Description

Quick Pitch: Amazon stock is attractive because it trades at ~30x my estimate of normalized, trailing earnings with a long growth runway. The business has not traded at this level at any point in the last five years. I believe the de-rating has occurred because Amazon is engaged in another investment cycle that is depressing the free cash flow generation of the business. I have seen the market misunderstand this historically and my Amazon investment history has coincided with similar investment cycles.

Company Background

Amazon is a business we all know on some level given its role in our everyday lives. The breadth of products and services offered by the company continues to rapidly expand and deserve a quick refresher on how the pieces fit together. The unifying theme that has enabled Amazon to reach its massive scale today is the company’s willingness to re-invest in new products or services that enhance what Amazon can offer customers. The 2020 Annual Shareholder letter highlights this way of thinking as Amazon aims to optimize decisions to ensure long term value creation for all parts of its ecosystem (customers, employees, 3P sellers), while earning attractive returns on capital.

Amazon has unlocked a huge amount of value through a few key decisions. Early on, the company rapidly expanded selection on Amazon.com by opening to third party sellers. By turning Amazon.com into a website where the company not only sold its own inventory, but also acted as a marketplace where other sellers could list products (like eBay), Amazon became the leading ecommerce platform in the US and Europe. The company offered a wider assortment of goods to customers, as well as a superior customer experience by investing in a logistics network that has reset customer expectations around the speed of delivery for e-commerce.

Amazon took its next major step when it rolled out Amazon Prime, an annual subscription that provided free shipping to customers on every order, as well as other ancillary benefits over time including Prime Video. This decision was pivotal as it massively increased customer loyalty and led the company to rapidly innovate its logistics network to eventually enable two-day delivery on millions of products while still generating a profit. Today, Amazon has more than 200 million Prime subscribers who spent ~$25b (in 2020) on Prime related subscriptions.

Amazon introduction of robust seller tools, particularly its Fulfillment by Amazon (FBA) offering, to third party (3P) sellers was the next layer of innovation. These services allowed anyone in the world to sell on Amazon and leverage Amazon’s infrastructure (think distribution centers and last mile logistics) to grow their business and improve their service offering to end customers. FBA allowed sellers to ship product directly to Amazon’s warehouses where they could store inventory and take advantage of the fact that consumers increasingly preferred to purchase products eligible for Amazon’s Prime 2-day delivery guarantee. Amazon has continued to introduce new third party seller services that have led to Amazon’s 3P business to account for a larger portion of GMV than its 1P business. Another benefit of building a marketplace business within Amazon is that it naturally created the opportunity for a large advertising business as 3P sellers look to promote their products within the Amazon ecosystem. The “Other” segment, which is comprised primarily of ad revenue, is the fastest growing part of the business today.

The final piece of Amazon that is worth discussing is Amazon Web Services (AWS). Amazon incubated this online infrastructure platform going back to 2000 when Amazon started to focus on building scalable infrastructure to allow its developers to run its rapidly growing e-commerce business more effectively, as well as think about building e-commerce platforms for other retailers. The team realized that the tools it built were likely to be in demand from start-ups around the globe. The company rolled out its first commercial product in 2006 that allowed customers to buy cloud compute capacity on-demand. AWS has added thousands of new services over the last 15 years and is now the leading provider of cloud infrastructure to enterprises and governments. The AWS portion of Amazon can be looked at in an independent way and our best guess is that this business will eventually be spun out and operated as a standalone company.

The graphic below does a nice job of showing how the company’s core business units fit together.

Figure 1. Amazon Business Map

My Amazon Investment History

I originally bought shares of Amazon in the spring of 2014.  Amazon was and is a tricky business to model because of the limited segment disclosures (especially before AWS was broken out) and because of the large investment cycles that they periodically go through.  As a result, I've always viewed an investment in Amazon as more art than science.  When I originally bought shares in 2014, I used a basic framework to value the business, which had two key, yet simple assumptions – Amazon had two primary businesses at the time, 1P retailing and 3P retailing. While Amazon was unprofitable on a GAAP basis (due to reinvesting through the income statement) you could approximate the underlying margins of these businesses using similar, more mature comps. This started with the idea that digital businesses were inherently more efficient than their analog equivalents.  As a result, I looked at the best in class for each of the two parts of Amazon’s business.  For the 1P business, I used Walmart’s EBIT margin (5.5%) and for the 3P business I used eBay’s EBIT margin (20%). This led to a rough estimate of the underlying profitability of the (then) two pillars of Amazon’s business despite management aggressively (and appropriately) reinvesting in future growth.

Amazon Business Update

Amazon’s original 1P business has grown substantially over the past five years, with revenue increasing from $91.4b to $238.9b.  A 2.6x increase over five years is nothing to sneeze at, but most of the incremental value creation has occurred from three other components.

First, Amazon has increasingly copied the Costco playbook, whereby traditional retail goods are sold at lower margins and the primary profits are generated through membership fees.  Amazon’s twist on this model has been to use its scale advantages to instead provide a variety of different services to not only consumers but also to its third-party sellers.  As a result, the profitability of the core retail business is increasingly driven by these 3P services, which generated just over $100b of commission revenue to Amazon over the past twelve months. These 3P services carry a much higher margin than the 1P business and enable the third leg of the retail stool, advertising.

Advertising is the biggest new area of value for Amazon.  Like the AWS playbook for the first 10 years of its existence, Amazon has deliberately buried the ads business in a line item called “Other.” This means making some educated guesses about the ultimate profitability of the business.  However, we do know from the footnotes that the ads business is the majority of the Other line item and that Other as a whole grew by almost 70% over the past year.  This has resulted in a business that generated ~$31b in revenue over the past twelve months and should continue to grow at a high rate for the foreseeable future. Understanding the steady state margins for this business is important for understanding future value.

The two ad platforms operating at Amazon scale are Facebook and Google.  Facebook has generated EBIT margins as low as 38% but typically well into the 40s in modern history.   During the past three years Google’s core business has generated low 30s EBIT margins; however, EBIT margins have been as high as 40% in the modern era.  The interesting part in thinking about Google as a comp for Amazon’s ad business is that Google must pay to acquire traffic, in the form of Traffic Acquisition Costs (TAC), and Amazon does not incur these costs for delivering ads on its own website. Google spends ~35-40% of its COGS on TAC.  In thinking about what kind of margin Amazon could generate on its ad business, a comparison to both Google and Facebook suggests that Amazon should be able to achieve a minimum of 30% + EBIT margins over time.

Lastly, AWS has emerged as a behemoth.  Although it has been discussed at Amazon for the past 15 years, it has become the largest cloud computing company in the world.  This has resulted in Amazon breaking out AWS into its own segment.  Some quick numbers – AWS generated over $57b in sales over the past twelve months with an average EBIT margin just under 30%.  It did this while growing at an average rate of 34%.  I expect AWS to continue to take share in the overall cloud market and operate near current levels of profitability.

Today’s price is undemanding.  Using the margin assumptions described above and estimated growth rates for each part of the business (see Exhibit A), I believe the stock will generate a ~20% annualized return. This return assumes no multiple expansion from today’s valuation, which arguably is conservative given how depressed the current multiple is relative to history.

Investment Risks

§  3P margins are driven by advertising growth durability: Amazon’s advertising platform has experienced incredible growth over the past four years and almost certainly carries the highest margin of the business lines within the 3P segment. The advertising business generated ~10x the revenue over the TTM vs four years earlier. At some point, Amazon will need to balance the conflict between pushing monetization and optimizing the customer experience. Google and Facebook growth suggest that this can be managed effectively if the overall ecosystem grows at a rate that allows ad load to only modestly increase because ad inventory growth and pricing increases can drive the majority of the growth. Below is a chart that shows the trajectory of Amazon’s ad growth vs Facebook starting from the point when both crossed $5b of annual ad sales respectively.

§  Capital Allocation: Amazon has re-invested every dollar since its IPO in 1997 to grow the business. Jeff Bezos is widely viewed as the greatest capital allocator of the last three decades, but he has stepped away from the business day-to-day. A bet on Amazon today assumes that the company will continue to excel at directing resources to projects that generate attractive returns. Andy Jassy clearly has a proven track record, having led AWS since its creation in 2003, but he is new to the CEO role, so his ascension is not without risk.  I will closely monitor Amazon’s decisions but given the long duration nature of the bets it will be difficult to quickly detect mistakes.

§  Regulatory hurdles: Amazon (as well as its “big tech” peers) over the last two decades has benefitted from being able to purchase innovative businesses before they become a real nuisance from a competitive standpoint. This will be less effective going forward because regulators have become less willing to allow consolidation among tech businesses. Amazon is best positioned among its peers to deal with this because of the capital-intensive nature of its two key business lines (e-commerce and Cloud Computing). Both verticals require huge amounts of capital and scale to effectively compete.

Exhibit A – Valuation

Catalyst

Amazon is intentionally increasing their capex spend, which will lead to an intentional reduction in operating leverage.  The primary reason for their increased capital investment is that demand outstripped supply which led to fullfillment center utilitization approaching 100% last year.  If Amazon had reached the end state, that would be a great outcome but given their overall levels of growth, that is clearly not the goal.  Andy Jassy talked about this in the Q3 call and as a result, I view this as clear evidence that they're entering into the latest investment cycle.  I expect the stock to be repriced once this increased capex/capacity is absorbed and Amazon returns to achieving higher levels of operating leverage.  We've seen this time and time again from Amazon, the idea of converting variable costs into fixed costs, and in my opinion, this is the latest example of it.  See below for Jassy's exact comments on the Q3 call.