We think Amazon AMZN is a compelling investment opportunity as the market continues to underestimate the secular shift toward e-commerce, the long-term opportunity for public cloud, and the growing importance of advertising within Amazon’s online presence. The company’s wide economic moat is driven by powerful network effects, cost advantages, high customer switching costs, and intangible assets, which we do not see being threatened. Amazon has been an obvious beneficiary of the pandemic-related lockdowns, and we believe that e-commerce trends have been pulled forward a couple of years. As the only true demand aggregator in the United States, this has strengthened the company in the near term, in our opinion, and leaves it in a better competitive position for the next five years.
We think online grocery shopping is here to stay, with the pandemic offering a proof of concept for many consumers who might not have been inclined to try it. While investors might have questioned the build-out of a captive transportation network two years ago, it proved to be a critical decision for the company and consumers and was the only real alternative for third-party sellers during the pandemic. We believe e-commerce will continue to take share from traditional retailers and forecast that Amazon will expand its e-commerce market share from 42% today to beyond 50% in 2025.
We foresee Amazon Web Services, advertising, and subscriptions growing faster than the company’s e-commerce business over the next decade. Further, each of these segments drives higher margins than the corporate average, which should allow both operating profit and earnings per share to outgrow revenue as margins continue to expand. Amazon is the clear leader in public cloud through AWS, which we believe is still in the early stages of mass adoption in a market that should ultimately be near $1 trillion. Its advertising business is already large and continues to scale, offering an attractive option for marketers looking to access a vast audience with a variety of proprietary data points about those consumers.
Wide Moat With a Stable Trend
We assign a wide moat rating to Amazon based on network effects, cost advantages, intangible assets, and switching costs. Given the breadth of the company’s offerings, moat sources are numerous and benefit all segments. We think the moat trend is stable.
Amazon has been disrupting the traditional retail industry for more than two decades and has emerged as the leading infrastructure-as-a-service and platform-as-a-service provider via AWS. This disruption has been embraced by consumers and has driven change across the industry as traditional retailers have invested heavily in technology in order to keep pace. COVID-19 accelerated this change, and given Amazon’s technological prowess, massive scale, and relationship with consumers, we think it has widened the company’s lead, which we believe will result in economic returns well in excess of cost of capital for years to come.
We believe Amazon’s retail business has a wide moat stemming from several areas:
- Network effects associated with its marketplace, where more buyers and sellers continually attract more buyers and sellers.
- Cost advantages tied to purchasing power, logistics, vertical integration (proprietary brands, owned delivery, and so on), and a negative cash conversion cycle.
- Intangible assets associated with technology and branding (this also applies to the advertising business).
We believe AWS is also a wide-moat business, thanks to the following:
- High customer switching costs.
- Cost advantages associated with economies of scale, where few competitors can keep up with the pace of Amazon's investments.
- Intangible assets arising from semiconductor and facility development.
- A network effect associated with a marketplace for software created to make AWS work better for users.
We would assign Amazon’s burgeoning advertising business a narrow moat based on intangible assets from its proprietary data on hundreds of millions of users and a network effect again focusing on buyers and sellers meeting in the largest available venues. We believe that the wide moat for Amazon’s entire business is greater than the sum of its parts, as the segments reinforce one another, and returns result in an unrivaled consumer experience.
Online Stores Are Amazon’s Bread and Butter
Amazon is the e-commerce leader in the U.S. with 42% share, representing approximately $374 billion in gross merchandise value. We see its e-commerce market share expanding by 200 basis points annually over the next several years, culminating with a share in excess of 50% by 2025.
According to the U.S. Census Bureau, the U.S. retail market was approximately $5.6 trillion in 2020 ($4.4 trillion excluding autos and auto-parts dealers), while the e-commerce market in the U.S. was about $800 billion. We do not believe that 100% of retail sales will move online in the long term (even in the really long term), but we do think there is room to comfortably double e-commerce within the broader retail mix over time. We also think that the pandemic has shifted consumer behavior more toward online and pulled the e-commerce demand curve forward by perhaps two years. The obvious conclusion on this alone is that Amazon’s online stores still have significant room to run.
Online stores consist of physical and digital merchandise sold directly by Amazon. Digital items sold on a transactional basis are included here, whereas such items sold on a recurring basis are included in subscription services. At 51% of total revenue in 2020, online stores represent the largest component of Amazon’s sales. The company benefited from a surge in COVID-19-related spending, which resulted in anomalous growth of 40% in 2020. We model a 12% compound annual growth rate for segment revenue over the next five years and double-digit revenue growth in 2025. We view this as a long-tail business that will continue to outgrow e-commerce more broadly while significantly outgrowing the broader retail industry over the next decade.
Third-Party Services a Natural Extension of Online Stores
Amazon also has a marketplace where third parties can sell items, which significantly expands the overall selection available to consumers on Amazon.com. With $80.4 billion in 2020 revenue, or 21% of the total, third-party seller services is Amazon’s second-largest segment by revenue and grew 50% last year. This segment experienced the same pandemic-related spike in growth as the online stores segment. We model a 20% CAGR for segment revenue over the next five years, as Amazon has demonstrated its power as the dominant demand aggregator in the U.S. We view this as a long-tail business that will continue to significantly outgrow the broader retail industry over the next decade.
Strategically, this is an extension of the company’s substantial footprint of distribution centers, which helps leverage its fixed costs with higher utilization and throughput. In third-party seller transactions, Amazon is not the seller of record. Rather than revenue being generated from the sales of an item, Amazon collects a commission on the sale, along with any related fulfillment and shipping fees.
Overall, we estimate that Amazon generates a blended commission of approximately 12% on third-party merchandise sold on its properties. We assume that Amazon makes a similar 12% on various Fulfillment by Amazon services, including storage, administrative fees, and shipping. Our total blended take rate is approximately 25%, which includes both the commission and the FBA services.
Prime Is the Secret Sauce in Subscription Services
Amazon’s subscription services include annual and monthly fees from Amazon Prime memberships along with digital video, audiobook, digital music, e-book, and other subscription services. Subscriptions are the company’s fourth-largest segment, with $25.2 billion in 2020 revenue, or 6.5% of total revenue, and grew 31% in 2020. We do not believe the pandemic had as positive an impact as it did on the company’s merchandise-driven business. We model an 18% CAGR for subscription revenue over the next five years. Like the company’s broader e-commerce businesses, we view this as a long-tail business that will continue to outgrow the company’s overall growth rate over the next decade.
Prime is both the glue that ties Amazon’s broader retail operations together as well as the grease that makes the whole business work. Subscriptions provide a recurring stream of high-margin revenue for Amazon to use to support expansion of its distribution and fulfillment capabilities.
Prime memberships generate high cash flow that can be reinvested to further improve the user experience on the technology, content, and delivery fronts. Prime customers are very sticky and tend to purchase from Amazon more frequently and across more retail categories. We think content combined with Prime subscriptions builds switching costs that consumers would need to overcome, although they might not last for the decades needed to deem them a moat source.
Prime memberships are not the only revenue stream with subscriptions. In fact, based on our estimates, other subscriptions accounted for approximately 45% of segment revenue in 2020. We believe this revenue consists predominantly of Amazon Music Unlimited and Amazon Kindle subscriptions. We see Amazon’s device strategy with the Kindle and Echo as a way to increase subscription revenue over time from both existing and new Prime members. We believe that Prime will continue to decrease in the mix of subscription revenue over the next five years.
Physical Stores a Work in Progress
Amazon’s physical stores include sales from traditional brick-and-mortar stores, including Whole Foods, Amazon Books, Amazon 4-Star, Amazon Fresh, Amazon Go, Amazon Go Grocery, and Amazon Pop Up. Physical stores are the company’s smallest segment at $16.2 billion in 2020 revenue, or 4.2% of total revenue, and contracted 6% in 2020. The pandemic had a negative impact on physical stores, as year-over-year comparable sales declined and even today some locations remain closed as a result of local conditions. We expect this to improve throughout 2021 as the U.S. in particular moves to open up as vaccine distribution proliferates. We model a 4% CAGR for physical store revenue over the next five years.
As of December 2020, Amazon operated 611 physical stores, the vast majority of which are Whole Foods Markets. Amazon closed its acquisition of the grocer in August 2017 and has continued to invest in the company. The most obvious changes have been lower prices at the stores, benefits (further discounting) extended to Prime subscribers, grocery delivery in as fast as one hour, more centralized procurement, and tighter inventory management at the store level. Amazon has also built out new Whole Foods stores, which we expect to continue at a measured pace as the company expands from 519 stores at the end of 2020 to 586 by 2025.
We would hardly call Amazon’s impact on Whole Foods revolutionary, but we do think the pandemic has led to a permanent shift in the way consumers approach grocery shopping. As lockdowns proliferated beginning in April 2020, consumers turned to e-commerce for many needs, including groceries. While other grocery delivery services like Peapod were around for years, the lockdowns served as a proof of concept for grocery delivery for many consumers that we think will grow from here.
In the meantime, Amazon continues to experiment with other grocery options as well as other physical store formats. For now, we think physical stores serve a triple mandate for Amazon in that they (1) reinforce loyalty and improve Prime member experience, (2) allow the company a mechanism to hone its physical store strategy, and (3) compete more directly with retailers and grocery stores such as Walmart.
We tend to think of Amazon’s physical stores not as a separate business, but rather another way to reinforce loyalty and offer benefits to Prime members. We also think they open up additional avenues through which to improve customer experience and better compete with retailers that are finding ways to defend their positioning against Amazon.
AWS Powers Growth
Amazon Web Services was a pioneer in public cloud infrastructure as a service and platform as a service and retains a substantial lead over its closest rival. AWS has driven profitability for the entire company; although it represented 11.8% of total revenue in 2020, it generated 59% of total operating profit dollars for Amazon. We model a 22% CAGR for AWS revenue and expect the segment to remain a key growth driver for the company over the longer term.
Enterprise customers rely on AWS for core IT infrastructure, which represents significant switching costs in terms of the time and expense involved in integrating applications with core software elements, such as a database, and dedicates a user to a specific set of software development tools. Ultimately, the operational risks of changing mission-critical technology infrastructure are high, which is why core elements such as enterprise resource planning systems and cloud providers are rarely changed.
We believe it is cheaper initially for companies to move workloads to the cloud, as there are fewer up-front costs and a lower bar to clear for maintenance and administration. Additionally, Amazon has devoted significant research and development resources to adding advanced features to the platform. AWS offers scale advantages to clients in that it is cheaper and faster to set up IT infrastructure in the cloud compared with undertaking the same effort independently. Customers also benefit from the ability to scale up computing power for burst requirements, paying for only what they need and having it available effectively on demand. Some of the largest technology companies in the world have simply failed to keep pace with Amazon’s massive investments over the years, and AWS’ cost advantage over its rivals is obvious.
Amazon has also amassed significant technology and process knowledge. The company expanded its distribution network by roughly 50% in 2020 while managing through a global pandemic. Given the size of its footprint, this is a monumental achievement and speaks to the company’s ability to quickly plan, construct, and expand facilities based on specific needs. The knowledge base to quickly and efficiently bring massive server farms on line for AWS is similarly impressive and only comes from the experience of previously building hyperscale data centers. Additionally, the company designs its own semiconductors that are used to power its server arrays and has also developed proprietary robotic automation technology used in its fulfillment centers.
Azure is widely perceived as the leader in public cloud--a view we share. We see Microsoft Azure as the only other legitimate competitor, with Google Compute Engine a distant third. We expect this market to eventually surpass $1 trillion as enterprise workloads increasingly move to the cloud. With AWS generating $45 billion in revenue in 2020 and Microsoft generating approximately $25 billion in revenue for the 12 months ended in December 2020, we believe the public adoption remains in its early stages.
Advertising Business Is a Hidden Gem
Amazon’s so-called other segment is tangentially related to its retail operations in that it is mostly advertising, which primarily occurs on Amazon’s own online properties, but also includes revenue from cobranded credit cards. Amazon can offer unique consumer insights, real-time shopping data, and an enormous audience.
We estimate that Amazon’s advertising business was at least $15 billion in 2020 and would be the third-largest Internet advertising model in the U.S. if it were a stand-alone company. Overall, the other segment generated $21.5 billion in revenue in 2020, which was up 52% over 2019 and represented just 5.6% of total revenue. The company has not provided disclosures, so we estimate that approximately 75%-80% of the other segment is advertising, which is growing rapidly and is likely the business with the highest operating margins in Amazon’s portfolio, in the range of 35%-40%.
We believe advertising dollars flow to where the eyeballs are and where information is known about the online user, which fits in very well with Amazon’s strengths. We therefore expect advertising to grow rapidly over the next several years and continue to boost the company’s overall profitability. For other segment revenue, we model a five-year CAGR of 25%.
Attractive Growth, Improving Margins, Solid Balance Sheet
Amazon grew 37.6% in 2020 to $386 billion in total revenue. We model a 16% CAGR for revenue over the next five years and see operating margins expanding by approximately 50 basis points annually to about 9%. We think that the secular shift to e-commerce will continue, including for new categories such as groceries, and that Amazon will benefit in an outsize manner. Relative to Visible Alpha consensus, our revenue forecast is approximately 2% higher in 2025, with upside in retail in general, AWS, and advertising.
We think Amazon will enjoy scale-driven margin expansion over time even as we acknowledge that history has shown such expansion will not always be smooth. More rapid growth in AWS and advertising, which carry higher margins, should also help increase overall margins.
Amazon maintains a solid balance sheet, with year-end 2020 cash of $84.4 billion and $31.8 billion in debt, which represents 0.7 times gross leverage on 2020 EBITDA. Outside of financial debt, $52.6 billion in long-term operating lease liabilities resides on the balance sheet as well. The long-term debt has maturities of $1.2 billion in 2021, $1.6 billion in 2022, and $2.3 billion in 2023. Because we project more than $26 billion in free cash flow in 2021 and material growth afterward, we do not foresee debt maturities as problematic.
Amazon carries $23.8 billion in inventory, representing 37 days of sales compared with a range of 45-52 days over the last 10 years. E-commerce spiked in 2020 as a result of the pandemic, leading to lower inventory, which we expect to trend back toward the company’s normal range over the next couple of years.
Periods of High Investment Can Drive Uneven Results
We believe that the uncertainty for Amazon is high, and that despite being an e-commerce leader, the company faces a variety of risks.
Amazon must protect its leading online retail position, which can be challenging as consumer preferences change, especially after the pandemic, as shoppers may revert to prior behaviors and traditional retailers bolster their online presence. Maintaining an e-commerce edge has pushed the company to invest in nontraditional areas, such as producing content for its Prime Video subscriptions and building out its own transportation network. The company must also maintain an attractive value proposition for its third-party sellers. Some of these investment areas have raised investor questions in the past, and we expect management to continue to invest according to its strategy, despite periodic margin pressure from increased spending.
The company must continue to invest in new offerings. AWS, transportation, and physical stores are three notable areas of investment. These decisions require capital allocation and management focus and may play out over a period of years rather than quarters.
Continued international expansion is likely to require similar investment and management attention but will also increase exposure to different regulatory environments. Some countries have instituted or may institute protectionist policies. Even in the U.S. over the last several years, lawmakers from both parties have increasingly focused on the amount of market power that large technology companies have accrued. Antitrust, data privacy, and Section 230 (liability for online content) issues have been repeatedly raised.
From an environmental, social, and corporate governance perspective, data breaches and service outages are a concern for any cloud service provider. As an online retailer, Amazon has personal information on hundreds of millions of consumers around the world, while AWS hosts proprietary mission-critical data for enterprises. Additionally, as a tech company, Amazon must attract and retain software developers, although this has not been a problem thus far.