What does a 94.7 composite earnings score mean?

The composite score, calculated by research firm Bespoke, aggregates revenue beats, EPS beats, guidance trends, and confidence indicators across all reporting companies. A score of 94.7 out of 100 places Q2 2026 among the highest-scoring seasons of the last two decades — comparable only to the post-COVID reopening boom of 2021. The closer you get to 100, the harder it becomes to sustain the level next quarter.

Are the earnings gains real, or just boosted by Anthropic and SpaceX investment gains?

Critics initially argued that the big tech beats were inflated by "other income" from unrealized gains on investments like Anthropic (Amazon, Google) and SpaceX. Analysis shows both are true: yes, investment gains contributed, but organic revenue and earnings growth — year-over-year operational improvements — are also strong and growing independently. The investment-gain tailwind, however, will not recur at the same magnitude next quarter, which is something to watch.

At current P/E multiples, is the S&P 500 expensive?

The S&P 500 forward P/E of ~20.9x sits slightly above the 5-year average (20.0x) and modestly above the 10-year average (19.8x), but well below the October 2025 peak of 24.2x. On price-to-sales and price-to-cash-flow, there is more richness — partly because the biggest companies are investing heavily in AI infrastructure and compressing near-term free cash flow. On earnings-based multiples, the market is not dangerously extended.

Q2 2026 Earnings Season Final Scorecard: 78% Beat Rate, 94.7 Score — One of the Best in 20 Years

By Yogurt · 2026-08-09 · Market Analysis

With 438 of 500 S&P 500 companies reported, Q2 2026 earnings season has earned an A+ grade: 78% beat on EPS, 74% on revenue, a 94.7 composite score, and 14% of companies raising guidance. Here's the full scorecard — and why the market's biggest risk now is its own success.

Before you dismiss the S&P 500's all-time high as pure sentiment or FOMO, look at what's actually underneath it. Q2 2026 earnings season is now in its final stretch — 438 of 500 S&P 500 companies have reported, and more than 1,100 US companies in total — and the numbers do not lie. This is one of the strongest earnings seasons in the past two decades.

The Scoreboard: 78%, 74%, 94.7

The headline numbers are striking across every dimension:

78% of companies beat on EPS — analysts underestimated profitability at a historic rate
74% of companies beat on revenue — demand-side surprises were broad and genuine
94.7 composite score (Bespoke Investment Group methodology) — one of the highest readings in over 20 years, ranking near the 96th percentile historically
10 of 11 S&P 500 sectors saw 82% or more of their companies beat estimates

That last point deserves emphasis. This is not a story about a handful of mega-cap tech names pulling the index higher. It is a broad-market outperformance. Energy, industrials, financials, health care, consumer discretionary — all beat at similarly high rates. When 10 of 11 sectors perform this well simultaneously, it reflects something real in the underlying economy, not statistical noise.

The Triple Play: 11.8% of Companies Did It All

A "triple play" in earnings parlance means beating on revenue, beating on EPS, and raising forward guidance in the same report. It is the hardest trifecta to hit, and typically only a handful of companies manage it in any given quarter. In Q2 2026, 11.8% of all reporting companies — roughly one in eight — pulled it off. The last time the triple-play rate approached this level was 2021, during the post-pandemic demand surge.

On the guidance front: 14.2% of companies raised forward guidance, placing the guidance-raise rate in the 90th percentile historically. Only 3.8% cut guidance. Management teams, who lived through the volatility of tariff fears and Fed uncertainty earlier this year, are now confident. CEOs and CFOs across industries are signaling they expect Q3 to be at least as strong as Q2.

Analysts Are Chasing — and That Is Bullish

When 78% of companies beat EPS estimates, there is a simple explanation: analysts were too conservative. Wall Street has spent the past six months systematically underestimating corporate earnings power, and now they are scrambling to catch up.

The clearest evidence: 2027 EPS estimates for the S&P 500 have risen from $350 in December 2025 to approximately $408 today, with many desks now tracking toward $425. At the start of Q2 reporting, the consensus was near $400. Three weeks of strong results pushed it to $408. This matters enormously for the index math.

Apply a 20x price-to-earnings multiple — the midpoint of the five-year average — to $407–408, and you get an implied S&P 500 fair value of ~8,150. The S&P currently trades near 7,700–7,800, suggesting roughly 200–400 points of additional upside on fundamentals alone, before any multiple expansion.

Are the Gains Real? The Anthropic Question

One pushback has gained traction on Wall Street: are the mega-cap earnings inflated by "other income" — specifically unrealized gains from investments in private AI companies like Anthropic and SpaceX held by Amazon, Google, and others?

The answer is nuanced. Yes, "other income" from investments contributed to Q2 EPS at several large companies. But analysis of the actual financials shows a clear pattern: organic revenue growth — year-over-year operational improvements — is strong and independent of investment gains. The Anthropic tailwind for Google and Amazon was real in Q2, but it will not recur at the same magnitude next quarter; Anthropic is already valued at over $1 trillion and is unlikely to double again in a single reporting period. Investors should mentally separate the organic story (strong) from the one-time investment gain story (fading).

Valuations: Less Stretched Than You Think

The S&P 500 at all-time highs sounds expensive by reflex. The data is more forgiving:

Forward P/E: ~20.9x — slightly above the 5-year average of 20.0x, above the 10-year average of 19.8x, but 20% below the October 2025 peak of 24.2x
Price-to-EBITDA: near the 5-year average
Price-to-Sales and Price-to-Cash-Flow: elevated — because the biggest companies are aggressively deploying free cash flow into AI infrastructure, compressing near-term reported cash flow without reducing intrinsic value

The market is not cheap. But on the metric that matters most — forward earnings — it is not in bubble territory either. Companies delivering this level of performance arguably deserve a premium multiple, not a discount.

The Paradox of an A+ Season

Here is the uncomfortable flip side of a 94.7 composite score: being this good makes you fragile.

When 78% of companies are beating estimates, analysts are forced to raise their targets and their forward estimates. The bar for Q3 2026 just got dramatically higher. If AI capital spending begins to decelerate, if consumer demand softens, or if geopolitical events introduce earnings uncertainty — even a 65% beat rate in Q3 would feel like a disappointment relative to today's expectations.

Look at the historical pattern: the only times the composite score approached these levels and then sustained were periods of genuine structural transformation — the post-dot-com recovery and the post-COVID demand surge. Outside of those inflection points, elevated scores tend to mean-revert within two to three quarters as analysts catch up and set the bar higher.

This is why you have Tom Lee and Michael Burry agreeing on the near-term outlook, despite their very different long-term views: a correction of 10–15% in the coming weeks is probable — not because the fundamentals are bad, but because when you are this good, any hiccup looks catastrophic. One soft guidance from a major name, one surprise macro data point, and the market will re-price the entire "inevitable path to 8,000" narrative.

What to Watch for Q3

The three variables that will determine whether the A+ season extends or reverts:

AI CapEx demand signals: If the hyperscalers (Amazon, Google, Microsoft, Meta) continue to signal accelerating AI infrastructure spending in their Q3 calls, the 2027 EPS estimates will keep rising. If they pause or reduce, the EPS math reverses.
Fed policy: A rate hike — even a 25bps move — would compress P/E multiples immediately. The current 20.9x multiple assumes rates stay flat or fall. A hike shifts the arithmetic against equities.
Organic growth vs. investment gains: As the Anthropic/SpaceX one-time gains fade from year-over-year comparisons, can companies sustain beat rates without that tailwind? The answer will be visible in Q3 reports starting October.

The scorecard is as strong as any in a generation. The S&P 500's all-time high is built on real earnings, real guidance confidence, and real demand. What is less certain is whether a market this good can stay this good — and that uncertainty is exactly where the next 10% move, up or down, will be decided.