August 2026 Market Review: Stocks Rose, Metals Surged, and the Economy Split

By Yuriy Matso · The Trading Tools · August 31, 2026 · 15 min read

Research note: built from month-end prices across US sectors, countries, crypto and major assets; our market-breadth series; and every release-dated economic series cited below. Computed 2026-08-31. How we use AI.

The index was strong; the confirmation was mixed

The S&P 500 gained +2.68% and the Nasdaq 100 gained +4.18%. Volatility drifted lower through the month, with the VIX closing at 14.92 against 15.99 at the end of July. By the standards of the month’s own history that is a good August and nothing more dramatic: the 9th best of 34, around the 76th percentile of the month’s record.

The headline return was not a false signal: volatility confirmed it and the equal-weight index also rose. But equal weight lagged, short-term breadth fell, and the winners were concentrated in technology, energy and health care. That makes this a narrower advance. It was neither broad participation nor a confirmed breakdown.

Seasonality was the least interesting thing about August

Since 1993 the S&P 500 has closed August higher 22 times in 34. That 64.7% hit rate ties July for 4th of the twelve, ahead of October, December and January. The median August gains +0.98%. Neither of those numbers describes a dangerous month.

The average tells a different story at +0.04%, and the average is what gets quoted. Both figures are correct. They disagree because 5 Augusts fell more than 5%: 1998 at -14.4%, 2015 at -6.1%, 2001 at -5.9%, 2011 at -5.5%, 1997 at -5.2%. Removing only the four worst of them lifts the month’s average from +0.04% to +1.11%, which would put it in the upper half of the calendar.

My read is that August is a normal month wearing the scar tissue of a handful of crisis-heavy outliers. The Russian default and collapse of Long-Term Capital Management in 1998, the US credit-rating downgrade in 2011 and the Chinese devaluation in 2015 happened to land in August. Treating those events as a property of the calendar is the error.

S&P 500 calendar-month returns since 1993. The gap column is the average minus the median: where it is most negative, a few deep losses are doing the work. Updates as each month completes.

The index rose faster than the average stock

A +2.68% month sounds broad. It was not. The share of stocks trading above their own 10-day average fell from 45.6% at the end of July to 34.5% at the end of August, and the equal-weight S&P returned +2.04% against the cap-weighted +2.68%.

August 27 is the session that shows it whole. Nvidia rose +8.74% and Salesforce +22.58%. The cap-weighted S&P gained +0.66% on the day. The same index equally weighted fell 0.30%. Two results carried the index while the average member went the other way. The earnings context is documented in the August 27 market report.

The longer averages tell a calmer story, and both belong on the page. Participation above the 200-day line actually improved slightly, from 51.5% to 52.5%. What that leaves is a market where the index gained close to three percent and roughly half of its members are above their own long-term trend.

Does a narrow advance mean trouble?

This is the point where the seasonal story usually turns bearish, so it is worth testing rather than asserting. I took every month since 2010-10 in which the index gained at least 2% while the share of stocks above their 50-day average fell, which is what August 2026 did. That gives 22 months out of 190.

The result runs against the intuition. A month later the index was higher 77% of the time and gained +1.47% on average, against 67% and +1.07% for every month in the sample (22 instances against 190). Three months out the advantage disappears: +3.01% against +3.12% across 21 instances and 188. Narrow leadership has been a poor short-term warning.

A sample this size is not a statistic, and the window starts after the financial crisis, which covers one long bull market and two short bear markets. I would treat this as evidence against the bearish reading rather than evidence for a bullish one.

The economy split between companies and households

The broadest August shift was not recession versus expansion. It was who still had momentum. Businesses were forming, investing and earning at unusually strong rates. Households were still spending more than a year ago, but almost not at all in July after inflation, with a thin saving cushion and much weaker hiring.

These are the observations published during August, not an internal score. Applications and capex are Census releases; profits and real spending are BEA releases. The quarterly profit figure is the local after-tax measure and can differ from a headline pre-tax series.

The split matters because the index is a claim on large-company earnings, not a survey of household comfort. It helps explain why stocks could rise through soft consumer data without requiring the data to be ignored: the corporate side of the same economy was accelerating. The primary releases are the Census business-formation report and the BEA profit release.

Hiring weakened, but layoffs did not break

The month’s central contradiction was a labor market that weakened without handing the Fed a reason to cut. The employment report released on August 7, 2026 put July payrolls 21,000 above June and the three-month average at only 38,000. The safety valves were still intact: unemployment was 4.1% and initial claims ended the month at 204,000, close to the cycle floor. That is a weak hiring market, not a layoff wave. The BLS release supplies the original print.

Inflation did not cooperate. The PCE release later in the month put headline prices 3.7% above a year earlier and the core measure 3.3%, both well clear of the 2% target. A softening labor market alongside inflation above three percent is the combination that gives a central bank the least room, and the index gained +2.68% through it. That is worth separating from a macro all-clear, because the data did not deliver one. See the BEA income and outlays release.

Housing kept deteriorating

Almost every housing series that reported during August moved the same way. New-home months of supply reached 9.6, the 96th percentile of a record that starts in 1963. Readings that high have otherwise clustered in the 1970s and in the years around the 2008 bust.

Each row shows the reading published during August 2026 against the one before it. Housing data is reported with a lag, so these describe July activity. Building permits are the honest exception in the table: they rose. Permits lead starts, so that is the row to watch for a turn.

One caution on the supply number. It has now printed near this level four times in eighteen months, so August did not discover a new problem. It confirmed a standing one. I covered the fuller question in whether the housing market is going to crash.

Energy entered the Hormuz shock with little slack

The cleanest physical constraint of the month sat inside the energy complex. Brent and WTI both fell through most of August, to their local readings of . Retail diesel went the other way and finished its reading at the 0th percentile of its entire history. Those local price series are dated where they are because our copies end a few days before the month does. Underneath those prices, refinery utilization reached 97.4%, around the 98th percentile since 1990. Distillate inventories fell to 103.4 million barrels, down 9.5% from a year earlier and near the 9th percentile of their history. The 200-day average 3-2-1 crack spread was $44.48 per barrel.

That was the starting point when fighting returned to the Strait of Hormuz on August 31. Brent jumped on the final session. The event did not create August’s refining constraint; it arrived on top of one. With plants already near full utilization and middle- distillate stocks unusually low, the September risk is less about adding refinery throughput than about whether crude availability and diesel inventories can absorb another disruption.

Where the money actually went

The sector table explains the narrow tape better than the index number does. Energy and technology led, and the defensive and rate-sensitive corners lost ground while the index gained.

Calendar-month price returns for August 2026, computed from the same daily closes the tools use.

Precious metals were the month’s clearest cross-asset signal

Gold’s +9.93% gain ranks around the 97th percentile of 200 monthly GLD observations. Silver rose even more, and the equity leverage embedded in miners turned the metal move into still larger gains for both miner funds. The full precious-metals complex participated; gold did not move alone.

The timing rules out an easy Hormuz explanation. Most of the advance was already in place by August 14, and second-half gains were smaller across all four rows. Gold also surged while equities rose, the VIX fell and long Treasuries moved only +0.33%. That combination looks more like demand for inflation, currency or disruption protection than a conventional flight from risk. The miners’ much larger gains add a speculative component to the hedge.

Crypto did start moving—and the move was broad

Every crypto proxy in the review finished August higher. Solana led at +42.53%, while the broad crypto index fund gained +27.50% and Bitcoin gained +25.34%. Ether and XRP also rose more than Bitcoin. This was participation across the group rather than a one-token squeeze.

Unlike precious metals, crypto’s move began in the second half. Bitcoin was essentially flat through August 14 before gaining +25.37% over the rest of the month; Ether gained +32.02% in that same window. The caution is in the final column: every proxy still finished the year to date below zero. August looks like the start of a broad rebound, but one strong month has not repaired the longer drawdown.

Global leadership and the company event of the month

Country figures are investable US-listed ETF proxies, not local-currency index returns. The selected companies are the largest verified August anomalies used in this review; the table does not claim to rank every stock.

The single-stock event of the month was Moderna. Its shares gained +156.00% after Moderna and Merck reported the first positive phase 3 readout for an individualized neoantigen therapy, intismeran autogene with Keytruda, in melanoma. The result matters beyond one stock because it moved personalized mRNA cancer therapy from a promising platform to a phase 3 success. The companies’ joint trial announcement is the primary source.

Policy stayed boxed in

July payroll growth weakened, but headline PCE inflation was 3.7% and core PCE was 3.3%. That mix denied the Federal Reserve an easy response. In his August 28 speech, Chair Kevin Warsh described strong business capital spending and profits alongside strain in housing and agriculture, while keeping the predominant focus on prices. That is almost exactly the split in the month’s data.

Treasury also announced on August 19 that the maximum size of its long-end liquidity-support buybacks would at least double beginning September 9. This is a market-functioning operation, not quantitative easing, but it is still relevant when long-duration liquidity is thin. The Treasury announcement supplies the terms.

What August says—and does not say—about September

It does not. Across the 14 years since 1993 in which August gained more than 2%, September finished higher 50% of the time and averaged -0.69%. The unconditional September averages -0.86% with 53% positive. Momentum from August carries no information I can find, in either direction.

What does survive is September itself. It is the only month with a negative average since 1993 at -0.86%, the only one that finishes higher barely half the time at 52.9%, and it carries the widest mean-to-median gap on the calendar. Its median is still +0.26%. September has the same shape as August, drawn harder: a normal middle and a heavy left tail.

So the honest seasonal statement is narrower than the one usually made. The calendar does not forecast returns. It describes the distribution of outcomes, and in these two months that distribution has more weight in its left tail than elsewhere in the year. Position sizing is a reasonable response to that. Selling in anticipation is not.

What would change my mind

Three specific markers, each on a page where you can watch it move.

• Participation breaks below 40%. The share of stocks above their 50-day average has held near half the market all year. A sustained move under 40% while the index holds up would turn the narrow-advance reading from benign into a genuine divergence. Market Breadth
• The business-household gap closes. A rebound in real spending and payroll growth would make August’s split constructive; a rollover in applications, core capex and profits would make it defensive. The gap matters more than either side alone. Economy Health
• Energy slack returns—or fails to. Falling utilization with rebuilding distillate inventories would relieve the constraint. Crude rising while utilization remains above 95% and inventories keep falling would confirm a physical squeeze rather than a price anomaly. Refinery Utilization

How we checked it

Frequently asked questions