S&P 500 Best Quarter in Six Years: AI Rally Drives Historic Gains

What You'll Learn

• Why Q2 2026 became a standout quarter for US equity indexes.
• How July shifted leadership from AI momentum toward value and equal weight.
• Why strong earnings do not remove valuation, concentration or execution risk.
• Which index returns, breadth measures and AI-capex signals matter next.

Introduction

S&P 500 AI rally 2026 is not one uninterrupted technology boom. The second quarter delivered a powerful advance, with Nasdaq's Market Intelligence Desk saying the S&P 500 and Nasdaq-100 posted their best quarters since Q2 2020. The Philadelphia Semiconductor Index rose 87.8 percent in Q2, its best quarter since its 1994 inception. By July, that leadership had reversed sharply.

The Nasdaq July review reported a 20.6 percent monthly decline for the Philadelphia Semiconductor Index and a 6.6 percent fall for the Nasdaq-100. It also showed a different part of the market holding up. The S&P 500 Equal Weight Index gained 1.0 percent in July and the Dow Jones Industrial Average gained 0.6 percent. Large-cap value rose 3.8 percent while growth fell 5.9 percent.

That is a rotation, not a simple market collapse. The official S&P Dow Jones Indices page showed the S&P 500 at 7,641.16 on August 20, with an 11.62 percent price return year to date, a 1.89 percent quarter-to-date return and a 2.02 percent month-to-date return. The same page showed a July 31 one-month price return of -0.13 percent.

The numbers point to a more useful question than whether AI is a bubble. Which companies can turn capital spending into revenue, margins and cash flow, and how much of that outcome is already reflected in the index price? This article separates Q2 performance from July breadth, then weighs earnings, valuation and the next risks.

What Made Q2 2026 Such a Strong Quarter

The quarter began after a difficult first quarter shaped by the US-Iran conflict, trade concerns and sharp price swings. Nasdaq's July 1 review said the S&P 500 and Nasdaq-100 recorded their best quarter since Q2 2020. The rebound started with large-cap growth and hyperscalers, then widened into small-cap, microcap, equal-weight and value benchmarks by quarter end.

Semiconductors were the clearest expression of the trade. The Philadelphia Semiconductor Index, often called the SOX Index, gained 87.8 percent during Q2 according to Nasdaq. Memory and chip companies became a larger part of the AI investment story as buyers looked beyond the dominant graphics-processing-unit suppliers.

The market also received help from a change in geopolitical pricing. Nasdaq said crude oil reversed its earlier Iran-conflict rise after the United States and Iran moved toward a negotiated peace. Lower oil prices reduced one immediate inflation concern and supported a relief move in risk assets. That effect can be powerful, but it is not a permanent earnings driver for every technology company.

Q2 expectations helped as well. Nasdaq's July 1 review cited FactSet expectations for 23.1 percent year-over-year S&P 500 earnings growth and 12.3 percent revenue growth at that point. Those were forecasts, not realised quarter results. The distinction matters because a price rally can begin before companies report the revenue and margin performance investors expect.

Our June market review covers the wider relief-trade setting. The S&P 500 story fits that setting, but it also depends on company-level earnings and the cost of building AI capacity.

Why July Changed the Leadership Group

Nasdaq's July 2026 review described a sharp change in the market's preferred theme. The SOX Index fell 20.6 percent during July after reaching a June high, and Nasdaq-100 fell 6.6 percent. The S&P 500 Equal Weight Index rose 1.0 percent, while the Dow rose 0.6 percent.

That contrast matters because a cap-weighted index can stay firm even when many members lag, while an equal-weight index gives each company a similar influence. When equal weight beats cap weight, participation is often broadening. When a few very large companies pull cap weight higher while the median stock lags, the headline index can hide concentration.

Large-cap value gained 3.8 percent in July while growth fell 5.9 percent. Nasdaq attributed the change to questions about when and how AI spending will produce returns, along with valuation levels, competition and the risk that AI services become more interchangeable. It said the pullback appeared driven more by positioning and valuation multiple compression than by a material deterioration in business fundamentals.

That last distinction is important. A share price can fall because investors pay a lower multiple for the same earnings outlook. It can also fall because the earnings outlook weakens. July's data, as summarized by Nasdaq, point first to a repricing of expectations. Future company reports must determine whether the market's concern remains a valuation issue or becomes an earnings issue.

What the Latest S&P 500 Data Says

The S&P Dow Jones Indices page provides the cleanest current reference point for the broad benchmark. As of August 20, the index stood at 7,641.16. Its price return was 11.62 percent year to date, 1.89 percent quarter to date and 2.02 percent month to date. A one-year price return of 19.47 percent was also reported.

The table combines two source bases. S&P DJI supplies the current S&P 500 index data. Nasdaq supplies the July relative-performance figures. They should not be read as one index methodology or one return series. The source and date are part of the number.

Our Bitcoin and Ethereum correction analysis offers a useful comparison. A headline index can remain positive while the market's leading group experiences a deep correction. That is why return dispersion matters.

How Strong Were Q2 Earnings?

By July 31, 61 percent of S&P 500 companies had reported Q2 results, according to Nasdaq's July review. Among those reporters, 86 percent exceeded EPS estimates and 77 percent exceeded revenue estimates. Nasdaq said the blended Q2 earnings-growth rate had reached 47.4 percent year over year and blended revenue growth had reached 14.1 percent at that point.

These figures are encouraging but incomplete. A reporting-season beat rate can change as more companies report. It can also reflect the gap between company guidance, analyst estimates and the actual result. A high beat rate does not automatically mean that the market is cheap or that future growth will match the latest quarter.

The sector distribution is also important. Nasdaq said 10 of 11 sectors had reported year-over-year profit growth and eight had double-digit gains. Energy, Communication Services, Consumer Discretionary and Information Technology led the cited earnings-growth figures. The article should not reduce that broad earnings picture to a story about only seven technology stocks.

At the same time, earnings quality must be examined. AI-related revenue can grow while spending on data centers, power, networking, chips and engineering also rises. Investors need to ask whether operating cash flow is keeping pace, whether depreciation is accelerating and whether new competition will reduce margins.

The Hexaware AI partnership analysis shows why company-specific announcements require separate checks on commercial scope and financial impact. An index theme can be real without every related company receiving the same economic benefit.

AI Capex Is Both an Opportunity and a Cost

Goldman Sachs Global Investment Research reported in June that the S&P 500's trailing four-quarter return on equity had reached 22 percent, a record in its series. It said the seven largest technology stocks had a collective ROE of 44 percent and that their ROE had grown by 9 percentage points over three years. Aggregate index profitability therefore helps explain why investors have accepted a high market multiple.

But the same research identifies a cost side. Goldman said the AI-capex boom had made hyperscalers more asset-heavy and that depreciation and amortisation could rise from 7 percent of revenue in 2022 to 12 percent in 2027. It also said consensus estimates pointed to an average seven-percentage-point decline in mega-cap technology ROE next year.

Those are estimates, not outcomes. AI investment may produce enough revenue and productivity to offset the larger asset base. It may also create excess capacity, price competition or a slower payback period. The market has to assign probabilities to both outcomes.

The semiconductor group has its own version of the issue. High demand and limited supply can lift pricing power and margins. New capacity and new competitors can later push in the opposite direction. A 20.6 percent July SOX decline does not prove that the AI buildout is finished. It shows that expectations and positioning can move faster than the underlying construction cycle.

Our digital-asset capital-allocation review applies a similar discipline to a different market. Growth stories need cash-flow evidence, not only a large addressable market or a powerful narrative.

Valuation and Concentration Risk

Goldman Sachs said in its June research that the S&P 500 traded at 21 times forward earnings, in the 87th percentile since 1980. Nasdaq's July 1 review cited a forward 12-month P/E of about 20.1 times, compared with 5-year and 10-year averages of 19.9 and 19.0 times. The difference reflects different dates, data sets or calculation conventions. It is not a contradiction that should be forced into one number.

A high multiple can be justified if earnings, margins and returns on capital remain high. It becomes vulnerable when rates rise, earnings estimates fall or investors decide that future AI returns will take longer. That is why July's shift from growth to value matters even though the S&P 500 remained positive year to date.

Concentration adds another risk. The largest technology companies have a greater influence on a cap-weighted index than the median constituent. Goldman said the aggregate S&P 500 ROE had risen while the median company's ROE had declined in recent years. That gap suggests the index can look stronger than the typical company.

The right response is not to assume that concentration will end immediately. It is to track whether earnings breadth, equal-weight performance and sector participation continue to improve. A broad rally has a different risk profile from a rally carried by a few large companies.

Our XRP price-target analysis is a separate example of why a target must be read with its assumptions. Market forecasts are scenarios, not promises.

What Could Move the Market Next

The next phase depends on whether the July rotation continues and whether earnings can support the valuation. Investors will watch Q2 reports still to come, forward guidance, data-center orders, cloud spending, semiconductor pricing, capital expenditure and free cash flow.

Interest rates are another channel. The July 29 FOMC held the federal funds target range at 3.50 to 3.75 percent. A higher long-term Treasury yield can reduce the present value of distant cash flows, which tends to pressure long-duration growth shares. Our 10-year yield guide explains why the benchmark is useful but not sufficient by itself.

Market breadth should be read with a date. If equal weight continues to beat cap weight, the Q2 concentration may be giving way to broader participation. If AI and semiconductor shares recover while equal weight lags, leadership may be returning to the same group. Neither outcome is guaranteed.

Valuation can also change through time without a price fall. If earnings grow faster than the index, the forward P/E can decline while the index rises. If earnings estimates weaken, the same price can produce a higher multiple. That simple denominator effect is why guidance matters as much as the latest quarterly headline.

The Bottom Line

The S&P 500's Q2 2026 performance was a real market event, and AI infrastructure was a major part of it. Nasdaq reported the best Q2 for the S&P 500 and Nasdaq-100 since Q2 2020, while the SOX Index rose 87.8 percent. July then delivered a sharp reset in AI leadership, with the Nasdaq-100 down 6.6 percent, the SOX Index down 20.6 percent and value outperforming growth.

The broad benchmark did not break. S&P DJI reported the S&P 500 up 11.62 percent year to date as of August 20, with a 19.47 percent one-year price return. The better conclusion is that the market changed internally. The question moved from whether AI demand exists to whether AI spending will produce durable returns at a valuation already above long-term norms.

Strong earnings can support the index. They do not remove concentration, valuation, rate or execution risk. Watch the equal-weight index, earnings breadth, forward P/E, free cash flow, depreciation and the next round of company guidance together.