Case Study: Amazon's Reinvestment Machine

For years Amazon reported little profit because it reinvested nearly everything it earned into new capacity, logistics and businesses like AWS. That reinvestment, funded by cash flow and a growing moat, is why it compounded. The lesson is that low accounting profit can hide a superb capital-allocation machine.

Where this fits: this is a Five Criteria case study centred on capital allocation and Criterion 1, business quality. Read it alongside the hierarchy of capital allocation and the Five Criteria pillar.

A note on figures: specific numbers here are illustrative and used to explain the mechanics, not to report exact financials. The well-established facts — Amazon’s founding in 1994, its 1997 listing, the launch of AWS in 2006 and Prime in 2005 — are history.

The Puzzle That Confused a Generation of Investors

For much of its life, Amazon presented investors with a puzzle. Revenue grew relentlessly, the business was clearly expanding, yet reported profit stayed thin or vanished altogether. On a traditional price-to-earnings basis it looked absurdly expensive, or impossible to value at all.

Many sensible investors passed on it for exactly that reason. Judged by the tools in our price-to-earnings ratio guide, it failed the test year after year.

They were using the right tool on the wrong business. Understanding why is one of the most valuable lessons in modern investing.

Reinvestment, Not Profit, Was the Point

Amazon’s thin profits were not a sign of a weak business. They were a deliberate choice.

The cash the business generated was not being lost. It was being ploughed straight back into the company: new warehouses, delivery capacity, technology, lower prices and entirely new businesses. Management chose growth and durability over reported earnings.

This is the single most important idea in the case. Reported profit is what is left after a company has spent on growth. A company that spends aggressively on profitable growth will, by design, report less profit today — and may be worth far more tomorrow.

To see through this you have to look past net income. Our guides on free cash flow and how to read a cash flow statement explain why cash generation and where it is deployed tell you more than the bottom line.

The Flywheel

Amazon’s management described its strategy as a flywheel — a self-reinforcing loop that gets easier to turn as it speeds up.

In simple terms:

• Lower prices and wider selection attract more customers.
• More customers attract more third-party sellers to the marketplace.
• More sellers widen selection further and spread fixed costs.
• Greater scale lowers costs, which funds still lower prices.

Each turn of the wheel strengthens the next. This is scale economics working as an advantage, the same idea we explore in Nick Sleep’s scale economies shared. Crucially, Amazon chose to hand much of the benefit of scale back to customers as lower prices, rather than taking it as margin — which fed the flywheel instead of the income statement.

Why the flywheel is a moat

A rival cannot easily copy a flywheel that has been spinning for years. The scale, the logistics network, the customer habits and the seller ecosystem reinforce one another. By the time a competitor matches one element, the leader has extended the others.

That is a durable moat built from operations and scale rather than a single patent or brand.

AWS: The Engine Inside the Engine

The reinvestment story has a second act that is easy to miss: Amazon Web Services.

Launched in 2006, AWS sold computing power and storage to other companies. It grew into a large, high-margin, high-return business — quite different in economics from low-margin retail.

AWS mattered to the investment case in two ways.

• It changed the quality of the group. A high-return cloud business lifted the blended economics of the whole company and demonstrated that Amazon could build, not just buy, new franchises.
• It funded further reinvestment. The cash AWS threw off helped pay for growth elsewhere, reducing the need to rely on outside capital.

This is why looking at Amazon as a single blended number was so misleading. It was really several businesses with very different economics stitched together, each at a different stage. Our guide to different business models is a useful companion here.

A Five Criteria View

Here is how the reinvestment machine maps onto our framework. Verdicts describe the economics, not precise published figures.

Criterion	What we test	Amazon, the reinvestment years	Verdict
1. Great business	Returns on capital deployed	Core retail thin, AWS very high; cash reinvested at attractive rates	Pass, with nuance
2. Durable moat	A named, lasting advantage	Scale, logistics, Prime habit, AWS lead	Pass
3. Aligned management	Long-term, owner-minded capital allocation	Explicit long-term focus, reinvestment over reported profit	Pass
4. Sound balance sheet	Survive a bad year	Funded largely by operating cash and float, not fragile debt	Pass
5. Reasonable price	A sensible price	Almost never cheap on headline multiples	Hard to judge, often demanding

The management signal

Criterion 3 is where Amazon scored highest and where the lesson for investors is strongest. Management told shareholders plainly, from the first letter onwards, that it would optimise for long-term cash flow per share and would accept lumpy, low or absent reported profits to do it.

That kind of candid, consistent, long-term framing is exactly what we look for in how to identify great management. The willingness to be misunderstood by the market for years is rare and valuable.

The Crucial Caveat: Reinvestment Must Earn a Return

It would be dangerous to take the wrong lesson from Amazon. “Low profit plus heavy spending equals future greatness” is false. For most companies that spend heavily and report no profit, the spending is simply destroying value.

The test is the return on the reinvestment. Reinvesting a pound at a high rate of return compounds wealth. Reinvesting a pound at a poor rate of return burns it.

Ask these questions of any would-be Amazon:

• Is the cash being reinvested actually earning high returns? Look at return on invested capital in the parts of the business that are mature.
• Is there evidence the flywheel exists? Growing customers, repeat usage, widening selection, improving unit economics.
• Does the spending build a moat, or just chase revenue? Revenue bought with permanent losses is not a business.
• Is management honest about the economics? Beware companies that use “we are investing for growth” to excuse a model that simply does not work.

Many companies have invoked the Amazon playbook to justify years of losses that never turned into anything. The difference is whether a genuine high-return engine is being built underneath. For the warning signs, read quality of earnings and red flags in a business.

What You Can Take From This

You do not need to find the next Amazon. You need to avoid being fooled by accounting in either direction — missing a great reinvestment story, or being seduced by a bad one dressed up as one.

• Look at cash, not just reported profit. Where is the cash going, and what is it earning?
• Judge each business within a company separately. Blended multiples hide the truth when segments differ.
• Reward honest, long-term management. But demand evidence that reinvestment earns high returns.
• Respect price even so. A wonderful reinvestment machine bought at an extreme valuation can still be a poor investment, as overpaying for stocks explains.

Amazon rewired how investors think about profit. Used carefully, that rewiring sharpens your analysis. Used carelessly, it becomes an excuse for owning anything that loses money.

Frequently Asked Questions

Why did Amazon report such small profits for so long? By choice. Management reinvested almost all the cash the business generated into warehouses, logistics, technology and new businesses such as AWS, prioritising long-term growth and durability over short-term reported earnings.

What is AWS and why does it matter? Amazon Web Services, launched in 2006, is Amazon’s cloud-computing division. It became a high-margin, high-return business that generates substantial profit and cash, helping to fund reinvestment across the wider group and lifting the quality of the whole company.

Was Amazon ever cheap on a P/E basis? Rarely, and the P/E was often misleading because reported profit was deliberately suppressed by reinvestment. Free cash flow and the economics of each segment told a more useful story than the headline multiple.

Does heavy reinvestment always mean a great business? No. For most loss-making companies, spending simply destroys value. Reinvestment only compounds wealth when it earns high returns. The presence of spending tells you nothing; the return on that spending tells you everything.

What is the main lesson of the Amazon case? Low accounting profit is not the same as a poor business. When a company reinvests cash at high rates of return, suppressed earnings can be a sign of strength rather than weakness. The key test is whether the reinvestment earns good returns.

Your Next Step

Learn to read the numbers that actually matter: start with free cash flow and the hierarchy of capital allocation, then estimate value properly with the DCF intrinsic value calculator. For a contrasting study of a mature, cash-returning franchise, read our case study of Apple as a consumer business.

---

About the author: Cameron Hayes is a senior investor relations and finance professional who has worked in IR at Nasdaq Copenhagen-listed companies including Nilfisk and Maersk Drilling. He holds an MSc from Copenhagen Business School and a BCom from the University of Ottawa, and founded Gingernomics to teach individual investors the Five Criteria framework. More about Gingernomics.

The content on Gingernomics is for educational and informational purposes only and does not constitute financial advice. Always do your own research and consult a licensed financial advisor before making any investment decisions. Past performance is not indicative of future results.