. Users have responded to price elasticity for years now by consuming an ever-increasing quantity of instances, by storing more data and by analyzing more of the data that they store. They will do so again this year, and next year and the year after that.
If Amazon had absolutely no competitors, it would still be profit maximizing behavior to cut prices. Profit maximizing competitors, in a market with visible price elasticity, will find that to maximize profits, pricing should be at levels where the marginal price equals the marginal cost. Sounds academic, I suppose but the formulas do work. Amazon doesn’t quite compute the formula of MC=MP to determine the exact prices it charges – its pricing models are more sophisticated than that and the last time I saw a marginal cost was in a text book. What it does instead is that it prices to achieve a margin on average cost – which gets to the same place. Look carefully and see how the price changes are different in different availability zones. That is all a function of marginal costs.
So why is growth slowing? Because at a $14 billion/year run rate, the law of large numbers is in evidence. I have looked at AWS for some time now and it has grown at prodigious rates for all the time I’ve been an observer. But it is no longer a business with developers as its primary customers. Just in order to achieve constant dollar growth in revenues requires selling larger deals, primarily to larger organizations. And that is what is happening. But that is an environment in which percentage growth becomes harder to maintain.
Does someone find it either surprising or worrisome that AWS will take longer to move revenues to $30 billion from the current $14 billion run rate, then it took to move revenues from $7 billion up to the just reported level. Will it be surprising to see that it takes AWS longer to move to $60 billion a year of revenues from $30 billion than it will take to move from $14 billion to $30 billion. I certainly hope there isn’t that kind of expectation and I have yet to see a professional analyst who would ever publish such expectations.
AWS was said to have missed expectations by about 3%-$100 million in aggregate for the quarter. Much of that was the FX headwind. More than all of that was the higher level of deferred revenues as contract signings become weighted toward multi-year cross-product agreements – the function of the company selling larger and larger deals to larger enterprises.
About the only way I could imagine AWS growing at its current rates for a year or two longer would be if it chose to enter the SaaS market to complement its IaaS and PaaS offerings. Will that happen? I think the odds are against that, but it is a possibility. There are other ways for Amazon to achieve the same effect through the agreements it has reached with Workday and with Salesforce.
I think it is reasonable to assume that percentage growth rates continue to moderate over the next several years and that is, I imagine, a pretty consistent theme in the expectations of both analysts and investors. Moderate is a subjective word. My thoughts are that percentage growth rates will decline 5-10 percentage points a year on average for several years to come but will still show growth that would make almost any other business green with envy. AWS is likely to be growing its revenues by more than 20%/yr., 5 years from now – by which time its revenue run rate ought to be approaching $50 billion.
There are, to be sure, many ways to value the business results and the cash flow that AWS is going to generate. If it grows more slowly than I think most likely, then the same economic formulas that have led it to reduce prices consistently, will suggest a slower cadence of price cuts and prices per unit will tend to stabilize. In turn, that is likely to lead margins to achieve an upward trajectory. As a shareholder, I don’t want that to happen. I want to see high price elasticity, higher growth and lower operating margins. It is the way to maximize long-term profits.
In addition, if AWS grows more slowly, then of course, capex will decline substantially and free cash flow will rise significantly more rapidly than profitability. So the specifics to use in a DCV calculation aren’t entirely clear to me.
But if AWS achieves a CAGR of 25% for the next 5 years, a conservative estimate, I believe, and if operating margins continue to improve over time and capex falls dramatically, then its terminal revenues will be approaching $45 billion, with free cash flow likely reaching close to $20 billion/year. What the present value of that might be can be determined in many ways but all of them suggest that the AWS component of Amazon, by itself supports a very substantial valuation. (I think it is more than a bit pointless to try to project a reasonable weighted average cost of capital and terminal growth rate for this company. Many other people go through the exercise and that is how most of the published price targets are constructed. The bias in most of these exercises is reversion to the mean, and by their nature, they simply have no way of incorporating newer business initiatives at a company like Amazon that has a ton of them, some of which have yet to be developed. The average mean price target for 12-month performance of $930 was reasonable given the methodology used in those analyses. But those analyses cannot incorporate the success of Amazon newest ventures whose future is unknown at this point.) But let’s turn to holistic valuation to see if the shares are ahead of themselves or if the runway for performance still persists.
OK, we all know that Amazon is a highly-valued stock and some commentators on this site believe that have discovered its over-valuation. I will just present a few facts to start this section. Currently, the published First Call consensus price target for Amazon shares is $929. Amazon shares are well covered with 43 analysts updating First Call with price targets and 46 doing so for rating.
In the wake of Thursday’s earnings release, it is likely that the average price target will be reduced by 2%-3%. There actually have been a few increases in target prices, but I think the course of east resistance is marginally down.
The price target cuts are mainly a product of reductions in earnings estimates and EBITDA. The models animating these estimates are what they are. By that I mean they are very mechanical in construction and do not try to look at qualitative aspects of margin development. Many analysts see the cadence of margin development slowing materially because of the commentary of management on the call and the rests of the quarter. But I really think forecasting that Amazon’s operating margins will 20% less in 2018 than previously forecast is a bit much.
When readers look at analyses that suggest Amazon shares are expensive, that kind of mechanical thinking constitutes much of the though process. At some point, date uncertain, Amazon will see a breakthrough in its fulfillment expense ratio. The ratio of marketing, general and administrative expense and content and technology costs which ballooned in Q4 will either drive incremental revenue growth or they will be managed. This is not an eleemosynary organization, perhaps contrary to the implications of some observers. There are many hundreds of basis points of operating expense that will either be used to drive revenue growth or will be managed down. When that happens, many of the newly issued projections for Amazon EPS and EBITDA will be consigned to the dust heap.
On the other hand, Amazon shares are highly rated by analysts and no one changed their rating in the wake of the company’s Q4 earnings release. There is a material gap between the qualitative assessment of analysts and the quantitative methodology used in putting together price targets. It would be well to consider that in looking at valuation.
But the estimates on First Call are the estimates we have to consider so they have to be used, much as they might mislead and have misled some commentators.
At the end of last year, Amazon had 486 million fully diluted shares outstanding. It’s closing price yesterday was $810/share. That is market capitalization of $394 billion. The company has net cash on the balance sheet of $11.6 billion, although there are substantial lease obligations not included in that calculation. But using standard definitions, Amazon has an enterprise value of $382 billion.
The current analyst consensus carried by First Call is for revenues is $168 billion. That number will almost certainly be adjusted lower when First Call recalibrates all of the new estimates by perhaps 2% or so. But at the current published level, the EV/S is 2.3X.
The issue of course is profitability. Amazon is not expensive on the basis of EV/S, especially with a growth rate of 20% or more. It is expensive on profitability levels given the enormous investments it has chosen to dedicate to ensuring future levels of growth.
The consensus earnings per share forecast for Amazon had been $8.62 for 2017 which itself is down from an estimate of $8.94 3 months ago. The estimate will go lower, to a greater or lesser extent. In any event, the P/E on the currently published First Call consensus is 94X. At this point, looking at that number is a waste of time. No one owns these shares because of their evaluation of the company’s P/E for 2017 or really for 2018 (notionally 55X).
As mentioned earlier in this article, Amazon suggests that free cash flow is a better way to measure its financial success. Indeed, it is the first financial result table that the company publishes in its earnings release. Last year cash flow from operations grew by 38% to $16.4 billion. While stock-based comp is not an insubstantial element in CFFO at 18% of the total, other changes to the cash flow statement were significantly greater. Depreciation grew by 29% and the net of deferred revenue balance grew by 50%, although the net change in deferred revenue balances only contributed $1.9 billion to CFFO.
In Q4, the net increase in the deferred revenue balance grew from just $100 million in 2015 to $700 million, suggesting a hidden strength not well publicized, particularly as it applies to AWS for the most part.
As mentioned earlier, payables grew by $5 billion, while inventories grew by only $1.2 billion, an increase of less than 10% and one reason to suggest that Amazon’s retail operations haven’t cost it in terms of writedown. Receivables grew by $3.4 billion, an increase of more than 50% year on year and something to look at more closely if the trend continues. Overall, the increase in payables was offset by the increase in receivables and inventories.
The company continues to increase capex at significant rates. Last quarter capex increased by 54% and the metric increased by 47% for the full year.
Putting together a free cash flow forecast involves loads of guesses. To ease the task, I am going to throw out trying to figure out some of the items involved with assets and liabilities. They tend to be set-offs against each other. Using current trends for the other major categories, I think CFFO will increase by $6 billion to $22 billion in 2017 and I expect that capex will grow another 40% to $9.3 billion. That yields a free cash flow projection of $12.7 billion and a free cash flow yield of 2.6%.
The issue regarding valuation isn’t whether a 2.6% free cash flow yield is a high valuation. Of course it is. The issue is how fast that number might be able to grow. Free cash flow increased by 27% last year and I am forecasting it will grow by almost 30% this year. At that cadence, the free cash flow yield will support far higher valuations than current for the shares.
I think most shareholders for Amazon have the perspective that free cash flow is going to see step function increases at over the coming years. Given the level of capex and its growth, that seems to be a very reasonable hypothesis. The company is not likely to need to expand its floor space by 30%/year unless growth expectations are seriously less than reality. The capital intensity of AWS will decrease at scale even if it could grow at current rates. At some point, as percentage growth decreases and data center efficiency rises, the growth in capex from that source will moderate close to zero. In the meantime, depreciation is rising at very rapid rates and grew by about the same amount in dollars as did capex in 2016.
I have no idea if Amazon will populate this country or the world with automated groceries or bricks and mortar book stores, or drones or air freight facilities or other capital intensive endeavors. But if that were to happen, it would mean that growth and operating margin progression has been seriously underestimated.
I think those commentators who feel that Amazon shares are overvalued haven’t really looked at the constituents of the business and the innovations that Amazon has pioneered. This is a company with the opportunity to grow free cash flow very rapidly and to support significantly higher valuations for some time to come. The potential for positive alpha, while no doubt less in percentage terms this year than last year, remains substantial.
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.