[hristov]

There are some issues with this explanation. The main issue is that the rules of accounting have a very good provision to take into account investing into the future. It is called capitalization.

Thus, if a company spends money to build or acquire a new asset, it is called capital spending and it is not subtracted from the profits. Thus, for example, if a company had a million dollars of profit and decided to spend these million dollars on a new fulfillment center, they could spend the money for their fulfillment center and still report a million dollars in profit.

So it is not quite clear-cut to say that Amazon's desire to build fulfillment centers around the world is costing them their profits. Those things should be capitalized and once they are capitalized they should not affect the profits. Amazon did in fact report significant capital spending (as one can see on their cash flow statement).

However, things are not that simple. Sometimes some expenses which are about building for the future and investing into new growth are not capitalized. This is the case because for some expenses the benefits are so uncertain and difficult to quantify that the SEC requires that they are reported as ordinary expenses instead of capital spending. These types of expenses tend to involve R&D and may include certain administrative expenses associated with growth initiatives.

Therefore, many companies that are trying to grow do report lower profits because they have those expenses that are associated with investment into future growth but are not capitalized. This may be the case for amazon. But it is a question to what extent it is the case for amazon. For example, they do capitalize software and website development for new products and websites. So one cannot simply say that they are showing losses because they are spending all the money on making great new products. But then again, they expense software development for existing products. So perhaps the losses are associated with new growth features that are built into existing software.

So all in all it is a big muddle and it is not at all clear whether amazon is an inherently highly profitable company that happens to be investing in the future, or they are wasting money, or their business model is just not that profitable.

[pmarca]

The accounting rules (GAAP) have only a loose correlation to how most modern large companies actually operate the levers of their businesses. Management teams of well-run companies spend very little time thinking about the formal financial statements.

Accounting bears the same relationship to actually running a business that the Efficient Market Hypothesis does to actually effectively investing -- which is to say, almost nothing.

[zero_intp]

While I agree that GAAP only vaguely represents reality, Management teams of all well run companies /DO/ think about formal financial statements.

These statements require thought or liability to potential jail time. In addition, the impact of these statements on financial markets (ie: stock price) is tremendous. Management damn well should be thinking about shareholder value.

[mistermann]

> liability to potential jail time

Are there any examples of this? It seems to me the SEC didn't make much use of all the archived emails (guaranteed to be in place due to SOX) during the 2008 banking crisis which makes me generally distrustful of these formal written rules.

[asmithmd1]

Yes, WorldCom leaps to mind

http://www.accounting-degree.org/scandals/

They misstated costs as investments and were able to show big profits - for a while.

[mistermann]

While Worldcom is one incident where one(?) person was punished, it is from 2002, and from that page, it was following that incident that they enacted SOX, which afaik hasn't been used since. Or if it has, the incidents have been few and far between. I will continue to believe the system is rigged.

[cynicalkane]

GAAP book value and GAAP earnings are definitely correlated with stronger returns for shareholders. Which seems to call into question the efficient market hypothesis, but says good things about GAAP.

[amitagrawal]

"Management teams of well-run companies spend very little time thinking about the formal financial statements."

Not true at all. Well run companies do spend a lot of time in preparing their financial statements. First, it is required by law, see - Dodd-Frank, Sarbanes-Oxley, etc.

Second, it can get well-run companies with good intentions into a lot of trouble with the SEC and investor lobbying groups.

[c2]

Sure, it's something the CFO thinks about, but it isn't really something that plays a factor into decisions about how to run the business for other C-level executives.

[chrisgd]

The majority of the article talks about free cash flow which takes capex into account. And capex does impact the p&l once it is built and being depreciated.

[fauigerzigerk]

That's all true, but Amazon is free cash flow positive. So it's not like they're making huge "losses" after deducting capex.

[fspeech]

The fcf for retailers can simply come from paying suppliers later than getting paid by customers. While a cheap source of capital (even this is doubtful), not exactly a winning strategy in the long run. The fcf will stop once revenue growth slows. As capital it is only cheap in the sense that you don't have to constantly raise equity for growth. Without earnings your equity base can't grow w/o getting more from shareholders.

[epa]

Only ignorant investors can get fooled by large account payables. It's not a source of capital..

[fspeech]

It is if it is stable enough (that is the negative working capital). In other settings (such as insurance) it is also known as float. Negative working capital is great if you can get it and keep it. The problem with the associated fcf is that the fcf is the derivative, literally. So if the working capital stops going more negative the associated fcf goes to zero.

[craytheon]

Free cash flow is falling in the last couple of years.

http://i.imgur.com/sJWam5L.png

[pbreit]

Can you explain how capital expenditures do not affect profits? Doesn't capitalization just mean that expenses are applied over time? They don't disappear, correct?

[hristov]

Capital expenses do not affect profits for the reporting period they happen in. It is true that after that reporting period there is a depreciation cost applied to take account of loss of value of an asset.

For example, if a company buys a distribution center, it will not expense the cost of the distribution center as an expense. But as time goes on it will expense a depreciation expense that accounts for the loss of value of the distribution center as the building gets older, less useful, etc. But buildings last for a long time, and a building with the land it is on never actually goes down to a value of zero. So not all expenses associated with the distribution center will be applied over time.

So capitalizing something definitely helps you get more income.

[PeterisP]

Capitalizing doesn't help you "get" more income, it simply allows you to write some of the same income in an earlier year.

[pbreit]

So all capital expenditures reduce income. That's what I thought. I've never understood why folks don't think "pay later" = "pay".

[mseebach]

Profit is how much your total assets have increased in value. Cash is an asset and a warehouse is an asset. If you spend $1m of your cash that would otherwise have been cash profits to build a warehouse worth $1m, your total assets remain the same and thus your profits remain the same.

But, unlike cash, a building doesn't retain its (dollar) value forever, it must be written off. I'm not sure how buildings are written off since they have a rather long "shelf" life, but laptops are generally written off over three year. So if you buy a $1500 laptop with cash in year 0, your profits in year zero are unaffected, but you must book a depreciation (a reduction in the value of your assets) worth $500 in years 1, 2 and 3. Hopefully you will, as with your warehouse, use your new asset to book profits in each of these years in excess of your write-offs.

[MaysonL]

Last quarter, Amazon had 834M of depreciation, and 281M of stock-based compensation: 1.1 billion of non-cash deductions from revenue. They made capital expenditures of 1.03 billion. They ended the quarter with about $100 million more cash than they started with.

[cperciva]

281M of stock-based compensation ... ended the quarter with about $100 million more cash than they started with.

This looks like a loss to me. They could have sold $281M of stock and gotten $281M in cash; instead they ended up $181M short of that.

[michaelt]

In general when you're a CEO or board member you'll make more money by granting yourself $x million in stock than you'll make by paying a dividend of $x million which has to be shared among all investors.

[encoderer]

Not to mention, there are tax benefits to stock based comp. First, you often have the choice of when to exercise the stock which can make a world of difference especially with AMT, and second, if you hold for a year you'll pay the long term cap gains rate. For lower income people (hehe in this case that would be people who make less than $400k) that's still 15% which is an incredible discount. For higher income folks it's now about 23% but, again, that represents a tidy discount against their marginal tax rate.

[dlgeek]

Bezos doesn't take any stock-based compensation and takes only a minimal salary (less than 90K/year).

He makes money by increasing the value of his existing stock.

[moca]

Yes. Capital expenditures don't disappear. You can buy a piece of land to build a warehouse. You can sell the land to get money back if needed, so you can't report the money for the land as loss.

[001sky]

Amazon has ~no profits before taking into account the ways that Capex (might) impact profits.