This article below is from my Seeking Alpha post earlier this week. I’ve studied AMZN’s financials and business model for several years. I’m probably one of the few analysts who bothers to scour the footnotes of AMZN’s financials. I was taught by the best at University of Chicago to start with the footnotes and work “up” when pulling apart GAAP financial statements. I can say with 100% certainty that the “Free Cash Flow” that Jeff Bezos promotes with ardent zeal is a fictional number, if not fraud. The SEC looks the other way. Suffice it to say that AMZN’s true trailing twelve month free cash flow based on strict GAAP is nearly negative $4 billion. I demonstrate this below.

Amazon Perfects the “Beat the Street” Game – Amazon (AMZN) reported 52 cents per share “earnings” on October 26th vs. the consensus 2 cent estimate after the market closed. The stock soared 7.8% after hours as hedge fund algos and retail daytraders chased the stock higher on the headline report. AMZN “walked” Street analysts’ estimates down to a number that was easy to “beat.” Ninety days ago the consensus estimate for Q3 was $1.09, with one estimate as high as $1.59. cents. By the time AMZN was about to report, the consensus estimate was two cents. This is how the game is played.

The graphic below from Yahoo Finance shows a 3-month timeline of this “walk-down” process for AMZN’s consensus earnings forecast for Q4 2017, Q1 2018 and the full-year 2017. The current estimates were again revised after the Company’s Q3 report (source: Yahoo.com/finance w/my edits):

Make no mistake: the company knowingly “guides” analysts down in order to engineer a “headline” surprise. The “beat the numbers” game is one of the many games connected with corporate earnings reports. That said, AMZN’s actual EPS in Q3 2017 was the same as Q3 2016 – zero EPS growth. Bear in mind that GAAP acquisition accounting is heavily at play here. Acquisition accounting enables a company to boost revenues and hide expenses.

[Note: All numbers are taken directly from AMZN’s Third Quarter 10-Q]

Here’s a fact that Wall Street or Bubblevision won’t report: in Q3 2016, AMZN’s GAAP tax rate was 47% vs 18% in Q3 2017. Anyone who has taken a basic accounting course knows that the GAAP tax rate is highly arbitrary and a major source of EPS manipulation. If AMZN had simply used a constant GAAP tax rate in Q3, its net income in Q3 would have declined to $200 million this year from $252 million in Q3 last year (remember these are GAAP earnings, not actual cash earnings). On this basis, AMZN’s EPS would have shown a drop from 53 cents last year to 41 cents this year. Anyone paying the current price of AMZN at a PE of 290 is likely ignorant of the fact that AMZN’s operating income is declining and its debt outstanding is increasing.

AMZN’s operating income plunged yr/yr for Q3 by $228 million, or nearly 40%. Operating income in its North American e-commerce business plunged $143 million, or 56%. AMZN’s e-commerce business lost $824 million on an operating business in Q3 (see p. 26 from the 10-Q linked above). YTD AMZN’s e-commerce business has lost nearly $1 billion). It likely would have been worse without Whole Foods numbers in mix. This is because, when AMZN acquired WFM, WFM’s operating margin was 4%. AMZN’s has been running near zero – it was 0.7% in Q3. Acquisition accounting, among other things, allows AMZN to present its numbers “as if,” meaning “as if” AMZN owned WFM since AMZN’s inception.

One of the primary reasons that AMZN’s operating margins decline continuously is the cost of fulfillment. “Fulfillment” is the cost of getting a product from the warehouse to the customer’s doorstep. In Q3 2016, AMZN’s fulfillment costs were 19.4% of product sales. By Q3 2017, it had jumped to 22.3%. Fulfillment is a cornerstone of AMZN’s e-commerce model. Offering free shipping to Prime members is a guaranteed money-loser. They are so efficient because they have practically perfected the art of warehousing. Along with a competent team of warehouse workers, their facilities are also equipped with state of the art rolling ladders and mobile platforms to help them get to and reach items faster and easier – head to this Platforms and Ladders link for more info on these kinds of warehouse solutions. Amazon has a lot of different warehouses, due to its huge amount of products on offer. They seem to constantly be adding more warehouses to their site by looking for steel buildings for sale. Those buildings are ideal for warehouses as they offer so much space to store the many products that will be moved around and leaving the facility in any given day. Amazon has a lot of these buildings and they’re all extremely well-organized, thanks to things like pallet racking systems, to make sure warehouse workers can do their jobs efficiently.

In general and on average, AMZN loses money on every e-commerce sale. AMZN’s e-commerce/consumer products operating margin will continue to decline because the Company is implementing an aggressive price-cut program at Whole Foods. This will drive the WFM business margins toward zero.

AMZN’s only source of operating income is the AWS (cloud services) business. The revenue growth rate from 2016 to 2017 for Q3 was 41%. This is down from the 55% growth rate that occurred year over year from Q3 2015 to Q3 2016. Part of this is a function of “scale.” As the business grows in overall size, the growth rate will tend to decline mathematically. But the AWS revenues are just 10% of AMZN’s total revenues.

Furthermore, AMZN’s AWS business is now under heavy attack. Google (NASDAQ:GOOG) (NASDAQ:GOOGL) and Cisco (CSCO) announced that they are teaming up to go after AWS’ cloud territory. More ominously for Amazon, Microsoft (MSFT) is quickly moving into and taking away AMZN’s market share in the commercial cloud space. Based on its FY Q1 numbers released Thursday, MSFT’s commercial cloud revenue annualized now exceeds AMZN’s AWS revenues annualized. AMZN historically has held the largest market share in cloud computing services. Given the new competition from dedicated tech companies, the profitability and growth of AMZN’s AWS business segment is at risk.

AMZN’s deceptive presentation of free cash flow – Every quarter AMZN presents an earnings slideshow, the first slide of which prominently shows trailing twelve month free cash flow. But this presentation of FCF is highly deceptive. On the first slide, AMZN shows its latest trailing twelve month FCF to be $8.050 billion. But that is a cherry-picked, non-GAAP derivation of actual free cash flow. Here’s AMZN’s actual GAAP FCF as derived from its Q3 10-Q (source: AMZN 10-Q, with my edits):

Free cash flow is technically defined as operating cash flow less capex expenditures and debt payments, the latter of which is negligible for AMZN – for now. Note the difference claimed to be $8.050 billion in “free cash flow” by Jeff Bezos and the negative $3.969 bullion actual GAAP FCF. Here’s the deal. Jeff Bezos conveniently omits the amount of cash AMZN spends to acquire property and equipment using capital leases and build-to-suit leases. To the extent that these expenditures are non-recurring, that presentation of FCF is valid. However, not only are AMZN’s expenditures under capital leases serially recursive, the payments increase every quarter and have been for several years. In 2014 AMZN’s full year cap lease expenditures was $4.9 billion. Thru Q3 2017, AMZN’s trailing twelve-month expenditure was $12 billion.

Furthermore, a “build-to-suit” property is built specifically for AMZN’s purposes. It likely is not easily sold re-leased for a next best use. Because of this, the lease functions as debt used to fund this capex. As such, the payments under build-to-suit leases should be treated as capex and not excluded from the derivation of free cash flow. Again, it’s an accounting sleight-of-hand employed by Bezos for the purposes of deception.

The use of capital leases to manipulate financials is not uncommon. However AMZN intentionally uses this financing techniques as mechanism to manipulate its numbers. Among other superficial accounting “benefits,” using capital leases rather than debt to fund expenditures enables keeps the appearance of debt off the balance sheet. It also allows AMZN to keep the cash used to fund capital leases out of the “Financing Activities” section of AMZN’s Statement of Cash Flows. AMZN is required to disclose the amount spent on cap leases, which it accomplishes in the footnotes. Very few analysts or investors bother to read the footnotes.

AMZN’s debt load – AMZN used $16 billion in near-junk bond rated debt to finance the Whole Foods acquisition. Its long term debt is now $24.7 billion. At the end of 2007, its long term debt was $1.2 billion. AMZN’s debt-load has grown by over 20x. However, at the end of Q3 2017, AMZN also had $18.8 billion in “other long-term liabilities.” This is almost entirely the capitalized leases used to fund property and equipment acquisitions. At the end of 2007, this number was $292 million. Use of cap leases has grown by a factor of 64x. Now, imagine if AMZN were forced by GAAP to include cap leases as part of its long term debt – not an unreasonable standard in this case. AMZN’s debt load would be $43.5 billion – nearly double the current disclosed level of debt.

See how this works? If AMZN were forced to consolidate cap leases into “long term debt,” its recent $16 billion bond deal would have been rated as non-investment grade – aka junk. The average cost of the $16 billion issued is 3.56%. If AMZN had been rated junk, it would have raised the cost of this deal by at least 100 basis points (1%) and likely more. Assuming an added cost of 1%, this would have added $160 million in interest expense. It might look like a smart move for Bezos to exploit GAAP accounting like this but it serves to pull the wool over the eyes of the investors who bought the bonds. This is because the true credit quality and ability to service the debt is significantly lower than that assumed by these investors.

The point here is that every facet of AMZN’s financials is highly misleading. AMZN is not what it appears to be. Yes, the stock has done remarkably well considering the ugly nature of the underlying truths. Note that AMZN did have a brush with insolvency in 2003-2004, but Warren Buffet bailed out AMZN by loading up on junk bonds Amazon had outstanding at the time. This was a temporary stay of execution that was followed up with the rapid inflation of the mid-2000’s credit and stock bubble, which enabled AMZN to refinance the junk bonds Buffet had bought. This gave AMZN plenty of cash to keep spending money to generate sales. AMZN also was bailed out by the bond market a couple of years ago, as it issued $3 billion in debt in 2012 and $5 billion of debt in 2014. If AMZN is truly generating free cash flow, why does it continuously have to issue debt to fund its operations?

Amazon has thus been given a free pass by the financial markets for most of its existence. Make no mistake, AMZN can do this only for as long as market bubbles inflate. If the current credit/stock bubble is in the process of deflating or has popped when it comes time for AMZN to start paying down its heavy debt load, including the capital leases, it’s highly likely that the market won’t enable AMZN to continue kicking the can down the road by refinancing the debt payments. AMZN clearly does not generate free cash flow that can be used to make the debt payment obligations. Thus, in this scenario, there’s a strong probability that AMZN would hit the wall, inconceivable as that may seem right now.

AMZN’s stock has had a remarkable run this year in defiance of the true underlying fundamentals (click to enlarge):

Amazon is a difficult stock to short because of its correlation with the overall stock market. However I’ve been able to scalp profits on an intra-day basis using near-money weekly puts. Anyone who is willing to manage a short position on a daily basis will eventually be rewarded. When AMZN surprised the market by missing its Q2 earnings, the stock sold off $140 from top to bottom over 2 months. If AMZN misses Q4 earnings the stock could, minimally, fill the gap in the graph above ($980) – a $160 decline using the closing price on November 21st.

If you are interested in short-sell ideas like AMZN, please visit this link: Short Seller’s Journal, where I offer a weekly newsletter that focuses on shorting the stock market.