Why "the market is up" is misleading
TL;DR — An index is an average, and averages hide dispersion. Month after month, the gap between the strongest and the weakest sector is around 11 percentage points at the median — in July 2026 it was 14.9. Anyone looking only at the index misses that movement, even though it is larger than the index's own.
The report is freely accessible, no registration required. Every figure in this article comes from the same data basis; the values for the latest full month can be traced in the report.
The average is not the story
"The market gained in July" is a sentence that can be true and still explain almost nothing. An index condenses hundreds of individual stocks into a single number. That number describes the middle — not the movement.
We measured how far the sectors actually diverge. The basis is 79 months from January 2020 to July 2026, measured across liquid US stocks, equally weighted and taken as the median per sector. Equally weighted means every stock counts the same, so a single heavyweight does not determine the picture.
The result is clearer than many would expect:
| Measure | Gap between best and worst sector |
|---|---|
| Lower quartile of months | 8.7 percentage points |
| Median | 10.7 percentage points |
| Upper quartile of months | 15.1 percentage points |
| Most extreme month | 35.4 percentage points |
In a typical month, then, roughly eleven percentage points separate the strongest sector from the weakest.
The real point becomes visible in a direct comparison. Taking all liquid stocks together, the overall market moved by 3.6 percentage points in a typical month — so the spread between sectors, at 10.7, was around three times as large. And that is no isolated case: in 76 of the 79 months examined, the sector spread was larger than the movement of the overall market. In 96 percent of all months, there was more movement in the dispersion than in the number that was reported.
All figures describe completed periods. Past performance is not a reliable indicator of future results.
Where this shows up in practice
- Comparing a portfolio against the index — A portfolio concentrated in a weak sector lags the index without the stock selection necessarily having been poor. Conversely, a lead may simply come from having been overweight the right sector.
- Reports about "the market" — Coverage almost always refers to the index. What happened across the breadth of the market is rarely mentioned alongside it.
- Looking back at individual months — A flat index month can have been a very eventful time for individual sectors. In July 2026, for instance, the overall market barely moved while almost 15 percentage points separated energy and technology.
- Comparing two periods — Two months with the same index return can have completely different sector pictures. Without the second layer they look identical.
- Putting your own positions in context — Without sector context it is unclear whether a price move was typical of the environment or specific to the stock. That is the most common reason a price move is read wrongly.
- Assessing fund results — A fund with a sector focus can only be judged sensibly if you know how that sector performed.
The second layer: how uniformly does a sector move?
The sector median answers how the middle of a sector performed. It says nothing about how closely the individual stocks sit together. That is precisely the second measure we report: the gap between the upper and the lower quartile of the individual stocks.
In July 2026 it looked like this:
| Sector | Dispersion within the sector |
|---|---|
| Technology | 26.5 pp |
| Materials | 12.3 pp |
| Utilities | 4.8 pp |
Among utilities the stocks moved closely together — there the sector figure describes almost every individual stock well. In technology, 26.5 percentage points lay between the upper and lower quartile. There the sector average is a very rough guide to any single stock.
That is why we show both numbers side by side. A sector return without a dispersion figure invites you to reason from the sector to the stock. Depending on the sector, that inference holds — or it does not.
Can anything be derived from the ranking?
The obvious follow-up question is this: if a sector led one month, does it lead the next one too? We recalculated that across all 79 months and measured the relationship between one month's ranking and the next month's.
The result is a weak but measurable relationship. In figures: the rank correlation is around 0.13. For context — a value of 1.0 would mean the ranking stayed exactly the same; 0 would mean it was entirely random. At 0.13, the previous month's order explains around two percent of the following month's rank differences.
In practice that means the previous month's ranking is a very weak signal and not a road map. Over two to three months the relationship disappears entirely in our data. Anyone reading the table as a prediction is over-interpreting it.
That is precisely why we present the ranking as a description of the past month and not as a forecast. The value of the report lies in context: seeing the environment a stock moved in, and whether a price move was typical for its sector or not.
How we calculate
So that the figures stay verifiable, here is the methodology in brief:
Equally weighted median instead of average
Every liquid stock in a sector counts the same, and we take the median rather than the mean. An average can be pulled by individual extreme values; the median describes the actual middle of the sector.
Full months only
The current month is left out. Otherwise partial months would be compared against full ones, and the ranking would drift over the course of a month without anything having changed in the prices.
A data filter that is necessary
Price databases contain retroactively adjusted prices — after a reverse stock split, for example. That data is correct, but unusable for a monthly return: individual series then jump by a factor of several hundred. Without a filter against this, the measured sector spread jumped to over 80,000 percentage points in our first run. We therefore cap the permissible price range within a month and discard stock-months that exceed it.
What does not count as a sector
Catch-all buckets such as "Unknown" and investment vehicles are left out. The price of an investment vehicle reflects other portfolios, not a business of its own — in a sector ranking that would be misleading. We likewise do not report sectors with fewer than five stocks in a month: a median over three values is not a dependable statement.
Why the liquidity tier is taken as of the time
Whether a stock counts as liquid is determined afresh for every month — using what was known back then, not what is known today. Otherwise a distorted picture would emerge: stocks that only became liquid later would appear retroactively in months in which they were in fact not tradable.
Frequently asked questions (FAQ)
Does high dispersion mean a sector is riskier?
High dispersion initially means only that the stocks within the sector performed differently. For placing an individual stock in context, it means the sector figure says little about it. Whether higher risk follows from that depends on the individual company and cannot be derived from the dispersion figure alone.
Why the median and not the average?
An average reacts strongly to individual extreme values. If one stock in a sector rises by 300 percent, that lifts the average noticeably even though the remaining stocks are unaffected. The median describes the middle and therefore stays closer to what held true for the majority of stocks.
How often does the sector ranking change?
Frequently. Across the 79 months we measured, the previous month's ranking is only a very weak indication for the following month. That is why we present the ranking as a description of the past month and not as an extrapolation.
Is this investment advice or a recommendation?
No. The report and this article describe how sector prices have developed in the past. No statement about future prices follows from that, and no recommendation to buy or sell individual stocks or sectors.
How does this differ from the sector analysis?
The sector analysis report evaluates the filings of institutional investors and shows where their capital is flowing. The sector rotation report shows how the prices of the sectors have developed. One report describes capital flows, the other price movements — both can point in the same direction or in different ones.
Why US stocks only?
The analysis rests on the mandatory filings to the US Securities and Exchange Commission and the price series attached to them. For this body of data, sector classification, trading volume and price history are available consistently and at the same quality. For other markets, coverage is currently uneven — a mixed analysis would show differences that stem from the data situation and not from the market.
Which stocks feed into the calculation?
Liquid US stocks above a minimum trading threshold, with a price above one US dollar and at least 15 trading days in the month in question. We do not report sectors with fewer than five stocks in a month, because a median over so few values is not dependable.
Get started
For every sector, the report shows the rank for the last full month, the return over one, three and twelve months, the price history of the past 24 months, and the dispersion within the sector. No registration, no cost.