Amazon Review!
Okay, I did this one about two weeks ago, before this recent crash. However, I’m not worried at all. Their numbers were good, buy more.
Amazon:
Too often, investors look for excuses to leave equity or to at least look for options that appear safer. This is only human nature, but this nature is often exacerbated by talking heads and general media discourse that is, more often than not, excessively pessimistic. Right now, however, the market is volatile in many aspects, too volatile for my taste, with a significant part of this volatility coming from the White House. From inauguration day on January 20th 2025 to April 8th 2025, the S&P 500 fell 18%, and the Nasdaq fell 21%. Since then, the S&P 500 has increased by 27% and the Nasdaq 37%, bringing both markets up to about 6% YTD.
This end result is average, but the path traveled is not. The tailwinds coming out of the Biden administration were immense, particularly around AI, clean energy, and tech in general, but every facet has been slowed in some way, and just now, has caught up to returns on the year that could be expected. Last year, the market was up 21% by late July, but this year, just 6%. The market is paying a price for volatility and uncertainty. One can see it in nearly every earnings report. Companies that I/Pax Romana Capital (PRC) hold like Equifax and ASML have tanked on earnings that were objectively positive in terms of past/current results but equally uncertain in terms of outlook due to artificially created unreliability, and companies that PRC does not hold like GMC have become depressed due to similarly created headwinds. It is not so much that these companies have changed intrinsically but more that the field of expected outcomes has expanded massively on the negative side due to actions from the US, particularly around tariffs.
Basically, if a healthy, 8/10 company remains ceteris paribus, but tons of uncertainty is thrown in about its ability to sell its product, the company could become a 5/10 company or stay as an 8/10 company. The market has become frothy, meme stocks (an indicator that the market is topping) are back in vogue, and tariffs continue to fly around and depress earnings potential in some spheres. In these moments, stability needs to be found, and Amazon (AMZN) provides that stability while giving immense upside.
There are three criteria that a company needs to meet to be considered a buy. First, it needs to be an intrinsically healthy company. This means good leadership and healthy books. Second, revenue and earnings (but revenue in particular) need to be growing at a clip that is more than healthy, with revenue coming in at double digit growth y/y preferably. And finally, the company needs to have a significant moat that prevents competition that could potentially bleed the company out or cap its ceiling and potential to grow. Amazon meets all of these.
REVENUE:
Let’s start from the bottom. Amazon’s revenue and EBITDA growth is very solid and is almost ETF like in terms of diversification.
Revenue growth of about 11% y/y more than enough, especially when the revenue streams are so diversified. Here are Amazon’s top six revenue streams.
Online stores–these are what you think of when you think of Amazon. Amazon’s online presence is the section of the business that ships whatever you want to your door (39% of total revenue). I have very few concerns on this front. Chinese companies like Shein and Temu are a bit pesky, as they peel away the cheapest of Amazon’s sales, but as anti-Chinese sentiment continues to ramp up, these companies will fall by the wayside as regulators tighten up on them (closing of the de minimis loophole for example). Besides, Amazon launched a program to directly compete with these companies. Outside of these Chinese companies, Walmart is the only other real competition, with their same-day pickup programs, but fundamentally, Walmart has a ceiling that prevents mass competition with Amazon. Yes, their omnichannel strength and huge network is impressive, but Amazon and Walmart still don’t truly compete with each other. Walmart does not have as many goods as Amazon, you still have to drive to Walmart, and prices are marginally lower. Smaller competitors like Ebay, Depop, Etsy and others exist on the edges, and they are not worth considering.
Physical stores-these are the brick and mortar stores that Amazon has built into. Amazon bought Whole Foods a few years ago, they have Amazon Fresh as well, and a host of others (3% of tot. revenue). The brick and mortar is a nice to have, more than anything too significant. I appreciate the diversification that these stores bring, but they aren’t worth much focus.
Third-party sales-These are commissions that Amazon picks up, as payment for acting as the intermediary between two entities, a seller and a buyer (25% of tot. revenue). Whenever one purchases something on Amazon that is not from Amazon, but it is instead from a company or person selling, Amazon picks up a decent chunk of the sale in the form of a commission. These commissions are incredibly cheap profits for Amazon, and they are as much a recognition of Amazon’s top-dog status in the e-commerce field. These profits will stay in Amazon’s hands and continue to grow for as long as Amazon can keep their position.
Advertising-Amazon sells advertising all over its products (9% of tot. revenue). These include the ads sold on Prime Video, on the website, and generally sponsored product ads. This revenue will also be present and continue to grow for as long as Amazon can continue to be at the forefront of the societal consciousness.
Subscription services-Amazon Prime accounts for most of this revenue, but there is also Kindle related revenue, as well as music and movie purchases/rentals (7% of tot. revenue). These are the aspects of the business that provide substantial diversification. Amazon Prime Video is the second largest streaming service in the world and Amazon Music is the third largest music service in the world. These parts of the business are where elements of stability come from.
AWS-AWS is where the rest of the stability and diversification comes (17% of tot. revenue). All of these revenue streams are growing at a very solid rate, usually in the high single-digits, low-teens, but AWS is growing at about 19% per year, which is significant because it makes about $110 billion in revenue per year ($200 billion in commitments). Revenue of $110 billion would make AWS the 66th highest-revenue company in the world were it a separate entity. Operating margins for AWS are 40% because it is just cloud storage on a massive scale, and these revenues are nearly guaranteed for the future. Nobody is leaving the internet to go back to paper-and-pencil, and at this point. For competition, AWS’s primary competition is Azure and Google’s Cloud Program. These companies exist in a bit of a triopoly, but AWS is the largest, and they have carved out a significant niche. AWS’s scale and maturity far exceeds that of its competitors, but Azure’s is a legitimate competitor. Google is a distant third, but they still present a threat on the analytics front. Displacing either company will prove challenging, but obviously, significant capital investments are being made to shore up the AWS moat.
Other-The other aspects of Amazon’s business are possible aspects of the business that could grow substantially. For example, Ring, Echo/Alexa, TVs, Kindles, and Amazon Pay are all verticals under which Amazon could see revenue growth and added diversification.
For Amazon, their revenue streams are all so diverse and so healthy, that stability and growth are the inevitable by-products.
EARNINGS:
Amazon’s earnings look odd on an FCF basis, which we’ll use as a primary measure for Amazon because it's such a capital-intensive, fast-growing business (despite being closer to a trillion in annual revenue than $200 billion in annual revenue).
Initially, Amazon’s FCF, especially when combined with SBC, which they use substantially, looks pretty weak, but to contextualize, you have to look at re-investment: Amazon spends a stunning amount on CAPEX and R&D.
As one can see below, Amazon spent almost $170 billion, or the total annual revenue of BP or Ford, on growing future revenue streams through CAPEX and R&D. Amazon’s earnings are so low compared to their revenue and size because they put all of their excess capital back into the company, which is how the company continues to grow their revenue at the rate of a company 1/100th their size. Last year, Amazon made, in revenue, the entire worth of the country of Belgium, but they continue to compound and diversify instead of cashing in, which is where stability in the long-term comes from.
We’ll cover this more later, but Amazon’s CAPEX is worthwhile. Amazon is not just spending money to spend money. They claim to operate as “the world’s largest startup,” and that is not an inaccurate claim. Amazon expands their CAPEX to expand the top line primarily with the (correct) assumption that the bottom line will follow. Many startups get caught in this step because they eventually run out of cash, or their ventures don’t materialize as fast as they should, but Amazon doesn’t struggle with either of these aspects. Amazon has enough cash to burn for as long as it takes until a product becomes worthwhile, or it gets binned like the Amazon phone.
Amazon has been spending their FCF primarily on AI related developments. Building out AWS for example. Scaling AWS for AI, chips, specialized servers, high‑bandwidth networking, etc… is intensively capital-hungry. This sort of investment depresses current FCF but secures future added revenues and will lock AWS into its current market-leading position. Then, there are investments in Anthropic, an AI company, new robots for all the fulfillment centers, funding development of Amazon’s AI chips, GPU clusters (basically a mega-computer), fintech expansions (Amazon Pay), and significant Prime Video outlays.
Basically, you should not be worried about earnings. This moment is one of significant spending to prepare the company better in the face of rapidly changing markets. These investments are guaranteeing future revenue growth, and these assets will return results relatively soon. Amazon’s operating-cash-flow base was up 15% in Q1 to $114 billion (TTM), but the FCF number was so depressed because the company is just spending so much money (which is more than healthy) to re-orient its infrastructure for AI.
The AI Perspective:
Every single facet of Amazon previously mentioned in this report guarantees revenue growth at an 11%-ish rate for the next…decade? In the long-term, however, AI will take Amazon’s E-Commerce business up another level.
The most expensive part of any delivery company is what is known as “the last mile.” The last mile is the distance between the distribution/fulfillment center and the end-goal, a customer’s house, a business, whatever. Another aspect of Amazon’s business that is quite expensive is the massive number of employees it has: 1,500,000 of them. In 2024, (and this is an estimate, since it’s not clearly laid out), Amazon spent around $200 billion on paying its employees. About $90 billion for fulfillment (staffing for warehouses, customer service, sortation & delivery centers), $80 billion for tech and infrastructure (data centers, corporate IT, AWS payroll), $40 billion for S&M, and $10 billion on general admin (corporate, finance, HR, legal, etc.). By far, employees are Amazon's largest annual expense, and last mile shipping is not far behind. In both fields, AI can eliminate inefficiencies.
Amazon already uses a million robots in their warehouses, and the number of packages shipped per employee per year is at 3,870, compared to 175 a decade ago. Right now, despite the enormous number of people employed in shipping products, Amazon has the fewest number of employees present per facility, at 670. These figures represent an easing of pressure. Those unions that were popping up a few years ago have not been mentioned in the media in years, turnover has slowed, and employees are being effectively re-trained into systems that they already have a grasp on, given their experience. But in totality, the number of employees Amazon has will continue to decrease in the coming years, boosting FCF, efficiency, and keeping revenue streams at the same level of growth.
The company claims that their goal is to have robots working alongside humans, but either way, total employment of humans will decrease in favor of robots. This phenomenon is inevitable because of robots. The physical appendages and motion of robots has been something that humanity has had a great grasp on for a while now, but getting robots to think on their own has been a massive challenge, one that has stymied attempts at total automation of large warehouses. But AI solves that problem by potentially giving robots the ability to process their surroundings, commands, and actions. Currently, Amazon is running trials on how AI works in their warehouses, how it improves efficiency.
These warehouse robots will be actualizations of AI agents like Claude (Anthropic’s LLM). These robots will have the capabilities to make real‑time pick‑and‑place or routing decisions, rather than relying on pre‑programmed paths, which is helpful, especially when combined with the increased physical strength that they possess. These agents that can make rational decisions based on their surroundings are being trialed right now, but when they are eventually borne out as a success (likely after working through some issues), headcount will flatline or decrease, even as sales and revenue increase.
Then, there is the earlier mentioned last mile. Right now, Amazon has been confirmed to be running tests on robot deliveries of packages. A human would drive the van around, and the robot would effectively ride shotgun, then get out and deliver the package. The robot would basically use all of the technology of an autonomous car to navigate landscapes and neighborhoods. Each robot is projected to cost around fifteen thousand dollars in total. Now, this does not sound very efficient on the surface. Amazon would be paying fifteen thousand for a robot, and they would still be paying around sixty thousand a year for the driver, but because the robot+driver combo is so much faster than the solo driver, the robot eventually pays for itself. If a driver usually can handle 100 deliveries a day, he could then handle about 130 because of the 30-40% speed up. This figure translates to a reduced cost of 23% per package, and that is just at the beginning.
Early on, the driver would likely have to walk with the robot to the porch, but as time goes on and the artificial intelligence and robot become more acclimated to their surroundings and responsibilities, the driver would become more of a robot-supervisor. The goal is to totally automate the last mile, eventually with autonomous vehicles as well, but that is likely much further away.
Bank of America stated that a robot buddy as a driver future is likely three to five years away, but whether it is two years (which would match with the Anthropic CEO’s predictions), or as late as 2030, the robot delivery-helper will cut costs and is generally representative of a striving for efficiency that makes Amazon a company that will continue to grow.
In totality, Amazon will save about $14 billion per year from their last mile savings and about $12 billion from their warehouse savings per year, conservatively. Meaning that one can expect earnings increasing by $25 billion alone from these tailwinds.
In short, artificial intelligence will allow Amazon to become even more efficient in terms of raw manpower needed to complete tasks, while increased usage needs for AI services will make AWS more necessary and profitable.
Valuation:
Valuation is rarely a concern of mine, because my general basic thesis is that it is both extremely difficult to time an individual stock, and that companies that meet my criteria are too high-quality for a steep valuation to dissuade me. However, valuation of a stock can be helpful when it comes to seeing an opportunity in a moment.
For Amazon, we’ll look at the price/fcf ratio, and the company’s forward p/e (for market sentiment).
On a trailing basis, the p/fcf yield looks completely out of wack, at about 118x, or a yield of 0.9%. However, this figure is merely present because of Amazon’s CAPEX spend. On a forward basis, which assumes that Amazon’s CAPEX will eventually moderate once this transitional phase, particularly in AWS, has completed, comes in at around 22x, or a 4.5x yield. These figures are derived from the projections over the next 12 months, and they are about a third of Amazon’s usual figures. Amazon’s 10-year trailing p/fcf yield is about 60x, 5-year, 59x, and 3-year, 82x. To be able to look forward and see a ratio of 22x is to see an Amazon that is cheaper on this basis than it has been for a decade, and one that will have the added boost of the investments that have caused this depression in FCF. Amazon’s tech peers, Apple and Microsoft, trade on a 29x and a 52x basis themselves, meaning that Amazon really is pretty cheap.
The company’s forward p/e ratio is about 35.6x, which looks very steep, but one must look at depreciation and amortization factors in this instance and create their own, adjusted, forward p/e ratio. I’ll put the math below, and then I will explain why and how I got to this conclusion.
Forward EPS=$231.78 (current price)/35.6 (current forward p/e)=$6.51
$52.795 (2024 D&A expenses in billions, proxy for 2025)/10.793 (billion shares)=$4.89 D&A per share
$6.51 (forward EPS)+$4.89 (D&A per share)=$11.40
$231.78 (share price)/$11.40 (adjusted forward EPS)=20.3x
That looks much better because 20.3x forward earnings is actually quite cheap. This figure better represents the underlying earnings power that Amazon possesses that is being disguised by CAPEX spend. Now, as a reader you might be confused by some of this math, particularly as to why I added D&A per share back into the company’s earnings.
D&A expenses weigh down on a company’s earnings, even though they’re technically non-cash charges. Basically, the company’s earnings are becoming less valuable on the books, even though technically, no cash is actually leaving the company. Amazon has been investing tons of money recently, and not so recently. When Amazon makes a massive investment, the worth of that investment decreases on the books over time, in a process called depreciation for physical assets (think a giant warehouse) or amortization for intangible assets (a brand). This happens in accounting to reflect a general pessimism, that the investment will wear out its usefulness as the asset becomes obsolete or its finite life expires. This D&A is not the company actually spending money, but the company technically becoming worth less. Amazon loses $53 billion per year in profit from D&A drag, lowering operating and net income, even though, again, no cash is actually leaving the company. Usually, D&A drag would be marginal for calculating p/e ratios, but because of Amazon’s colossal investment figures, it adds up annually. High D&A makes p/e look much higher than it actually is, distorting earnings and making them look excessively depressed.
That said, this method is too quick and dirty, and it likely overstates earnings power. In this case, observing future p/fcf is a better valuation metric. By adding back all of the D&A, actual, worthwhile depreciation that captures a lessening of value is ignored. I merely included the adjusted forward p/e ratio to show that the true earnings power, relative to the price of the stock, is much cheaper than the 36x would lead one to believe.
Risks:
Amazon both faces many and few risks due to its sheer breadth. The biggest risk is regulation. With liberals likely soon to return to power in America, and the EU maintaining its anti-huge American tech company stance, Amazon will likely have to continue handing out big damages. For example, Amazon is fighting in the UK to avoid paying around $4.5 billion in damages for trivial matters around third-party sellers. Then, there is a major case in the UK
The more worrisome legal case against Amazon is in the US. The FTC+17 US states have accused Amazon of monopolistic behavior, and that case is set to go to trial in late 2026. The worst-case scenario is obviously the b-word/s-word (breakup/spinoff), but more likely, Amazon is looking at a hefty fine. If contracts had to be changed, that would also potentially be problematic, but this suit is in a wait-and-see phase.
Unionization was touched on earlier, but as mentioned, increased robotics should gradually lessen the pressure on employees. In 2022, when all of the unionization worries were being kicked up, the catalyst was that employees were being overworked. There were so, so many employees, but simultaneously, there were too many goods that needed to be moved. Now however, the pressure has been lessened because there are the same amount of employees doing better work. Unionization should not be a concern for any investor, as the odds of it coming to fruition are slim and the consequences would theoretically be small.
The largest issue in my view is tariff issues, or really, the general gumming up of international trade. Amazon acts as the global market for so many items, and third-party sellers, particularly in China have already had to raise prices. Amazon cannot control these prices directly, so these changes could dampen their commissions. In the long-term, I do not see tariffs as a genuinely troubling development, but they could be irksome in the short-term.
Final word:
To return to the criteria at the beginning.
Good leadership? This was not touched on much, but Andy Jassey has proven to be a very solid CEO. He took over in 2021, and has delivered record revenues, significant stock appreciation, and record profits. He has done so while dealing with the after-effects of COVID, of a hyper-expensive AI re-orientation, and a shifting regulatory environment. He has made solid, aggressive bets, particularly on AI, that seem to be paying off in some aspects already. His leadership, a very important element of the company, is a plus.
Healthy books? Debt is in a perfectly manageable position, and as I’ve been explaining for 3,000 words, there is only one direction for revenue to go.
Revenue/earnings? Revenue is growing at around 11% per year and while FCF is so volatile that it can be hard to know what is happening (-$37 billion in one year, $8 billion the next), both revenue and earnings seem to have a high floor and a growing ceiling.
Moat? This is arguably the most important part of a business, and Amazon’s moat is stellar. Amazon has a moat compared to their competition by proxy of being part of every growing industry in the country. Their competitive advantage in so many fields is an ability to attack problems with enough money and intelligence to eventually circumvent issues. In E-commerce, Amazon is by far the largest, with clear plans to expand into a market with plenty more room. In the cloud, Amazon has a clear lead, and they are positioned to maintain or accelerate revenue growth again, with the introduction of all their AI chips, all the server installations, and their general preparation for a new wave of technology. Even where Amazon does not have a moat currently, like video, advertising, or fin-tech, one could make a reasonable bet that Amazon will be able to leverage their network to pull in new customers and keep old ones. Amazon continues to integrate valuable new markets as they continue to move beyond just being a shipping and computing power company. In short, Amazon has a definite moat in their most important fields, E-commerce, compute, and third-party sales for a myriad of reasons, but they create a competitive advantage in the fields they enter by having more resources and a more defined network to leverage than their competition, and this is enough to eke out enough market share to justify continued funding.
Final, Final Word:
Amazon seems too overplayed, or at least, too obvious. When I was telling my friend about a company I was researching, his mom overheard and asked me which one. When I told her Amazon, she sort of scoffed and then said, “now?” While the investing community does not hold the same view of my friend’s mom, shout out Anne, it kind of does. Too often, investors fall into the fallacy of believing that every mega-cap company like Amazon is an Apple. They were great ten years ago, now their revenue is flat, and even if it’s not, they could not possibly continue to grow at the same rate. “Their market cap is over two trillion. If everything happened as you predict, the market cap would probably triple over the next few years, and that’s too much optimism.” To this straw-man, I think the obvious answer is that we are in a new age of mega-cap companies. The rich get richer, and beyond an imagined stone wall that the community expects companies of Amazon’s size to run into, there is nothing standing in their way. Amazon has more revenue on hand than any of their competitors, except for Walmart, and even they will fall behind Amazon within the year. Amazon has a mentality that isn’t too conservative like Apple, and their leadership isn’t as questionable and weak as Google’s. Microsoft and Nvidia are great, but PRC/I already have a sizable position. Meta is not as big as Amazon, Netflix is not as big as Amazon, Walmart is not as big as Amazon. The list goes on and on.
And again we return to the assumption that because Amazon is so big, eventually, their momentum will reverse and inertia will take its toll. Eventually, investors and my friend’s mom assume Amazon will hit its ceiling, and that will happen pretty soon because all of the explosive growth happened years ago, and now the magic is gone. However, the culture of the company is different than other mega-caps, and it is through this culture that investors can find legitimate stability in revenue that is rock-solid, combined with a ceiling that will continue to match the ambitions of the company.
