US Second Quarter PCE Inflation Was Revised Up to 5.3 Percent, but June and July Ran at About 0.6 Percent Annualised

By The Edge Research Team

Three official United States releases landed on Wednesday and they do not tell the same story. The Bureau of Economic Analysis left second quarter growth at 1.5 percent annualised in its second estimate but revised both quarterly inflation measures up, the personal consumption expenditures price index to 5.3 percent and the measure excluding food and energy to 3.6 percent. The same morning it reported that in July the price index rose 0.2 percent on the month and 3.7 percent on the year, that real consumer spending was flat, and that households saved most of a 0.5 percent gain in disposable income. The Census Bureau reported new orders for manufactured durable goods up 1.1 percent, almost all of it transport.

The quarterly revisions describe April to June. The monthly data describes July. On the evidence published on Wednesday the two point in opposite directions, and the monthly data is the more recent.

The quarterly price revisions cover a quarter that ended two months ago

The personal consumption expenditures price index for the second quarter was revised up 0.2 of a percentage point to 5.3 percent annualised, the measure excluding food and energy up 0.2 of a percentage point to 3.6 percent, and the gross domestic purchases price index up 0.1 to 5.8 percent.

Set that against the Bureau’s own monthly series. The price index fell 0.1 percent in June and rose 0.2 percent in July. Compounded, those two months come to about 0.1 percent, which is roughly 0.6 percent at an annual rate on our calculation. The core measure rose 0.1 percent in June and 0.2 percent in July, about 1.8 percent annualised on the same basis, against a twelve month rate of 3.3 percent.

Both recent monthly readings are substantially softer than the second quarter annualised rate and than the twelve month rates. The quarterly figure cannot be decomposed into implied April and May monthly readings, because the Bureau calculates it from average price index levels across quarters rather than by compounding three monthly changes, and The Edge makes no such reconstruction here. What can be said is narrower and still material: the June and July observations show considerably less inflation momentum at the margin than the second quarter figure suggests, while the twelve month rates of 3.7 percent headline and 3.3 percent core continue to incorporate earlier price increases.

The headline growth rate and the private demand figure measure different things

Real gross domestic product increased at an annual rate of 1.5 percent in the second quarter against 2.1 percent in the first, unrevised from the advance estimate. Real final sales to private domestic purchasers, which is consumer spending plus gross private fixed investment, was revised up 0.3 of a percentage point to 4.2 percent, and current dollar gross domestic product was revised up to 8.0 percent.

The 4.2 percent figure is a real and useful measure of demand from households and private fixed investment. What it is not is an alternative estimate of gross domestic product, and the 2.7 percentage point gap between the two should not be read as evidence that the private economy outproduced the headline. Real final sales to private domestic purchasers excludes government spending, changes in private inventories and net exports, so a movement in any of those components can open a substantial gap. Rising imports can widen it, because imported goods and services may be counted inside consumption or investment and are then deducted in calculating domestic production. The Bureau names two of those components for this quarter: a downturn in government spending, and imports rising faster in the second quarter than in the first.

Imports are deducted from gross domestic product because they have already been counted once inside consumption and investment, not because they reduce output. The deduction removes a double count. In this estimate the Bureau said an upward revision to consumer spending was partly offset by an upward revision to imports, leaving headline growth essentially unchanged. That means the positive revision to consumption was counterbalanced in the arithmetic by a larger subtraction for imports. It does not establish that the additional consumer spending itself consisted of imports, and The Edge does not claim that it did.

The two independent measures of what the economy produced are closer to each other than either is to 4.2 percent. Real gross domestic income rose 2.2 percent against 1.2 percent in the first quarter, and the average of gross domestic product and gross domestic income, which the Bureau publishes because the two can diverge, rose 1.8 percent.

Profits from current production increased 400.9 billion dollars at an annual rate in the quarter, after an increase of 74.4 billion in the first. Profits enter gross domestic income directly, so that increase is one candidate explanation for the income side coming in above the output side, but the Bureau does not attribute the gap and establishing it would require a full decomposition of the statistical discrepancy. The Edge does not express the two increases as a ratio, because the value of such a ratio is set by the size of the smaller base rather than by anything in the economy.

The price measure that describes what United States buyers actually faced is the gross domestic purchases index, revised up to 5.8 percent. It is not the same thing as the price of what the country produced, which is a separate index the Bureau publishes and which The Edge did not obtain from this release; no figure for it is asserted here.

July income rose, real spending did not, and the saving rate went up

Personal income rose 0.4 percent in July and disposable personal income rose 0.5 percent, an increase of 125.9 billion dollars, which the Bureau attributes mainly to compensation, government social benefits and income receipts on assets. Personal outlays rose 36.6 billion dollars and current dollar consumption rose 36.3 billion, or 0.2 percent.

Income therefore outran outlays by about 89 billion dollars on our calculation, and personal saving finished the month at 712.0 billion with the saving rate at 3.0 percent. The rate rose over the month. What can be said from the release is that July’s income gain did not translate proportionately into consumption. Why is not established by the data: caution, timing and ordinary monthly variation are all consistent with it. The rate remains low relative to longer run historical norms, which leaves households a relatively small current income buffer if spending growth again begins to exceed income growth.

In real terms consumption was flat. Real personal consumption expenditures increased 1.3 billion dollars, which the Bureau describes as less than 0.1 percent at a monthly rate, after 0.4 percent in June. Averaged, the two months run at about 0.2 percent a month, roughly 2.4 percent annualised on our calculation, which is a step down from the second quarter rather than a halt. One month of flat real spending in an advance estimate is not by itself evidence of a turn.

The composition moved sharply. Current dollar spending on services rose 86.2 billion dollars while spending on goods fell 49.9 billion. Those are nominal figures and cannot on their own account for a real outcome, since goods and services prices move differently, and The Edge did not obtain the real split. What they do show is a rotation from goods into services rather than an absence of demand.

Durable goods orders rose on transport, and defence orders fell

New orders for manufactured durable goods increased 3.6 billion dollars or 1.1 percent in July to 339.3 billion dollars, up in four of the last five months and following a 0.5 percent increase in June. Transportation equipment rose 2.6 billion dollars or 2.3 percent to 116.2 billion after two consecutive monthly declines, contributing about 0.8 of the 1.1 percentage point rise on our calculation, with everything else contributing about 0.3.

Orders excluding transportation rose 0.4 percent, to 223.1 billion dollars on our calculation from the two published levels. Orders excluding defence rose 1.3 percent, which is above the 1.1 percent headline, and that is only arithmetically possible if defence orders fell on the month. The Census Bureau does not state the defence figure in the advance summary and The Edge does not estimate it. The composition that follows is non defence transport equipment carrying the increase, defence orders declining, and the remainder of manufacturing close to flat.

This is an advance report and is revised in the full report. It also does not, in its summary text, carry orders for non defence capital goods excluding aircraft, which is the series that proxies business equipment investment; The Edge could not extract that figure from the release tables and therefore draws no conclusion about the capital spending cycle.

Gross domestic product, second quarter 2026, percent change at an annual rate

Personal income and outlays, percent change from the preceding month

PCE price index, recent pace against the twelve month rate

Durable goods, new orders, July 2026

Treasury par yield curve, 25 August 2026, official

United States markets, intraday 26 August, 19:32 GMT

Why it matters

The Federal Open Market Committee meets on 15 and 16 September. It held the federal funds target range at 3.50 to 3.75 percent on 29 July, but it did so on a nine to three vote, with Beth Hammack, Neel Kashkari and Lorie Logan all preferring a quarter point increase at that meeting. Reuters reported on Wednesday that futures put the probability of an increase in September at 38.1 percent on the CME FedWatch tool. Three standing dissents in favour of tightening, and a minority but substantial probability priced, mean the argument in front of the committee is about whether to raise, not whether to cut.

Wednesday’s releases give evidence to both sides of that debate. The hawkish case rests on the upward revisions to second quarter inflation, on twelve month inflation of 3.7 percent headline and 3.3 percent core, and on private final demand growing at 4.2 percent annualised, and it follows a meeting at which three policymakers already preferred an increase.

The case for patience rests on the more recent monthly momentum. Headline prices over June and July compound to roughly 0.6 percent annualised on our calculation and the core measure to about 1.8 percent, real consumption was flat in July, and the saving rate rose. Two months are too short to establish a new trend, but they are considerably softer than the quarterly figures, which describe a quarter that ended two months before they were published.

A committee that has held since December can reasonably conclude that the July moderation needs confirming before it changes anything.

The test will not come from this statistic. The Bureau publishes August personal income and outlays on 30 September, two weeks after the committee votes, so the price series that produced Wednesday’s disinflationary signal will not be updated before the decision. What the committee will have is the August employment report on 4 September, the August producer price index on 10 September and the August consumer price index on 11 September. If those confirm softer inflation and demand, the case for holding strengthens and the 38.1 percent priced probability will look too hawkish. If they show renewed price pressure or stronger activity, the argument made by July’s three dissenters gains weight.

Reuters quoted Ellen Zentner of Morgan Stanley Wealth Management saying the inflation print was not enough to shift the balance for September but that the Federal Reserve may feel more pressure to move if later data point the same way. On the evidence of the July income release, the later data so far points the other way.

Outlook

Federal Reserve Chair Kevin Warsh speaks at the Jackson Hole symposium on Friday, per Reuters, which is the first read on how the committee is weighing the quarterly revisions against the monthly data. The Bureau of Economic Analysis has said the 2026 annual updates of the national, industry and regional accounts will all begin on 30 September, the same morning it publishes August personal income and outlays, so the historical series against which the September decision is judged will itself be revised two weeks after that decision is taken.

The Treasury had not published the 26 August par yield curve at the time of writing. On 25 August the two year stood at 4.17 percent and the ten year at 4.64 percent, each about six to seven basis points lower on the day, a near parallel shift down rather than a repricing concentrated at the front end.

Nvidia reports second quarter results after Wednesday’s United States close, which is a larger near term influence on the equity levels in the table above than any of Wednesday’s macroeconomic data.

Sources: the United States Bureau of Economic Analysis, the United States Census Bureau, the Board of Governors of the Federal Reserve System, the United States Department of the Treasury, Reuters and CNBC. Gross domestic product and personal income and outlays are the Bureau of Economic Analysis releases of 26 August 2026; durable goods orders are the Census Bureau advance report of the same date and are subject to revision. The federal funds target range, the nine to three July vote, the three dissents and the 15 and 16 September meeting dates are from the Federal Reserve’s own statement of 29 July 2026 and its published meeting calendar. The September release dates for the August employment report, producer price index and consumer price index are from the Bureau of Labor Statistics schedule. This report does not cover the Bureau’s contributions to growth by component, the change in private inventories, real imports and exports, orders for non defence capital goods excluding aircraft, or the implicit price index for gross domestic product, none of which The Edge obtained from the release tables, and it draws no conclusions that would require them. Compounded two month rates and percentage point contributions are our calculation from the published figures. The Treasury curve is the official 25 August series. Market levels were captured at 19:32 GMT on 26 August, are intraday rather than closing prices, and were taken while the United States session was open.

Get Complimentary Insights

Follow us