The Magnificent Seven Stopped Leading and Something Healthier Happened

The Rotation

During the first half of 2026 several members of the group known as the Magnificent Seven including Meta Tesla and Microsoft lagged the broader market. Leadership shifted toward semiconductor and memory chip makers and toward financial industrial and other sectors that had been overlooked during years of concentrated technology leadership

Still the major indices finished the half near record levels. The gains simply came from a broader set of contributors

Why Breadth Matters

Market breadth measures how many stocks are involved in a move. When an index rises on the strength of a handful of huge companies the index can look healthy while most of its constituents do not

Narrow leadership creates fragility. If a small number of names generate profitability the index depends on those specific businesses continuing to perform and any disappointment is transmitted directly to the entire market. That was the structural concern during the concentrated years and it was legitimate

Expanding leadership reduces that dependence. When the financial industrial and semiconductor sectors contribute alongside the biggest technology names the index relies on more independent engines

A Worked Example: Measuring Breadth With Two Numbers

Breadth sounds like a vague quality of a market. It's arithmetic and you can calculate it from figures published every day. Here is an illustrative half-year calculation

Start with both versions of the index. Let's say the cap-weighted index returned 8 percent over the period and the equal-weighted version of the same index returned 11 percent

Those two numbers already answer the question. The cap-weighted index tells you what the money made. The equal-weighted index tells you what the typical company made. When a typical company beats the money the profits come from the middle of the market and not the top which is the definition of widening. A three-point difference in favor of an equal weight is a significantly large gain

Now let's remember again what the giants really did. Suppose the top ten names account for 38 percent of the index and returned 2 percent during the half. Their contribution is 0.38 times 2 which is equivalent to 0.76 percentage points

The index returned 8 points in total so everything else contributed 8 minus 0.76 or 7.24 points out of a weight of 62 percent. Divide and the rest of the market returned 7.24 divided by 0.62 which is approximately 11.7 percent

Megacaps gained 2 percent and the other 490 companies gained nearly 12 percent.That's the rotation described at the beginning of this article expressed as a number rather than a feeling and it required subtraction and division

Compare this with the restricted case to see why the distinction is important. Reverse the figures. If the largest ten had returned 18 percent and everything else 1.7 percent the index would have produced 0.38 times 18 plus 0.62 times 1.7 which is 6.84 plus 1.05 or about 7.9 percent. Almost the same headline. The same record on the chart and a market in which the average company went virtually nowhere

That's the whole argument for calculating breadth rather than reading the index. Two markets with opposite internal structures generate indistinguishable headlines

What Drove the Shift

High interest rates were the main mechanism. High rates weigh more on companies whose value depends on cash flows in the distant future because they are the ones that are discounted the most. The longest duration stocks are exactly those with high multiple growth

Finance works differently. Banks earn a spread between what they pay on deposits and what they earn on loans and securities and that spread often benefits from higher rates. Industrialists tend to follow physical economic activity rather than the mathematics of discount rates

The strength of semiconductors and memory reflected demand tied to AI infrastructure where memory in particular had become a bottleneck rather than a commodity

The duration point deserves a number because it is typically stated rather than shown. A company that is expected to deliver most of its cash flow within twenty years has a valuation dominated by the discount factor applied to those distant years and the discount factors compound. If the rate used to discount a twenty-year cash flow is moved from 7 percent to 9 percent its present value falls by about a third since 1.07 to the twentieth power is about3.87 while 1.09 to the 20th power is about 5.60. Do the same with a cash flow that arrives two years from now and the present value will fall less than 4 percent. Nothing changed in either business. The arithmetic of waiting did

The Earnings Underneath

The rotation was not purely a discount rate story. Corporate profits grew at a rate well above historical averages during the period which is the fundamental support that made the record levels defensible rather than speculative

The distinction is important. An index that rises with multiple expansion meaning investors pay more for the same earnings is a bet on sentiment. An index that rises along with earnings growth is a different proposition. The first half was characterized by significant earnings growth which is why the advance was maintained despite real obstacles such as conflicts volatile oil and persistent inflation

You can split any index move into those two parts without much effort and it's worth doing so before accepting anyone's characterization of a rally. Price equals earnings times multiple so the percentage change in price is roughly the change in earnings plus the change in multiple. An index that rose 8 percent while earnings grew 9 percent actually saw a slight contraction in its multiples meaning investors paid lessper dollar profit at the end of the period than at the beginning. This is almost the opposite of a speculative advance and is invisible in the index number alone

Case Study: The Breadth Divergence That Preceded 2000

The reason analysts see breadth is a specific episode and it's worth knowing in detail because it's the strongest evidence the measure has

Throughout 1998 and 1999 the major US indices continued to reach new highs. Below them the leading decline line which simply adds the number of stocks rising each day and subtracting those that fall had stopped confirming. It peaked in the spring of 1998 and declined for almost two years as the major indices rose to their March 2000 high

What that divergence described is exactly the reverse case in the previous example. A shrinking group of very large technology and telecom names led the index while mid-sized stocks were already in decline. In 1999 most stocks on the stock market were below their own moving averages;In one year the Nasdaq rose more than 80 percent

When the top finally came the damage was distributed accordingly. The Nasdaq fell about 78 percent from its March 2000 high to its October 2002 low. Many of the ordinary companies that had been falling quietly since 1998 did so much less because they had never participated in the final stretch

A more recent version occurred in 2021. The Nasdaq Composite peaked in November 2021 but by that time the average stocks in the index were already substantially below their own high and a large portion of the most speculative names had peaked in February. The index looked good for nine months while its components fell apart

Two episodes twenty years apart the same firm: the incumbent held by a shrinking group while participation deteriorated. That's the pattern that breadth analysis should detect and it's why a wide advance is actually better news than a narrow one at the same level of the index

Where Broadening Is Not Automatically Good News

Having made the case I want to be careful how far it extends because breadth analysis tends to be interpreted as clear-cut

The shift toward defense is a warning not a celebration. Breadth improves when money moves out of the leaders and into a broader set of names and that broader set is hugely important. The spread of money into the industrial and financial sector in the face of increased economic activity is one thing. The money spread toward utilities commodities and healthcare because investors are going on the defensive is also broadening and has often preceded weakness rather than strength. The 2026 case involves financials and semiconductors which is thehealthier version but the metric alone cannot distinguish them

The amplitude was improving in 2007. The measure is not a reliable timing tool in the positive direction. Divergences have preceded highs and there has also been broad participation shortly before big declines. It is much better evidence of fragility when it is absent than evidence of safety when it is present and those are not symmetrical statements

The superior performance in equal weight is due in part to size exposure. An equal-weighted index keeps a lot more money in smaller companies than a cap-weighted one. When small-cap companies rally for reasons that have nothing to do with participation equal weight does better and breadth appears to improve. One of what is celebrated as breadth is a size factor that carries a different label

Half a year is not a trend. Leadership is constantly rotating and constantly reversing. Interpreting a regime change in six months of relative performance among three mega-cap names is the kind of conclusion that seems obvious in retrospect and is close to a coin flip in advance

My own reading is that the 2026 extension is genuine and that its importance is being exaggerated not least because commentators have spent three years warning about the concentration and are happy to have something to point out. This is opinion rather than analysis

How to Measure It Yourself

Two techniques are useful. Compare an equal-weighted version of an index with the standard cap-weighted version. When equal-weighted outperforms the average stock is outperforming the giants and breadth is improving

Second track the percentage of constituents trading above their two hundred day moving average. A rising index with a falling percentage is the classic warning of a narrowing advance

How I Actually Use Breadth

I treat amplitude as a description rather than a signal and the distinction changes how I use it

The first thing I do with any index security is subtraction from the previous example. Cap-weighted returns minus equal-weighted returns in the same window. It costs nothing and converts a number that describes money to a number that describes companies. Most market commentary quietly mixes them up

Second when the percentage of voters above their two-hundred-day average falls as the index rises I stop treating the level of the index as informative. That's the 1999 setup and its value is not that it predicts a top but that it tells me that the headline has stopped summarizing the market

Third I look at which sectors are expanding not just whether the breadth improved. The share of the financial and industrial sectors along with semiconductors is a different message from the share of utilities and commodities and the aggregate breadth number is identical in both cases

Fourth I try not to update six-month data too much. Breadth is most useful in multi-year periods where the 1998-2000 divergence and the 2021 divergence are visible and persistent. The semi-annual readings are mostly noise disguised as structure

That's how I read it. It's a description of the method rather than a market call and none of this is investment advice

The Bottom Line

Leadership expanded from a handful of mega-caps to sectors benefiting from higher rates and real activity and it did so with underlying earnings growth. A track record built on broad participation is structurally stronger than one built on seven names

Arithmetic is what makes that statement testable and not rhetorical. An index that's up 8 percent with its top ten names up 2 percent implies that the other 490 companies got about 12. Reverse the two and you get a nearly identical headline of about 7.9 percent that describes a market where the average company went nowhere. The table can't tell you which one you're in.A subtraction can

Consider 1999 as the reason this all matters. The early decline line stopped confirming in the spring of 1998 while the indices rose for another two years and then the Nasdaq fell about 78 percent. Breadth is a much better way to tell you a market is fragile than it is to tell you it's safe and the 2026 broadening is real welcome and worth about six months of evidence