Jeff Bezos was a master at GAAP accounting manipulation back when Elon Musk thought that “GAAP” was a clothing store chain. AMZN’s numbers are just as manipulated as Tesla’s. But the difference between Jeff Bezos and Elon Musk is that, whereas Musk is a pure caveman with his fraud, Bezos is clever about disguising and hiding the accounting manipulation. About six years ago I spent a considerable amount of time deep-diving into AMZN’s financials going back to 2004, which is when AMZN’s business really began to takeoff. After about two weeks of tedious but intensive study of the footnotes in 10Q’s and K’s, I pieced together a lot of the GAAP manipulation tools embedded in AMZN’s financials.

I will note that in 2018 Amazon denied a request from the SEC for more information about the Prime business, including disclosing in its financials the amount of sales attributable to Prime members. I’m certain Bezos rejected this request because Prime is a money-losing proposition and does not want to provide the evidence of that by breaking out the numbers. I recall sometime around 2013 or 2014 Bezos admitted in an interview with Bloomberg that Prime lost a couple billion per year.

AMZN pulled a lot of of the usual GAAP tricks to generate this quarter’s net income “beat.” Bezos slashed marketing expenses by a considerable amount as a percentage of revenue, as the marketing expense was essentially flat vs Q2 2019. Historically the marketing expense has grown YoY at a healthy rate. He may have just figured out a way to justify capitalizing some that expense – i.e. throwing some amount of the marketing expense into an asset account and amortizing it over time. This would reduce the amount of marketing expense shown in the income statement, thereby increasing operating and net income. Too be sure during Q2 a lot of companies cut back on web-based advertising, but if this was the case with AMZN, the cost-improvement is one-time, non-recurring.

Though AMZN reported EPS of $10.50 vs. $5.32 in Q2/19, several red flags for me point to the improbability of net income nearly doubling YoY. The operating income margin in the North America product segment (e-commerce + whole foods + sundry other small businesses) declined again to 3.7% from 4.1% in Q2/19. For the first time he showed a tiny operating profit in the International e-commerce business. I’m certain there were accounting games to accomplish this but I can’t prove it with just the publicly available numbers.

AWS (the cloud business) continues to experience slowing sales growth and declining margins. AWS contributed to 59% of the Q2 operating income but just 12.1% of the total revenues. And the percentage of revenues represented by AWS sales declined.

AMZN’s overall operating margin was 6.5% but the Products (online + WF) operating margin was just 3.1% vs 4.9% in Q2/19. This decline is attributable I believe to declining margins in the Whole Foods business. Again, AMZN offers fat discount specials to Prime members on many products at WF, which drives sales growth at the expense of profitability. Unfortunately, AMZN does not break out the sales and income attributable to the WF business – yet another layer of opacity on AMZN’s financials. I predicted when AMZN acquired WF in mid-2017 (at the time WF was 5% operating margin business being folded into a 3% operating margin business) that Bezos would drive margins lower at WF in an effort to generate revenue growth.

The cost of fulfillment rose – again – to 26% of product sales vs 25.6% last year. The Company generates sales by subsidizing the selling price of online products with 2-day free
delivery for Prime members. This is a money losing proposition and it enables predatory
pricing to drive out competition. Bezos is being grilled by Congress about the possible use of predatory pricing strategies to drive out competition, along other anti-trust issues. Rest soundly that this is nothing more than political theatre and nothing will be done to curtail AMZN’s effort to put the competition out of business.

AMZN’s debt increased again to $33 billion (41%) in Q2 but the Company is not using the funds to buyback shares. If the business really is generating free cash flow, why issue more debt? AMZN has to issue debt from time to time to fund cash needs. Without going into the complicated calculus here, AMZN’s free cash flow claim is an accounting mirage. At the end of Q2 2012, AMZN had zero debt. It had $24 billion in debt after closing the WF’s deal. Now it has $33 billion in long term debt. To my knowledge, unlike most other big companies that issue debt for the sole purpose of buying back shares, AMZN has rarely if ever repurchased shares. This is because it needs the debt funding to cover expenses.

Finally, AMZN used to disclose the amount of cash it spends every quarter for operating and finance leases plus that amount cash used to acquire PP&E under operating and finance leases at the bottom of the State of Cash Flows. No more. Now it discloses this information in the footnotes in a section titled “Supplemental Cash Flow Information.” This may sound trivial but the cash used for the PP&E purchases is not included in Bezos’ definition of free cash flow. In addition, very few analysts and investors ever bother to look at the footnotes.

Bottom line: If I gross up the the first 6 months of 2020 operating income and add a couple billion for growth, I get full-year estimated operating income of $20 billion. This stock is trading at 80x estimated operating income for a business that generates a 6.5% operating margin and said margin declines almost every quarter. 88% of the business model generates just a 3% margin and that margin is declining. Right now it doesn’t matter. The stock algos, Chinese retail gunslingers and Robinhood idiots will chase anything that moves. You make a dead skunk carcass move and the Robinhood morons will chase it.