Amazon (AMZN) DCF Analysis

 

Introduction:

AMZN has built a strong business model, creating a global logistics network to connect consumers to sellers all over the world. This behemoth of a business inadvertently makes AMZN irreplaceable in our day-to-day life, the convenience and speed at which AMZN offers its services is unparalleled by many companies. Because the company takes a long-term view on returns, it chooses to forfeit short-term gains for longer-term returns. On top of its logistics business, AMZN is also involved in cloud computing the next frontier with large growth potential, all these show the forward-thinking acumen its leaders have.

Investment Thesis:

Established Logistics Network

AMZN’s most compelling competitive advantage is that it has an established global logistics network that no other company can replicate without substantial monetary and time investment, through controlling its own logistics network by purchasing its own cargo ships and leasing planes (SOURCE). AMZN can control where the ship docks which allows them to be more efficient with their cost through avoiding busy ports.

Strong Reverse Logistics

Consumers increasingly perceive easy & free returns as essential (SOURCE). A very large C.A. for AMZN is its established reverse logistics as evidenced by AMZN being able to offer free return shipping for such a large volume of goods due to them expanding the scale of their reverse logistics that it makes economical sense for them to be able to offer these services for free. AMZN has also increased the ease of returning products, for example, by using UPS stores as a physical retail location where customers can drop off their parcels. UPS is reliant on AMZN as well, accounting for 11% of UPS revenue which gives AMZN large leverage over UPS (SOURCE)

How retailers manage their reverse logistics is also important as it is an additional outlet of significant revenue source. A few notable strides that AMZN is taking in this front is FBA Grade and Sell, where AMZN grades the item that buyers returned and grades it. Afterward, 3P Sellers (FBA Sellers) set the price based on the condition of the product.

Branding Goodwill

Even though in recent years a few large e-commerce platforms have appeared such as Taobao, Shoppee, and Alibaba. AMZN has managed to maintain its reputation as a place that consumers will turn to if they want to get high-quality, safe, and reliable products that would not be flimsy or poorly made. The main reason why people use Taobao or Alibaba is because it is a website to get quasi items at a fraction of the price (SOURCE). Consumers have grown to accept the compromise in quality in exchange for a discount, this may be great in an economic downturn but during an average or booming macro environment, it ends up becoming a disadvantage. On top of that, by competing on price Taobao and Alibaba in the long run is unable to earn much margins and reinvest substantially to grow their business at a pace comparable to AMZN.

Revenue:

Retail Sales
Retail sales refer to AMZN’s online marketplace and physical stores.
AMZN has such large economies of scale that they can offer higher quality products for cheaper. “Amazon prices are 13% cheaper than other retailers for essential items.” (SOURCE). On top of that, the increasing convenience of AMZN through speedier delivery increases the appeal of AMZN. “more than 90 percent of consumers expect two- to three-day delivery as the norm, while 30 percent expect same-day delivery.” (SOURCE).
AMZN can achieve higher rates of same-day delivery due to their strategy called “regionalization”, where they store items closer to customers. This results in customers choosing AMZN to deliver their everyday shopping needs as AMZN can get them their items significantly faster. Regionalization also allows AMZN to save on cost which in turn allows AMZN to offer a wider variety of items at a lower Average Selling Price (ASP) profitably. On top of that, AMZN has also released RUFUS an interactive AI model that allows the user to chat with it to get more personalized recommendations based on the consumer's diverse requirements.

AMZN's next frontier of growth comes from India, where it is building out a logistics network to be able to serve rural customers in India. However, AMZN has recently invested heavily in its fulfillment network by doubling it. (SOURCE) So, I assumed that AMZN would slow down growth plans to capitalize on their new fulfillment network fully. India represents a very strong growth opportunity given the high smartphone penetration rate of India, the rapid growth of the Indian economy, and their large population (SOURCE), this culmination of factors makes India a very lucrative market for AMZN to expand in.
When forecasting Average Monthly Traffic, I took into account (SOURCE). I assumed that as AMZN’s fulfillment network gets more comprehensive their monthly traffic increases as AMZN’s value proposition increases. I assumed that AMZN expands its fulfillment network significantly every 5 years.
When forecasting Sales/Monthly Traffic, as AMZN begins targeting more price-conscious countries, I assumed that AMZN’s sales/monthly traffic did not grow due to AMZN raising prices and only grew in line with the perpetual inflation rate. However, in the longer run as AMZN becomes more entrenched in these countries’ consumer habits I assumed that AMZN has larger price-setting abilities.

Third Party Sales (3P)
Third Party Sales (3P) refers to the %sales AMZN earns as commission from third party sellers that sell on AMZN’s website.

The appeal of AMZN to retailers is that retailers can reach a massive audience very easily and can glean valuable insights on their products from the tools that AMZN platforms provide. Listing on AMZN also adds a layer of credibility and convenience, especially for Fulfillment By Amazon (FBA) sellers.

When forecasting 3P Gross Merchandise Value, I took into account (SOURCE). Opting for less granularity I forecasted GMV as a % of historic averages. As AMZN expands into more markets I expect this GMV to increase as there are now more sellers that are attracted to the services provided by AMZN.

When forecasting %GMV Earned, I assumed that AMZN slowly begin taking up a larger portion of the pie as AMZN would build a very strong value proposition to sellers and sellers would grow increasingly reliant on AMZN’s platform. However, I assumed modest growth as AMZN has to remain attractive to sellers by not cutting too much into their margins.

Advertising
Advertising refers to ad revenue AMZN earns from advertisers to advertise on AMZN’s platform.
When forecasting Advertising, there is a lack of suitable data to forecast advertising, and the success of advertising hinges on the success of retail. Opting for less granularity I forecasted it as a % of retail revenue.

Subscription
Subscription refers to AMZN’s subscription service, Amazon Prime.
When forecasting number of subscribers, taking into account (SOURCE). I believe that NFLX is a very strong competitor having dedicated their entire business model exclusively to content acquisition and marketing compared to AMZN which has a lot of other businesses that they need to pay attention to. I believe a significant portion of subscribers are purchasing Prime Video as an add-on to Amazon Prime rather than just for Prime Video alone. On top of that, in foreign countries where local broadcasters are still in a strong position, it’s tougher for AMZN to break into so I believe that Amazon Prime subscribers will not grow at an aggressive pace. In the past few years, with the backdrop of a poor macro-environment AMZN’s subscribers numbers Y/Y growth rate declined evident that Amazon Prime is perceived as a luxury good or an item placed in low priority.
When forecasting churn rate, taking into account (SOURCE). Opting for less granularity I forecasted it as a % of historic averages. However, I believe that AMZN’s churn rate is higher than it should be due to the influx of new subscribers trying out Amazon Prime for the first time. So, over time I believe that as consumers get used to using Amazon Prime and develop a reliance on the service I believe that the churn rate tends downwards. So I assumed that as AMZN enters into newer markets the value of Amazon Prime increases which incentivizes consumers to stay subscribed to Amazon Prime.
When forecasting Annual Revenue/Member, as AMZN begins targeting more price-conscious countries, I assumed that AMZN’s Annual Revenue/Member did not grow due to AMZN raising prices in these countries and only in countries where AMZN has an established presence. So, this leads to the overall Y/Y growth rate being slower in the short run.

AWS
AWS refers to AMZN’s cloud computing services. AWS offers SAAS and IAAS.

AWS has a strong first movers advantage, they started in 2006 whereas Google Cloud started in 2008 and Microsoft Azure started in 2010. I believe that AWS has had strong growth in recent history in part due to the chip shortage and a mass rush into AI Training that demanded a lot of chips, causing some companies to resort to renting more cloud GPUs from AMZN.

AMZN’s Gen AI strategy has 3 layers. The first layer is targetting the training and inference market for AI. The second layer (Amazon Bedrock) offers large numbers of foundational LLM that businesses can use and develop according to their needs, allowing customers to experiment with different models to optimize experience and cost. The third layer is (Amazon Q) which functions similarly to ChatGPT.

Others
AMZN’s other revenue sources.

When forecasting Others, I forecasted it as a %retail sales as the success of Others hinges on the success of AMZN’s retail arm of business. Opting for less granularity I forecasted it as a % of historic averages.

Reality Check:

As a reality check for my forecast, I looked at the world’s GDP to figure out if the GMV for AMZN is in line with the total growth in goods and services in the world. As of 2023, AMZN’s GMV as a %World GDP is 0.5%. By 2040, the world GDP is at 153T (SOURCE), I assumed that AMZN’s GMV as a %world GDP is 1.8%. This is realistic as over time, it would make sense for smaller players to sell their products on AMZN rather than rebuilding a new logistics network from the ground up. An example of smaller players using a much larger platform to sell is live stream commerce, which is prominent on platforms such as Douyin (SOURCE).

The total world population as of 2023 is at 8.05 billion (SOURCE), and monthly Traffic as of 2023 as a % of the total world population is at 32.92%. By 2043, the world population will grow to 9.36 billion (SOURCE), I assumed that AMZN’s monthly traffic as a %world population is 49.67%.

I believe that these assessments are realistic given that I am expecting AMZN to expand its logistics network about 6x over 20 years, increasing its reach far beyond its domestic reach in the US. The speed of expansion is logical too given that AMZN was able to double its logistics network 2x over 2 years.

Cost:

COGS
When forecasting COGS, Regionalization has reduced the distance needed by AMZN to deliver their item to customers, in turn saving cost.

G&A
When forecasting G&A, I took into account the number of employees. AMZN has been reducing headcount over the past few years, as AMZN overhired during COVID-19 to keep up with the raging demand.

When forecasting Cost/Employee, as AMZN is in money-saving mode I assumed that they continued the historic trend of cutting down Cost/Employee for the next 2 years before growing in line with the perpetual inflation rate.

R&D
When forecasting R&D, R&D was capitalized. To determine the useful life I looked at what the majority of the spending was on, “Software development costs capitalized were not significant for the years presented.” and “The weighted average remaining life of our capitalized video content is 3.5 years.” - 2023 Q4 10-K. So, I assumed that the main expenditure came from capitalized video content, and applying the conservatism principle, the amortization was rounded down to 3 years.

Change in NWC:

When forecasting change in NWC as AMZN recently rolled out their regionalization efforts in FY23, the impact on Day Inventory may be very apparent and should be assumed as the status quo going forward.

D&A and CapEX:

When forecasting Amortization and Content Amortization, opting for less granularity I forecasted it as a % of historic averages.

When forecasting Content Acquisition and CapEX, I assumed that as AMZN expands its reach the CapEX spending increases before regressing to below historic averages.

WACC:

“The weighted-average remaining term of the financing obligations was 17.9 years and 17.0 years and the weighted-average imputed interest rate was 3.1% as of December 31, 2022 and 2023.” - 2023 Q4 10-K

10Y T-Bond Yield (1M Avg) = 4.31%
Beta (SOURCE) = 1.17
Stable Market ERP (SOURCE) = 4.60%
COE = 9.69%

AMZN is rated “AA-” (SOURCE)
AA Bond Yield (1M Avg) = 5.03%
Marginal Tax Rate = 21.00%
AT-COD = 3.97%

Stock Price (5D Avg) = $173.97
Shares O/S = 10387.38M
Market Value of Equity = 1807092.50M
Weighted Average Maturity = 17.0 Years
FY23 Debt Obligation = 67150M
Interest Expense = 2081.65M
Market Value of Debt = 60026.19M

%Debt = 3.21%
%Equity = 96.79%
%WACC = 9.48%

Conclusion:

Ultimately, in my base case, I value AMZN at $155.67 per share. I believe that AMZN is still overvalued despite it being a good company. This overvaluation of AMZN is a direct result of AMZN posting high free cash flow in recent quarters which gave investors the confidence that AMZN is capable of deriving positive free cash flow from its business. However, most of this positive cash flow was derived from their high margins AWS and advertising increasing substantially Y/Y. This could mean that positive cash flows are a cyclical occurrence as cloud computing is reliant on the volatile AI market and advertising is a cyclical industry that relies on macro-conditions. I believe that there is a strong case for AMZN to break beyond its current market valuation, but for it to be successful will be dependent on how the Indian market reacts to it and if AMZN is able to restart its push in the China market.

Base Case: [INSERT]
Best Case: [INSERT]
Worst Case: [INSERT]
Sanity Check: [INSERT]
Revenue Model: [INSERT]
Revenue Model Part 2: [INSERT]
Cost Model: [INSERT]
Change in NWC Schedule: [INSERT]
Capitalised R&D: [INSERT]