Amazon's Shareholders Are Swimming Naked, And The Tide Is About To Go Out

| About: Amazon.com, Inc. (AMZN)

Summary

In the first half of 2014, Amazon's working capital has constricted more than expected and its capital intensity has gone up alarmingly.

If these trends continue into the future, I estimate Amazon will turn out to be overvalued by around 450%.

Amazon's share price is levitating against the pull of economic gravity because of easy money. As yields normalize, its cost of capital will rise. Caveat emptor.

In my previous four articles, I set out and defended a valuation thesis; even on optimistic projections, Amazon (NASDAQ:AMZN) is overvalued by 60%.

Every investor should ask themselves: where could I be wrong? In my first article, I set out seven numbers that I said will be central to Amazon's fate. I could be wrong in any of the seven. In my first article, I presented an analysis showing how sensitive the valuation was to changes in these seven assumptions, and I pointed out that, rather than sales growth or profitability, it was working capital and capital expenditure that really swung the needle on Amazon.

In this final article, I will mark my assumptions to market. I will compare how Amazon has actually performed in the first two quarters of 2014 against them, and the implications this might have for the future. I will close with a discussion of Amazon's cost of capital, and a last look (for now) at what a sensible investor should be willing to pay for an Amazon share today.

The table below shows Amazon's performance in the first six months of 2014 against my assumptions:

My 2014 assumptions

Amazon's performance in 2014 (H1)

Net Sales Growth Rate (%)

22%

23%

EBITDA margin (%)

5.7%

5.7%

Depreciation (% of sales)

5.7%

5.4%

Taxes (% of profit before tax)

31.8%

179%

Working capital (% of sales)

3%

-9.5%

Capital expenditure (% of sales)

4.6%

11.2%

Click to enlarge

In 2014, Amazon is performing broadly to my expectations on sales and profitability. I am perhaps a bit pessimistic on the depreciation charge, but accounting rules can mean that the half-yearly charge is distorted so one should not read a great deal into this just yet. The provision for taxes as a proportion of sales looks frightening at first, but it is not meaningful because Amazon is making advance provision for taxes, which distort this ratio for the half-year.

The two numbers that should most worry Amazon shareholders are working capital and capital expenditure.

I define working capital as all operating cashflows less net income and depreciation. I therefore capture the effects not just of Amazon's core working capital cycle (inventories, payables and receivables) but also its stock compensation arrangements, which help it to keep cash expenses down.

In the first half of 2014, working capital has constricted more than I expected, reducing Amazon's operating cash flows by more than $3.7bn. This is to some extent an artifact of a half-yearly reading; for the same period in 2013, Amazon's working capital arrangements had sucked out $3bn from cash profits. It's perhaps more meaningful to consider the trailing twelve months ending June 30, 2014. In this period, Amazon's working capital cycle has produced about $1.2bn of positive cashflow, but this was still a sharp reduction (of nearly 40%) compared to the $2bn it generated a year ago. When you take into account the fact that sales have actually increased in this period, you can see immediately that the working capital cycle is spinning a lot slower than I expected, which is bad news for shareholders.

The working capital number however pales into insignificance compared to the outturn on capital intensity. I define capital expenditure to include assets acquired on capital and finance leases, since this has an identical effect on future cashflow to assets acquired on-balance sheet. This capital expenditure relative to sales (or what I call Amazon's capital intensity) has increased to 11.2% (from 8.5% in the same period in 2013). Even if we ignore capital and finance leases and go by Amazon's own definitions, capital expenditure has still increased to more than 6% in the first half of 2014 (compared to 4.8% in the same period in 2013). In absolute terms, Amazon's capital expenditure program has increased by more than 60% compared to the same period in 2013, without a proportionate increase in sales. This should be a source of considerable anxiety to shareholders: all else equal, the higher the capital intensity, the greater the drain on Amazon's free cashflow.

Based on these outturn numbers for the first half of 2014, I feel justified in retaining my assumptions on sales growth, profitability and taxes; moderating down my assumptions on depreciation (from 47.5% to 45% of net fixed assets, which produces a charge for 2014 in line with the half year result); turning down my assumptions for working capital efficiency (from 3% to 1.6% of net sales, so generating cashflow of about $1.5bn for the whole of 2014); and dialing up my expectations of capital intensity from 4.6% to about 6% for 2014 and beyond.

This produces a valuation of around $60 a share for Amazon, which would suggest that Amazon is overvalued by nearly 450%!

Some readers will argue that extending ratios from half-year results out indefinitely into the future is excessively pessimistic. For instance, capital expenditure may be higher in the first half of 2014, but could still decline thereafter as Amazon's "investment cycle" comes to an end. Working capital may have been tighter in the first half of the year, but could still loosen by the winter holiday season.

If you like, you could characterize this last valuation as a pessimistic case. The valuation results presented in my first and this last article would then be the bookends of Amazon's valuation, ranging from a pessimistic reading of $60 a share to an optimistic one of $195 a share. What should really concern Amazon investors is that wherever you fall on that spectrum of optimism or pessimism, you cannot avoid the conclusion that Amazon is hopelessly overvalued - by 60% in the optimistic case, and by 450% in the pessimistic one!

Some investors argue that valuation does not matter for Amazon - its status as a momentum stock will keep propelling it forward. If you're an investor who bought into Amazon at the start of this year, you will already be skeptical of this argument. The stock is down from an all time high of $408 at the start of 2014 to about $330 in August (a correction of nearly 20%).

Even if you're a long-term buy-and-hold investor, and you're confident that Amazon will eventually "grow into its valuation," you should reflect on what that is going to take. On my optimistic case, Amazon will not grow into a valuation of more than $230 a share in the next ten years. To 'grow into its current valuation of $330', Amazon would need to improve its EBITDA margin to over 9% while growing sales rapidly and keeping capital intensity constant, or slash capital intensity to a third of its current level while maintaining sales growth and profitability. The odds are overwhelmingly against this.

Why then do Amazon's investors tolerate such high share prices? This brings me to the final part of my series: Amazon's cost of capital. I don't believe Amazon's shareholders are making silly assumptions on sales, profitability or capital expenditure - it is a heavily analyzed stock, and even bullish Wall Street stockpushers have reputations to protect.

Instead, what I think has been happening in the past five years is that Amazon's investors are discounting future cash flows (implicitly or explicitly) at artificially low discount rates. All my valuation calculations in these articles have been done using a discount rate of 10%. But on my optimistic case for instance, you could derive Amazon's current valuation if you were discounting its cashflows at around 7.3%.

Is 7.3% a credible estimate of Amazon's long-term weighted average cost of capital? The short answer is no. It is a short-term distortion created by a low yield macro environment sustained by the Federal Reserve's quantitative easing program. When that program ends and yields begin to normalize, so will discount rates and Amazon's cost of capital will go up.

My estimate of a reasonable long-term cost of capital for Amazon is around 10%. I get this by starting with a long-term leverage assumption of 15%, against Amazon's current 3.2%; I believe Amazon will issue more debt in the future, but will probably never be a highly leveraged company.

Its current weighted average cost of debt is around 2.3%, although its long-term notes are trading at yields of 2.7% (or about 30 basis points more than 10 year Treasuries). I assume as yields normalize, Treasuries will rise to about 3.0%, and so a fair estimate of Amazon's long-term cost of debt is probably around 3.3% before tax, and around 2.2% net of tax at a marginal rate of 35%.

The long-term equity risk premium is around 6.3%, as estimated by Professor Damodaran in his 2014 paper Equity Risk Premiums - the 2014 edition. With the long-term (15-year) beta for Amazon against the S&P 500 index at 1.3, I estimate its cost of equity is likely to be c. 11.5%.

A leverage of 15%, cost of debt at 2.2% and cost of equity at 11.5% give me a weighted average cost of capital for Amazon of around 10%.

This to my mind is the clear and present danger staring Amazon's shareholders in the face. A sensible long-term 'buy and hold' investor today should not be willing to pay more than $195 a share for Amazon, and probably a lot less. It is currently trading at more than $330.

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.