Today’s Market Takeaways

The Forex Economic Calendar Guide explains how to read release schedules, while the Forex Market Hours Guide sets out the relationship among Tokyo, London and New York trading hours.

1. The final US session of July: a rebound with sharp internal divergence

The closing figures can make the day look like a uniform US equity advance. Yet the Nasdaq Composite rose 1.0% while the Russell 2000 fell 0.5%, a gap of 1.50 percentage points. The divergence suggests capital concentrated in large companies that demonstrated earnings growth rather than pricing an even improvement in economy-wide growth. When large-cap indices rise while smaller companies fall, the index direction alone can overstate the breadth of participation.[S1]

The S&P 500 gained 1.0% for the week, its first positive week in three, while it was slightly lower for July as a whole. A daily rebound, a weekly gain and a small monthly decline are not contradictory. Oil, interest rates, semiconductors and monetary policy were repriced substantially during the month, and the final-day rally did not erase all of that variation. The useful distinction for the week ahead is which inputs persisted and which were specific to one session, rather than simply extending Friday’s direction.[S2]

Intraday turning points: early optimism, a reversal, then a recovery into the close

The move in memory and semiconductor shares was larger still. Micron Technology went from an early 6.4% gain to a 6.5% intraday loss and closed down 5.9%, a high-to-low swing of about 12.9 percentage points. That does not establish that the long-term AI demand theme disappeared. It indicates that, once demand growth is broadly recognized, supply constraints, capital expenditure, pricing, customer concentration and margins can separate company-level returns.[S2]

The gap between Amazon’s 15.3% gain and Apple’s 7.4% decline was 22.70 percentage points. Even among large technology companies, the market rewarded evidence connecting cloud growth and AI investment to earnings, while it penalized the possibility that supply constraints could limit next-quarter growth. Treating the index gain as uniform optimism about AI would miss that selection. Revenue growth matters, but so does the path from investment to operating income, contracted demand and cash flow.

Friday’s path also shows why identical closing returns can embody different degrees of conviction. A steady rise throughout the day is not the same as a 1.3% advance that disappears and is then rebuilt. Friday followed the latter pattern. The early earnings response did not settle the direction; the market absorbed higher long-term yields, higher oil, weakness in Apple and sharp semiconductor selling before establishing the close. The next observation is therefore whether the buying behind the morning high or the selling behind the reversal reappears, not a mechanical continuation of the close.

For the week, the Nasdaq Composite gained 1.6%, ahead of the 1.0% increases in both the S&P 500 and Dow, while the Russell 2000 rose by less than 0.1%. The large-growth-company advantage was therefore visible on both the daily and weekly horizons, although the S&P 500 remained slightly lower for July. Across daily, weekly and monthly windows, Friday is better described as a partial recovery from a volatile month than confirmation of a new long-duration trend.[S1][S2]

Index construction also matters. The S&P 500 and Nasdaq Composite are especially sensitive to companies with very large market capitalizations. Smaller companies in the Russell 2000 tend to be relatively more exposed to domestic demand, refinancing conditions and borrowing costs. Large-cap strength alongside small-cap weakness does not by itself prove economy-wide deterioration, but it shows that the benefit of earnings growth was uneven. Whether stronger data bring small companies into the advance or higher long-term rates leave them further behind will help describe the quality of participation.

2. Rates, oil and weekend crypto: the discount-rate burden remained

Long-term yields rose alongside equity prices, so Friday was not a broad risk rebound caused by falling interest rates. Companies whose earnings surprise outweighed the higher discount rate advanced, while uncertainty about future profits or supply constraints weighed on others. Equities can rise when rates rise, but the same yield level becomes a heavier valuation input if earnings upgrades stop.

Oil acts on both inflation and corporate earnings. Higher prices can support energy-company revenue and profit while raising costs for transport, chemicals, airlines and consumer goods. For households, fuel and freight costs can reduce disposable purchasing power; for central banks, the key questions are inflation expectations and second-round price transmission. One day of higher oil cannot determine policy, but the 74-basis-point increase in the 10-year yield from its pre-war level shows that energy and inflation uncertainty has not disappeared from bond pricing.[S2]

Bitcoin continues to trade over the weekend, but spot crypto is not a substitute for US equity futures: liquidity, participants, regulation and venue structure differ. Its weekend direction therefore cannot be copied mechanically into Monday’s equity direction. The limited observation is that price discovery continued in a separate market while cash equity markets were closed.

Treasury analysis requires a distinction among the policy rate, short-term yields, long-term yields, real rates and expected inflation. The confirmed closing figure used here is the 10-year yield. Reliable published July 31 closes for the two- and 30-year yields were unavailable, so no substitute values are presented.

3. Monetary policy and macro data: slower headline growth, resilient demand

The Federal Reserve kept the target range for the federal funds rate at 3.50% to 3.75% on July 29. The vote was 9-3, with Beth Hammack, Neel Kashkari and Lorie Logan preferring a 25-basis-point (0.25 percentage point) increase. The dissents represented 25% of the voting group. The unchanged decision matters, but three preferences for an increase also show that concern about inflation was substantial within the committee.[S3]

Second-quarter real GDP grew at a 1.5% annual rate, down from 2.1% in the first quarter. Private domestic final sales, which remove some of the volatility from inventories and net exports, accelerated to 3.9% from 1.7%. The 2.40-point gap produces a different picture from the 1.5% headline alone: overall growth slowed, while final private demand remained firm.[S4]

In the same GDP release, the PCE price index increased at a 5.1% annual rate and the index excluding food and energy rose 3.4%. In June alone, the headline PCE price index declined 0.1% from May, while the core index rose 0.1%. From a year earlier, headline and core inflation were 3.7% and 3.3%, respectively. A monthly decline driven by volatile components does not establish that underlying inflation has returned sustainably to target.[S4][S5]

Real personal consumption expenditures increased 0.4% in June, faster than the 0.3% increase in nominal spending, while the personal saving rate was 2.7%. Resilient consumption supports company revenue, but a low saving rate reduces the buffer if employment or real income weakens. The employment report therefore needs to be read across payrolls, unemployment, participation, wages and revisions rather than through the payroll headline alone.[S5][S10]

The Employment Cost Index rose 0.9% in the second quarter. Wages and salaries increased 0.9% and benefit costs rose 1.0%. From a year earlier, total compensation was up 3.4%, wages and salaries 3.2%, and benefits 3.8%. Labor-cost growth may be moderating, but it remains relevant to service prices. Assuming rapid easing from slower GDP alone would miss the persistence of private demand and compensation.[S6]

Nominal and real strength should also be kept separate. Revenue and wages can rise at the same pace as prices without increasing the quantity households can buy or the real profit companies retain. In June, real spending increased 0.4% while nominal spending increased 0.3%, as the aggregate monthly price level declined. Yet year-over-year inflation remained above 3%, leaving the cumulative increase in household costs visible on a longer horizon. A low saving rate makes slower income growth more consequential.[S5]

The difference between GDP and private domestic final sales also illustrates how inventories, government spending and trade can move a quarterly headline. Inventories and net exports can be revised substantially; stronger imports reduce GDP arithmetically even when they reflect firm domestic demand. The August 4 trade release and August 6 wholesale inventory report will help assess how the composition behind the initial 1.5% estimate may evolve.[S4][S11]

Policy reaction depends on consistency across a sequence, not a single number. Strong payrolls accompanied by slower wages and higher participation would have a different inflation implication from strong payrolls alone. Weak payrolls with upward revisions and resilient openings would not necessarily show an abrupt break. Likewise, an ISM headline below 50 combined with a rising price component would join growth concern with inflation concern. The policy question is whether the reason to maintain the 3.50%-3.75% range becomes stronger or weaker.

4. Corporate results: from the size of AI spending to its link with revenue and profit

Amazon reported $200.6 billion of revenue and $27.5 billion of operating income. AWS produced $42.2 billion of revenue and $16.6 billion of operating income. These figures put AWS at 21.04% of company revenue but 60.36% of operating income. The direct contribution of cloud growth to the profit mix helps explain the 15.3% share-price gain. At the same time, trailing-12-month free cash flow was negative $7.6 billion, primarily because of greater AI-related capital spending. Profit growth and cash outflow therefore coexist, keeping the pace of return on investment central.[S9]

Microsoft generated $90.0 billion of revenue, including $59.3 billion from Microsoft Cloud, a 65.89% share. Azure revenue grew 43% and commercial remaining performance obligations reached $678 billion. When revenue growth, contracted demand and operating income rise together, the market can treat heavy capital expenditure as an advance investment in future revenue. Remaining performance obligations nevertheless combine different contract durations, service mixes and recognition dates, and are not all near-term profit.[S7]

Apple’s $109.4 billion of revenue and $2.02 in diluted EPS set records for its June quarter. The 50.1% gross margin included an approximately two-percentage-point benefit from tariff refunds, and the refunds added $0.11 to EPS. The 7.4% share-price decline despite strong reported results showed that the market placed greater weight on next-quarter growth and component availability. The temporary refund benefit and the possibility that AI demand constrains component supply need to be separated.[S2][S8]

Across the three companies, AI-related demand supports revenue while also increasing the burden on equipment, power, semiconductors, memory and supply chains. The burden appears in different places: capital expenditure and free cash flow for cloud providers, component availability and gross margin for device companies, and pricing and capacity for chip suppliers. A single statement that AI demand is strong cannot describe those distinct profit sensitivities.

5. This week’s economic, policy and corporate calendar

The Treasury’s tentative schedule includes auctions of 13- and 26-week bills on August 3 and six- and 52-week bills on August 4. Bill demand reflects expectations for the policy rate and money-market conditions, but it does not move for exactly the same reasons as long-term yields. On August 5, the Treasury is also due to announce terms for the following week’s three-, 10- and 30-year auctions, making expected supply and demand relevant to the term premium.[S14]

The order of releases matters. Manufacturing and construction on August 3 are followed by trade, openings and orders on August 4. Services on August 5 adds information on prices and employment; productivity on August 6 addresses whether companies can absorb wage growth; and the employment report on August 7 updates household income and policy expectations. An early-week move in yields can be challenged by a later release, so no single figure fixes the full week’s interpretation.

Each release also needs to be separated into its headline, components and revisions, and into nominal and real measures where applicable. A higher manufacturing composite paired with weaker new orders and higher prices would look more like cost pressure than broad demand recovery. More job openings alongside slower hiring and quits could show companies preserving vacancies while becoming cautious about actual staffing. A narrower trade deficit can reflect either stronger exports or weaker imports, which carry different signals for US and foreign demand. In the employment report, payrolls, unemployment, participation, hourly earnings, hours worked and prior-month revisions need to be read together.[S10][S11][S12]

Release timing creates additional differences. Survey indicators can capture current business judgment relatively quickly, while construction and trade data describe an earlier period. The household and establishment surveys in the employment report cover different populations. A weak leading measure beside a strong lagging measure may represent an economy partway through a slowdown rather than a contradiction. Conversely, better sentiment without stronger orders, hours or real spending may show expectations moving ahead of realized activity.

The US cash-equity market is scheduled for normal hours on August 3. Nasdaq states regular hours of 09:30 to 16:00 Eastern time, corresponding to 22:30 to 05:00 Japan time while the United States observes daylight saving time. The session closes at 05:00 Japan time on August 4.[S13]

Omitting unverified closing figures does not remove Europe as a comparison point. For US multinationals, European revenue translation, regional demand and energy costs can still appear in company guidance. Sector reactions in European banks, capital goods and consumer companies may also provide context before US cash trading begins when confirmed data become available. The direction or magnitude of an index move is not inferred in the absence of a verified cash close.

The Macro Analysis Guide provides a framework for joining different releases, while the Macro Scenario Analysis Guide explains conditional branches without forcing them into one direction. The useful question is which hypothesis each new observation strengthens and which it weakens.

Analysis & Commentary