On Thursday afternoon, Amazon.com (AMZN) reported Q3 earnings. Amazon announced a loss of 23 cents per share (excluding a one-time charge related to Amazon's investment in LivingSocial) on revenue of $13.81 billion. This missed the analyst consensus for an adjusted loss of 8 cents per share on revenue of $13.9 billion. However, it fell within Amazon's typically large guidance window for revenue between $12.9 billion and $14.3 billion, and the operating loss of $28 million modestly beat guidance for an operating loss between $350 million and $50 million.
In addition to falling short of analysts' relatively bullish revenue and earnings estimates, Amazon also forecast Q4 revenue and earnings well below consensus. Amazon expects revenue between $20.25 billion and $22.75 billion and operating income between a loss of $490 million and a gain of $310 million. The high end of the revenue forecast was below the average Q4 analyst estimate of $22.8 billion. Even at the high end of this wide range, Amazon's forecast implies EPS of about 42 cents, below consensus. While the operating income forecast seems very conservative, a positive surprise on the scale of Q3 would only boost EPS by a few pennies (leaving it below the current consensus estimate). The relatively weak forecast was in line with my expectations, while Q3 was a little weaker than I expected due to foreign exchange losses.
Typically when a company misses on revenue and earnings, and projects the next quarter's revenue and earnings below consensus, analysts express concern. However, Amazon continues to catch a break from the analyst community. Marketwatch assembled a series of comments from analysts at RBC, Lazard Capital Markets, and Needham & Co. Each managed to find a silver lining in the results. One analyst pointed to strong growth in high-margin segments, while the other two suggested that the Q4 forecast is probably conservative and sets up Amazon for an earnings beat in January. Meanwhile, in a research note this morning, analysts at Credit Suisse were even more bullish. While revenue missed their forecast by nearly half a billion dollars, they described these figures as "close" to their estimate. Moreover, while the Credit Suisse analysts lowered revenue and margin targets for the next couple of years, they maintained that this will have no impact on long term free cash flow generation and therefore kept a $301 price target.
These analyst opinions are all supported by some level of data, so they are not entirely implausible. However, in my opinion, they are all overly credulous. There is plenty of data to support a bearish thesis as well, and the bear case is much more straightforward at this point. First of all, as my fellow Seeking Alpha contributor Paulo Santos has pointed out on a number of occasions, Amazon's EPS has been declining for two years. I am not quite as bearish as Mr. Santos (who argues that this decline is structural and permanent). However, even if Amazon manages to reverse its recent margin decline in future years, the nature of Amazon's business implies that operating margins will always be lean (5% is probably as good as it gets). At 5% operating margin and projected 2012 revenue of approximately $62 billion, Amazon would only show EPS of roughly $4. Amazon trades at more than 50X earnings in this rosy hypothetical scenario. Of course, Amazon bulls would respond that revenue is growing rapidly, justifying the high multiple. This points to the second element of the bear case: slowing revenue growth.
As I have written before, I am very skeptical about Amazon's ability to maintain its revenue growth rate. While growth was negatively impacted by a couple of percentage points by the stronger dollar and strong comps, worldwide revenue growth dropped from 44% in Q311 to 27% in Q312. Revenue growth also dropped sequentially and in every geography. Furthermore, Amazon has continued to be very cagey about the impact of sales taxes on its business. On the earnings call, CFO Tom Szkutak answered a question about sales taxes as follows: "The only thing I could point to is, we collect in over -- either sales tax or equivalent value-added tax. We collect in over 50% of our revenue today. We have very good businesses in those states and geographies that we do that in long ago [sic]." Stating that Amazon sells a lot of merchandise in sales tax jurisdictions is a far cry from stating that collecting sales tax has no impact on revenue. If the change was as innocuous as many analysts seem to think, Amazon could (for instance) provide data on sales growth in New York vs. the rest of the U.S. since imposing sales tax there. The company has declined to do so.
This is a major issue because Amazon began charging sales tax in Texas, Pennsylvania, and California during Q3. I have previously estimated that Amazon derives roughly 15% of sales from these three jurisdictions. As Mr. Santos noted in his article (referenced earlier), Best Buy (BBY) saw relative outperformance in Texas after Amazon began collecting tax there in July. Sales tax avoidance has been a competitive advantage for Amazon in most of the U.S., and the headwind from losing this advantage on 15% of sales could depress company-wide growth by 300 basis points. Moreover, many Amazon customers in California apparently went on "buying binges" before the imposition of sales taxes on Sept. 15, which likely pulled some Q4 sales into Q3. This headwind could depress growth by an additional 50-100 bps in Q4.
Furthermore, I believe the closing of the sales tax gap in additional states was a major reason that Best Buy and Target (TGT) decided to match online prices (from Amazon and others) this holiday season. Some analysts have made the counterintuitive argument that matching Amazon's prices could drive more customers to buy from Amazon, because people would have to look at the item on Amazon in order to get the price match. Some customers may do that, but the "theory" of showrooming is that customers want to go to the store to look at the product or try it out. Once they are there, why would they order online rather than taking the product home right away for the same price? The only plausible reasons I can think of are: 1) long checkout lines, and 2) waiting for home delivery might be preferable for extremely bulky items. In most cases, these downsides will be outweighed by the benefit of instant gratification.
Turning to the longer-term, Amazon will have difficulty maintaining a high growth rate as comparables become progressively more difficult. It's one thing to grow revenues by 30% from a base of $48 billion, and another thing to grow at that rate from a base of $100 or $150 billion. Credit Suisse's $301 price target assumes revenue growth above 20% through the end of the forecast period in 2017 (from 2016 to 2017, revenues are projected to grow from $154 billion to over $185 billion). The historical revenue trend of Wal-Mart (WMT), a tremendously successful retailer, suggests that revenue growth will decelerate much more quickly. From 1991-1992, Wal-Mart posted 35% growth to $44 billion in sales. The company posted better than 20% sales growth in each of the next three years, nearly doubling revenue in the process. But after reaching $82 billion in sales in 1995 (roughly in line with analyst estimates for Amazon's revenue next year), Wal-Mart's growth rate permanently dropped below 20%, with the exception of the year 2000. For the past twelve years, the revenue CAGR has been in the single digits.
In my opinion, the drop-off in Wal-Mart's growth rate has had little or nothing to do with management mistakes. It becomes very difficult to generate strong revenue growth past a certain level of sales because of market saturation. It should be noted that much of Wal-Mart's growth over the past 15 years or so has been in the international segment. Amazon has much broader international exposure today than Wal-Mart did at the comparable period in its development, leaving less upside here. I therefore see the recent decline in revenue growth as the beginning of long-term deceleration rather than a blip. Amazon will continue to show double digit growth over the next five years to be sure. But I expect the growth rate to drop from close to 30% this year to around 20% in 2014, and to 15% or below by 2017.
Thus, analysts are missing the broader story in picking out individual positive data points to support their bullish theses. While analysts are correct to note that Amazon's growth rate is still strong, they seem to overlook the fact that this growth rate is slowing dramatically. Many are overly eager to dismiss management's Q4 guidance as being ultra-conservative, even though Q3 revenue came in $500 million below the high end of the initial guidance. I believe that analysts' supportive comments played a major role in Amazon's relief rally on Friday; the stock ended the day up nearly 7%. However, based on the fairly weak results and guidance, I maintain my short/sell recommendation and I am setting a one year price target of $160. I expect the Q4 results in January to undermine the last vestiges of the bull case, as the impact of sales taxes and competitors' price-matching campaigns is revealed.