Using systems theory, i think margin is a buffer for profitability that (to your point) allows for shock absorbing (as all buffers do.) Higher margins are also a risk of losing market share to a price-competitor, and you can suggest that market share is also a buffer for consumer engagement -- the ability of a firm to successfully raise prices (in the short term) is proportional to its market share.

Does the buffer of market share outweigh the buffer of margin?

Interestingly, the "what companies are valued on" is an increasingly surprising thing when you think about how real estate development works. Properties are valued on a multiple of their Net Operating Income, which creates a scenario where capital improvements that have low ROI in terms of NOI are actually high ROI when you include the effect on equity (future sale price.) In the current environment of 4% cap rates (NOI / Sale price) each additional dollar of income increases the value of the property by $26 (one dollar for the income and 25 dollars for the sale price bump.) This is why every landlord is very incentivized to borrow as much as possible to improve property as much as possible to raise rents as high as the market will reliably bear (vacancy kills your NOI and scares off acquirers) and then sell the property.

There's also an asymmetry of information and of control. In a downturn, Bezos has full visibility into where all of Amazon's COGS is going, and a number of levers he can turn to axe some of that spending, reduce future investment, or shed unprofitable business lines. Wall Street has virtually zero of that information (they get the consolidated financial statements, but none of the detailed breakdown) and zero of that leverage. Therefore, it's significantly less risky to Bezos to run on razor-thin margins than it is to an Amazon shareholder.

Same reason that founding a startup is significantly less risky for a world expert in its problem domain than for a random investor in that startup. The founder has full knowledge of the domain and a lot of levers they can pull to adjust strategy; the investor has none of that. That's also why many VCs require terms to reduce their risk (liquidation preferences, participation, anti-dilution, warrants) or increase their control (board seats, voting rights, information rights).

> when the company isn't doing well and investors are scattering to safety and liquidity is nowhere to be found a higher margin gives your company a buffer when you have to cut

And how has that worked out for 440 companies that used to be in the F500 that no longer are there?

> In a downturn, a slight decrease in margins across the board will push a company on the razor's edge from profitability to unprofitability.

Except the reason their margins are so low is because they keep feeding capital into the business to fund new projects.

> Of course when revenues are growing like crazy and the economy is gangbusters it sounds great to cut margins for absolute growth but that's a booming economy. Things are very different in a downturn.

Yup, you're right, things are very different in a downturn. Great listen about how NetSuite navigated downturns: https://a16z.com/2015/05/15/a16z-podcast-why-saas-revenue-is...

This seems like a dilemma between a superior strategy, but only if executed with discipline.

In reality the majority of businesses which run on large margins get lazy, allocate capital ineffectively, and underinvest in R&D.

Thin margins force you be lean. Thick margins afford you the opportunity (and risk) not to be.

Amazon retail has gross margins of 37% and net margin of, lets say 3.7%.

That means it cost it 63% to buy products(on credit like terms), and 33.3% to fullfil orders.

So from a risk terms we can say their margin is equivalent to 3.7/33.3 = about 10%. Which isn't bad.

Furthermore, a lot of those costs are variable costs, like shipping, which go down if demand drops.

Amazon has massive control over their margins. Therefore in a downturn, they just turn a dial and are profitable again.

They're just choosing to reinvest profits into growth, and can stop doing that whenever they want.

I think we mostly agree, but not about how much risk they're exposing themselves to in a downturn.

ANd what better what to obfuscate actual company footing - and real power than to prevent people from looking at, analyzing or even knowing what level of error margins you're capable of floating in a downturn.

What better way to additionally hedge against that than to have a many-billion-dollar contract for hosting with the government.

Exactly. Your comment is the most astute so far in this thread.

Amazon is a cash cow that simply feeds a relatively decentralized innovation engine. Moreover, the strategic moves they make (intentionally keeping GAAP margins thin, not separately disclosing AWS in the financial statements for years, etc.) should always be assumed to come from multiple, well-considered motivations.

Someone says, "It's just customer-obsession." Yes, that's the simple myth/story they want you to believe.

That is of course looking only at Net margins. Assuming they are actually on thin margins from there Gross Margins which is still a respectable 30+%. Amazon does go into very deep to try and hide their net margins or competitive advantage. They continue to invest and optimise their operation.

Sure but who will have lower margins than Amazon in a downturn; also higher margins also means more competition from people with lower margins so I can’t really agree with your analysis.