Amazon's Sacred 7 Numbers Show It's Overvalued

Aug. 7.14 | About:, Inc. (AMZN)


Amazon bulls say a heavy investment cycle is depressing profits; bears say the business is fundamentally unprofitable. The truth is in between.

I show that even relatively optimistic assumptions about profitability cannot justify Amazon's current valuation. I describe Amazon's "sacred seven" numbers, and estimate it is 60% overvalued.

Amazon's current valuation can only be justified by significant increases in working capital cashflows, and significant reductions in capital expenditure (both relative to net sales).

Amazon's sacred seven numbers

I've been long Apple (NASDAQ:AAPL) and short Amazon (NASDAQ:AMZN) the past three years.

I've made no money from it - the gains from the Apple trade have almost exactly offset the losses on the Amazon one. But in the process, I've learnt some interesting things about valuing technology companies.

This article is about Amazon. I want to highlight some important numbers that investors should pay close attention to in the next 12 months.

The first thing you see when you look at Amazon as an investor is its vertiginous price-earnings ratio, which currently stands at 486, compared to about 16 for Apple.

Amazon bulls say this is caused by a heavy investment cycle artificially depressing earnings per share. Amazon bears say the business is fundamentally unprofitable. The truth - as always - is actually somewhere in between.

Amazon's earnings per share number is a function mainly of four parameters - the net sales, the operating profit (or EBITDA) margin, depreciation and taxes. An investor who can understand how these four parameters will behave over time will be able to accurately predict the long-term earnings per share trend for Amazon.

On the other hand, the price of an Amazon share is the present value of discounted free cashflows to equity. This number depends on three parameters - (in addition to operating profits above), the cash Amazon is able to generate from working capital movements, the capital expenditure required by the business and the discount rate (or Amazon's cost of capital). An investor who can correctly estimate these three parameters should be able to predict the future share price trends for Amazon.

These seven parameters should be, in my view, Amazon's "sacred seven" numbers. Before I continue, let me set out my current view of these seven parameters, against Amazon's actual historical record, based on an analysis of each of its audited financial statements between 2007 and 2013.

1. Net Sales: have grown at c. 30% per annum in the last five years, but are likely to grow at between 20-22% in the next two years, and gradually decline as Amazon grows bigger to 15% per annum by 2018 and 3% per annum by 2023.

2. EBITDA margin: has declined from just over 6% five years ago to a low of about 4% in 2011 before recovering to 5.4% in 2013. Consensus estimates for Amazon suggest a continuing improvement in margins in 2014 and 2015 to about 5.7%. I believe margins are likely at best to stabilise at that level over time.

3. Depreciation: has remained fairly consistent at about 45-50% of the previous year's value of property and plant on the balance sheet (this includes un-expensed software development costs). However, due to the large increase in capital expenditure in recent years, depreciation as a proportion of net sales has shot up from 1.5% to 4.4% in the last five years, and this is why Amazon's net profit margins have collapsed from 3.7% in 2009 to 0.4% in 2013!

4. Taxes: Apart from an aberrant year in 2012 (which Amazon blamed on forex losses, acquisitions and changes in tax laws), taxes as a proportion of pre-tax profits have gradually increased from c. 21% to c. 31% over the last five years, and are likely to remain at that level in the future.

5. Cash from operations: In addition to cash profits from operation (net profit plus depreciation), Amazon has consistently generated another $1.5-2bn per year of cash through its working capital and stock compensation arrangements. As a proportion of sales, this cashflow has gradually reduced from 8% to c. 3% over the previous five years. I believe this working capital cashflow has now stabilised and will remain at c. 3% of sales in the future. However, stock-based compensation will also progressively dilute the equity base over time.

6. Capital expenditure: has increased rapidly as a proportion of net sales from 1.5% in 2009 to over 6% in 2012, before moderating to 4.6% in 2013. I believe the capital to sales ratio is likely to stabilise at the 4.6% level from 2014 onwards.

7. Discount rate: I believe long-term investors in Amazon will want to see a nominal return of at least 10% per annum, and so should discount free cashflows at this rate.

What do the seven assumptions above suggest for earnings per share and share price? They imply that Amazon will likely report a loss for 2014 of c. $0.10 per share, followed by a steady improvement to $2.10 per share in 2015, rising to $5.37 by 2018 and $5.80 by 2023.

They also suggest that Amazon will continue to generate a lot of cash from operations, rising from $5.4bn in 2013 to c. $9bn in 2015, $14bn by 2018 and nearly $21bn in 2023.

However, over this period, Amazon will continue to demand rising capital expenditure, which will eat into free cashflow. I estimate that Amazon's capital expenditure programme will increase from $3.5bn in 2013 to about $5bn a year by 2015 and continue rising to nearly $8bn a year by 2018, before stabilising at a steady state of around $12bn a year by 2023. Assuming the further current spend on acquisitions of c. $300m per year continues into the future, this means that free cashflow will improve at a more muted pace, from about $1.8bn in 2013 to about $3.5bn in 2015, $5.3bn by 2018 and $8.5bn by 2023.

The free cashflow above discounted at 10% over the ten years from 2014 to 2023 produces a market capitalisation of c. $36bn. But since this ignores cashflows beyond 2023, we should add an estimate of the discounted free cashflows beyond 2023. This is called a terminal value, and I have used the perpetuity growth method (assuming a constant 3% growth rate of free cashflows beyond 2023) to estimate a discounted terminal value of c. $54bn. This gives us a stock capitalisation of c. $90bn for Amazon.

With c. 460mn shares outstanding on a diluted basis in 2013, and assuming the flow of stock-based compensation to Amazon's employees continues to dilute the share base at about 3% per year, the stock capitalisation above translates to a value of roughly $195 per share.

Amazon is currently trading at nearly $315 per share, so in my view, it is roughly 60% overvalued.

To get a feel for what really swings the valuation, it's interesting to consider how sensitive it is to each of the seven parameters. I have summarized this in a short table below:

Valuation Parameter

(Base Case value)

Upside value

Share price

Downside value

Share price

Net Sales growth rate (20% declining to 15% by 2018 and 3% by 2023)

Stays up at 20% a year through to 2018


Declines to 10% a year by 2018


EBITDA margin


Rises to historical high of 6.2%


Falls to historical low of 4%



(47.5% of fixed assets)

Falls to 40% of fixed assets


Rises to 55% of fixed assets



(31% effective rate)

Falls to historical low of 21%


Rises to 35% federal statutory rate


Operating cashflow

(3% of net sales)

Rises to 5% of net sales


Falls to 1% of net sales


Capital expenditure

(4.6% of net sales)

Declines to historical low of 1.5% of net sales


Rises to historical high of 6.5% of net sales


Discount rate






Click to enlarge

The table starkly shows just how sensitive Amazon's valuation is to its working capital cashflow and its capital expenditure. Many investors fret about Amazon's shipping costs and product mix - these are of course important in driving profitability, but even significant improvements in profitability won't move the needle much towards Amazon's current valuation.

Instead, to justify today's valuation, investors must believe that working capital cashflows will rise sharply back to historical highs, and capital expenditure will fall back to historical lows.

As new information comes in over time from Amazon's reported results, I will consistently update my view of these seven numbers, reporting where I was right and wrong.

In subsequent posts, I will dissect each of my seven valuation assumptions and explain my reasoning.


The current valuation of Amazon is bullish, and appears to factor in substantial improvements to operating cashflows and significant reductions in capital expenditure. These do not however accord with the historical trend - Amazon's working capital position has been getting tighter rather than looser, and its capital expenditure relative to sales has generally been rising rather than falling.

Disclosure: The author is short AMZN. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.