Amazon's (NASDAQ:AMZN) third-quarter results make very difficult reading for its shareholders. True, there are some aspects of the report that optimists could clutch at: for the nine months ending September 2014, net sales have grown at 22% (with the category encompassing Amazon Web Services clocking in growth over 40%), and gross profit margin has expanded by over 2 percentage points. Moreover, Amazon's own favoured metric - free cash flow - has, according to its calculations, increased from $388m to $1.1bn in the 12 months preceding September 2014 (compared to the same period in the previous year).
But any financially literate reader of the report will see through these positives pretty quickly. Rapid sales growth and increasing gross margins are pointless if they do not translate into profits. And this third quarter shows up very clearly this central problem in the Amazon income statement: while gross margin has expanded, operating profit (or EBITDA) margins have contracted due a sharper rise in fulfilment, marketing and technology and content expenses, leading to an operating loss rather than a profit. EBITDA margins have fallen from 5.4% in 2013 to just under 5% in the first nine months of 2014, with the company declaring an operating loss of $412m for this period, compared to a profit of $745m for the same period last year.
The problems for Amazon don't end there. Let's scrutinize the company's claim above that free cash flow (a non-GAAP measure) has increased dramatically in these nine months. Amazon defines free cash flow as cash flow from operations less purchases of property and equipment, including software and website development. Even though Amazon reported a net loss in the twelve months ending September 2014, cash flow from operations has increased from just under $5bn to $5.7bn in the same period. Meanwhile, purchases of property and equipment have risen from $4.5bn in 2013 to $4.6bn in 2014. That is the basis of management's claim that free cash flow has increased.
However, consider for a moment that the "cash flow from operations" above of $5.7bn consisted principally of depreciation ($4.3bn) and stock-based compensation ($1.4bn). Neither of these fit very comfortably within a definition of free cash flow - certainly, neither is available for distribution to shareholders, as they are both intrinsic costs of carrying on the business, and do not represent distributable cash. With depreciation, this effect can be seen directly, as Amazon's purchases of property and equipment of $4.7bn more than offsets the depreciation charge. In steady state, you would expect depreciation charges to be offset by capital expenditure on maintenance of assets. And as for stock-based compensation, as I have argued before, this is an indefensible inclusion in any calculation of free cash flow to shareholders. Paying employees in stock rather than cash does not confer any economic benefit to existing shareholders - what they gain in terms of lowered cash expenses, they lose dollar-for-dollar in dilution of equity.
Perhaps this is a good juncture at which to discuss another curious transition highlighted by this quarter's report. For some years, Amazon's management (and media reports) have repeated the claim that Amazon has "a cash-generating operating cycle" (because it turns over its inventory quickly, and receives income from customers much faster than it has to pay its suppliers). I pointed out in my previous articles that I thought this working capital machine was spinning a lot slower, and this quarter's report illustrates the point well. In the twelve months ending September 2014, core working capital (changes in inventory, receivables and payables) was actually a drain of around $600m on operating cash. Since inventory turnover appears to have been constant between 2013 and 2014, this drain on cash seems to arise from the evaporation of liquidity provided previously by a favourable receivables and payments cycle. In plainer English, Amazon no longer appears able to generate a cash "float" by receiving money faster from customers than it pays its suppliers. Even if it did, though, I would question the logic of including working capital cash flow in computing free cash flow to shareholders - a bit like stock-based compensation, this is not distributable cash, but simply a promise to pay someone in the future rather than in cash now. Working capital "floats", however, do have some advantages - they can save the business money by reducing the need for debt to fund working capital or capital expenditure, and so increase profitability compared to an identical business with less "float". That said, even if Amazon benefited from this effect in the past, it appears no longer able to do the same.
The absence of any meaningful cash being pumped out from Amazon's operating heart and the constriction of its working capital arteries makes it even more alarming that it seems to be haemorrhaging cash on capital expenditure. Its accounting of expenditure on property and equipment shows a modest increase; but it does not include the acquisition of assets on long-term capital or build-to-suit leases. Almost all its warehouses, for instance, are acquired in this manner. In the twelve months to September 2014, Amazon spent over $4.2bn on these off-balance sheet purchases, compared to $2.1bn in the same period of 2013! It is true that the $4.2bn is not a cash cost in 2014 - the interest and principal repayment under these leases will be spread out as cash charges to the income statement over time. However, their effect on Amazon's future earnings (and distributable cash to shareholders) is identical to a cash spend of $4.2bn in 2014 financed by equivalent long-term borrowing.
Properly accounted, therefore, the computation of free cash flow to shareholders should take net income, add back depreciation but not stock-based compensation, and subtract all capital expenditure, including off-balance sheet acquisitions. If we do this for Amazon, we get what I think is a truer view of its free cash flow position: a worsening from a negative $3.7bn in the twelve months ending September 2013 to a negative $4.7bn in the same period ending September 2014.
Amazon optimists will point out that if its cash position was really deteriorating this badly, we should expect to see a large increase in debt. It is true that long-term debt on the balance sheet has actually reduced slightly (from $3.2bn at the end of December 2013 to $3.1bn at the end of September 2014). But other long-term liabilities have increased over this period from $4.2bn to $6.1bn, reflecting the large increase in its off-balance sheet commitments to repay capital and financing leases. Moreover, in September 2014, Amazon entered into a new revolving credit facility of up to $2bn. The need for a working capital credit facility is, in my view, driven directly by the company's increasingly tighter working capital situation.
So this is my conclusion from the third-quarter report: Amazon's contracting margins, constricting cash flow and rising long-term liabilities mean that even after the sell-off on Thursday night, the stock remains grossly overvalued. Meanwhile, the supportive macro policy environment created by quantitative easing and low interest rates is rapidly coming to an end. Shareholders have patiently waited for a pot of gold at the end of Amazon's rainbow; they might find, instead, the beginnings of a perfect storm.
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.