In 1998 Tim Cook moved to Apple. The company was a breath away from bankruptcy, with $400 million of unsellable computers in the warehouse — nineteen warehouses of them. Electronics lose value fast, and every day those machines sat, they got lighter.
The first thing Cook did was not build a new product. It was to make that stock disappear.
He closed ten of the nineteen warehouses. With fewer warehouses there is nowhere to pile goods, so you can only buy less. He cut core suppliers from 100 to 24, and made the survivors move next to Apple's factories. Parts counted as the supplier's inventory until Apple took them; Apple paid for what it used.
That was not enough. Before the iMac launched in 1998 he did the thing the industry still tells as legend: he bought up $100 million of air freight in advance. Nobody in the world had seen what the machine looked like. He had already bought out its air corridor.
Why not ship by sea and save the money? Because Cook was doing different arithmetic. A month on the water and a month on the warehouse floor are the same month. He would rather burn freight than let goods sit.
In one year, Apple's inventory turnover went from 30 days to 2.
What does "30 days" mean? That from entering the warehouse to being sold, your goods sit for 30 days on average. Thirty days to two means the same money that used to turn 12 times a year now turns 180 times. The goods are the same, the margin is the same — but the same capital earns fifteen times the profit over a year.
And what is two days? The industry likes to compare it with McDonald's. McDonald's turns its burger inventory in 2.5 days; Apple, selling phones and computers, got to 2. A hardware company turning faster than a fast-food chain.
Cook summed it up in a line the business schools still quote:
"Inventory is fundamentally evil. You want to manage it like you're managing milk — milk has a shelf life, and inventory has a shelf life."
But the truly brutal part was not "hold no inventory". It was "not even fronting the money".
Apple collects on a phone 32 days after selling it on average, and pays its suppliers more than 115 days later. In and out, that gives Apple a cash conversion cycle of better than minus seventy days — the customer's money is already in Apple's pocket and Apple has not paid the supplier a cent. The supplier is effectively lending to it and paying for the privilege.
The same trick got inventory down to two days: push both the material and the money onto the supplier's side of the line.
A man who hated inventory, hated funding it, hated every idle dollar, took Apple from the edge of bankruptcy to a $4.6 trillion market value.
And then, in his last year before retirement, that same man changed his mind.
In 2026 Apple began to stockpile on purpose. Why?
Because the situation has reversed — and the thing rising hardest this round is memory.
Memory chips went from 10% of the bill of materials on an iPhone 18 Pro a year ago to 34% now, overtaking the processor for the first time as the most expensive component. Samsung's and SK Hynix's memory inventory has fallen below ten days of supply, and channel spot fill rates are under half — money alone no longer guarantees you can get the parts.
Once memory moved, peripheral material like screens was dragged up with it. Apple had to stock ahead: OLED panel inventory went from four weeks of cover to six.
On the July earnings call, Cook called this market "a hundred-year flood".
He squeezed inventory when goods were cheap, bulky and losing value by the day. He builds inventory now because material is getting more expensive, may run out, and is cheaper today than tomorrow.
The same two words, "inventory" — one time it should be less, one time it should be more.
So the answer was never "less inventory is always better".
Inventory turnover is not warehouse management in the end. It is a judgement call: is this pile of goods making money for you, or quietly burning it?
Material worth holding is a gain at six weeks of cover. Goods you cannot sell are a loss at one day.
On September 1, Cook stepped down as Apple's CEO. That line of his, "inventory is evil", was never really saying inventory is a crime. It was saying that every unit sitting in a warehouse is quietly charging you interest. The only question is whether you can afford the interest — and sometimes you cannot.
Next, I will work out the arithmetic of "how much should I order at a time" — the same money, so why does one company turn it 12 times a year and another 180?