You think you own 500 companies. You own nine.

Inflation is not 2 percent. It is 10.

Statistics agencies measure inflation with a basket of goods. We measure it with gold, because gold is the only money nobody can print more of. Since January 1971 the dollar price of gold has risen by 8.8 percent a year. That is how much the dollar has lost against a fixed yardstick.

But gold is not entirely fixed. Every year new gold is mined, roughly 1.5 percent of everything already above ground. Think of it as a cake that grows by 1.5 percent a year. If your slice still gets 8.8 percent more expensive every year, it is not the cake that has become more valuable. It is the money that has become less valuable. Add the two together and we get the real inflation. We call it true-flation.

In kronor the bar is 11.6 percent, because the krona has also lost against the dollar since 1971. Anything growing slower than that loses purchasing power, even if it looks like a gain on the account statement.

Two of five indices clear the bar. Neither is Swedish.

We took the five indices Swedish savers see most often: the S&P 500, Nasdaq 100, OMXS30, the whole Stockholm exchange (OMXSPI) and Germany's DAX. All calculated in dollars so they can be compared, ten years back, excluding dividends.

The Stockholm exchange has returned 6 to 7 percent a year in dollars. That sounds decent. Measured against true-flation it is a loss of around 4 percent a year, for ten years. Measured in gold, OMXSPI has lost 41 percent of its value since 2016.

The index does not tell you what the companies do. Most of them fall short.

That the S&P 500 returned 13.6 percent a year does not mean the companies in the S&P 500 returned 13.6 percent. We took every company that sits in the indices today and was listed ten years ago, 837 companies, and asked: did you beat 10.3 percent a year?

In the S&P 500, 208 of 465 companies answer yes. That is 45 percent. More than half the companies in the world's strongest index have therefore lost purchasing power. On the Stockholm exchange, 21 percent of companies clear the bar. In the DAX, 35 percent.

But look at the right-hand bar for each index. The 208 American companies that make it weigh 77 percent of the index. The 48 Swedish ones weigh 46 percent. That difference explains how an index can clear a bar that most of its companies miss.

Nine companies account for half the return.

We calculated how much each company contributed to the S&P 500's return over the past ten years, based on how much it weighed when the period began. Then we sorted. Nine companies are enough to reach half. Nvidia alone accounts for 11 percent of the entire index's return, from a starting weight of 0.2 percent.

What happens to the index if we take them out? We rebuilt each index with today's members and their weights ten years ago, and then removed the companies that contributed the most. It is an experiment, not a real index. But it shows how thin the margin is.

The index fund buys more of what has already risen.

This is how a market-cap-weighted index works. The bigger the company, the larger its share of the index. When a company rises it gets bigger, and the index automatically owns more of it. When a company falls the index owns less of it. The index never sells the winners and never buys the losers. It does exactly the opposite.

The ten largest contributors in the S&P 500 weighed 14 percent ten years ago. Today they weigh 43 percent. In the Nasdaq 100 the same group went from 50 to 74 percent. Whoever buys an index fund today is therefore buying the winners, after they have won, at the size they reached by winning.

This is not a flaw in the index fund. It is how it is built. But it is not the picture most people have. "Broad diversification across 500 companies" describes the list, not the money. The money sits in ten.

The Stockholm exchange: three companies and a long tail.

Sweden is the clearest picture of all this. Of 358 companies on the Stockholm exchange, 226 were listed ten years ago. 48 of them beat the bar. 91 of them, four in ten, have fallen in dollar terms over ten years. Not fallen against gold. Fallen.

Three companies account for half of the whole exchange's return: AstraZeneca, ABB and Investor. AstraZeneca does it with 9.5 percent a year, below the bar, purely by virtue of being the largest.

What beats money is a technology wave, not the market.

The Nasdaq 100 is the only index where breadth holds. There, 70 percent of companies beat the bar. The median company returns 16.4 percent. The index holds up even without its ten biggest winners.

That is not surprising. The Nasdaq 100 is a technology index and we are living in the middle of a technological wave where capital and profits gather in exactly those companies. But read it correctly. It does not say that equities beat inflation. It says that one sector does, right now. The same sector that sits at the top of the S&P 500 and carries that index too. It is one and the same wave, counted twice.

The new perspective

If you take one thing away from this analysis, take these three.

This is not a call to sell the index fund. It is a description of what you actually own. Past performance is no guarantee of future returns. But it is a fairly good description of what has happened.