Amazon’s $6bn Man
In last week’s article, the first in this series on Amazon’s Cash Flow Statement, I looked at the Cash from Operations section. This week, I look at Cash from Investing and in the next article, I will conclude by looking at Cash from Financing and what the Cash Flow Statement doesn’t tell you.
Cash from Investing is straightforward in principle – it looks at the amount of capex and cash spent on investments in other businesses, equity investments, associates and acquisitions – acquisition spend includes earnout payments on past deals. There are some subtleties here in that the value of the acquisition under IFRS only includes the cash spend (less cash in the acquired business). Paying a vendor in stock, for example, does not appear in the cash flow.
I shall cover what goes in the cash flow statement and what is omitted more in the next article, but it’s a common misconception that the Cash Flow Statement is a reliable indicator of the disposition of funds within a business; it isn’t. Moreover, there are subtle differences between IFRS and US GAAP which can further confuse. I intend to explain this more in an upcoming online course where I walk through a 10-K. Watch out for that in a future letter.
Talking of courses, at the end of this article, I shall flag our brand new Forensic Analysis Bootcamp – 8 sessions of 90 minutes over Zoom will cover all the tricks you need to become a better analysing financial statements. You can get on the priority list via the link at the end.
Amazon Q3 2022 TTM Cash Flow Statement
Source: Behind the Balance Sheet from Sentieo data
Cash from Investing
So last week we looked at the $39.7bn cash generated from operations and now we turn to the $39.4bn of cash spent on investments. Let’s just zoom into that part of the statement in the table below:
Amazon Q3 2022 TTM Cash from Investing
Source: Behind the Balance Sheet from Sentieo data
The first thing to note that in investing we have the purchase and sale of marketable securities. This is simply Amazon investing some of its cash pile in Treasuries to gain some interest income and to manage its cash reserves efficiently – this is common among the big tech stocks. But it doesn’t really add to our understanding of the business – it’s like moving part of the contents of your wallet to another pocket in case you are mugged (I do this when I travel). It’s easier to visualise what’s happening by taking this out of the analysis:
Amazon Q3 2022 TTM Cash from Investing Reframed
Source: Behind the Balance Sheet from Sentieo data
This is a better way of looking at the underlying economics – Amazon has actually spent $66bn on capex and has sold $6.6bn of assets, giving a net capex of $59.4bn vs $52.1bn in the prior period, up 14% year on year. In addition, it has spent $8bn on acquisitions. And these two data points are extremely important because they can help us understand if Amazon is deploying its capital efficiently.
Even to question Amazon is seen as heretical in some quarters, but I think it’s necessary to examine this for every stock and the purpose of this series is to look at how to do the analysis as much as examining Amazon, where I do not profess to have any analytical edge (although we receive such a high frequency of deliveries at home that my wife is on first name terms with the drivers).
At this stage, I generally like to have a look at the Fixed Assets note. This is one area where the IFRS treatment is much more helpful than US GAAP where fewer details of the movements are disclosed.
Fixed Assets
Fixed assets net have increased from $147.2bn one year prior to $177.2bn at Sep 30, 2022, an increase of $30bn. Depreciation for the 12 month period was $38.6bn, so I estimate the increase in gross assets at roughly $68.6bn. Note that the $7.9bn spent on acquisitions will likely also deliver an increase in PP&E as the acquired company’s assets are consolidated. Hence, these numbers can only be rough-cut.
Amazon Balance Sheet
Source: Behind the Balance Sheet from Sentieo data
Capex in the cash flow above was $66bn and again note that this will not tie up exactly with the increase in the balance sheet because of
1) Assets added in the balance sheet but not yet paid for at period-end
2) Assets added in the balance sheet as a result of acquisitions
3) Disposals on the same basis as 1 and 2 above
4) Assets acquired under finance lease
This makes the reconciliation of capex in the cash flow with asset movements in the balance sheet more complicated than most non-accountants would expect. Sorry, but it’s important to understand these intricacies of the cash flow statement as how companies invest is critical to shareholder returns – this is important stuff.
As we are looking at Fixed Assets, Amazon’s trend in Capex:Depreciation is interesting. The chart shows that it hit peak levels in 2021 before tailing off somewhat in the 12 months to September, 2022.
Amazon Capex: Depreciation
Source: Behind the Balance Sheet from Sentieo data
Part of the reason for the increase is higher capex, part lower depreciation. Here is the chart of depreciation as a percent of sales:
Amazon Depreciation:Sales
Source: Behind the Balance Sheet from Sentieo data
One quirk of Amazon’s depreciation policy is that it has revised the life of its servers, not once but twice.
Here is what Amazon said in its 2020 10-K:
“We review the useful lives of equipment on an ongoing basis, and effective January 1, 2020 we changed our estimate of the useful life for our servers from three years to four years. The longer useful life is due to continuous improvements in our hardware, software, and data center designs. The effect of this change in estimate for the year ended December 31, 2020, based on servers that were included in “Property and equipment, net” as of December 31, 2019 and those acquired during the year ended December 31, 2020, was a reduction in depreciation and amortization expense of $2.7 billion and an increase in net income of $2.0 billion,or $4.06 per basic share and $3.98 per diluted share.”
And here is their 2021 10-K update:
“…in Q4 2021 we completed a useful life study for our servers and networking equipment and are increasing the useful lives from four years to five years for servers and from five years to six years for networking equipment in January 2022, which, based on servers and networking equipment that are included in “Property and equipment, net” as of December 31, 2021, will have an anticipated impact to our 2022 operating income of $3.1 billion. We had previously increased the useful life of our servers from three years to four years in January 2020.”
So Amazon increased the useful life of servers from 3 years to 4 years to 5 years, with a reduction in depreciation of $2.7bn then $3.1bn. My question would be, did something miraculous happen in 2021 which they hadn’t expected when they did that review in 2020? You would expect that they would have thought this through quite carefully and if they thought the life was 5 years, they would have stated that at the time.
More likely I suspect, they felt uncomfortable about making a change which would have impacted operating profit by $6bn in one go as that might have attracted a lot of attention. To be fair, they are not alone – Microsoft has done the same thing and reported this in its 2021 10-K:
“ In July 2020, we completed an assessment of the useful lives of our server and network equipment and determined we should increase the estimated useful life of server equipment from three years to four years and increase the estimated useful life of network equipment from two years to four years. This change in accounting estimate was effective beginning fiscal year 2021. Based on the carrying amount of server and network equipment included in property and equipment, net as of June 30, 2020, the effect of this change in estimate for fiscal year 2021 was an increase in operating income of $2.7 billion and net income of $2.3 billion…….”
And then in its 2022 10-K, it said this:
“In July 2022, we completed an assessment of the useful lives of our server and network equipment. Due to investments in software that increased efficiencies in how we operate our server and network equipment, as well as advances in technology, we determined we should increase the estimated useful lives of both server and network equipment from four years to six years. This change in accounting estimate will be effective beginning fiscal year 2023. Based on the carrying amount of server and network equipment included in property and equipment, net as of June 30, 2022, it is estimated this change will increase our fiscal year 2023 operating income by $3.7 billion. We had previously increased the estimated useful lives of both server and network equipment in July 2020.”
So Microsoft has also increased profitability by $6bn in two years. It at least explained that there was a reason that the life was lengthened a second time – investments in software that improved efficiency. And it suggested that two years had elapsed between the reviews.
Amazon and Microsoft Depreciation Policies Compared
Source: Behind the Balance Sheet from 10-K Filings
What’s interesting about this comparison is that Microsoft used to be much more conservative than Amazon but is now less conservative. Microsoft used to be one of the cleanest large cap stocks when it came to accounting. No longer.
(Satya Nadella’s compensation was reported as over $300m in 2021. I haven’t checked and I don’t care. But if I were an institutional shareholder, and there was an eps growth factor in the calculation, I would I would go hmmmm).
You know what’s strange? I cannot recall the last time I saw a company saying that it was shortening the life of an asset because it had over-estimated its life. It may be that accountants are all honest, decent folk who take an overly cautious and prudent view of the world and then are pleasantly surprised when things don’t turn out that bad.
Or it may be that they are short of earnings growth.
Because the change in rate is not retroactive – the historical numbers are not adjusted so these changes boost profits and eps growth. Often there is a geared effect on earnings as the tax paid doesn’t change, although it looks in the case of Amazon and Microsoft that there is some impact on the deferred tax charge.
Expect more of this type of shenanigans as earnings become tougher to grow in the real world.
I should emphasise that I have no reason to believe that Amazon or Microsoft are doing anything sinister. A life of 2-3 years sounds short. But reported earnings growth is boosted by the change.
There are likely no moving parts in any of this kit and who would ever replace their iPhone or laptop after 3 years because the phone wont synch to the desktop or the laptop has started to run slow? (Guess who).
Presumably big tech has top engineers working on their own servers to make them last longer and lengthen their life – don’t worry about the customers, guys.
The title of this piece is The $6bn Man which I think would be appropriate for Amazon CFO, Brian Olsavsky, who has conjured that out of thin air. It alludes to the TV show, The Six Million Dollar Man. Lee Majors played former astronaut, Steve Austin, who after an accident is rebuilt with bionic parts, giving him superhuman strength, speed and vision.
(Had I been writing about Microsoft, I would have liked to entitle it The $6bn Woman, but the spin off TV series was The Bionic Woman).
The Six Million Dollar Man
Source: IMDB (an Amazon company!)
Conclusion
When we drill down further into how Amazon has invested this cash, we will see that part of the discrepancy between the movement in fixed assets in the balance sheet and the capex in the cash flow statement is the result of leasing assets – Amazon does a lot of this. But as this is critical to the calculation of free cash flow, which Amazon presents in three different formats, I am going to cover that next week, as I think this article has been long enough.
I have given some clues as to my approach to the Cash from Investing section here and will return to the Cash from Financing section and calculating Free Cash Flow next time.
Paying subscribers can read on for an interesting analysis of Amazon by Will Nutting in his bi-weekly publication, aimed at institutional investors.
Will’s product is not cheap – it’s a multiple of what I charge, but he is offering actionable ideas and a real life portfolio. He described me in his missive last Tuesday as “a grumpy Scotsman”. Harsh, but at least half true.
And a reminder about our upcoming new course:
If you want to improve your analytical skills and be able to analyse financial statements like a professional, I am launching the first cohort of my Forensic Analysis course in February. You can pre-register here for early notification. If you have already emailed me, you are already on the list.
There are limited spaces and I am offering this first cohort a reduced rate so I would encourage those interested to sign up now. The course is based on my Forensic Accounting Course which has had 450+ students from some of the largest and most successful funds in the world - names like Schroders, Wellington, Baillie Gifford and many more.