Description

As is widely known, Amazon has struggled over the last few years. They overbuilt during Covid and took a series of missteps. Retail margins collapsed from ~5% to negative territory, returns on invested capital declined, FCF turned negative as cash flows declined and capex remanined high, etc. On top of that, many thought AWS would grow +20% through the cycle and clearly this assumption was wrong, as growth has declined to a halt and margins have been hit. Now management is cutting costs in order to drive efficiencies and return the business to historic profitability.

Rightly so, investors are questioning management´s focus on returns, the real profitability of the business and many other relevant factors. These concerns are real and I don´t think they should be brushed aside, but this has handed us the opportunity to become shareholders in what I believe is a great business at a reasonable price.

For this investment to work, efforts to increase profitability have to start showing up in the financials. If that occurs to the extent I believe possible, shares might double over the next 4 years.

To show how bad sentiment currently is, shares have declined +40% over last year and have returned close to 0 since 2018. EV/Sales multiple has declined from 3.2x to 2.0x. Basically, valuation has collapsed while the business has shifted towards higher margin and quality businesses. The chart shows how segments that “should” have high margins and profitability like AWS, 3P, Ads and Subscriptions have been growing much faster than the traditional 1P retail.

A business with a durable competitive advantage that earns elevated returns over long periods of time and has ample room for growth, should be considered great.

Think there is little controversy that Amazon has durable competitive advantages in a market that should keep growing. The only controversial point is profitability. So this is where I have a variant perception and will be the focus of this writeup.

Retail

Amazon dominates online retail with +40% share of US online commerce and is now larger than Walmart in total retailing. Market should keep growing as total retail grows and online takes share from the current 20% level.

Amazon is in an enviable position in which customers trust the brand and have created a habit of buying at the site. Amazon has built a dominant logistics network that will help them sustain this dominant position in online commerce for years to come. Customers care deeply about fast and free delivery, and major amounts of money have been spent on efficiently delivering packages. Amazon has kept reinvesting and strengthening this position by adding features to its Prime offering such as media, free and fast delivery, etc.

In a sense, Amazon has a strong base of recurring revenues from customers that buy from Amazon consistently and this bond is strengthened through Prime, logistics network, etc. Over time, my guess is they will figure a way to get their share of value. 

Online retail has some significant advantages over traditional retail. There is limitless selection, people can shop from the comfort of their home saving time and money, lower inventory needs compounded by the fact that 3P is 60% of total volume, no need to pay for high rent locations, etc. 

AWS

AWS has the characteristics of a great business as well. The company dominates the large and growing cloud market. AWS has run rate revenues of $85bn, and according to management, we are still early innings. There are only 3 scale players and there are significant barriers to entry given capex requirements and time to profitability. AWS is investing more than 40% of revenues or close to $35bn in annual capex. Another characteristic that should make this a good business is that customers tend to be sticky as changing providers is a highly complicated endeavor. For these reasons, seems highly unlikely we will see many competitors in the field. AWS should remain dominant in a market that overtime should be much larger than its current form.

In the past, some have characterized AWS as a commodity supplier of compute and storage. This is a highly technical operation that includes chip design, applications and other solutions. This is less controversial now as the business has grown with different offerings. AWS has also averaged close to 30% EBT margins in the recent past while growing 30%, validating some of the positive characteristics.

Retail Profitability

Historically, North America retail has done something close to 5% margins while reinvesting in new ventures and growing at a quick pace. Profitability has declined in the past few years as they overinvested, made some changes to their logistics operation and probably misallocated capital into unrelated ventures. Management has been focused on driving efficiencies to return profitability to a healthy level.

If we assume that by 2027, retail as a whole is doing $650bn in revenue and that comes at a 5% margin, this segment would do $32bn EBT. This includes sales outside US, but seems reasonable as the majority of revenue would be at the “mature” stage. 

While 5% margin is a reasonable target, I believe there is significant upside. Main reason is segments such as ads, 3P and subscriptions are growing much faster than 1P retail. By 2027, revenues from ads should be at $75bn. This line was close to 0 when the retail business was doing 5% margins in the early 2010s. How much of tailwind that is to margins is unknown but should be a big one.

If margins ex-ads return to 5% by 2027, that would yield $29bn EBT. Let´s assume Ads come in at 35% margins (avg of Google and Meta), that would add another $27bn EBT. Total retail EBT would be $56bn at that point or $45bn NI.

AWS Profitability

There is less speculation on AWS profitability as in recent years margins have been close to 30%. Recently these have shrunk to 25%, which might be due to the slowdown in the sector and some overinvestment. Jassy has mentioned how we are still early innings in the cloud migration as 90% of workloads are still on premise. He has also talked about how the training and inference of LLMs/AI should help in the near future as the compute and storage requirements are large.

By 2027, I expect revenues should be close to $150bn as margins return to the 30% level. That would leave us at $45bn EBT and $36bn NI.

Valuation: margins inflect higher

As shown before, under these assumptions, by 2027 retail would be earning $45bn and AWS $36bn. Amazon as a whole would be earning $80bn. Question now is what would the right multiple be at that point?

My bet is retail will keep investing and strengthening its position in online commerce. By 2027 we would still have a dominant retailer, with high porifitability and growing high single digits. 25x seems reasonable, 4% earnings yield growing say 7% is not expensive.

Walmart which is a comparable, yet growing slower and does not have some of the positive business attributes, trades at similar valuation. Costco at a much higher 35x.

At 25x, retail would be worth $1,125bn.

AWS, by 2027 we would have $36bn in earnings. At that point, the business should hold a dominant position in the growing cloud market, with a sticky customer base, significant profitability and scale advantages that should proof durable.

If we assume at that point, revenues are growing 10% with high returns on additional invested capital, what is the right multiple? Think 30x is not excessive given the qualities mentioned before and the profitable growth ahead.

At 30x, AWS would be worth $1,075bn

That gives us a total value for Amazon, assuming margins inflect higher, of $2,200bn by 2027. This would be $210/share or 2x in 4 years for 18% IRR. There is additional upside if company returns some capital in form of repurchases as FCF inflects higher during this period, which I believe will happen.

Valuation: no significant improvement

Besides the 2x potential, what I believe makes this investment attractive is there is some protection in case things don´t turn exactly as expected. Investors should do “Ok” even if there is no significant improvement in margins and profitability.

If we assume that all the benefit from high revenue streams such as Ads, 3P and subs do not show in the numbers and retail margins stay close to 5%, then this business would earn close to $25bn NI by 2027. At 25x, retail would be worth $625bn then.

In the case of AWS, if margins remain depressed at 25% then by 2027 this segment would be at $30bn NI. At 30x, AWS would be worth $900bn.

In this scenario in which margins do not inflect higher, by 2027 Amazon would be worth $1,525bn or $145/share. That is a decent return, but no spectacular.

Downside Case

There are always risks that these assumptions remain too optimistic. Also, things can go bad when you are playing around with multiples and assumed amrgins a few years out. This is necessary for this type of investment, but incorporates complexity and some dissapointing scenarios.

Business quality might erode and multiples might be too high. Other unknowns might come up over time and we should take into account all these possibilities.

As in every investment, there are risks and we will see how it plays out. Think there is some downside protection as this is a quality business with a self-help story.

Catalyst

-AWS returns to growth with help from the economy, LLMs, etc and margins inflect higher

-Growth in ads boosts retail margins

-Capital return program announced as Amazon returns to sustainable FCF generation