May 2026 Was a Tech Rally Disguised as a Bull Market. The Numbers Behind the Narrowness.

The Headline Numbers Were Misleading

If we looked only at the performance of the S&P 500 for May 2026 up 5.3% for the month and bringing its quarter-to-date gain to 16.3% we would have come away thinking this was a broad and powerful bull market with participation throughout the economy.what was happening in American companies

The information technology sector rose 19.76% in May doing essentially all of the heavy lifting for the index. In fact eight of the eleven S&P 500 sectors fell during the month. Energy declined 5.63%. Utilities fell 5.19%. Consumer staples fell 1.66%. The financial sector lost 1.06%. The real estate industrial and materials sectors also fell. Only Healthcare with aincrease of 2.38% and Consumer Discretionary with an increase of 2.13% joined Technology in positive territory. Two sectors kept company with technology. This is not a broad demonstration. That is a sector that does the work of eleven

Growth over stock divergence was equally stark. The Russell 3000 Growth Index rose 7.16% in May more than double the Russell 3000 Value Index's 2.94% gain. The Bloomberg Magnificent 7 Index which tracks Apple Microsoft Google Amazon Meta NVIDIA and Tesla added 6.64% as large-cap AI leadership reasserted itself after the volatility it sparked.the peak of the Iran war in April. The breadth by market cap size told a slightly less alarming story than the breadth by sector: The Russell 2000 which tracks small-cap stocks rose 4.37% and the Russell Top 50 the fifty largest stocks on the market gained 5.03%. So the rally was dominated by technology but it wasn't purely a mega-cap story either. The small caps also participated. A little less than thegiants

Why Technology Dominated

The May tech boom had three distinct catalysts and they reinforced each other rather than acting alone

The first was earnings. May marked the end of the first-quarter 2026 earnings season. Five of the Magnificent 7 companies Amazon Meta Alphabet Microsoft and Apple reported in late April and early May and all five posted cloud and AI-related revenue growth that exceeded expectations. NVIDIA had not yet reported when the month began it reported in late May but the anticipation alone added speculative energy to semiconductor names duringweeks

The second catalyst was oil oddly enough. The fragile ceasefire between the United States and Iran announced in late April held for most of May pushing Brent crude back from its peak of over $119 and easing the energy inflationary pressure that had been weighing on growth assets. Lower energy prices are a direct tailwind for the profits of technology companies specifically as their data centers are huge consumers of electricity. Cheaper energy is cheaper computing

The third was pure excitement. SpaceX filed for its IPO in late May announcing plans to raise $75 billion which would make it the largest IPO in history. That filing didn't directly affect most companies' profits but it added a wave of enthusiasm for transformative technology investments across the market. Momentum has a way of spreading to names that had nothing to do with the original news

What Breadth Actually Means: Two Ways to Weigh a Market

Before we get to the arithmetic I want to define the terms because amplitude is used loosely and actually refers to a few different things

The S&P 500 that you see quoted everywhere the one that returned 5.3% in May is weighted cap.Each company's influence on the index is proportional to its market capitalization that is its share price multiplied by its outstanding shares. A company worth $3 trillion moves the index about a thousand times more than a company worth $3 billion dollar move after dollar move. That's why a handful of huge tech companies can single-handedly carry the entire S&P 500 even when most of the 500 companies that comprise it arefalling. In reality the index does not follow 500 companies equally. It is following 500 companies that are overwhelmingly in the largest group

a equally weighted The index corrects for that by construction. All stocks get equal weighting typically rebalanced quarterly to about 0.2% each for a 500 stock index regardless of whether the company is worth $3 trillion or $30 billion. If you want to know how the average company really did regardless of size an equal-weighted index is a much better indicator than the cap-weighted number everyone cites in the news. The gap between a company's performanceA capitalization-weighted index and its equally weighted counterpart in the same month is one of the clearest signals of breadth there is. A small gap means a wide participation. A large gap means a few giants sweeping away the rest

the advance and descent line is a different tool that measures the same underlying question. Each trading day you count how many stocks in a given universe closed higher those that advanced and how many closed down those that moved down. Subtract those that fell from those that advanced and you get that day's net reading. Add each day's net reading to a running total and you get the leading fall line a long-running series that in a genuinely healthy bull market should trend upward along with the index.prices.When the price index continues to rise to new highs while the advanced decline line flattens or lowers that divergence is one of the oldest warning signs in technical analysis. It means that fewer and fewer stocks participate in each new high even when the headline figure looks good

A Worked Example: How an Index Can Rise While Most Stocks Fall

Here's the mechanism in its purest and most simplified form using a small hypothetical index that I'm building solely to illustrate the math. None of these are real companies or real returns. Call them Stock 1 through Stock 5

Let's assume our illustrative index has a total market cap of $1 trillion split into five parts. Stock 1 call it MegaTech is worth $600 billion a 60% weight. Stocks 2 through 5 are worth $100 billion each a 10% weight each. That concentration is not unreasonable. It's close to how weighted the actual top positions of the stock have become.S&P 500

Now let's say MegaTech has a spectacular month and earns a positive 20% return. The other four stocks have tough months: Stock 2 returns negative 5% Stock 3 returns negative 4% Stock 4 returns negative 3% and Stock 5 returns negative 2%

Values	Weight	Return	Weighted contribution
Action 1 Megatechnology	60%	+20%	+12.0
Value 2	10%	-5%	-0.5
Value 3	10%	-4%	-0.4
Value 4	10%	-3%	-0.3
Value 5	10%	-2%	-0.2

To get the cap-weighted index return multiply each stock's return by its weighting and add the five results. That gives 12.0 minus 0.5 minus 0.4 minus 0.3 minus 0.2 which equals positive 10.6%. Our illustrative cap-weighted index is up 10.6% for the month. Anyone looking solely at the index's performance would say it was a strong month and based on the numbercap-weighted it really was

Now calculate the equal-weighted version of the same five stocks. Instead of weighting by size simply average the five raw returns: 20 plus minus 5 plus minus 4 plus minus 3 plus minus 2 equals 6. Divide that 6 by 5 and the equal-weighted return will result in a positive 1.2%. It's already a very different picture an index that's up just over a percentage point instead of double digits built from the exact same five stocks inexactly the same month

Then look at the median the return of the middle stock once you rank the five from worst to best. In order the returns are negative 5 negative 4 negative 3 negative 2 and positive 20.The median value the third is negative 3%. The median stock in our illustrative index fell 3% in a month and the cap-weighted index rose 10.6%. Four of the five stocks lost money. Only one gained and it gained so much that it dragged the entire index into strongly positive territory almost by itself

That gap a cap-weighted index that's up 10.6% an equal-weighted version of the same stock that's up just 1.2% and an average stock return of negative -3% is exactly the shape of what happened in the actual S&P 500 in May 2026 simply compressed from 500 stocks to five and expanded for clarity. The actual May 2026 was less extreme than my illustrative numbers as eight ofEleven sectors fell rather than four of five individual stocks but the mechanism that produces the gap between the index holder and the typical company experience is identical. The leading decline line for this illustrative index would also have been unambiguous for the day: one advance four declines for a net reading of -3.A month constructed entirely of days like that would produce a leading decline line pointing firmly downward even as the index itself was setting new highs

The Bond Market's Warning

While stock investors celebrated technology's performance in May the bond market told a more uncomfortable story

The 10-year Treasury yield ended May at 4.45% and the 2-year at 3.98% leaving the yield curve upward sloping. But the move was driven by fear of inflation not confidence in growth and that distinction is very important. The April CPI rose 0.6% month over month and 3.8% year over year the highest reading since May 2023 and energy was responsible for more than40% increase.The April PCE report told the same story: headline PCE at 3.8% core at 3.3%.In fact real average hourly wages declined as inflation outpaced wage growth for the first time in three years.The personal savings rate fell to 2.6% its lowest level since mid-2022

The Consumer Squeeze Underneath the Rally

The consumer is absorbing pressure from three directions at once. Energy prices remain elevated even after the ceasefire pushed them back. Tariff shifting is still working its way through the system. Real wages are eroding for the first time since the 2022-2023 inflation peak

Stock markets are pricing in the strength of corporate earnings. Bond markets are pricing in the persistence of inflation. That divergence between the two signals is one of the defining tensions of 2026 and May's narrow tech-led rally in no way resolves it. Let's look again at which sectors fell in May: energy utilities commodities real estate and financials. Those are precisely the sectors most sensitive to interest rate level risk. Their underperformance was not arandom noise. It was a signal about what the bond market's inflation reading implies for sectors that compete with fixed income for investors' dollars or depend on it for financing

What This Means for Portfolio Construction

A portfolio that simply tracked the S&P 500 index in May 2026 performed strongly. A 5.3% return in one month is a very good month by any historical standard. A portfolio that was underweight technology significantly underperformed the same benchmark. A portfolio that tilted toward value energy or dividend-paying sectors the traditional defensive stance in an inflationary environment also performed poorly

This is market concentration in its purest form. When a handful of companies drive the majority of an index's performance index investing still produces excellent results but sector tilts or individual factors even reasonable ones can vastly underperform the number everyone cites. Understanding the gap between index performance and what's really happening across the economy requires looking at sector attribution and breadth not just the top number. May 2026 is the clearest example of thatdistinction you can find

Case Study: The Late 1990s Narrowing Into the Dot Com Crash

The clearest historical precedent for a narrow tech-led rally like the one in May 2026 is the race to the dot-com peak in the early 2000s

Throughout 1998 and 1999 the S&P 500 and the Nasdaq Composite continued to reach new highs driven overwhelmingly by a relatively small group of technology and telecommunications companies. Cisco Systems then the company that sold the routers and switches that literally built the Internet's infrastructure briefly became the most valuable company in the world in March 2000 with a market capitalization of more than $500 billion ina time when that was an almost unthinkable figure. Microsoft Intel and a wave of new public Internet companies with barely any revenues took on the rest of the load. Many ordinary profitable companies outside that group remained stable or fell for a long period while the major indexes rose

Measures of market breadth caught the warning long before the crash. According to several widely cited accounts the early decline line for broad market stocks turned and remained trending lower long before the S&P 500 and Nasdaq reached their final highs in March 2000 exactly the kind of divergence I described two sections ago: Fewer and fewer stocks actually participate in each new index high. The index continued to rise thanks to a shrinking group.of leaders even as the underlying tape was quietly deteriorating

Then it broke. The Nasdaq Composite peaked just above 5,000 points in March 2000 and fell about 78% when it bottomed in October 2002. The S&P 500 also fell although much less as its damage was more concentrated in the technology names that had made all the momentum on the way up. Companies with real profits which had participated less in the mania generally held up better.than historic stocks that had provided none of the market breadth on the ups and all of the pain on the downs

I want to be careful here because it's tempting to read May 2026 as if it were 2000 and I don't think that's a responsible statement. The concentration in 1999 was based in part on companies with no profits and in some cases no real revenue model which is a genuinely different setup than the large-cap technology companies that today are seeing strong real cloud and AI revenue growth. Low breadth alone doesn't predict a decline. It's an entry not aforecast.What the dotcom period really teaches is more limited and more useful than a prediction: a growing cap-weighted index can mask a severely deteriorating average stock for a long time sometimes years before anyone outside the breadth of the data notices

Where This Model Breaks

I just dedicated an entire section to drawing a line from May 2026 to 1999. Let me argue against my own framework because it's easy to overstate the connection and intellectual honesty requires saying it outright

First cap weighting is not a distortion that needs to be corrected. It is in a real sense accurate. A $3 trillion company genuinely represents more economic value more employment more earnings power than a $30 billion one and weighting the index by market capitalization reflects that reality rather than obscuring it. An equal-weighted index is not the true market nor is one cap-weighted. It is a different lens with its own bias one thatIt gives a small illiquid company the same influence as a giant which is its own kind of distortion

Second a narrow breadth can persist for a long time without leading to a crisis. The early dip divergence in the case study above lasted for a year or more before the market actually broke and many periods of narrow breadth in market history never broke at all.once had it

Third some of the current concentration is structural and not a warning sign of anything specific. Passive investing has grown enormously since 2000 and the dollars flowing into the S&P 500 index funds mechanically buy more of the largest weightings reinforcing the cap-weighted concentration regardless of the underlying fundamentals. This is a different mechanism than the story mania of 1999 and may behave differently when it finally unfoldsif it develops

Fourth and this is the point mentioned earlier in this article that I think is underweight: Breadth was actually good in May 2026 but not in the same country. Emerging markets Korea specifically were participating broadly in a global rally driven by technology and AI. A narrow US sector story next to a broad international story is a very different setup than a narrow rally with nowhere else for the money to go.US amplitude data at this time this is probably where the error lies

How I Actually Read a Breadth Report

When a month like May 2026 comes around this is the order I actually work in

If the gap is wide as it clearly was in May 2026 I treat that gap itself as the story rather than the return of the headlines

From there I look at sector attribution which sectors rose and which fell because that tells me whether the divergence is a company an industry or something broader like growth versus value. Eight of eleven sectors fell while one sector gained almost 20% is about as narrow a signal as sector attribution can give. I also compare small caps to megacaps like the Russell 2000 versus Russell Top 50 numbers did here because that tells meIt tells whether the concentration is really about the size of the company or a specific industry

I don't use any of this to time a top. My honest read is that breadth divergence describes what's already happened it's not a reliable prediction of what will happen next and I've already been wrong once. A few years ago I read too much into a tight month and stayed in an overly defensive position for too long while the market kept going up anyway. In my experience what breadth data is really for is to make sure I understand what I have. If my portfolio passively follows the index aA month like May 2026 is a great month period and I shouldn't convince myself that I'm nervous about a number that's treating me well. If I hold active tilts value dividends small cap whatever the tilt breadth data is the way I explain why that tilt underperformed this particular month without concluding that the tilt itself was a mistake. These are two very different situations carrying the same 5.3% headline

The Bottom Line

May 2026 produced a really good return for the S&P 500 based on a really tight breadth. Technology gained almost 20% and took down eight other sectors that were losing money during the month. The gap between a cap-weighted index and an equally weighted version of the same stock and the gap between the index's return and the median stock's return are the clearest ways to see that split and the worked example above shows exactly how a single dominant stock can drag down an entire index.The narrowing of the late 1990s to the dot-com peak shows that this pattern can last a long time before resolving and also shows why the narrow breadth alone is not a forecast as many narrow periods simply widen again. Bond markets were pricing in the persistence of inflation at the same time that stock markets were celebrating the strength ofprofits and that tension underlies everything else in this article. My conclusion is that the S&P 500's 5.3% is completely real and completely true and it's also not the whole story and both things can be true about the same number at the same time