Consumer Sentiment Report: Current Conditions Sour as Expectations Improve
Consumer Sentiment Report: Current Conditions Sour as Expectations Improve
The University of Michigan index fell 1.8 points to 46.3 in October, 1.5 points above its May record low, as current conditions dropped 6.2 points and expectations rose 1.0 point.
Early Signal
- Sentiment fell 1.8 points to 46.3 in the preliminary October survey. The index was 48.1 in September and is 1.5 points above the record low of 44.8 set in May. Economists had expected a reading near 47.7.
- Current conditions fell 6.2 points to 44.7. The index of current economic conditions was 50.9 in September, and forecasters had looked for a reading near 51.
- Expectations rose 1.0 point to 47.3. That is ahead of the 45.5 consensus and follows a 5.2-point decline in September.
- Inflation expectations rose for a second straight month. The year-ahead expectation is 4.7%, up from 4.6%, and the long-run expectation is 3.5%, up from 3.4%. Both are the highest since May, and the long-run reading is above the 2.8% to 3.2% range of 2024.
- The hard data on household spending have not confirmed a pullback. The September flash composite PMI of 58.4 was the strongest since 2021, and we hold our forecast for real consumer spending growth of 2.3% in 2026. The split between lower- and upper-income households, however, is widening.
Overview
The University of Michigan's Index of Consumer Sentiment fell 1.8 points to 46.3 in the preliminary October survey, from 48.1 in September. That is below the consensus range of 47.6 to 47.8 and 1.5 points above the record low of 44.8 set in May. Current Economic Conditions dropped 6.2 points to 44.7 and the Index of Consumer Expectations rose 1.0 point to 47.3. The final October reading is due October 23.
A quick reminder about what these numbers measure. Each question behind the Michigan indexes is scored as the percent of households answering favorably less the percent answering unfavorably, plus 100, so the indexes report the breadth of sentiment and not its intensity. A household that moves from “good” to “bad” counts the same whether its view slipped a little or a lot. We try to use breadth language throughout our analysis and reserve words like growth and spending for the hard data.
September set the baseline. Sentiment fell 3.6 points to 48.1, the lowest reading in four months, and the damage was in expectations, which fell 5.2 points to 46.3 while current conditions eased 1.0 point to 50.9. The University of Michigan pointed to elevated fuel prices, high prices that keep climbing and renewed trade disputes. The decline was shared across political parties, with Republican sentiment 20% below January and Democratic sentiment 13% below.
What Changed In October
Current conditions took the hit. Buying conditions for durable goods deteriorated as high prices and borrowing costs discouraged purchases. Sentiment weakened the most among lower-income households and households with smaller stock portfolios, which have fewer resources to absorb higher prices. This is one of the primary transmission mechanisms of monetary policy and suggests the Fed will need to do less. Table 1 lists the five headline readings.
| Indicator | October (prelim.) | September | Change |
|---|---|---|---|
| Consumer sentiment | 46.3 | 48.1 | -1.8 |
| Current economic conditions | 44.7 | 50.9 | -6.2 |
| Consumer expectations | 47.3 | 46.3 | +1.0 |
| Year-ahead inflation expectation | 4.7% | 4.6% | +0.1 pt |
| Long-run inflation expectation | 3.5% | 3.4% | +0.1 pt |
Table 1. October's preliminary readings against September's final readings. Source: University of Michigan, Surveys of Consumers.
Expectations improved a little. Households are more dissatisfied with their present circumstances and somewhat less pessimistic about the coming year. That points to affordability pressures today as the source of the weakness. The expectations index beat the 45.5 consensus even as the current conditions index missed the 51.0 consensus by more than six points.
Inflation Expectations
Both inflation expectation measures rose for a second straight month. The year-ahead expectation rose to 4.7% from 4.6% and the long-run expectation rose to 3.5% from 3.4%, the highest readings for each since May. The long-run expectation stayed between 2.8% and 3.2% throughout 2024.
The long-run expectation is the number to watch. If households now treat higher inflation as a lasting feature of the economy, the long-run expectation is where it will show first. Energy prices and trade-related costs can lift expectations for a time without lifting underlying inflation. A persistent rise in the long-run measure would complicate monetary policy, with inflation still above the Fed's 2% objective.
Sentiment And Spending
Spending has not followed sentiment lower. Households are dissatisfied with the economy, yet many are still spending, supported by employment income and wealth. The Michigan survey measures perceptions, and those are shaped heavily by prices, politics and media coverage. Spending depends more directly on employment, income, wealth and access to credit.
The Conference Board survey carried the same message in September. Its index fell 6.7 points to 81.9, and fuel prices dominated what respondents wrote in. The September flash composite PMI of 58.4 was the strongest since 2021, and the labor market differential in the Conference Board survey stood at +1.7%, just above zero.
OUR CALL
The consumer sector is slowing without contracting. The sharp drop in buying conditions for durable goods is a downside risk to discretionary spending, particularly for households with limited financial buffers. Even so, sentiment has been an unreliable guide to spending for most of this expansion, and employment, real disposable income and household wealth are better guides to the near term. We hold our forecast for real consumer spending growth of 2.3% in 2026 and 2.0% in 2027 (September 23 vintage), and we leave the Fed call unchanged: a hold at the October 27-28 meeting, a second quarter-point increase on December 9 to 4.00%-4.25%, and at least one more in 2027.
The test is the Conference Board's October report, released October 27, the first day of the FOMC meeting. If the labor market differential turns negative, with more households calling jobs hard to get than plentiful, we will take our consumer spending forecast lower. If it holds above zero, we read the Michigan weakness as a response to prices and keep the forecast where it is.
Bottom Line
Households are less comfortable with the economy they have than with the one they expect, which is an uncomfortable combination but short of a recession signal. The Michigan index is built differently from the Conference Board's, and the 80 threshold in our midterm research does not carry over to it. The Conference Board's October report on October 27 is the last before the November 3 election, and it is the one we will use to test our spending forecast.
What We Are Watching
- Final October reading, October 23: the preliminary survey is subject to revision, and the split between current conditions and expectations is the part to check.
- Long-run inflation expectations: the 3.5% reading is above its 2024 range, and a third consecutive increase would add to the case that households see higher inflation as lasting.
- Durable goods buying conditions: the deterioration in October is the main downside risk to discretionary spending.
- Labor market differential, October 27: the Conference Board's +1.7% reading sits just above zero, and a move below zero would change our consumer spending forecast.
- The FOMC meeting, October 27-28: our call is a hold, and the inflation expectations in this survey are part of the case for patience.
Mark P. Vitner
Chief Economist, Piedmont Crescent Capital
mark.vitner@piedmontcrescentcapital.com | 704-458-4000
Sources: University of Michigan, Surveys of Consumers, preliminary October 2026 (released October 9) and final May through September 2026; investingLive, consensus forecasts; The Conference Board, Consumer Confidence Survey, September 2026; S&P Global, flash U.S. PMI, September 2026; Piedmont Crescent Capital, Consumer Confidence Report (September 29, 2026) and forecast vintage of September 23, 2026.
© 2026 Piedmont Crescent Capital, LLC. For informational purposes only. Not investment advice.
More Measured Guidance, Not No Guidance
A Piedmont Crescent Capital Special Report
More Measured Guidance, Not No Guidance
Chairman Warsh wants monetary policy off the emergency footing it has held since 2008. We expect a Fed that says less, owns less and sets a more normal policy rate. It will still look little like the Fed of 2006, but will be striving for a more normal operating process.
Last week’s reset of October rate expectations showed what Chairman Warsh is trying to do and how far the Committee can take it. Warsh has said he wants investors to watch the economy and not the Fed. At his July 29 press conference he put it as “market participants are learning to play the ball, not the referee.” After the September 16 increase to 3.75%-4.00%, the market-implied odds of another in October climbed to about 70%. Then New York Fed President John Williams said a further increase “late this year” might be appropriate but that there was “no need for urgency,” Vice Chair Philip Jefferson struck a similar tone, and the odds fell to about 25% before the September jobs report pushed them lower still. A week earlier Williams had told an audience in London that the era of forward guidance was over.
The episode marks the edge of the experiment, and the edge is a useful guide to the destination. Eric Rosengren, who ran the Boston Fed until 2021, told the Journal’s Nick Timiraos that “you don’t have to promise a path of rates for people to find an implied path of rates.” Loretta Mester, who led the Cleveland Fed for a decade, observed that “nature abhors a vacuum” and that only the chair can speak for the Committee. EY-Parthenon’s Gregory Daco has described the same pattern from the market side, with Williams and Governor Christopher Waller filling a communication void and investors reading individual data points as policy tests. Each of those officials described a reaction function, a statement of what he or she would do if nothing else changes, and something else always changes. Investors treated the conditional as a plan.
Warsh’s own words describe a smaller role for forward guidance and stop short of abolishing it. At Jackson Hole on August 28 he said the practice “was adopted by my colleagues and me during the Global Financial Crisis,” that it was “essential at the time” and that it “has overstayed its welcome.” He then said, “In normal times, the role of forward guidance should be limited and circumscribed.” His objection, as he told senators at his April 21 confirmation hearing, is that the Fed tells the world its forecasts and then holds “on to those forecasts longer than they should.” The problem is commitment, and an explanation of how the Committee reads the economy does not create one. Warsh appears to agree, having introduced his Jackson Hole remarks as an outline or a trail map and added, “just don’t call it forward guidance.” Reuters called the result a “smidgen of guidance.” We prefer measured, or disciplined, forward guidance, which means less of it but not none.
The Fed used to say much less, and that was probably too little. I remember when the Fed kept to itself. Until February 4, 1994, the Committee issued no statement after its meetings, and the market had to work out what it had done by studying the Desk’s open market operations and, after 1979, the weekly money supply numbers. In 1988 the funds rate target changed 13 times, and only 3 of the changes came at or around a scheduled meeting, while 4 were relayed to the Committee on impromptu calls and 6 were tied to no recorded FOMC communication. The policy directive was published with a lag, about 90 days at first and later 45 days, and the lag survived a Freedom of Information Act suit filed in 1975. Observers weighed those scraps, along with the economy’s recent performance, against the projections the Chair laid out in the semi-annual Humphrey-Hawkins testimony. The Committee tended to move on the back of key reports, mostly the employment report and occasionally the purchasing managers’ survey when it held a major surprise. On January 3, 2001, a day after that survey fell to 43.7, its lowest since April 1991, the Committee cut the funds rate 50 basis points to 6.00% on an unscheduled conference call, nearly four weeks before its January 30-31 meeting. William Greider’s 1987 account of the Volcker era, Secrets of the Temple, described a Fed run like a temple whose priests held the secrets. Allan Meltzer, whose two-volume History of the Federal Reserve drew on the Committee’s own records, argued that discretionary policy had failed in 1929-33 and 1965-80 and called for more rule-like behavior, in part because a Fed that can tell Congress whatever it wishes is hard to hold to account. Silence matched a time before 24-hour news channels, dedicated financial channels and the internet. Even then, economists scrutinized the Fed’s statements and testimony for how it would react if growth or inflation ran above or below its projections. That is a reaction function, and it is the kind of guidance we think the Committee should keep.
The pre-crisis Fed was not silent, which is the case for calling this a normalization. The Committee has announced its funds-rate decision after every meeting since 1994, and in the years before the crisis it described its leanings in plain language. In August 2003 it said policy accommodation could be maintained for “a considerable period,” in January 2004 that it could be “patient” and in May 2004 that accommodation would be removed at a pace “likely to be measured.” Those phrases carried no date, no threshold and no projected path. What followed the crisis was different in kind. The Committee tied the funds rate to a calendar in August 2011 (“at least through mid-2013”) and to a 6.5% unemployment threshold in December 2012, began publishing the dot plot in January 2012, held a press conference after every meeting from 2019 and in 2020 promised to hold the rate at zero until inflation reached 2% and was on track to moderately exceed it. Returning to the pre-crisis template means keeping the first kind of language and retiring the second.
Forward guidance and large asset purchases were emergency tools because the policy rate had no room left to work with. When the funds rate sits at zero, a central bank that wants to ease further can change what investors expect for the rate in the future, which is guidance, or buy long-term securities, which is quantitative easing. Waller drew the same line on September 3, describing forward guidance as a tool for the effective lower bound and agreeing with Warsh that it is not appropriate now. The effective funds rate averaged less than 0.25% in 36 quarters between the start of 2009 and early 2022, in two separate episodes. It is 3.88% today, and each tool built to substitute for the rate should recede as the rate returns.
The list of emergency measures is longer than the three that get the headlines. The Committee cut the funds rate to a range of zero to 0.25% on December 16, 2008, and then bought assets in four waves, starting with up to $600 billion announced in November 2008 and enlarged by $1.05 trillion in March 2009, followed by $600 billion of Treasury securities in November 2010, $40 billion a month of mortgage-backed securities beginning in September 2012 with $45 billion of Treasury securities added that December, and open-ended purchases in March 2020. Operation Twist, which sold short-dated Treasury securities to buy long-dated ones, began in September 2011 at $400 billion and was extended by $267 billion in June 2012. Alongside came interest on reserves, emergency lending facilities, swap lines with foreign central banks, the dot plot, average inflation targeting and, as late as March 2023, the Bank Term Funding Program. Several of those are retired already, among them the unemployment threshold, Twist and the term funding program. Interest on reserves, the dot plot and a very large balance sheet remain.
The research on what these programs accomplished supports measured guidance. Much of the effect arrived on the day of the announcement. Gagnon, Raskin, Remache and Sack of the New York Fed found that longer-term rates fell by as much as 150 basis points around eight key announcements between November 2008 and November 2009, including 91 basis points on the 10-year Treasury yield. Krishnamurthy and Vissing-Jorgensen concluded that signaling, the market’s reading of purchases as a promise to keep the funds rate low for longer, played the primary role in the second round of purchases, and they asked whether guidance alone could have done the job. Eric Swanson, then at the San Francisco Fed, found that six announcements in the original 1961 Operation Twist moved long-term Treasury yields by about 15 basis points. The researchers disagree on the mechanism, since Gagnon and his coauthors attribute most of the decline to a lower term premium and not to lower expected short rates, and Greenlaw, Hamilton, Harris and West argued in 2018 that the effects of Fed announcements tended not to persist. Event studies also have a limitation, because purchases were often announced alongside bad economic news. Markets look forward, and they priced in what the Committee said it would do before it did it. Words carry real power, which is the reason guidance should be measured, and a balance sheet reduction that is announced and gradual should move yields by little. Jonathan Wright estimated in a 2022 Brookings paper that running off Treasury holdings would raise the 10-year term premium by only about 10 basis points.
The balance sheet is the hardest of the three to normalize, and we do not expect it to return to its pre-crisis size. Total assets were $6.74 trillion at the end of September against $0.87 trillion in the first quarter of 2007 and a peak of $8.94 trillion in the first quarter of 2022. Runoff ended on December 1, 2025, and purchases of Treasury bills to manage reserves lifted assets about $100 billion before the Desk paused them, and the September minutes released today say the pace of any resumption is “not on a preset course” and will be set to keep reserves ample. Jon Faust of Johns Hopkins writes that most of the growth since 2008 reflects the 2019 shift to an ample-reserves system, which most economists consider superior to the scarce-reserves system it replaced. Regulation has also raised banks’ demand for reserves, which now stand at about $3 trillion, roughly 9% of GDP against less than 1% before the crisis (First Trust; MetLife Investment Management). Stanford’s Darrell Duffie argues that shrinking the balance sheet without first reducing demand for reserves will not work, and he lists regulatory changes, payment-system redesign and tiered interest on reserves as ways to reduce it.
Even the most ambitious proposal in circulation leaves a balance sheet nearly three times its 2007 size. Donald Luskin of TrendMacro has proposed a new Fed-Treasury accord that would take Fed assets down to the $2.4 trillion of currency outstanding, which is 2.8 times the 2007 level, and Jai Kedia and Norbert Michel of the Cato Institute would exit mortgage-backed securities and hold only short-term Treasury securities. Allianz expects a hybrid, with a leaner balance sheet backed by standing repo facilities, and ABN AMRO expects a gradual shortening of the Treasury portfolio, a Twist in reverse. The portfolio’s weighted average maturity is a little over eight years against 3.25 years before the crisis, and the $1.9 trillion of mortgage-backed securities is running off at about $20 billion a month, which takes eight years. Warsh announced task forces in June, one of them on the balance sheet, and hopes “most, if not all of them” conclude by year end. We expect a portfolio that is shorter in maturity, lighter in mortgage-backed securities and somewhat smaller, with the size coming down only as fast as the demand for reserves does.
A smaller balance sheet adds supply of long-dated paper to a bond market that is already pushing back. The 10-year Treasury yield climbed from about 5% after the September 16 increase to around 5.25% by the start of last week, part of a global rise in long-term rates, and Jefferson cited the rise in his remarks. The minutes released today add that nominal yields rose about 35 basis points across the two- to ten-year range in the weeks before the meeting, and they note market commentary linking higher term premiums partly to uncertainty about the Treasury buyback program and heavy private debt issuance for AI infrastructure. Our August view was that the long end, and not the funds rate, is the binding constraint on the economy, because federal borrowing, an energy shock and the debt issuance behind the AI buildout are all competing for the same pool of capital. Warsh has said that unconventional policies should otherwise be “used sparingly, if at all” outside genuine crises, and a Fed that runs off duration into that market would add a fourth source of supply. That is the reason we expect the Committee to change the composition of the portfolio before it tries to change the size.
The policy rate also needs to look more normal, though it will not return to its pre-crisis norm. The effective funds rate held at 5.25% from mid-2006 to mid-2007 and averaged 4.4% over 1990-2007 by our calculation. The Committee’s longer-run median in the September projections is 3.2%, up from 3.1% in June, with individual participants ranging from 2.9% to 3.9%. A May note from the St. Louis Fed averaged several alternative models and put the real neutral rate at about 1.4% in the fourth quarter of 2025, which at 2% inflation corresponds to a funds rate a little above 3.4%, and it found that the Committee’s own estimate tends to adjust toward that average with a lag. We believe the AI capital buildout, the persistence of large federal budget deficits and rising risks to global supply chains, which are encouraging more investment in the U.S., argue for a neutral rate at least half a percentage point above the Committee’s 3.2% median and in the neighborhood of 4%. One reason we are not raising our estimate of neutral more aggressively is that so much of the recent growth has come from capital spending, which expands the economy’s ability to grow and so works mainly on the supply side. Demand for housing, the traditional transmission mechanism for monetary policy, remains quite soft, with real residential investment about 3% below its level a year earlier. The unusually wide dispersion in the Fed’s dot plot suggests the Committee is deciding whether to agree, and the minutes released today show that a couple of participants have already raised their estimates of neutral.
The real policy rate tells the same story. We have long found it useful to compare the real federal funds rate with the gap between actual and potential GDP, and the comparison is informative now. From 1990 to mid-2007 real GDP averaged almost exactly CBO’s estimate of potential, and the real funds rate, measured against core PCE inflation, averaged 2.2%. From the third quarter of 2009 through the third quarter of 2015 the real rate averaged about -1.4% while output ran about 2.6% below potential. In the second quarter of 2026 output is 1.4% above potential and the real rate is only 0.5%. The pre-crisis relationship is crude, with one variable and an R-squared of 0.45, but it implies a real rate of about 3.3% for a gap of that size. Our neutral estimate of about 4%, or roughly 2% in real terms, is close to the pre-crisis experience, so the distance shows that policy remains accommodative for an economy that has moved beyond the post-crisis and post-pandemic periods. Several participants said in the minutes released today that they view the current policy rate as not restrictive or only mildly restrictive. It also bears on the path below, since a funds rate of 4.25%-4.50% with core inflation near 3% would put the real rate at roughly 1.25% to 1.5%, close to the St. Louis Fed’s 1.4% estimate of neutral and still short of the roughly 2% implied by ours.
Our own path takes the funds rate to the old average and no further. We expect an increase on December 9 to 4.00%-4.25% and another in the first quarter of 2027 to 4.25%-4.50%. The minutes released today say most participants judged that another increase “would likely be appropriate by year end,” and many favored a higher path as insurance against inflation staying above target. The midpoint of 4.375% matches the 1990-2007 average and sits nearly 0.9 percentage point below the 2006-07 plateau. The Committee’s own median path then eases to 3.75%-4.00% in 2028 and 3.50%-3.75% in 2029 as inflation returns to 2%. A policy rate that settles near 4% leaves roughly 350 to 400 basis points of room before zero, and that room is the real argument for normalization, because the less often the rate sits at zero the less often the Committee needs guidance and asset purchases to do the work of a rate. Quantitative easing stays in the toolkit for genuine crises, as Warsh’s own formulation allows, and Faust expects “extremely strong pushback” inside the Committee to any move to restrict it once rates reach zero.
The dots show a Committee that has not settled on where the rate ends up. The longer-run projections of the 18 participants span a full percentage point, from 2.9% to 3.9%, with the largest cluster at 3.0% and a median of 3.2%. Our estimate of neutral in the neighborhood of 4% sits just above the top of that range. The near-term dots lean our way. Eight of the 18 participants place the funds rate at 4.25%-4.50% at the end of 2027, double the 4 who do at the end of 2026, and that is where our path puts it. The median still eases to 3.9% in 2028 and 3.6% in 2029, but the spread widens from half a point in 2026 to 1.25 points in 2027 and a full point in 2028 and 2029. We read that spread as a debate about neutral.
Three things could make this wrong, and the first is the calendar. The October 27-28 meeting falls six days before the midterm elections, and Rosengren notes that a one-meeting delay “won’t make a big difference economically but could make a big difference politically.” The second is the bond market, because if long-term yields keep climbing a Committee that says little has few ways to calm them. The third is the communications task force. Robeco’s analysts suggest the review could lead to a scenario-based projection format like those at the European Central Bank and the Bank of England, and a change of that kind would fit our view as long as the Committee still explains how it reads the data.
Most of our calls in this piece can be scored by the end of the year, and the balance sheet call by the end of 2027. The October 28 statement should carry no calendar, threshold or path language, as in every statement since June. The minutes released today, which record a 12-0 vote, show a Committee that approaches each meeting with “an open mind” and ties its decisions to incoming information, with no promised path. By year-end, the task forces should report, and we expect the communications review to leave a projection release in place in some form, so an outright end to the dot plot would mean we have the destination wrong. We expect the December 9 increase, which Williams has called “a reasonable way of thinking about it.” On the balance sheet, we expect total assets to be lower at the end of 2027 than today’s $6.74 trillion and no lower than $5 trillion, with composition changing first. The median respondent in the New York Fed’s July survey of market expectations sees assets of $6.90 trillion at the end of 2027, so the market does not yet agree with us on direction.
The end state we expect is a Fed that says less, owns less and sets a more normal policy rate. It describes how it reads the economy and stops short of promising a path, holds a portfolio that is shorter and smaller than today’s and still several times its 2007 level, and runs a funds rate that peaks near its old average and settles below its old plateau. Our call would break if long-term yields or credit spreads tightened financial conditions faster than the Committee intends, because a Fed that has just spent a year teaching markets to expect silence would then have to decide whether to speak.
Appendix A. The emergency toolkit, 2008 to 2026
| Tool | Introduced | Scale | Status, October 2026 | Our expectation |
|---|---|---|---|---|
| Zero policy rate | Dec. 16, 2008; again Mar. 2020 | Effective funds rate below 0.25% in 36 quarters | Retired; effective funds rate 3.88% | Peaks at 4.25%-4.50% in Q1 2027; settles near 4% |
| Large-scale asset purchases | Nov. 25, 2008; Mar. 18, 2009; Nov. 3, 2010; Sept. 13, 2012; Mar. 2020 | Assets from $0.9T in early 2008 to $4.5T at end-2014 and $8.9T in Mar. 2022 | Runoff June 2022 to Dec. 1, 2025; assets $6.74T; bill purchases to manage reserves, now paused | Smaller than today, above $5T by end-2027 |
| Maturity extension (Operation Twist) | Sept. 21, 2011; extended June 20, 2012 | $400B plus $267B, $667B in all | Completed in 2012; Treasury portfolio maturity a little over 8 years against 3.25 before the crisis | Portfolio shortens |
| Calendar and threshold guidance | Aug. 9, 2011; Dec. 12, 2012 | "At least through mid-2013"; 6.5% unemployment threshold | Retired | Not revived |
| Outcome-based guidance and average inflation targeting | Aug. and Sept. 2020 | Hold the rate at zero until inflation reached 2% and was on track to moderately exceed it | Dropped in the 2025 framework review | Not revived |
| Summary of Economic Projections (dot plot) | Jan. 25, 2012 | Median and range of participants’ funds-rate paths | Published Sept. 16, 2026; Chairman Warsh submitted no projection | Survives in some form |
| Press conference after every meeting | Jan. 2019 | Chair explains the decision at each meeting | Held at the June, July and September 2026 meetings, with little narrative | Continues, with less narrative |
| Interest on reserves and ample reserves | Oct. 2008; ample-reserves system from 2019 | Reserve balance rate 3.90% since Sept. 17, 2026; reserves about $3T | In place | Stays; size depends on reducing demand for reserves |
| Emergency lending facilities and swap lines | 2007-09; Mar. 2020; Bank Term Funding Program Mar. 2023 | Facility sizes varied | Emergency facilities closed; standing repo facility in place since 2021 | Held in reserve for crises |
| Reserve management purchases | Dec. 2025 | Treasury bills | Paused in September; pace set month to month to keep reserves ample | Pace revisited after the task force reports |
Appendix B. How far back toward the 2006-07 norm
| Measure | 2006-07 level | Post-crisis extreme | Today | Share reversed | Where we expect it to end |
|---|---|---|---|---|---|
| Effective funds rate | 5.25% | 0.125% (zero-bound midpoint) | 3.88% | 73% | Peaks near 4.4% in 2027; settles near 4% |
| Fed total assets | $0.87T (Q1 2007) | $8.94T (Q1 2022) | $6.74T (Sept. 30, 2026) | 27% | Lower than $6.74T by end-2027 and above $5T |
Sources and Notes
Sources. Federal Reserve Board (FOMC statements, press release of January 3, 2001, Summary of Economic Projections of September 16, 2026, minutes of the September 15-16, 2026 meeting released October 7, 2026, H.4.1, Chairman Warsh’s Jackson Hole remarks of August 28, 2026); Allan H. Meltzer, A History of the Federal Reserve; William Greider, Secrets of the Temple (1987); Money and Banking (FOMC communication history, 2019); CNN Money (January 2-3, 2001); Marvin Goodfriend, “Monetary Mystique: Secrecy and Central Banking,” Journal of Monetary Economics, 1986; Federal Reserve Bank of St. Louis (FRED; “Comparing the FOMC’s Estimate of R-Star with Alternative Estimates,” May 2026); Federal Reserve Bank of New York (Survey of Market Expectations, July 2026; Gagnon, Raskin, Remache and Sack, “Large-Scale Asset Purchases by the Federal Reserve: Did They Work?,” Staff Report 441); Federal Reserve Bank of San Francisco (Eric Swanson, “Let’s Twist Again,” 2011); Brookings Papers on Economic Activity (Krishnamurthy and Vissing-Jorgensen, 2011; Jonathan Wright, 2022); National Bureau of Economic Research (Greenlaw, Hamilton, Harris and West, Working Paper 24687); Congressional Budget Office and Bureau of Economic Analysis (via FRED); Kitco (Chairman Warsh’s July 29, 2026 press conference); The Wall Street Journal (Nick Timiraos, October 6, 2026); Reuters; American Banker; Johns Hopkins Krieger School of Arts & Sciences (Jon Faust); Axios (Darrell Duffie); TrendMacro (Donald Luskin); Cato Institute (Jai Kedia and Norbert Michel); First Trust Portfolios; MetLife Investment Management; ABN AMRO; Allianz; Robeco; Piedmont Crescent Capital calculations. © 2026 Piedmont Crescent Capital. All rights reserved.
Mark P. Vitner
President & Chief Economist, Piedmont Crescent Capital
mark.vitner@piedmontcrescentcapital.com · 704-458-4000
Piedmont Special Reports are published by Piedmont Crescent Capital for informational purposes only and do not constitute investment, legal, or tax advice. Views expressed are those of the author as of the date of publication and are subject to change.
September Employment Report: The Calendar Nets Out
September Employment Report: The Calendar Nets Out
Payrolls Rose 29,000 and July and August Were Revised Down a Combined 60,000, Leaving the Five Months From May Through September Averaging 49,000. The Unadjusted Summer Roll-Off in Leisure & Hospitality Was Unusually Small, but Re-Estimated Seasonal Factors Took Most of the Payoff, Breadth Narrowed and Our Call for the Fed Is Unchanged
Early Signal
- Nonfarm payrolls rose 29,000 in September, well short of a consensus near 85,000, and revisions took 60,000 jobs out of July and August. July now shows a decline of 10,000 from a gain of 21,000 and August a gain of 133,000 from 162,000, so the three-month average is 51,000 against 71,000 a month ago. The five months from May through September, a stretch bent by the earliest Memorial Day and the latest Labor Day, averaged 49,000 a month, close to the 45,000 average of the twelve months through August. Private payrolls rose 46,000 and government lost 17,000.
- The seasonal argument we made in August held in the unadjusted data and lost most of its payoff to the seasonal factors. Before seasonal adjustment leisure & hospitality shed 454,000 jobs from August to September, fewer than in any September from 2017 to 2025 outside the pandemic years, and restaurants and bars account for about four-fifths of the shortfall against the 514,000 average of the past three Septembers. The seasonal factors, however, implied a roll-off of about 464,000, between 52,000 and 80,000 less than in each of the last three years, so the adjusted gain was 10,000 against the 80,000 that the average of those years' factors would have produced.
- Our call for a strong August or September remains right, with August the hot month and September softer as the unwind arrived through the revisions. August cleared the 125,000 bar we set on August 23 at 133,000 even after revision. Leisure & hospitality accounts for 42,000 of the 60,000 downward revision to July and August, and local government education, where we looked for part of August's gain to come back out in September, gave back only 1,800 after an August gain revised up to 49,300.
- Set aside leisure & hospitality and local government education and payrolls rose about 21,000 in September and 47,000 in August, an average of 34,000 that sits inside our breakeven range of zero to 40,000. Hiring breadth narrowed, with the one-month private diffusion index at 49.0 from 57.6, its first reading below 50 since December, and the manufacturing index at 46.5 from 63.9. Both are one-month readings, and the six-month indexes, at 54.8 for private industries, unchanged, and 52.8 for manufacturing, up from 51.4, held up.
- Manufacturing payrolls rose 9,000 and are up 72,000 from December's low, and construction added 11,000. Machinery (+4,500) and plastics & rubber products (+4,600) led the factory gains, the ISM employment index rose to 52.7 for a third month above the 50.3 threshold ISM associates with rising factory payrolls, and nonresidential specialty trade contractors added 12,300 jobs while residential specialty trades lost 7,900.
- Health care added 17,000 jobs, about half its 33,000 average of the past twelve months, and the losses were concentrated in white-collar industries. Information lost 10,000 jobs, professional & business services 9,000 and financial activities 7,000, which leaves financial activities 129,000 below its May 2025 peak, 90,000 of it in insurance carriers. Local government outside education lost 10,600.
- The unemployment rate rose to 4.2% from 4.1%, and all of the rise came from reentrants and new entrants, as job losers declined. Reentrants rose 152,000 and new entrants 116,000, together 268,000, while job losers fell 45,000 to 3.2 million and job leavers fell 173,000. Unrounded the rate moved to 4.18% from 4.14% on a labor force that grew 485,000. Participation rose to 61.8% from 61.6% and 61.4% in July, and household employment rose 406,000.
- Average hourly earnings rose 5 cents, or 0.1%, to $37.81 and are up 3.0% over the past year, the slowest pace of the year. The private workweek held at 34.4 hours and the index of aggregate weekly hours was unchanged after rising in August, so hours added nothing further this month. Initial claims fell to 197,000 in the week ended September 26, the lowest since July. We hold our call for a pause on October 27-28 and a quarter-point increase on December 9. Bond yields fell following the report but a stronger economy in 2027 may see yields retest the highs reached over the past two weeks.
- Hours plus productivity point to third-quarter GDP growth of 2.7% to 3.5%. Hours worked rose at a 1.3% annual rate over the past three months, with the longer workweek doing most of the work, and productivity growth of 2.2% over the past year puts our proxy near 3.5%, or 2.7% at the second quarter’s slower productivity pace, against GDP growth of 2.2% in the second quarter.
The Calendar Argument, Scored
We said on August 10 to expect a surprisingly strong employment report in one or both of August and September, and on August 23 we put the bar at 125,000. August cleared it even after revision, at 133,000 against the 162,000 first reported, and September did not, at 29,000. We score the call as made, because August cleared the bar and September carried the unwind through the revisions, so the October data should be clear of this summer’s distortions. Payrolls added 152,000 over July through September, an average of 51,000 a month, and 246,000 over the five months from May through September, an average of 49,000, a span in which the earliest Memorial Day and the World Cup pulled hiring forward and the latest Labor Day delayed the roll-off. A distortion that nets out over a few months is what seasonal factors are built to do over a full year, and most of this is now behind us.
The mechanism we described in August has two halves, and the unadjusted data support both. Restaurants and bars hired more people in May and did not hire as many people as they usually do in June and July. They therefore have far fewer people to let go as the summer winds down in August and September. The seasonal factors for leisure and hospitality jobs in August and September compensate for jobs that are expected to decline. Since fewer jobs were added during the prior two months, the end-of-summer decline will be less than usual, resulting in a larger seasonally adjusted job gain. The first half held, since restaurants and bars added just 16,000 jobs from May to August before seasonal adjustment, against 124,000 in 2023, 55,000 in 2024 and 69,000 in 2025. The second half held in September, when restaurants and bars shed 148,400 against 179,300 last year, 198,500 in 2024 and 209,100 in 2023.
The shortfall in the roll-off was real, and almost all of it was in restaurants and bars. Leisure & hospitality shed 454,000 jobs from August to September before seasonal adjustment (from 17,551,000 to 17,097,000), fewer than in any September from 2017 to 2025 outside the pandemic years of 2020 and 2021 and 60,000 fewer than the 514,000 average of 2023 through 2025. Restaurants and bars lost 148,400, or 47,000 fewer than their three-year average of 195,600, about four-fifths of the shortfall, while arts, entertainment & recreation lost 221,000 against 224,000 to 233,000 in each of the past three Septembers and accommodation lost 84,400.
The seasonal factors, however, expected a smaller roll-off this year, and that is where the payoff went. The roll-off the factors implied for leisure & hospitality, the unadjusted change minus the adjusted one, was about 464,000 this September against 516,000 in 2023, 544,000 in 2024 and 541,000 in 2025. Had the factors expected the 534,000 average of those three years, the unadjusted loss of 454,000 would have printed as a gain of about 80,000 instead of 10,000. Restaurants and bars show the same pattern, with an implied roll-off of 159,000 this year against 209,000, 224,000 and 218,000 and an adjusted gain of 10,800. BLS re-estimates the factors with each release, and the August factors went the same way, taking the adjusted August gain in restaurants and bars to 33,800 from the 59,200 first reported. A factor that has seen one unusual summer appears to treat part of it as normal, which shrinks the adjusted payoff in the months where we expected it to be largest.
The unwind arrived through the revisions, and local government education gave back nothing. Of the 60,000 downward revision to July and August, 42,000 is in leisure & hospitality, where July is now a decline of 38,000 and August a gain of 37,000 against first prints of -21,000 and +62,000, and BLS attributes the monthly revisions to additional reports and the recalculation of seasonal factors. Employment in the industry is 16,969,000, 45,000 below May's 17,014,000, so the jobs lost over June and July have not been recovered. We said on September 4 to look for part of local government education's August gain to come back out in September, and it did not, with a decline of 1,800 after an August gain revised up to 49,300. Before adjustment districts added 723,000 workers from August to September, in line with 726,000 in 2025 and 723,000 in 2024.
A Narrower Base of Hiring
The September weakness was spread across industries, with leisure & hospitality and local government education adding only 8,000 jobs between them. Health care and social assistance supplied 23,000 of the 29,000 gain, construction 11,000 and manufacturing 9,000, while information (-10,000), professional & business services (-9,000), financial activities (-7,000) and local government outside education (-10,600) took away about 37,000. Private payrolls rose 46,000 and government lost 17,000, with federal employment down 1,000, state down 3,000 and local down 13,000.
The one-month diffusion index for private industries fell to 49.0 from 57.6, the first reading below 50 since December. Remember that a diffusion index measures the breadth of hiring and not the magnitude, so the move says that fewer industries added workers in September than in August, when the index was revised up from 55.6 to 57.6, and says little about how many jobs they added. The manufacturing index fell to 46.5 from an upwardly revised 63.9 (61.1 as first published), a swing of 17 points, though June's 57.6 followed a May reading of 47.9, so swings of that size are not unusual in this series. The six-month indexes, which smooth that noise, were 54.8 for private industries, unchanged, and 52.8 for manufacturing, up from 51.4, so the broader trend is intact even if September's breadth was not. This is a key variable to watch next month.
Manufacturing payrolls rose 9,000 and are up 72,000 from December's low, so the factory recovery remains intact in a narrower month. Durable goods added 8,000 and nondurables 1,000, led by machinery (+4,500) and plastics & rubber products (+4,600), while paper (-2,600), furniture (-2,000) and printing (-1,800) gave back jobs. Printing is one of the two industries that ISM reported as contracting in September. July's gain was revised to 20,000 and August's to 15,000, so factory payrolls have added 44,000 over three months, about 15,000 a month. The ISM employment index rose 1.5 points to 52.7 in September, a third month above the 50.3 threshold ISM associates with rising factory payrolls, so the survey and the payroll data agree on direction. The factory workweek held at 40.6 hours with overtime at 3.0, and aggregate weekly hours in manufacturing rose 0.1%. Hiring follows orders by several months, and the ISM new orders index rose to 55.3 in September from 53.7, so we look for small factory gains through the fall.
Construction added 11,000 jobs, and the nonresidential side did the hiring. Nonresidential specialty trade contractors added 12,300 jobs and heavy & civil engineering 2,600, while residential specialty trade contractors lost 7,900. Over the past year nonresidential construction employment (building, heavy & civil engineering and specialty trades) is up 142,000 and residential employment is down 33,000. Heavy & civil engineering, the category that builds data center campuses, substations and the site work for large plants, is up 26,300, or 2.2%, over the year, and the industry's 11,000 gain compares with an average of 10,000 over the past twelve months.
Health care added 17,000 jobs, about half the 33,000 monthly average of the past twelve months. Ambulatory health care services added 13,400 and hospitals 12,000, while nursing and residential care facilities lost 8,700, including 7,000 at continuing care retirement communities and assisted living. Social assistance added 6,300.
The losses were in white-collar industries, led by information, professional & business services and financial activities. Information lost 10,000 jobs and is 120,000, or 4.2%, below a year ago, with publishing down 4,000, broadcasting 3,000 and computing infrastructure and data processing 1,600, and computer systems design lost 4,400 and is 32,200 below a year ago. Professional & business services lost 9,000, with temporary help services down 10,900 and administrative and support services down 17,600. Financial activities lost 7,000 and is 129,000 below its May 2025 peak, with insurance carriers and related activities accounting for 90,000 of the decline, according to BLS.
The Unemployment Rate Rose on Entrants
The unemployment rate rose to 4.2% from 4.1%, and all of the increase came from people entering or reentering the labor force. Reentrants rose 152,000 to 2.29 million and new entrants rose 116,000 to 818,000, together 268,000, while job losers and people who completed temporary jobs fell 45,000 to 3.2 million and job leavers fell 173,000 to 741,000, reversing most of August's 121,000 increase. Household components are rounded and seasonally adjusted independently and do not sum to the 78,000 rise in the number of unemployed, but the direction is clear, since the people added to the unemployment count were looking for work and fewer of them had lost a job. Unrounded the rate moved to 4.18% from 4.14%, within the 4.1% to 4.3% range it has occupied since March.
The labor force grew 485,000 and participation rose two tenths to 61.8%, matching its level in April and May. Participation is up four tenths from July's 61.4%, household employment rose 406,000 and the employment-to-population ratio rose to 59.2% from 59.1%. The number of people not in the labor force fell 346,000 and the marginally attached fell 236,000 to 1.5 million. U-6 fell to 7.6% from 7.7% even as the number working part time for economic reasons rose 111,000 to 4.5 million, and the long-term unemployed were 1.9 million, 27.1% of the unemployed.
Prime-age workers drove the increase, the reverse of August. Participation at 25 to 54 rose three tenths to 83.7% and the prime-age employment-to-population ratio rose three tenths to 80.7%, while participation at 55 and over fell three tenths to 36.9%, back to July's reading. A cohort with a participation rate near 37% keeps taking a larger share of the adult population from cohorts whose rates exceed 83%, and Peak 65 adds roughly 4.1 million people a year to it through 2027, so the aggregate rate can fall even if no one changes their mind about working.
The immigration story is less clear than it looked a month ago. Before seasonal adjustment the foreign-born labor force was 32.28 million in September, 206,000 larger than a year earlier, against a gap of 379,000 the other way in August, while the native-born labor force was 1.13 million smaller, at 138.02 million against 139.14 million. Foreign-born participation was 66.0% against 66.3% a year ago and the foreign-born unemployment rate was 3.2% against 4.1%. These are single-month readings from a small subsample, and the January population controls affect year-over-year levels, so we would not rebuild the supply argument on one month, but the larger decline in the labor force is in the native-born population, which fits retirement.
Our September 28 View put the payroll pace consistent with a stable unemployment rate at the mid-teens to the high 80,000s a month, and a gain of 29,000 sits inside that range, with the rate moving four hundredths of a point. It also sits inside our breakeven range of zero to 40,000. A labor force that rebounds this quickly will not keep that breakeven near zero, and if participation holds near 61.8% the payroll pace needed to hold the jobless rate steady rises toward the middle of the range. We expect a modest rebound in foreign-born workers following the midterm elections and look for stronger overall job growth in 2027.
Wages, Hours and the Fed
Wage growth slowed again, and hours did not add to it. Average hourly earnings rose 5 cents, or 0.1%, to $37.81, and the year-over-year rate eased to 3.0% from 3.1% in August and 3.7% in February, the slowest pace of the year. Production and nonsupervisory pay rose 7 cents, or 0.2%, to $32.60. The private workweek held at 34.4 hours, the index of aggregate weekly hours was unchanged after a gain in August and aggregate weekly payrolls rose 0.2%, so August's jump in hours held in place and added nothing further. A wage gain of 3.0% is consistent with the Fed's 2% inflation objective once productivity growth of 2.2% over the year through the second quarter is counted, and unit labor costs rose just 1.4%.
Hours worked and productivity put third-quarter GDP growth between 2.7% and 3.5%. Our GDP proxy adds the growth in hours worked to the growth in productivity. The index of aggregate weekly hours for all private employees averaged 117.0 in the third quarter, 0.3% above the second quarter and a 1.3% annual rate, with the longer workweek, 34.4 hours against 34.3, adding 0.8 points and payroll growth 0.4. Productivity rose 2.2% over the year through the second quarter, which puts the proxy at about 3.5%, against real GDP growth of 2.2% in the second quarter and the Atlanta Fed’s GDPNow estimate of 3.7% for the third. Productivity grew at a slower 1.4% annual rate in the second quarter, however, and the proxy at that pace is 2.7%. The payroll-based hours measure also ran ahead of BLS in the second quarter, growing at a 1.2% annual rate against the 0.3% that BLS reported for hours worked in the nonfarm business sector, so we lean toward the lower half of the range, which is still a faster pace than the second quarter and one that rests on a longer workweek more than on hiring.
The report lowers the pressure on the Committee for October 27-28 and leaves our December call intact. Before the release futures put the odds of a hold at the October meeting near 72%, the 2-year Treasury yield was 4.80%, the 10-year 5.24% and the 30-year 5.60%, after reaching its highest level since 2004 on September 24, so the long end was already tightening financial conditions ahead of the labor data. The market reaction was a relief rally. By 9:10 a.m. the 2-year yield was 4.74%, down about 5 basis points from Thursday's close of 4.79% after touching 4.69%, and December S&P 500 futures were up about 1%, near 7,800. A payroll gain of 29,000 against a consensus near 85,000, a 0.1% wage gain and a jobless rate that rose on entrants give the doves what they need for October. The report does not settle the question the hawks are asking, because September produced soft payrolls and a rate that rose on entrants, which is neither the falling rate that would have sharpened our labor-force argument nor the reacceleration in hiring that would have called it into question.
We are maintaining our call. The Committee pauses on October 27-28 and raises the target range a second time on December 9, to 4.00% to 4.25%, with at least one further increase in 2027 carrying the terminal range to 4.25% to 4.50% or possibly slightly higher. Every reaction function is ceteris paribus, and the conditions that have moved since the Committee last met are an unemployment rate of 4.2%, wage growth of 3.0% and long yields at multiyear highs, which point in different directions. We will test the call against the October report on November 6. Payrolls between 20,000 and 80,000 with the private diffusion index back above 50 keep December on track, and a negative print with the diffusion index still below 50 would weaken it.
Our Call · The Calendar Nets Out, and Our Call Holds
Payrolls rose 29,000 against a consensus near 85,000 and the unemployment rate rose to 4.2%, which will read as a weak report, but the five months from May through September averaged 49,000 a month, close to the 45,000 average of the past twelve months and just above our breakeven range of zero to 40,000. The seasonal argument we made in August worked in the unadjusted data, where restaurants and bars shed far fewer workers than usual after hiring far fewer over the summer, and the re-estimated factors then took most of the payoff. We score our call for a strong August or September as made, with August the hot month and September carrying the unwind, and we do not look for more help from the calendar. Job losers fell, the entire rise in the jobless rate came from entrants, prime-age participation rose and claims fell to 197,000, so the labor market looks small and in balance, not weak. Breadth narrowed in a single month, with the private diffusion index at 49.0, and we want it back above 50 in the October report on November 6. We hold our call on the Fed: a pause on October 27-28, a quarter-point increase on December 9 to 4.00% to 4.25% and at least one more in 2027 that carries the terminal range to 4.25% to 4.50% or higher. That call is ceteris paribus, and the readings that matter most between now and December are core PCE and the long end, since a labor market adding 30,000 to 50,000 jobs a month with a labor force that is growing again does not by itself argue for an earlier or later move.
Mark P. Vitner – Chief Economist, Piedmont Crescent Capital
mark.vitner@piedmontcrescentcapital.com · (704) 458-4000
Sources: U.S. Bureau of Labor Statistics, The Employment Situation, September 2026, including the Current Employment Statistics and Current Population Survey tables; U.S. Bureau of Labor Statistics, Productivity and Costs, Second Quarter 2026; U.S. Department of Labor, Unemployment Insurance Weekly Claims; Institute for Supply Management; Board of Governors of the Federal Reserve System; Federal Reserve Bank of Atlanta, GDPNow; Dow Jones and Bloomberg consensus estimates; CME Group. Disclaimer: This report is provided for informational purposes only and does not constitute investment, legal, tax, or accounting advice, nor an offer or solicitation to buy or sell any security. Information is drawn from sources believed to be reliable, but its accuracy and completeness are not guaranteed. Views expressed are those of the author as of the date of publication and are subject to change without notice. Past performance is not indicative of future results.
ISM Manufacturing Report: The Order Book Reopened as Prices Snapped Back
The Order Book Reopened as Prices Snapped Back
The Manufacturing PMI came in slightly below expectations, edging a tenth of a point lower to 54.5 in September, but new orders, backlogs and hiring improved while production, inventories and deliveries eased, the Prices Index jumped 6.8 points to 77.9, and the regional surveys, led by a 58.8 reading in Chicago, provided confirming evidence of a strengthening factory sector.
Early Signals
- The expansion held its pace. The Manufacturing PMI registered 54.5 in September, a tenth below August’s 54.6 but still marking the ninth consecutive month above 50. ISM notes that reading is consistent with 2.4 percent real GDP growth on an annualized basis, the same as August.
- The order book reopened. The New Orders Index rose 1.6 points to 55.3 and the Backlog of Orders Index jumped 4.6 points to 56.4, its highest reading since February. Together they recovered just over half of August’s decline in new orders and all of the decline in backlogs.
- Hiring improved and production eased. The Employment Index rose 1.5 points to 52.7, a third month above 50. The Production Index fell 1.6 points to 56.7, an 11th month of growth, and the Inventories Index slipped into contraction at 48.6.
- Prices jumped. The Prices Index rose 6.8 points to 77.9, the largest one-month increase since March and the highest reading since May, a 24th month of rising input costs. The share of respondents paying more rose to 58.6 percent from 46.2 percent.
- This past month’s regional surveys were strong but uneven. The Chicago Business Barometer, the old NAPM-Chicago, jumped 11.7 points to 58.8, while the average of five regional Fed surveys eased to 13.4 from 18.7. Four of the five districts were positive, led by Philadelphia at 37.8, and Richmond, where building products, furniture and home textiles are key industries, slipped to a negative 2.
Key Takeaways
| Key Concept | Findings |
|---|---|
| Manufacturing PMI | Slipped to 54.5 from 54.6, a ninth straight month of expansion following a ten-month contraction. The 12-month average is 52.3, with a high of 55.6 in July and a low of 47.9 in December. ISM puts that reading in line with 2.4 percent real GDP growth on an annualized basis. |
| New Orders and Backlogs | New orders rose 1.6 points to 55.3 and backlogs jumped 4.6 points to 56.4, the highest since February. Customers’ inventories fell to 41.6 from 42.8 and have been judged too low for 24 consecutive months, which historically points to future production. A Chemical Products panelist called that industry’s improvement temporary. |
| Production | Down 1.6 points to 56.7, an 11th consecutive month of growth and the lowest reading since June. The unadjusted responses were essentially unchanged from August, so the decline reflects the seasonal adjustment. |
| Employment | Up 1.5 points to 52.7, a third month in expansion after 17 straight months below 50. The ratio of panelists hiring to those managing or reducing head counts widened to 1.5 to 1 from 1.3 to 1. Most of the gain in the published index came from the seasonal adjustment. |
| Prices | Up 6.8 points to 77.9, a 24th month of increases and the largest one-month rise since March. The share paying more rose to 58.6 percent from 46.2 percent, while the share paying less fell to 2.8 percent from 4.0 percent. |
| Supplier Deliveries | Slower for a tenth month, at 59.0, down 0.3 point from August. Electronic components have been in short supply for 19 consecutive months, electrical components for 15 and memory for nine. |
| Trade | New export orders fell 2.3 points to 50.9 and the Imports Index fell 1.5 points to 51.0, both still in expansion. Panelists cited a Canada tariff and the broader trade war. |
| Breadth | Twelve of 18 industries grew and two contracted, Printing & Related Support Activities and Textile Mills. By GDP weight, the contracting share fell to 2 percent from 22 percent, almost all of the change coming from Chemical Products, which flipped to growth. Five of the six largest industries expanded. |
| Regional Surveys | Chicago jumped 11.7 points to 58.8, the highest since May. The average of five regional Fed surveys eased to 13.4 from 18.7 but was above zero for a third straight month, with Philadelphia, Dallas and Kansas City strong, and Empire State and Richmond weaker. |
| Policy Signal | Reaccelerating orders and hiring alongside input costs near 78 argue against easing. We continue to look for the Fed to hold on October 27-28 and to raise the target range on December 9 to 4.00 to 4.25 percent, with at least one further increase in 2027 carrying the terminal range to 4.25 to 4.50 percent or higher. |
The Overview
The Manufacturing PMI slipped a tenth of a point to 54.5 in September, its ninth consecutive month above 50, and the headline hides a rearrangement underneath it. Two of the five components that feed the index improved, new orders and employment, and together they added about six tenths of a point. Production, supplier deliveries and inventories took away a little more than that. The gains came in the orders-and-hiring part of the survey and the declines came in the output-and-inventory part. Inventories fell 2.0 points to 48.6 and subtracted the most, four tenths of a point on their own, which is as likely to mean demand is drawing stock down as that firms are pulling back, since customers’ inventories, at 41.6, have been judged too low for 24 consecutive months. Remember that the ISM index is a diffusion index and provides a reading on the breadth of the strength in the factory sector, not the magnitude of that strength. That said, orders are the most forward-looking component of the ISM survey, and the gains in orders suggest manufacturing activity is ramping up.
The breadth measures in this report swung by far more than the headline. Twelve of 18 industries grew in September and two contracted, Printing & Related Support Activities and Textile Mills, yet ISM reports that only 2 percent of manufacturing GDP was contracting, down from 22 percent in August. The swing is one industry. Chemical Products is the largest manufacturing industry by value added, $579 billion of the sector’s $2.9 trillion in 2025 on Bureau of Economic Analysis figures, close to a fifth of the total, and it flipped from contraction to growth in September. Printing, at $43 billion, and textile mills and textile product mills, at $18 billion, together come to about 2 percent of the sector, which matches the share ISM reports as contracting and implies that essentially all of it was in strong contraction. Five of the six largest industries grew.
The factory-cycle sequence we have cited all year runs from orders to backlogs to hours worked to hiring, and in September the front of that chain turned back up. Last month we said the next two order readings would settle whether the recovery was decelerating or ending. September is the first of the two. New orders recovered just over half of August’s decline, to 55.3 from 53.7, and backlogs recovered all of it, to 56.4 from 51.8. The back of the chain was already improving. Average weekly hours in manufacturing were 41.7 in August for a third month, the highest since June 2019, and manufacturing payrolls added 16,000 jobs in August after 14,000 in July and 13,000 in June. With orders and backlogs up, hours at a seven-year high and hiring rising, the chain is intact. One order reading does not make a trend, and October’s is the second of the two.
The Prices Index jumped 6.8 points to 77.9, and the way it got there is the mirror image of August. Last month the index held at 71.1 because the shares paying more and paying less both thinned. This month the share paying more rose 12.4 points to 58.6 percent, the highest since May, while the share paying less slipped to 2.8 percent from 4.0 percent. The movement came out of the middle of the distribution. The share reporting no change fell 11.2 points to 38.6 percent, so the standstill on costs we described in August did not last.
The Order Book Reopened
New orders rose 1.6 points to 55.3 in September and the Backlog of Orders Index jumped 4.6 points to 56.4, its highest reading since February. Both have now been in expansion for nine consecutive months, so September reverses most of what August gave back. Panel sentiment on demand did not improve with the indexes. The ratio of positive to negative demand comments narrowed to 1.7 to 1 from 2 to 1 in August and 3.5 to 1 in July, and across all topics 60 percent of the comments were negative. The panel is describing demand more cautiously than the index does.
The raw responses confirm the improvement. The share of panelists reporting higher orders rose to 23.4 percent from 19.0 percent, while the share reporting lower orders was essentially unchanged at 15.5 percent. The gain came out of the middle, where the share reporting no change fell to 61.1 percent from 65.4 percent. On the unadjusted responses the index rose about 2.3 points to 54.0, a little more than the 1.6-point gain in the seasonally adjusted series. Order books started growing again for part of the panel in September without shrinking for anyone else.
The rest of the demand block was mixed. New export orders fell 2.3 points to 50.9 and imports fell 1.5 points to 51.0, both still in expansion, with a Canada tariff and the broader trade war appearing in the panel comments. Customers’ inventories fell 1.2 points to 41.6, the 24th consecutive month judged too low, a condition that historically precedes restocking. Buyers committed a little further forward on capital equipment and maintenance supplies than in August, with capital expenditure lead times up 5 days to 176 and maintenance and repair supplies up a day to 49, and left production materials unchanged at 84 days. The comment that matters most for how to read the orders data came from a Chemical Products panelist, who called that industry’s improvement temporary and attributed it to customers bringing forward purchases, delayed raw-material price increases and reduced competitor capacity, with structural overcapacity, pricing pressure and protectionist trade policy still in place. We would not build a forecast on one month of that industry’s survey, although over time chemical producers should do better as overall manufacturing activity increases.
Hiring Strengthened, Though Less Than It Looks
The Employment Index rose 1.5 points to 52.7 in September, a third month in expansion after 17 straight months below 50, and the ratio of panelists hiring to those managing or reducing head counts widened to 1.5 to 1 from 1.3 to 1. ISM’s guidance is that an Employment Index above 50.3 is generally consistent with rising manufacturing payrolls in the Bureau of Labor Statistics data, and September is the third reading in a row above it.
The raw responses are less exciting than the change in the index. Of the panelists, 15.3 percent reported more head counts, 14.5 percent reported fewer and 70.2 percent reported no change, so the net between those adding and those cutting widened to 0.8 points from 0.3 in August, and the no-change share fell from August’s 76.7 percent. On the unadjusted responses the Employment Index was about 50.4, against about 50.2 in August, so most of the 1.5-point gain in the published series came from the seasonal adjustment. Hiring is real and improving, but the headline overstates how much.
The hard data have been catching up to the survey. Manufacturing payrolls added 16,000 jobs in August after 14,000 in July and 13,000 in June, 43,000 over three months, and now stand 58,000 above December’s trough. Employment is 23,000 above a year earlier, the first positive year-over-year change since September 2023. Payrolls are still 1.8 percent below their January 2024 level, however, while factory output is 3.5 percent above it and aggregate hours are 1.6 percent above it.
The workweek is still doing work that hiring has not. Average weekly hours in manufacturing were 41.7 in June, July and August, the highest readings since June 2019. Six tenths of an hour above December’s 41.1 is worth roughly 184,000 workers at December’s workweek, against the 58,000 the sector has actually added since then. Employers stretch hours before they hire when they doubt a recovery will last, because hours can be unwound quickly and hires cannot. Three straight gains in payrolls say some of them have started to commit, and Friday’s employment report for September will show whether that pace has held.
The Investment Cycle Is Still Running
Nothing in the data suggests the capital spending behind this recovery is slowing. Orders for nondefense capital goods excluding aircraft reached $87.6 billion in August, up 1.6 percent from July and 14.1 percent from a year earlier, and up 10.6 percent since December alone. Since the start of 2024 factory output is up 3.5 percent while aggregate hours are up 1.6 percent and payrolls are down 1.8 percent, which is the signature of a cycle that deepened capital before it added workers.
The Production Index fell 1.6 points to 56.7, an 11th month of growth and the lowest reading since June, but we read most of that decline as seasonal noise. On the raw responses 25.4 percent of panelists reported higher output and 11.9 percent reported lower, an unadjusted index of about 56.8 against about 56.6 in August. The seasonal factor added about 1.7 points to August’s published figure and almost nothing to September’s, so the unadjusted series was essentially unchanged. Industrial production shows a similar pause, with manufacturing output slipping 0.3 percent in August after a long run of gains.
Supplier deliveries slowed for a tenth consecutive month, at 59.0, 0.3 point less than in August. Slower deliveries usually accompany firming demand, and they can also signal genuine scarcity. With orders and backlogs rising in the same month, firming demand is a larger part of the story than it was in August, when orders were falling. Electronic components have been in short supply for 19 consecutive months, electrical components for 15 and memory for nine. The Chicago survey describes the same strain from another angle, with 27 percent of its respondents naming semiconductors and electronic components as a supply constraint and 23 percent naming metals and specialty alloys. Lead times came up in 21 percent of the negative ISM panel comments, against 46 percent in August.
Prices, Tariffs and the War Premium
The Prices Index rose 6.8 points to 77.9 in September, the largest one-month increase since March and the highest reading since May, and has now signaled rising input costs for 24 consecutive months. Aluminum has been up in price for 34 consecutive months, copper for 15 and steel for 11. Pricing volatility was the most common concern in the negative panel comments, cited in 46 percent of them, followed by tariffs at 34 percent, the Iran war at 30 percent and lead times at 21 percent.
The share of respondents paying higher prices rose to 58.6 percent, the highest since May, and the share paying lower prices fell to 2.8 percent, so the net share paying more is now 55.8 percentage points against 42.2 in August. Nearly three of every five manufacturers paid more for materials in September and almost none paid less.
The official price data have turned in the same direction. The producer price index for processed goods for intermediate demand, the series ISM’s Prices Index is most closely tied to, rose 1.4 percent in August after declines of 0.6 percent in June and 0.2 percent in July, to a record level above both the 2022 peak and May’s reading. Its year-over-year rate rose to 11.5 percent from 10.1 percent in July. Last month we argued that the deceleration in producer prices had less room left in it than two monthly declines suggested. The August producer price data settled that faster than we expected, and the September survey points the same way.
Three forces are behind the survey index, and they are separable. Tariffs raise costs through the value chain, and panelists cite a Canada tariff and the broader trade war, which are policy choices. The Iran war keeps a premium on petroleum-based products and freight, and it was named in 30 percent of the negative comments, unchanged from August. The third is scarcity in the categories the AI buildout is consuming, which will not ease materially until the buildout does.
The Regional Surveys Were Strong, but Uneven
The Chicago Business Barometer, the survey formerly known as the NAPM-Chicago, jumped 11.7 points to 58.8 in September from 47.1, against a consensus near 51 and its highest reading since May. Production rose 15.5 points to its highest since May, new orders rose 13.3 points, supplier deliveries rose 9.6 points for a 20th month above 50, and order backlogs rose 8.4 points, although they are still in contraction, for a third month. Employment fell 4.3 points and went back into contraction, inventories slipped 0.6 point, and prices paid fell 3.7 points to July’s level, with no respondent reporting lower prices for a seventh straight month. A one-month gain this large is unlikely to repeat. It does say that orders and output picked up together in the Midwest, which matches the ISM’s own detail, and the employment reading is a reminder that Chicago and the national survey did not agree on hiring.
The five regional Fed surveys were stronger than the cooling in their average suggests and more divided than it implies. The average of the five headline indexes eased to 13.4 from 18.7 in August, the third consecutive reading at 13.4 or higher, a stretch the series has not matched since the spring of 2022. Philadelphia fell to 37.8 from 47.4 but has now posted three straight readings of 37.8 or higher, which has no precedent in our data back to January 2022. Dallas slipped to 9.8 from 11.6 even as its production index rose to 29.5 from about 13 and its new orders index rose to 30.7 from 22.0, and Kansas City rose to 14 from 10, its highest in at least 13 months, with new orders at 24 from 16. The two weak spots were Empire State, where general business conditions fell to 7.6 from 20.6 and new orders to 2.0 from 17.3, and Richmond, whose composite fell to a negative 2 from 4, the first negative reading since February. The Richmond Fed survey has an unusually strong connection to housing-related manufacturing. The Fifth District includes important concentrations of furniture, home textiles, wood products and other building materials, making the survey particularly sensitive to residential construction, remodeling and housing-market conditions.
Hiring and prices tell the same story as the ISM. The five-district employment average was 8.9 against 8.6 in August, with Dallas at 15.1 from 8.0, Philadelphia at 11.8 from 27.9, New York at 10.6, Richmond at 7 from a negative 2 and Kansas City flat at zero. Prices paid rose in New York to 63.1 from 58.6, in Philadelphia to 48.6 from 40.9 and in Kansas City to 68 from 55, and Dallas’s raw materials price index of 52.2 is nearly twice its series average of 28.0. Delivery times lengthened in both New York and Philadelphia. The five-district average for new orders was 16.0 against 17.7 in August. The one soft spot was the outlook in Dallas, where the index of expected general business activity fell to 20.8 from 37.2. The five-district average has tracked the PMI closely since 2022, with a correlation of 0.82, and at 13.4 it is consistent with a PMI well above 50. The final S&P Global manufacturing PMI, at 55.9 from 53.9 in August after a flash estimate of 57.0, was stronger still.
Our Call
The recovery reaccelerated at the front of the factory cycle, and October’s order readings will say whether it lasts. Last month we said the next two order readings would settle whether the recovery was decelerating or ending. September answers the first half in the recovery’s favor. Orders and backlogs rose, hours worked are at a seven-year high and hiring has risen for three months. Hiring responds to orders with a lag of several months, so we look for manufacturing employment to post small gains through the fall and for the Manufacturing PMI to hold in the low to mid 50s.
Composition is again doing more work than the headline. The swing in GDP-weighted contraction from 22 percent to 2 percent is one industry, and that industry’s own panelist calls the improvement temporary. The seasonal adjustment is doing much of the work in both the employment gain and the production decline, so the unadjusted readings are the safer guide this month. This is still a capital goods recovery riding an investment cycle, with core capital goods orders up 14.1 percent from a year ago, and it stays vulnerable to anything that interrupts that cycle.
We hold our call on the Fed, and September’s data lean toward the upside. We continue to expect the Committee to pause at the October 27-28 meeting and to raise the target range a second time on December 9, to 4.00 to 4.25 percent, with at least one further increase in 2027 carrying the terminal range to 4.25 to 4.50 percent or higher. That call is ceteris paribus, and September’s data cut toward an earlier move, not a later one: a Prices Index near 78, a producer price index at a record, rising hiring and a regional average of 13.4. Friday’s employment report is the release most likely to settle the timing.
The risk to this call is that it is too patient. A Prices Index near 78 and a record producer price index would argue for acting in October if Friday’s employment report is strong. The risk in the other direction is that the reacceleration in orders is a pull-forward. The Chemical Products panelist described customers bringing purchases forward, and customers’ inventories have been too low for 24 months, which should support restocking, but if October’s ISM and Chicago surveys both give back September’s gains, the order book will look like it has been oscillating around 54 instead of building. In that case we would lose confidence in the December increase.
Mark P. Vitner – Chief Economist, Piedmont Crescent Capital
mark.vitner@piedmontcrescentcapital.com · (704) 458-4000
Disclaimer: This report is provided for informational purposes only and does not constitute investment, legal, tax, or accounting advice, nor an offer or solicitation to buy or sell any security. Information is drawn from sources believed to be reliable, including the Institute for Supply Management, the U.S. Bureau of Labor Statistics, the U.S. Census Bureau and the Federal Reserve Board, but its accuracy and completeness are not guaranteed. Views expressed are those of the author as of the date of publication and are subject to change without notice. Past performance is not indicative of future results.
Second-Quarter GDP by Industry: Capital Spending Now Shows Up in Industry Output
ECONOMIC INDICATOR REPORT · GDP BY INDUSTRY, SECOND QUARTER 2026
Capital Spending Now Shows Up in Industry Output
Six industry groups supplied 70% of real GDP growth over the past four quarters, up from 34% in 2015-19, although structures, defense and pharmaceuticals have yet to show up in the accounts.
Early Signal
- Growth of 2.2% is unremarkable, but its makeup is not. Real GDP rose at a 2.2% annual rate in the second quarter, matching both the average of the past four quarters and the gain over the year. Equipment investment contributed 0.60 percentage points to that growth (0.12 in 2015-19) and intellectual property products 0.48 (0.34), while structures, residential investment and government together subtracted 0.32.
- Six industry groups supplied 70% of the growth. The groups we map to the seven forces contributed 1.54 of the 2.2 percentage points over the past four quarters, compared with 0.87 of 2.56 (34%) in 2015-19. The AI buildout group alone supplied 0.66 points.
- Information is the largest single piece. The sector accounted for 0.57 percentage points of growth, or 26%, while making up 5.9% of nominal GDP, and data processing & internet publishing supplied 0.40 of that. State data trace 43% of the sector’s contribution to California and Washington, which hold 17% of national GDP.
- Factory-floor industries are growing several times as fast as the economy. Real value added rose 13.3% in fabricated metal products, 11.9% in other transportation equipment, 11.5% in machinery and 8.8% in electrical equipment over the past year, and the reshoring group’s contribution swung from -0.01 points in 2015-19 to 0.19.
- The drop in structures is the end of a federally funded plant boom. Real investment in manufacturing structures fell 25.7% over the past year and sits 34% below its third-quarter 2024 peak, and structures as a whole subtracted 0.21 percentage points from growth. Census data show computer, electronic and electrical plants, the CHIPS-era semiconductor and battery boom, account for 95% of the decline in factory construction, and construction of all other factories is roughly flat.
- Three forces are not yet visible in the aggregate data. National defense contributed 0.06 points (0.08 in 2015-19), chemical products, the industry that includes drug makers, contributed 0.03, and the group we use for Peak 65 contributed 0.46, a steady rather than accelerating pace.
- Imports absorbed part of the boom. Imports subtracted 0.73 percentage points from growth over the past four quarters, double the 0.36 drag in 2015-19, which fits a boom in which much of the computing hardware, along with the HVAC equipment for data centers, is imported.
The Overview
Real GDP grew at a 2.2% annual rate in the second quarter in the Bureau of Economic Analysis’s third estimate, the same pace as the average of the past four quarters (3.9%, 0.2%, 2.5% and 2.2%) and the 2.2% gain over the year. The headline is slower than both the 2.6% average of 2015-19 and the 3.0% average of 2023-24, but the composition has shifted in the direction our thesis of seven capital-intensive forces would predict. Consumer spending contributed 1.7 percentage points, about what it did before the pandemic, while equipment and intellectual property products together added 1.1 points, up from 0.5, and structures, housing and inventories held growth back.
Housing and public spending no longer carry their share. Residential investment subtracted 0.1 points over the past four quarters, structures subtracted 0.2 and government added essentially nothing (0.02), against a combined 0.6 points from the three in 2015-19 and 1.1 in 2023-24. The seven forces are the AI buildout, energy infrastructure upgrade, defense replenishment, civilian aerospace expansion and next-generation defense buildup, reshoring of critical industries and materials, next-era pharmaceutical investment and Peak 65. BEA’s industry accounts, published alongside this release, let us test how many of them show up in output rather than in announcements.
Business investment explains the change in growth, and imports explain why it is not larger. Imports subtracted 0.73 percentage points over the past four quarters against 0.36 in 2015-19, while exports added 0.63 points against 0.18, so the trade balance itself was a small drag (-0.10) even with a jump in foreign sales. Inventories subtracted another 0.13. Because equipment spending is measured gross of imports, and a significant amount of it, including the HVAC equipment that data centers require, is imported, the domestic industry data below are the better test of where the capital is actually being turned into output.
Where the Growth Came From
We sorted the industries in BEA’s contribution-to-growth table into six groups that map to the forces. The AI buildout group (data processing & internet publishing, software publishing, computer systems design and computer & electronic products) contributed 0.66 percentage points to growth over the past four quarters, up from 0.46 in 2015-19. Energy infrastructure (utilities, electrical equipment, pipeline transportation and support activities for mining) added 0.12 points after adding essentially nothing before the pandemic, and the reshoring group (machinery, fabricated metal products and primary metals) added 0.19 after subtracting 0.01. Aerospace and other transportation equipment contributed 0.08, up from 0.02, and chemical products, our pharmaceutical proxy, 0.03, up from 0.02. Health care and real estate, our closest industry proxy for Peak 65, contributed 0.46 against 0.38.
Together the six groups supplied 1.54 of the 2.2 points, or 70% of growth, compared with 34% in 2015-19. The composition matters, however. Nearly all of the increase comes from the AI buildout group and from the industrial groups, which together contributed 0.43 points, about 20% of growth, up from 0.03 (about 1%) in 2015-19. That is a large change from a small base, and it is concentrated in machinery, metals and electrical equipment rather than spread evenly across the forces. Health care and real estate contributed 0.46 points, more than the four industrial groups combined, but that pace was already largely in place before the forces took hold.
The Information Sector
The information sector deserves its own look, because it supplied 0.57 percentage points of the 2.2% growth over the past four quarters, or 26%, while accounting for only 5.9% of nominal GDP. In 2015-19 it supplied 0.41 of 2.56 points, or 16%. Data processing & internet publishing supplied 0.40 points of the 0.57, up from 0.13, and software and publishing added another 0.14. Broadcasting & telecommunications slipped to 0.07 from 0.18, and film, music & sound subtracted 0.03. Real value added in the sector rose 10.3% over the year, and in data processing & internet publishing it rose 22.1%, the fastest gain of any industry in Exhibit 4.
The state data released alongside the national estimates show where that growth is booked, and the list is short. Weighting each state’s information contribution by its share of nominal GDP, California and Washington, which together hold 17% of national GDP, account for 43% of the sector’s contribution over the past four quarters, and adding New York lifts the share to 60% with 25% of GDP. The 16 states we define as the South hold 36% of GDP and account for 22%. That geography tracks the headquarters of the largest software, internet and cloud companies, most notably in the Seattle area and the San Francisco Bay Area (including San Jose), but this release stops at the state level, so it cannot separate the Bay Area from the rest of California.
We would not read that map as a picture of where the activity takes place. Cloud services, software and online platforms are delivered to customers in every state, and the data centers, developers and support staff behind them are spread across the country. State GDP appears to allocate the value added largely to where the firms’ establishments and payrolls are located, so the sector looks outsized in a few headquarters locations and is hard to quantify anywhere else. The quarterly state figures are also volatile: Washington’s information contribution to its own growth was 2.2 percentage points in the third quarter of 2025 and has averaged about zero since (-0.4, 0.3 and 0.0 in the following three quarters).
What Industry Output Shows
Real value added points the same way. Of the 16 industries we track in Exhibit 4, the nine fastest-growing over the past year are all tied to the forces: data processing & internet publishing (+22.1%), fabricated metal products (+13.3%), other transportation equipment (+11.9%), machinery (+11.5%), electrical equipment (+8.8%), software publishing (+8.5%), computer systems design (+7.4%), primary metals (+6.0%) and utilities (+5.9%), against 2.2% for the economy as a whole. Retail trade (-0.7%), nondurable goods manufacturing (-3.2%) and transportation & warehousing (-4.0%) shrank.
One line in Exhibit 4 argues against reading the equipment boom as domestic manufacturing strength. Real value added in computer & electronic products fell 0.9% over the year, even though real investment in computers and peripheral equipment rose 43.3%. The gap is consistent with imports supplying a large share of the hardware (and of the HVAC equipment that data centers require), and it explains why the gains show up in software, data processing and computer systems design, the industries that install and run the equipment, rather than in the factories that make it.
The Investment Gap in Structures
The weak spot in the buildout is structures, and the weakness has a specific source. Real investment in computers and peripherals is 198% above its fourth-quarter 2019 level (an index of 298) and software is 98% higher (198), but manufacturing structures, which had more than doubled to an index of 222 by the third quarter of 2024, has since fallen 34% to 147, including a 25.7% decline over the past year. Power & communication structures are still 6% below their 2019 level (an index of 94). Structures as a whole subtracted 0.21 percentage points from growth over the past four quarters. The decline reflects the passing of the peak buildout of federally funded battery and semiconductor plants. Census data through July show computer, electronic and electrical manufacturing construction down 45.4% year to date, which accounts for 95% of the entire category’s decline, while construction of all other factories is down only about 2%, essentially flat, and chemical plants, which include pharmaceutical plants, are down 1.2%.
The forces we describe require plants, grids and pipelines as well as chips and code, so a turn in structures outside the chip and battery plants is the evidence we most want to see before calling the buildout broad. Utilities output is up 5.9% over the year and electrical equipment 8.8%, which suggests the demand is there, but the investment that expands transmission and generation capacity has not yet shown up in the accounts.
Three Forces the Data Do Not Yet Show
Defense replenishment appears in neither measure. National defense spending contributed 0.06 percentage points to growth over the past four quarters, no more than the 0.08 of 2015-19, and real value added of federal defense fell 0.9% over the year. Pharmaceuticals are not visible either, since chemical products, the industry that includes drug makers, rose only 1.2% and contributed 0.03 points. Peak 65 is real but steady, with health care & social assistance up 3.1% and the health care and real estate group contributing 0.46 points against 0.38 in 2015-19.
None of the three is a reason to abandon the thesis, because defense procurement and drug plant construction typically take years to reach output, and aggregate accounts are slow to register them. They are, however, the parts of the thesis a skeptic can fairly say have yet to arrive.
Our Call
The industry data support the thesis where it is easiest to measure and leave three forces unproven. Six industry groups now supply 70% of growth, and equipment and software investment have replaced housing and government as the swing components of GDP. We are not changing our forecast on the strength of this release. Here is the test, with dates: the advance estimate of third-quarter GDP should show equipment and intellectual property products contributing at least 1.0 percentage point combined (1.08 over the past four quarters), and structures should stop subtracting. If structures keep shrinking while computer investment slows, the buildout is narrower than we argue. And until national defense contributes more than the 0.08 points it did in 2015-19, we would not treat replenishment as visible in output.
President & Chief Economist, Piedmont Crescent Capital
Sources
U.S. Bureau of Economic Analysis, Gross Domestic Product (Third Estimate), Industries, Corporate Profits, State GDP, and State Personal Income, 2nd Quarter 2026; State Personal Consumption Expenditures, 2025 (released September 30, 2026), including GDP by Industry Tables 5, 10 and 13, National Income and Product Accounts Tables 1.1.2 and 5.3.6, and Regional Accounts tables SQGDP1, SQGDP9 and SQGDP11. Industry groups, state weights by nominal GDP and the four-quarter and 2015-19 averages are PCC calculations. Exhibit 1 averages quarterly annualized contributions; the 2015-19 and 2023-24 averages in Exhibits 2 and 3 use annual contributions.
This report reflects the analysis and opinions of Piedmont Crescent Capital as of the date shown and is provided for general informational purposes only. It does not constitute investment, legal, or tax advice, is not an offer or solicitation, and should not be relied upon as the basis for any investment decision. Forecasts and other forward-looking statements are inherently uncertain and actual outcomes may differ materially.
Consumer Confidence Report | Business Conditions Turn Negative
Consumer Confidence Report: Business Conditions Turn Negative
The Conference Board index fell 6.7 points to 81.9 as all five of its components deteriorated, and fuel prices dominated what respondents wrote in.
Early Signal
- Confidence fell 6.7 points to 81.9. The Consumer Confidence Index was 88.6 in August. The release follows two prior months of softening.
- The Present Situation Index fell 7.9 points to 109.3, the larger of the two declines. The Expectations Index fell 5.9 points to 63.6, its third consecutive monthly decline.
- Current business conditions turned net negative for the first time since September 2024. The net balance fell 3.4 points to -1.9%, and the move came almost entirely from more households saying conditions are bad.
- The labor market differential narrowed to +1.7%, from +4.2%. Households saying jobs are plentiful (23.6%) barely outnumber those saying jobs are hard to get (21.9%).
- Unfavorable answers rose in all five components. On average the unfavorable share rose 2.2 points while the favorable share fell 0.8 point.
- Inflation and rate expectations both rose. The average 12-month inflation expectation is 6.1% (median 5.1%), and 68.4% of households expect higher interest rates, up 5.2 points.
- Family finances turned net negative. Households calling their current finances bad now outnumber those calling them good, for only the second time in the four years the question has been asked.
- The headline index is 1.9 points above 80. Our research shows that a reading below 80 has typically preceded outsized midterm losses for the incumbent party. The October report, due October 27, is the last before the election.
Overview
The Conference Board's Consumer Confidence Index fell 6.7 points to 81.9 in September, from 88.6 in August. The Present Situation Index dropped 7.9 points to 109.3 and the Expectations Index fell 5.9 points to 63.6. Expectations sits well below 80, the level the Conference Board has long treated as a signal of recession risk. The preliminary survey ran September 1 to 23, so it captured the September 16 rate hike, a surge in fuel prices and the ongoing geopolitical tensions.
A quick reminder about what these numbers measure. Every component is built from the share of households giving a favorable, unfavorable or neutral answer, so the index reports the breadth of sentiment and not its intensity. A household that moves from “good” to “bad” counts the same whether its view slipped a little or a lot. We try to use breadth language throughout our analysis and reserve words like growth and spending for the hard data.
The Present Situation
Current business conditions turned net negative for the first time since September 2024. The share calling conditions good was 18.5%, nearly unchanged from 18.8%, while the share calling them bad rose to 20.4% from 17.3%. The net balance fell 3.4 points to -1.9%. One quick takeaway is that consumers’ perception of the economy appears to be more sensitive to interest rate hikes than the economy itself has been. Treasury yields stayed near their highest level since 2007 on the day.
The labor market differential narrowed to +1.7%, from +4.2% in August. Households saying jobs are plentiful slipped to 23.6% from 24.5%, and those saying jobs are hard to get rose to 21.9% from 20.3%. Roughly as many households now report jobs hard to get as plentiful. That fits the labor market we described in the September 28 issue of A View from the Piedmont, where a falling unemployment rate reflects a shrinking labor force more than strong hiring. Households describe a job market in balance, with little hiring to lean on. This morning’s August JOLTS report, with job openings down 256,000 to 7.1 million and hires little changed at 5.2 million, is also consistent with this view.
Expectations
Households expect both business conditions and jobs to weaken over the next six months. Net expectations for business conditions fell 3.2 points to -9.5%, and net expectations for jobs fell 3.1 points to -14.4%. The 28.4% of households expecting fewer jobs outnumber the 14.0% expecting more by roughly two to one. Income expectations stayed positive at +2.5%, down from +5.5% in August, and the share expecting a lower income rose to 15.4% from 13.5%. Consumers are clearly experiencing some stress from higher energy prices and are concerned about rising interest rates.
The shares show where the deterioration came from. In each of the five components, the larger part of the move came from more households giving an unfavorable answer and the smaller part from fewer giving a favorable one. Unfavorable shares rose by an average of 2.2 points, while favorable shares fell by an average of 0.8 point (Chart 3). The neutral middle thinned in every component too, most of all for current business conditions, where the share saying conditions are normal fell to 61.1% from 63.9%.
Prices, Rates And Family Finances
Fuel prices dominated the write-in responses. References to prices, the high cost of goods and services, and oil and gas prices in particular rose to new highs, reflecting September's surge in fuel costs. Comments about war and conflict eased but stayed elevated. On September 25, retail diesel averaged $6.50 a gallon (AAA) and November Brent traded near $104 a barrel.
Inflation expectations rose alongside. The average 12-month expectation rose 0.3 point to 6.1% and the median rose 0.3 point to 5.1%. The average sits a full point above the median, which tells us a tail of very high answers pulls up the mean. The University of Michigan's final September reading for the year ahead, 4.6% and up from 4.0%, uses different wording and a different sample but points the same way.
Households expect the Fed to keep going. The share expecting higher interest rates over the next 12 months jumped 5.2 points to 68.4%. They have reason to think so, with the 10-year Treasury closing at 5.24% on September 28, its highest level since 2007, and the 30-year mortgage rate at 7.03%. Consumers still expect stock prices to rise over the next year, though with less conviction than in August.
Family finances turned net negative. The share calling their family's current financial situation bad overtook the share calling it good, for only the second time since the question was introduced four years ago. The series sits outside the index, and it moved the same way as everything else.
The 80 threshold matters for the November 3 midterms. Our research shows that a headline Consumer Confidence reading below 80 has typically preceded outsized losses for the incumbent party, which fits our framework of reading midterms through real after-tax per capita income growth. At 81.9, September's reading is 1.9 points above the threshold. The October report, due October 27, is the last one before voters go to the polls, and a decline of two points or more would put the index below it.
The Conference Board and Michigan surveys locate the damage differently. In Michigan's final September survey, current conditions eased only a point to 50.9 while expectations fell 5.2 points to 46.3. In the Conference Board survey, the Present Situation Index fell more than the Expectations Index, 7.9 points against 5.9. The two are built from different questions, business conditions and jobs at the Conference Board, and personal finances and buying conditions at Michigan, and both pointed lower.
Spending Plans And Who Is Feeling It
Spending plans softened at the margin. On a six-month moving average basis, plans to buy autos and homes both declined slightly, and planned spending on services pared back again. Hotels for personal travel, movies, airfare and amusement parks were among the discretionary categories that moderated. Vacation plans held up, with 42.6% of households planning a trip in the next six months, up 0.5 point, but the gain was limited to domestic travel and planned foreign travel dipped. The link between spending plans and actual spending remains loose, however. The direction is probably more important than the magnitude.
Confidence weakened across age groups, income groups and political affiliations. On a six-month moving average basis, confidence trended lower in every age group and in nearly all income groups. Households earning $125,000 to $149,000 reported the largest decline over six months, and higher-income groups stayed the more optimistic. Gen Z and Millennials have the highest confidence, while Generation X, Baby Boomers and the Silent Generation continued to weaken. In September, confidence fell among Democrats, Republicans and Independents alike.
OUR CALL
September's drop reads as a reaction to the cost of living. The hard data on household spending have not yet confirmed a pullback, and the September flash composite PMI of 58.4 was the strongest since 2021. We hold our forecast for real consumer spending growth of 2.3% in 2026 and 2.0% in 2027 (September 23 vintage), and we leave the Fed call unchanged: a hold at the October 27-28 meeting, a second quarter-point increase on December 9 to 4.00%-4.25%, and at least one more in 2027.
The test is the October report, released October 27, the first day of the FOMC meeting. If the labor market differential turns negative, with more households calling jobs hard to get than plentiful, we will take our consumer spending forecast lower. If it holds above zero, we read September as a response to fuel prices and higher rates and keep the forecast where it is.
Bottom Line
Confidence deteriorated sharply in September after two months of softening, and the shares behind the index show households changing their minds in one direction, a more cautious one. Fuel prices, a rate hike and Treasury yields at their highest since 2007 gave them reasons. The labor market differential is the line to watch, because it remains positive but only slightly so. Our forecasts for consumer spending and the Fed path stand, and the October survey will tell us whether they should.
What We Are Watching
- Labor market differential: the +1.7% reading sits just above zero. A move below zero on October 27 would change our consumer spending forecast.
- The 80 threshold: a headline reading below 80 in the October report, released a week before the midterms, would put the index in the range that has typically preceded outsized losses for the incumbent party.
- Fuel prices and inflation expectations: if diesel and gasoline ease, the 6.1% average inflation expectation should be the first series to retrace.
- Wednesday's core PCE inflation and Friday's employment report: both arrive before the Committee meets again and will show whether the household mood in this survey has reached spending and hiring.
Mark P. Vitner
Chief Economist, Piedmont Crescent Capital
mark.vitner@piedmontcrescentcapital.com | 704-458-4000
Sources: The Conference Board, Consumer Confidence Survey, September 2026 (preliminary results, cutoff September 23, conducted by Toluna); University of Michigan, Surveys of Consumers, final September 2026; U.S. Bureau of Labor Statistics, Job Openings and Labor Turnover Survey, August 2026 (released September 29); U.S. Treasury, daily par yield curve, September 25, and market reports of Treasury yields, September 28 and 29; Freddie Mac, Primary Mortgage Market Survey, week ended September 24; AAA, national average retail diesel, September 25; S&P Global, flash U.S. PMI, September 2026; Piedmont Crescent Capital, A View from the Piedmont, September 28, 2026, and forecast vintage of September 23, 2026.
© 2026 Piedmont Crescent Capital, LLC. For informational purposes only. Not investment advice.
Tight Labor, Tight Money | A View from the Piedmont
A View from the Piedmont — Our Weekly Commentary on Money, Credit, Exchange Rates & Geopolitics
Tight Labor, Tight Money
New York Fed President John Williams said another rate hike by year end is a reasonable bet, and the bond market took him at his word: the 10-year Treasury touched its highest level since 2007. Underneath, a labor force that keeps shrinking raises the odds that a falling unemployment rate will make it exceedingly difficult to bring inflation down to the Fed’s 2% target.
Editor’s note: This issue covers the full data slate released since our last edition, New York Fed President John Williams’ September 24 remarks calling another rate hike by year end a reasonable expectation, and a diplomatic calendar that produced a signed U.S.-Denmark-Greenland framework and a Trump-Xi meeting that extended the trade truce without breaking new ground.
Market Dashboard
(as of Friday, September 25 close, unless noted)
| Rates & Credit | Rates & Credit | Equities & Commodities | Equities & Commodities |
|---|---|---|---|
| Fed funds target range | 3.75%-4.00% (unchanged since Sept 16 hike) | S&P 500 | 7,743.41 (+0.51% Fri) |
| 10-yr Treasury | 5.17% (Fri close, highest since 2007) | Dow Jones | 51,828.62 (+0.93% Fri) |
| 30-yr Treasury | 5.49% (Fri close, highest since 2004) | Nasdaq Composite | 27,068.72 (+0.48% Fri) |
| 2-yr Treasury | 4.81% (Fri close) | Russell 2000 | 2,837.55 (+0.07% Fri) |
| 2s-10s spread | 36 bp | Brent crude (Nov) | $104.32 (-2.1% Fri, +0.4% wk) |
| 30-yr mortgage | 7.03% (Freddie Mac, wk of Sept 24) | WTI crude (Nov) | $92.41 (-2.3% Fri, -7.9% wk) |
| Retail diesel | $6.50/gal (AAA) | Gold (Dec) | ~$4,300 (spot +0.1% Fri, -2% wk) |
| Core CPI (Aug, y/y) | 2.4% (core goods +0.1% m/m) | U.S. Dollar Index (DXY) | 100.97 (-0.3% Fri) |
| Initial jobless claims (wk ended 9/19) | 197,000 | UMich 1-yr inflation exp. (Sept, final) | 4.6% (up from 4.0%) |
| Chicago Fed NAI (Aug) | -0.04 | UMich Consumer Sentiment (Sept, final) | 48.1 (-3.6 pts m/m) |
Treasury yields are the official daily par yield curve for Friday, September 25, including a completed 30-year close. The mortgage rate is the Freddie Mac PMMS survey for the week ended September 24. Retail diesel is the AAA national average for September 25.
Summary
New York Fed President John Williams said Thursday that another rate hike by year end is a reasonable expectation, and the bond market has priced that in as merely a starting point. Speaking in London at a National Institute of Economic and Social Research conference, Williams called it “a reasonable way of thinking about it,” an unusually direct signal from a sitting FOMC voter this early in the intermeeting period. A weak $70 billion five-year Treasury note auction the same week added its own pressure. The 10-year Treasury touched its highest level since 2007, near 5.17% to 5.23% intraday, and Freddie Mac’s weekly survey put the 30-year mortgage rate at 7.03%, the highest since January 2025.
Underneath the rate story is a labor-market story that should make it more difficult to bring inflation down to the Fed’s target. The U.S. civilian labor force is smaller than it was a year ago. Immigration enforcement and falling participation have done the work a hiring boom usually does, and that arithmetic lowers the payroll growth needed to hold the unemployment rate steady. A falling jobless rate that merely tracks a shrinking labor force against soft demand is not inflationary; the same decline with GDP holding at or above potential is a different problem, and the September projections put growth a touch above the Committee’s own speed limit.
The diplomatic calendar mostly resolved loose ends rather than breaking any significant new ground. Trump and Xi met in Washington September 23-25 and extended the Busan trade truce to January 10, 2027, a technical extension Treasury Secretary Bessent called a placeholder rather than a breakthrough. The United States, Denmark and Greenland signed their framework agreement September 22, opening new U.S. sites beyond Pituffik. Two fronts we flagged last issue moved the wrong way: Pakistan struck ten targets inside Afghanistan on September 24, and Ukraine struck two more Russian refineries, taking roughly 370,000 barrels a day of capacity offline.
Two releases we expected this week did not arrive. The third estimate of second-quarter GDP and August core PCE inflation, the Fed’s preferred gauge, are due out September 30, landing alongside August JOLTS the day before and ahead of the ISM manufacturing report and the September employment report, four market-moving releases inside one five-day stretch a month before the Committee meets again.
This Week’s Argument
A falling unemployment rate, a shrinking labor force
The unemployment rate is 4.1%, down from 4.3% a year earlier, and that decline has not come from a hiring boom. The civilian labor force is down about 1.0 million from August 2025. Participation has slipped to 61.6% from 62.3%. Net immigration has collapsed, and the foreign-born labor force is smaller than a year ago; the pullback is not confined to workers without legal status, as tighter enforcement and new work-authorization restrictions have pulled some immigrants with legal work status out of the labor force as well. Prime-age participation has held up better than the headline rate, so the squeeze is demographic and policy-driven, not a collapse in core-age attachment. That mix has cut the payroll pace consistent with a stable unemployment rate to something like the mid-teens to as much as 80,000 a month, depending on the migration assumption and assumptions about the return of foreign-born workers with legal status. Soft job growth can now coexist with a falling or stable jobless rate. We lean toward higher estimates and look for foreign-born workers with legal status to return to the workforce during the coming year.
The inflation question is tidier than that. The risk is a falling unemployment rate while real GDP runs at or above potential. The September projections have 2026 growth at 2.3% and 2027 at 2.4%, against a 2.0% longer-run speed limit, with the jobless rate already a tenth below the Committee’s own 4.2% longer-run median. Potential labor input is nearly flat. Capital spending, especially AI-related equipment and intellectual property, is carrying a disproportionate share of growth, an employment-light expansion running a small positive output gap. Firms ask more of a smaller pool before they bid wages sharply higher. Unit labor costs are the transmission belt: nonfarm productivity is up 2.2% over the four quarters through the second quarter, and unit labor costs are up 1.4%. The Employment Cost Index is stuck near 3.4%. Those readings are compatible with a grind back toward 2% inflation if productivity holds. They are not a reason to assume the labor market will do the Fed’s work for it.
The payroll detail is not one story either. Manufacturing has picked up, adding 16,000 jobs in August and 58,000 since the December low, led by machinery and fabricated metals. Construction is still growing, but the mix has shifted: residential building is roughly flat from a year ago while nonresidential work, factories and data centers, carries the sector. Financial activities have been shedding jobs on the monthly payrolls, and this week’s selloff in banks, brokers and insurers, after Meta’s Muse agent hit No. 1 on the App Store, looks like the market pricing broader AI-disruption risk into consumer-facing service businesses rather than a narrow read on financial-sector hiring. We do not expect a 1970s-style spiral, but we would not rule out a Committee that stays on hold longer than fading energy and tariff effects would seem to justify if the jobless rate keeps edging down while output holds near 2¼% to 2½%.
Behind the Numbers
The Brent-WTI spread is the cleanest read on this week’s oil market, and it reflects two different themes. Brent, the November contract, held near $104 as Ukraine’s strikes on two more Russian refineries, Lukoil’s Perm plant and the Novoshakhtinsk facility in Rostov, took a combined 370,000 barrels a day of capacity offline, against a backdrop of fuel-export bans Moscow already had in place. WTI, also the November contract, fell nearly 8% on the week on hopes for a U.S.-brokered truce with Iran. The spread between the two, more than $11 a barrel, is the widest since May, and it is a geopolitical risk premium on the global barrel, not the domestic one, showing up in the number.
September’s flash PMIs were the strongest since 2021, and that, along with hawkish Fed comments, sent bond yields higher. The composite index rose to 58.4 from 56.0, manufacturing to 57.0 from 53.9, and services to 58.7 from 56.5, but input prices accelerated and suppliers’ delivery times lengthened in the same report, a combination the regional Fed surveys have been flagging all month: the Philadelphia Fed’s prices-paid index jumped to 48.6 in September from 40.9 and its delivery-times index surged to 26.5 from 3.7 even as its headline activity gauge fell to 37.8 from 47.4, and the New York Fed’s Empire State survey showed prices paid rising to 63.1 from 58.6 on a headline reading that eased to 7.6 from 20.6. Growth and renewed price pressure are printing together, which is the data-side argument behind Williams’ line that another hike by year end is a reasonable expectation.
We have focused on the manufacturing revival this year, noting that the factory sector provides the cyclical swing in economic growth. The resurgence in manufacturing is being led by the AI buildout and related energy grid buildout. AI is not the economy’s only strong suit, however. Commercial aircraft assemblies have ramped up this year, as Boeing has gotten the kinks out of its production process. Shipments of nondefense aircraft through August are up 5.0% from the same period a year ago. Defense replenishment is another key growth area, as is the private space industry. Reshoring of critical industries is also driving output, and the pharmaceutical industry is in the midst of a once-in-a-generation expansion, fueled by breakthrough medicines and the desire to bring the supply chain back into the U.S. All of this investment is driving interest rates higher, as private credit needs bump up against persistent large federal deficits and resulting Treasury borrowing.
New home sales rose 6.4% in August to a 684,000 annual rate, an eight-month high, even as the 30-year mortgage rate finished the week at 7.03%. Durable goods orders were flat on the headline in August, but the ex-transportation reading, the cleaner read on underlying business investment, rose 0.3%. The Chicago Fed’s National Activity Index slipped to -0.04 in August from an upwardly revised 0.08 in July, still consistent with growth near trend rather than a slowdown.
The final University of Michigan survey split cleanly between the present and the future. Current conditions eased only a point to 50.9. Expectations fell 5.2 points to 46.3, and year-ahead inflation expectations jumped to 4.6% from 4.0%, the highest reading since June. Consumers are not describing a weaker economy today. They are describing one they expect to cost more.
Bottom Line
We continue to expect the Committee to pause at the October 27-28 meeting and deliver a second quarter-point hike on December 9, taking the funds rate to 4.00%-4.25% by year-end, with at least one further quarter-point increase in 2027 carrying the terminal rate to 4.25%-4.50% or higher. New York Fed President Williams’ September 24 remarks, calling another hike by year end “a reasonable way of thinking about it,” are consistent with that path but do not pin down which meeting delivers it. That call is ceteris paribus, as it has been every issue since the hike, and less is holding still than it was a week ago: this week’s weak five-year Treasury auction and the strongest flash PMI reading since 2021 both cut toward an earlier move, not a later one. The September employment report, due October 2, and August core PCE inflation, due September 30, are the releases likely to settle the near-term timing before the Committee meets again. A jobless rate that keeps falling while growth holds near or above potential, the risk we lay out above, would argue for an earlier or larger move, not a smaller one.
A cut is not on our radar under this path, and we expect the Committee to hold once it reaches it. Only a weakening economy, paired with wider credit spreads and tighter lending conditions, would argue for a cut before then. The two-year Treasury is already pricing close to three hikes, consistent with the upside risk in our call, though that pricing can move quickly against a single month of weaker data. We also think the calendar flattered August’s numbers. The latest-possible Labor Day, September 7, extended the summer season and kept seasonal staff on payrolls through the survey week. September data may benefit the same way, but seasonal factors are built to net to nothing across a full year, so that flattering effect implies a payback later, not a free pass.
Hold the 10-year where it has moved this week, and we would take residential construction down a little more and start curbing business fixed investment in the second half of 2027. The level that creates real pain is not a fixed number; it is where the 10-year sits relative to nominal GDP growth. Nominal GDP has been growing near 6.6% on a four-quarter basis, and the 10-year, now at its highest level since 2007, has closed roughly a third of that gap since the second quarter. Once the yield actually reaches that growth rate, financing costs exceed the economy’s capacity to generate the income that services them, and investment is what gives first. That is the falsifiable test behind this call: a 10-year that keeps climbing toward the growth rate, or nominal growth that keeps decelerating toward the yield, is the signal to take both calls down further.
CFO and Treasurers Corner
- Refinance and term-out decisions should not wait for a cut that is not coming. This week’s bond market reaction, the 10-year Treasury touching its highest level since 2007, says the market already agrees. Lock in financing against a floor near 4.00%-4.25% by year-end, not a hoped-for easing.
- Budget the diesel crack, not the barrel. Russia’s fuel-export bans, already in force before this week’s refinery strikes, keep pulling barrels out of the diesel and jet-fuel market regardless of what happens next in Ukraine; a Politico report that Washington was weighing its own diesel-export limits was denied by the White House the same day it ran, so that risk is not live for now. Watch for it to resurface if domestic diesel prices keep climbing.
- Widen dollar hedge ratios, and do not assume the post-hike move is done. The dollar index posted a second straight weekly gain even after a Friday dip to 100.97, evidence the hawkish repricing since September 16 has legs. Revisit hedge ratios set before that meeting rather than waiting for a pullback that has not shown up yet.
- Budget labor costs off unit labor costs, not the unemployment rate. Unit labor costs are up 1.4% over the past four quarters against 2.2% productivity growth, and the Employment Cost Index is steady near 3.4%. A falling jobless rate driven by a shrinking labor force is not, by itself, a signal to budget for faster wage growth nationally; in the most recent state employment data, already two weeks old, South Carolina, North Carolina and Nevada stood out as states still adding workers at pace, and that is where the wage-cost risk concentrates.
- Power capacity remains the line item most treasurers still leave out of project underwriting. Nothing this week changes that; the capital-intensive, employment-light expansion described above is still running hardest through the sectors that supply its electricity.
- Treat Red Sea and Pakistan-Afghanistan risk as one supply-chain exposure, not two. Houthi missiles reached Taif and Yanbu inside Saudi Arabia on September 24, widening the target set beyond shipping itself, and Pakistan’s airstrikes into Afghanistan the same day raise the odds of a disruption reaching routes through South Asia or the Bab al-Mandeb. Compare current routing and reinsurance costs against Cape of Good Hope economics for any shipment where the math was close before.
Piedmont Perspective
Inflation Travels Through the Supply Chain, Not Just the CPI Basket
From the Cleveland Fed and European Central Bank’s joint inflation conference, September 24-25, 2026
The Cleveland Fed and the European Central Bank spent Thursday and Friday on a question that has dogged the expansion all year: why inflation has stayed above the 2% objective for more than five years, even after the first round of tariff and energy shocks was supposed to have washed through. The papers were technical. The implication for the outlook is not.
Cleveland Fed president Beth Hammack opened the conference by putting the current mix of shocks on the table: output growing at a solid pace, the labor market close to her definition of maximum employment, and inflation still elevated with risks tilted to the upside. She grouped the shocks the way a regional economist would: tariffs and oil on the supply side, the boom in AI-related capital spending on the demand side. On Friday’s policy panel she went further. The risk that concerns her most is not another one-off price spike; it is the chance that an inflationary mindset takes hold after the Committee has missed its target for more than five years. The ECB’s Boris Vujčić offered the contrast: the euro area had largely converged on its inflation target earlier this year before a second energy shock arrived; August euro-area inflation was back up to 3.2%, and the ECB’s own September projections now put 2027 inflation at 2.5%, up from 2.3% in June.
Several of the papers rested on a simple accounting fact standard models bury: firms sell to other firms, not just to households. Once that is written down, inflation is not one Phillips curve but a system of sectoral ones tied together by the input-output table, and two results follow that cut in opposite directions. An upstream shock, oil, a tariff on intermediates, a jump in semiconductor prices, does not hit the CPI on impact; it walks down the chain, and each sticky link delays how much shows up in core inflation months later. A demand shock does the opposite: intermediate prices rise gradually, downstream firms see higher sales but only modestly higher costs, and they have less reason to reprice, which is why a capital-heavy, employment-light expansion can keep generating growth while the inflation impulse from AI-related equipment and power demand is still working its way through producer prices.
A Cleveland Fed paper presented Thursday adds a piece firm surveys have been hinting at: firms in industries with larger sales relative to GDP and more flexible prices forecast CPI more accurately, and peripheral firms rationally pay less attention. That inattention travels downstream. If a supplier is not watching aggregate inflation, a buyer’s costs are stickier than its own price-setting rule would suggest. A few of the policy-design papers pointed toward a shared implication: optimal policy in a network still looks like stabilizing a price index, but the weights are not CPI or PCE shares. More weight belongs on larger, stickier and more upstream sectors, energy, chemicals, logistics, semiconductors, and less on the downstream services that dominate the consumer basket but sit far from where the cascade starts.
Map this year’s shocks onto the network and the data look less mysterious. Energy is a classic upstream commodity shock; delayed pass-through into core goods and transportation services is the model’s prediction, not a puzzle. Tariffs are a tax on a subset of imported intermediates, and how much shows up in producer prices depends on where the taxed good sits in the chain. AI-related capital spending is a demand shock concentrated in a handful of hubs, semiconductors, electrical equipment, nonresidential construction and utilities, and those hubs can move aggregate inflation by more than their weight in the CPI basket, with the impulse persisting after first-round equipment prices settle because the cost cascade is still moving downstream.
For the Carolinas and the rest of the Piedmont, the same map has a regional reading. The manufacturing and logistics belt from Greenville and Spartanburg through Charlotte, the Triad and into the Virginia Tidewater sits closer to those upstream nodes than the national average. Energy, chemicals, transportation equipment, electrical equipment and the data-center buildout are not rounding errors in this district; when those prices move, local producer and eventual consumer prices tend to move with them, on a lag easy to miss if the only series on the desk is metro CPI. We will get a cleaner test in the next few inflation prints. If oil stays off its spring highs and tariff pass-through continues to fade, core goods inflation should decelerate even if AI-related equipment prices stay firm. If core goods refuse to cool while utilities, electrical equipment and nonresidential construction keep climbing, the network is still transmitting, and the case for looking through the next relative-price surprise gets weaker, not stronger.
The Week Ahead
Tuesday: August JOLTS job openings; the Conference Board’s Consumer Confidence Index for September.
Wednesday: Third estimate of second-quarter GDP; August core PCE inflation, the Fed’s preferred gauge and the last one before the October 27-28 meeting; ADP’s September private-payrolls estimate. What would move us: a core PCE reading materially above 3% would argue for pulling the hike forward from December.
Thursday: ISM Manufacturing PMI for September; August construction spending.
Friday: The September employment report, the last one before the October 27-28 meeting. What would move us: payroll growth holding near this year’s soft pace alongside a jobless rate that keeps falling would sharpen the labor-force argument above rather than resolve it; a clear reacceleration in hiring would call the whole thesis into question.
Watch for further Fed speakers testing Williams’ language on whether October or December carries the next hike, the Danish and Greenlandic parliamentary ratification timeline on the framework signed September 22, and whether Pakistan-Afghanistan and the Houthis’ inland push toward Taiz and Marib stay contained to their own fronts.
U.S. Economic & Financial Outlook
The table below carries our September 23 forecast vintage, updated to reconcile the curve against the Committee’s September 16 projections and this week’s sharp back-up in longer yields. The 10-year Treasury path for the second half of 2026 is now built around 5.10% rather than the roughly 4.40% we had penciled in five days earlier, and the 30-year mortgage path moves with it, closer to 7.00%-7.30% than the high-6% range in our prior vintage. The 2026 year-end funds-rate range we had built in, 4.00%-4.25%, is unchanged from our prior vintage and matches the Committee’s own median dot for year-end 2026; the repricing here is concentrated in the long end, not the policy path itself. That repricing also trims our 2027 growth and housing forecasts modestly and nudges the consumer-spending path lower within the year, the ordinary consequence of a higher discount rate working through the model.
U.S. Economic & Financial Outlook, as of September 23, 2026
| Indicator | Units | 2024 Actual | 2025 Actual | 2026 Forecast | 2027 Forecast | Q1-26 Actual | Q2-26 Actual | Q3-26 Fcst | Q4-26 Fcst | Q1-27 Fcst | Q2-27 Fcst | Q3-27 Fcst | Q4-27 Fcst |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Output | |||||||||||||
| Real GDP | % q/q SAAR | 2.8 | 2.1 | 2.3 | 2.6 | 2.1 | 1.5 | 3.6 | 3.1 | 2.4 | 2.6 | 2.7 | 2.7 |
| 4th Qtr-to-4th Qtr Chg, Yr/Yr for Qtly Data | % YoY | 2.4 | 2.0 | 2.6 | 2.7 | 2.7 | 2.1 | 1.9 | 2.6 | 2.7 | 3.0 | 2.8 | 2.7 |
| Final Sales to Private Domestic Purchasers | % q/q SAAR | 3.0 | 2.6 | 2.7 | 2.9 | 1.7 | 4.2 | 4.5 | 3.9 | 3.6 | 3.3 | 3.2 | 3.2 |
| Consumer Spending | % q/q SAAR | 2.9 | 2.6 | 2.3 | 2.0 | 0.5 | 3.4 | 2.3 | 2.0 | 2.1 | 2.0 | 2.0 | 1.9 |
| Nonresidential Fixed Investment | % q/q SAAR | 2.9 | 4.1 | 6.2 | 4.6 | 10.6 | 8.5 | 9.0 | 7.7 | 6.6 | 5.9 | 5.7 | 5.4 |
| Residential Investment | % q/q SAAR | 3.2 | -2.2 | -1.8 | 0.5 | -7.8 | 0.0 | -3.8 | -1.3 | -0.5 | 0.4 | 0.8 | 1.2 |
| Industrial Production, Manufacturing | % YoY | -0.9 | 1.0 | 2.0 | 3.9 | 1.0 | 1.6 | 1.6 | 3.6 | 4.0 | 3.6 | 3.7 | 3.7 |
| Light Vehicle Sales | Mil. units, SAAR | 15.9 | 16.2 | 16.1 | 16.4 | 15.5 | 16.4 | 16.2 | 16.3 | 16.3 | 16.3 | 16.4 | 16.4 |
| Labor Market | |||||||||||||
| Unemployment Rate | % | 4.0 | 4.3 | 4.2 | 4.0 | 4.3 | 4.3 | 4.1 | 4.1 | 4.1 | 4.0 | 4.0 | 3.9 |
| Nonfarm Payrolls | Avg. mo. chg., thous. | 150 | 62 | 82 | 95.25 | 73 | 81 | 85 | 88 | 88 | 92 | 98 | 103 |
| Employment Cost Index | % YoY | 3.9 | 3.5 | 3.4 | 3.5 | 3.4 | 3.4 | 3.4 | 3.4 | 3.5 | 3.4 | 3.5 | 3.5 |
| Housing Market | |||||||||||||
| Housing Starts | Thous. units, SAAR | 1,370 | 1,356 | 1,354 | 1,370 | 1,418 | 1,337 | 1,320 | 1,340 | 1,350 | 1,370 | 1,380 | 1,380 |
| New Home Sales | Thous. units, SAAR | 684 | 679 | 636 | 665 | 622 | 650 | 630 | 640 | 650 | 660 | 670 | 680 |
| Existing Home Sales | Thous. units, SAAR | 4,060 | 4,060 | 4,076 | 4,188 | 4,043 | 4,120 | 4,040 | 4,100 | 4,150 | 4,180 | 4,200 | 4,220 |
| Case-Shiller National Home Prices | % YoY | 5.1 | 2.2 | 1.4 | 2.5 | 0.8 | 1.2 | 1.5 | 1.9 | 2.3 | 2.5 | 2.6 | 2.7 |
| Inflation | |||||||||||||
| Consumer Price Index (CPI) | % YoY | 3.0 | 2.6 | 3.3 | 2.7 | 2.7 | 3.8 | 3.4 | 3.2 | 3.0 | 2.3 | 2.6 | 2.6 |
| Core CPI | % YoY | 3.4 | 2.8 | 2.6 | 2.8 | 2.5 | 2.7 | 2.7 | 2.6 | 2.7 | 2.8 | 2.8 | 2.8 |
| PCE Deflator | % YoY | 2.6 | 2.6 | 3.3 | 2.8 | 3.1 | 3.9 | 3.2 | 3.0 | 2.9 | 2.5 | 2.7 | 2.7 |
| Core PCE Deflator | % YoY | 2.9 | 2.8 | 3.1 | 2.7 | 3.1 | 3.4 | 3.1 | 2.8 | 2.7 | 2.4 | 2.6 | 2.6 |
| Interest Rates & Markets | |||||||||||||
| Fed Funds Target Range | Avg./Range, % EOP | 5.14 | 4.21 | 3.63 | 4.06 | 3.50-3.75 | 3.50-3.75 | 3.75-4.00 | 4.00-4.25 | 4.25-4.50 | 4.25-4.50 | 4.25-4.50 | 4.25-4.50 |
| 2-Year Treasury | %, period end | 4.37 | 3.81 | 3.92 | 4.06 | 3.78 | 4.13 | 4.85 | 4.75 | 4.60 | 4.50 | 4.45 | 4.40 |
| 10-Year Treasury | %, period end | 4.21 | 4.29 | 4.74 | 4.89 | 4.35 | 4.40 | 5.10 | 5.10 | 5.00 | 4.90 | 4.85 | 4.80 |
| 30-Year Mortgage Rate | %, period end | 6.72 | 6.60 | 6.77 | 6.89 | 6.10 | 6.49 | 7.30 | 7.20 | 7.00 | 6.90 | 6.85 | 6.80 |
| Brent Crude | $/bbl, period avg. | 79.91 | 68.32 | 92.05 | 85.00 | 78.00 | 100.20 | 98.00 | 92.00 | 90.00 | 88.00 | 82.00 | 80.00 |
| Trade-Weighted Dollar Index | Period average | 122.9 | 122.7 | 119.1 | 119.9 | 119.0 | 119.5 | 119.3 | 118.8 | 119.1 | 119.6 | 120.3 | 120.6 |
| Trade-Weighted Dollar Index | % chg., period | 1.7 | -0.1 | -2.9 | 0.6 | -1.4 | 0.4 | -0.2 | -0.4 | 0.2 | 0.4 | 0.5 | 0.2 |
| Government, Trade & Inventories | |||||||||||||
| Government | % q/q SAAR | 3.8 | 1.1 | 0.0 | 1.2 | 4.4 | -1.0 | 2.4 | 1.8 | 1.5 | 1.3 | 1.1 | 1.1 |
| Net Exports | $ Bil., Chained 2017$ | -1,032.6 | -1,090.7 | -1,002.42 | -1,322.40 | -1,001.8 | -1,083.7 | -1,177.4 | -1,267.4 | -1,312.4 | -1,332.4 | -1,322.4 | -1,322.4 |
| Net Exports (percentage point impact) | Percentage points | -0.48 | 0.39 | -0.03 | -0.22 | -0.37 | -1.14 | -1.53 | -1.46 | -0.72 | -0.32 | 0.16 | 0.00 |
| Business Inventories | $ Bil., Chained 2017$ | 43.5 | 28.5 | -27.33 | 23.13 | -16.7 | -53.1 | 70.1 | 40.0 | 20.0 | 22.5 | 25.0 | 25.0 |
| Inventories (percentage point impact) | Percentage points | -0.17 | -0.21 | -0.12 | -0.06 | 0.23 | -0.72 | 1.93 | -0.41 | -0.16 | -0.12 | 0.04 | 0.00 |
Annual figures are full-year averages or year-over-year percent changes; 2024-2025 are actual, 2026-2027 are forecasts. Q1-26 and Q2-26 are the latest actuals. Source: Piedmont Crescent Capital, BLS, BEA, Census, Federal Reserve, EIA.
Sources and Notes
Sources. Bureau of Labor Statistics (Consumer Price Index; Initial Jobless Claims); U.S. Census Bureau (Advance Monthly Retail Trade; New Residential Construction; Advance Durable Goods); Federal Reserve Bank of New York (President Williams remarks, National Institute of Economic and Social Research conference, London, September 24, 2026; Empire State Manufacturing Survey, September 2026); Federal Reserve Bank of Philadelphia (Manufacturing Business Outlook Survey, September 2026); U.S. Treasury (5-year Treasury note auction results, September 23, 2026); Federal Reserve Bank of Chicago (National Activity Index); S&P Global (flash U.S. Composite, Manufacturing and Services PMI, September 2026); University of Michigan (Surveys of Consumers, final September 2026); Freddie Mac (Primary Mortgage Market Survey, week ended September 24, 2026); U.S. Bureau of Economic Analysis (GDP and Personal Income and Outlays release schedule); Cleveland Fed / European Central Bank (Inflation: Drivers and Dynamics conference, September 24-25, 2026, remarks of Beth Hammack and Boris Vujčić); European Central Bank (September 2026 macroeconomic projections).
Geopolitical. White House and NPR (President Trump’s UN General Assembly address, September 22, 2026); Al Jazeera, CNBC, CBS News and NPR (US-Denmark-Greenland framework signing and terms, September 22); Al Jazeera and NPR (Trump-Xi meeting, September 23-25, and the Busan trade truce extension); Al Jazeera and Reuters (Pakistan-Afghanistan border escalation, September 23-24); Reuters (Houthi missile attacks on Taif and Yanbu, Saudi Arabia, September 24); CNN (Houthi inland offensive toward Taiz and Marib, September 21); Rigzone and Bloomberg (Ukrainian strikes on the Perm and Novoshakhtinsk refineries, September 25); Reuters (Russia’s fuel-export bans, in force since summer 2026); Politico and the White House (reported and denied U.S. diesel-export restriction proposal, September 23).
Market data as of Friday, September 25, 2026 close, unless noted: TheStreet and EnergyNow (equity indices, Treasury yields, Brent and WTI); AAA (retail gasoline and diesel); USAGOLD (precious metals); TradingEconomics and CNBC (ICE U.S. Dollar Index, DXY, including week-over-week performance).
A View from the Piedmont is published weekly by Piedmont Crescent Capital for informational purposes only and does not constitute investment, legal, or tax advice. Views expressed are those of the author as of the date of publication and are subject to change. © 2026 Piedmont Crescent Capital, LLC. All rights reserved.
One Hike, Four Fronts | A View from the Piedmont
A View from the Piedmont — Our Weekly Commentary on Money, Credit, Exchange Rates & Geopolitics
One Hike, Four Fronts
The Fed moved for the first time since 2023. So did geopolitics: Yemen, Ukraine and Pakistan turned more negative, while an agreement among the U.S., Denmark and Greenland looks set to clear this week on good terms for all sides.
Editor’s note: This issue covers the full data slate released since our last edition, the Federal Reserve’s September 16 decision, and a geopolitical scan across four fronts we think investors are underpricing individually and mispricing as a group. The house forecast table carries our September 18 vintage, reconciled below against our call, and the Committee’s new dot plot appears in full in This Week’s Argument.
Market Dashboard
| Rates & Credit | Rates & Credit | Equities & Commodities | Equities & Commodities |
|---|---|---|---|
| Fed funds target range | 3.75%-4.00% (hiked Sept 16) | S&P 500 | 7,650.50 (+0.17% Fri) |
| 10-yr Treasury | 5.01% (Fri close) | Dow Jones | 51,682.64 (-0.18% Fri) |
| 30-yr mortgage | 6.95% (Freddie Mac, Sept 17) | Nasdaq Composite | 26,522.55 (+0.39% Fri) |
| 2-yr Treasury | 4.76% (Fri close, highest since June 2024) | 30-yr Treasury | 5.34% (Fri close) |
| 2s-10s spread | 25 bp | 2s-30s spread | 58 bp |
| IG credit OAS | 78 bp (ICE BofA, Sept 17) | HY credit OAS | 270 bp (ICE BofA, Sept 17) |
| Brent crude (Nov) | $104.33 (+0.3% Fri, +12% on the month) | Russell 2000 | -0.50% Fri (rate-sensitive small caps lagged) |
| Retail diesel | above $6.00/gal | Gold (Dec) | $4,368.60 (+0.6% Fri) |
| Core CPI (Aug, y/y) | 2.4% (core goods +0.1% m/m) | Silver | $67.18 (+3.0% Fri) |
| Initial jobless claims (wk ended 9/13) | 196,000 | Conference Board LEI (Aug) | 99.5 (-0.1% m/m) |
Treasury yields are the official daily par yield curve for Friday, September 18; the 2s-10s and 2s-30s spreads are computed from those par yields. Credit spreads are ICE BofA option-adjusted spreads through Thursday, September 17, the latest published. The mortgage rate is the Freddie Mac PMMS survey for the week ended September 17.
Summary
- The Committee raised the federal funds target range a quarter point to 3.75%-4.00% on September 16, the first hike since 2023, on a unanimous 12-0 vote. Our own technical rule, built on core CPI and core goods breadth, said hold: August core CPI cleared our 0.3% threshold but core goods rose just 0.1%, a quarter of the 0.4% bar the rule required. The dot plot itself shows just two members who see September as the last hike against sixteen who expect at least one more. We are going a quarter point further than the median: at least two more quarter-point increases, with a terminal rate of 4.25%-4.50% or higher by the end of next year.
- Last week’s full data slate was consistent with an economy expanding solidly albeit unevenly. Retail sales rose 1.2% in August, control-group sales rose 1.4%, and utilities output rose 6.2% over the year, the fastest-growing line in the industrial report and the clearest evidence yet of the capital-intensity theme we have carried all year.
- Four geopolitical fronts moved in the same week the Fed did, and we think markets are treating them as separate stories when they add up to one. A bombing in Kohat, Pakistan reopened its border conflict with the Afghan Taliban; the Houthis consolidated control of Yemen’s Red Sea coast; Ukraine struck the Moscow oil refinery itself; and the US and Denmark announced a Greenland framework we called in January. None is a US recession risk on its own. Together they are a standing bid under energy prices and the dollar at the exact moment the Fed has told us it isn’t done.
- Our call: pause October 27-28, hike December 9 to 4.00%-4.25%, and at least one more quarter-point increase in 2027 to a terminal 4.25%-4.50% or higher. That call goes a quarter point past the Committee’s own median dot, and it is ceteris paribus: it assumes the four fronts above do not escalate further.
This Week’s Argument
The rule said hold. The Committee said credibility.
Two weeks ago we set ourselves an explicit test for the September meeting: hike if core CPI cleared 0.3% month over month and core goods cleared 0.4%, hold otherwise. August’s report cleared the first bar. Core goods rose just 0.1%, a quarter of the 0.4% bar the rule required, a reading Oxford Economics’ independent core-goods tracker corroborated. On that test alone, the case for a hold was not close, and we said so in our last issue.
The Committee hiked anyway, 25 basis points to 3.75%-4.00%, unanimously. We think the credibility explanation is the right one, even though it is not what set our framework: a Committee that spent a month guiding markets toward a hike, and watched CME-implied odds run from 58% to roughly 90% over the following week, had more to lose from disappointing a fully-priced market than from a marginal miss on core-goods breadth. That is a credibility argument, not a data argument, and it beat our data-driven rule.
The new Summary of Economic Projections tells us the Committee is not finished. The median dot now shows the funds rate at 4.1% at both the end of 2026 and the end of 2027, 3.9% in 2028, 3.6% in 2029, and 3.2% longer run, all up from June’s projections of 3.8%, 3.6%, 3.4% and 3.1%. Core PCE inflation is projected at 3.4% this year, easing to 2.5% in 2027, 2.2% in 2028 and 2.0% in 2029. GDP growth is projected at 2.3% this year and 2.4% next, and unemployment is expected to hold near 4.1% through 2028. The dot plot below shows only two members who consider Wednesday the last increase; twelve see one more move before year-end and four see two. A Committee that just hiked, with sixteen of eighteen members penciling in at least one further move, is not a Committee we would call finished.
Our own call goes a quarter point further than the median: at least two more quarter-point increases, one later this year and at least one more in 2027, with a terminal rate of 4.25%-4.50% or higher by the end of next year. We think the near-term move lands at the December 9 meeting rather than October 27-28, on the same election-proximity logic that argued for an October pause even before Wednesday’s decision reshaped the path. Every dot on that plot, and every word of this call, is ceteris paribus. Something else always changes, and something else already has: Brent above $104, retail diesel above $6, a mosque bombing in Pakistan, and a Red Sea chokepoint drifting toward one armed faction’s control were not among the assumptions built into the September 16 projections.
Markets took the hike in stride. The S&P 500 closed the week at 7,650.50, up 0.17% on Friday, and the Nasdaq added 0.39% to 26,522.55, while the Dow slipped 0.18% to 51,682.64 and the Russell 2000 underperformed, down half a percent as smaller, more rate-sensitive balance sheets absorbed the higher-for-longer message first. The 10-year Treasury finished the week near 5.00%. Gold still rallied to $4,368.60, up 0.6% on Friday, as a late-week pullback in crude pulled real yields down even as the Fed tightened, a reminder that the inflation-adjusted rate, not the headline funds rate, is what moves the metal.
Four fronts, not four headlines
We have covered Pakistan’s border conflict with Afghanistan, Houthi expansion in the Red Sea, and Ukraine’s campaign against Russian energy infrastructure separately in past issues. This week we think they, and Greenland, belong in the same paragraph, because each is a version of the same underlying story: state authority fraying at a frontier, a chokepoint, or a resource claim, in four different regions at once.
Pakistan and Afghanistan. A suicide bomber struck inside the Kohat Police Lines compound during Friday prayers on September 18, killing police officers and civilians in a toll that different outlets put between roughly 20 and 30 in the attack’s first day and that is still moving as the count is finalized. A local group calling itself Ittehad-ul-Mujahideen Pakistan claimed responsibility, while the better-known Tehreek-e-Taliban Pakistan denied involvement. Islamabad’s response went straight at Kabul: the attackers had “roots in Afghanistan,” the Afghan Taliban regime “actively sponsors terrorism in Pakistan,” and its failure to act violates the 2020 Doha framework that obligated Afghan territory not be used against Pakistan. This follows a February and March stretch in which the two governments traded actual airstrikes and border clashes and each called the result open war. With the U.S. still absorbed by the war with Iran, first joined militarily in February and not yet concluded, we think Washington has less bandwidth than usual to manage a frontier dispute like this one, and similar flashpoints are more likely to surface elsewhere while that holds. A nuclear-armed state of roughly 250 million people escalating with an isolated, sanctions-bound neighbor is not, by itself, a market event. It is the kind of frontier-authority story that shows up first in reinsurance pricing and supply-chain routing around South Asia, and only later, if it escalates further, in a headline that moves financial markets.
The Red Sea. The Houthis have consolidated control over Yemen’s Red Sea coastline, including the port of Hodeidah and a string of islands, cementing their position astride the approach to the Bab al-Mandeb strait. Their targeting has widened from Israeli and US-linked shipping earlier this year to vessels and infrastructure connected to Saudi Arabia. Yemen’s anti-Houthi coalition has fractured further instead of consolidating in response, trading blame instead of coordinating. We think this is a meaningfully different risk than the episodic missile exchanges of 2024. An attack on a shipping lane is an insurance problem, priced ship by ship. A chokepoint under one faction’s durable control is a standing tax on the route, present in freight rates and crude time spreads whether or not a vessel is hit in a given week, and it is a real part of why Brent has held above $100 through a month with few other fresh bullish catalysts.
Ukraine and Russian energy. Ukraine struck the Yaroslavl refinery, roughly 15 million tons of annual processing capacity and one of Russia’s larger plants, along with a military airfield near Rostov-on-Don, on September 17. Overnight September 19-20, in what Moscow officials called the year’s largest drone attack on the capital, Ukraine reached the Moscow oil refinery itself, in the city’s Kapotnya district, a plant that processed 11.6 million tons of crude in 2024; the wider attack killed three people in the Moscow region and forced roughly 400 residents from a damaged apartment tower in Ramenskoye. Open-source tracking of the broader campaign now puts two Rosneft refineries, at Syzran and Saratov, fully offline, with three of Russia’s six largest diesel plants shut down or running near a quarter of capacity. This is a targeted campaign against Russia’s refining base specifically, layered on top of a Russian domestic fuel crisis now stretching across two winters. The read-through for us is less about Brent, where global spare capacity has absorbed the loss of Russian refined-product exports reasonably well, and more about the diesel crack, already the tightest margin in the energy complex this year and now facing another marginal buyer looking to replace barrels that used to come from Russia.
Greenland. On September 19, the United States and Denmark announced a framework agreement expanding the American military footprint in Greenland beyond the existing Pituffik Space Base, while both governments emphasized that Danish sovereignty and Greenlandic self-determination remain intact. The framework bars non-NATO bases on the island and blocks Chinese or Russian investment and installations. All three governments are expected to sign formally at the UN General Assembly this month, pending ratification by the Danish and Greenlandic parliaments. This is close to what we argued would happen in “Trump’s Greenland Gambit: The Panama Canal Lesson America Forgot”, published in this newsletter on January 26, when we compared Washington’s approach to the 1999 Panama Canal handover, where formal US withdrawal left a strategic vacuum that Chinese capital filled through ports, logistics and financing rather than through force. Our case then was that “Ownership has been replaced by control. Purchase has been replaced by permanence. Strategic drift has been replaced by alignment,” built on three pillars we can now see in this week’s announcement: a locked-in, permanent US military footprint, an effective American veto over foreign infrastructure and investment decisions, and preferred access for allied firms in resource development, all without a formal change in sovereignty. That is a fair description of what got announced this week. The question we raised in January, whether formal sovereignty is enough protection against a patient strategic competitor when it isn’t backed by permanence on the ground, is the same one Danish and Greenlandic lawmakers are now being asked to answer by ratifying it.
Taken as a group, none of these four is, by itself, a threat to US growth. Together they put a standing bid under energy prices and under the dollar’s safe-haven role at the exact moment the Federal Reserve has told us it intends to keep raising rates.
More power, less output, and the labor market holding the line
The rest of last week’s data slate is consistent with an economy that is still capital intensive, and last week’s numbers restated that pattern rather than revised it. Retail sales rose 1.2% in August against an 0.8% consensus, and the control group, the cut that excludes autos, gasoline, building materials and food service and maps most closely to the consumer-spending line in GDP, rose 1.4%. Those non-core items feed into the GDP calculation in other ways, so they are not excluded from our analysis. Building materials sales feed into the calculation for residential investment, which remains soft. Housing starts fell 2.6% on the month to a 1.275 million unit annual rate, with single-family holding up better than multifamily; permits told a similar story, down 2.7% on the month but up 3.5% over the year, with multifamily permits at 467,000 suggesting builders are still committing to rental supply even as for-sale demand softens at the margin. Part of the strength in apartment building reflects would-be homeowners remaining renters for longer. Completions, which reflect decisions made twelve to eighteen months ago, fell 11.9% on the month and 27.1% over the year, the sharpest move in the report and a lagging signal.
Industrial production was unchanged in August and manufacturing output slipped 0.3%, with capacity utilization at 76.3%, more than three points below its long-run average; there is real slack in the factory sector, and it is a meaningful reason the Committee does not see the revival in manufacturing as an inflation risk even while it raises rates elsewhere. Utilities output rose 1.8% on the month and 6.2% over the year, easily the fastest-growing line in the report and the plainest evidence of the capital-intensity story: the fastest-growing part of industrial production is the part that exists to power everything else, principally data centers and the broader electrification of business investment.
The two regional manufacturing surveys disagreed on direction and agreed on the labor market. The Philadelphia Fed’s index rose to 37.8 in September, well above the 30.5 consensus. The Empire State survey fell to 7.6 from 20.6 in August, with new orders holding narrowly positive at 2.0 but shipments slipping negative. Both surveys’ employment components stayed constructive, with Empire State’s employee count index rising to 10.6 and hours worked jumping to 17.0. Two regional surveys moving in opposite directions while both point to steady hiring reads more like a plateau than a turn in either direction. Initial jobless claims came in at 196,000 for the week ended September 13, still historically low. The Conference Board’s Leading Economic Index fell 0.1% in August to 99.5, its first monthly decline since March, with the six-month growth rate turning slightly negative; the Board trimmed its own 2027 GDP forecast to 1.8% from 1.9%, a soft revision, not an alarming one, and one more data point for a labor market and a consumer holding up better than the leading indicators suggest they eventually will.
Behind the Numbers
Retail sales’ control-group reading rose 1.4% in August, the cleanest read of underlying consumer demand in the report, and it came with a caveat worth keeping. It strips out the categories, autos, gasoline, building materials and food service, that swing on price and financing terms, not volume. Reuters’ own reporting on the release credited steady wage growth and recent stock market gains for the strength, but also noted consumers are becoming more selective and searching out lower-priced goods even as they spend more in aggregate. That is a consumer still willing to spend, not a consumer spending without discipline, and it is the more durable version of resilience.
Housing completions are a rearview mirror, and the mirror is showing a decision made under different conditions. A unit reflected in this month’s completions data typically broke ground twelve to eighteen months earlier, which places the projects finishing now squarely in the high-rate stretch of 2025. The 11.9% monthly and 27.1% annual declines in completions say more about builder caution then than about demand now. The more forward-looking number in the same report, multifamily permits at 467,000, tells a different story: builders are still willing to commit to new rental supply even as the for-sale, single-family segment cools at the margin, which is the right read for anyone financing or supplying that pipeline.
Slack in the factory floor, none in the power grid, and that asymmetry is doing real work in the Fed’s own reaction function. Capacity utilization at 76.3%, more than three points below its long-run average, is a big part of why the Committee can raise rates without worrying that it is choking off manufacturing capacity that’s already running hot; there is room to spare. Utilities output growing at 6.2% year over year, in the same report, is the opposite condition: demand for power, driven by data centers and the broader capital-intensity buildout, is running well ahead of the rest of industrial production, and grid capacity, unlike factory capacity, is not something that can be brought back online with a few weeks’ notice.
Bottom Line
We now expect the Committee to pause at the October 27-28 meeting and deliver a second quarter-point hike on December 9, taking the funds rate to 4.00%-4.25% by year-end, and expect at least one further quarter-point increase in 2027 carrying the terminal rate to 4.25%-4.50% or higher. That call goes a quarter point past the Committee’s own median dot, and it is ceteris paribus in the fullest sense: it assumes the four geopolitical fronts above do not escalate further or allow even more flashpoints to emerge. If any of them do, a wider Pakistan-Afghanistan exchange, a Houthi move to fully close Bab al-Mandeb rather than tax it, or a sharper jump in Brent or the diesel crack from further Ukrainian strikes, we think the risk runs toward an earlier or larger move, not a smaller one, since all three would show up first as energy-price and dollar pressure, the two channels most likely to move this Committee. Check this against how Chair Warsh frames the neutral-rate question in October, since there is no Summary of Economic Projections at that meeting, and against the vote count and new projections on December 9 itself.
A personal note. A little over three years ago, a friend recommended that I read The End of the World is Just the Beginning. The author’s central thesis was that as the day America could guarantee freedom of navigation around the world was drawing to an end, a new system of toll takers would likely emerge. I thought the book was well written but questioned why he felt the U.S. Navy would become so much less effective. The book turned out to be a timely read and the timetable for challenges to the post-World War II protection of the world’s sea lanes has clearly come into question. The book is still a timely read and well worth the time and effort.
CFO and Treasurers Corner
Stop planning around a near-term cut. Funding costs now have a floor near 4.00%-4.25% by year-end and higher in 2027 under our call. Lock in financing decisions against that path, not a hoped-for easing.
Budget the diesel crack above $6, not the barrel. The marginal driver is refining capacity lost to Ukrainian strikes and Houthi chokepoint risk, not global crude supply and demand. Hedge the crack spread specifically where distillate exposure is material.
Reassess Red Sea and Suez routing insurance now, not after the next incident. Houthi consolidation of Yemen’s Red Sea coast is a standing cost on that route, not an episodic one. Compare current premiums against Cape of Good Hope routing economics for any shipment where the math was close before.
Widen dollar hedge ratios on international receivables. A Fed that has told markets it intends to keep raising rates, against central banks elsewhere that are further along in easing, argues for a firmer dollar over the near term. Revisit hedge ratios set before September 16.
Treat grid interconnection and power capacity as a line item in any project with material electricity load. Utilities output is growing at more than six times the pace of the rest of industrial production. Capacity charges and curtailment terms belong in the underwriting, not as an afterthought.
Read the housing completions plunge as a lagging signal, not a forward one, when assessing exposure to that sector. Multifamily permits at 467,000 are the better forward indicator, and they still show builders committing to new supply. Higher mortgage rates are likely to persist throughout the forecast period. Yet we still see a modest recovery in existing home sales next spring and summer, driven by a steady decline in asking prices in major markets.
Piedmont Perspective
The Committee Signals One, We Expect Two
Abridged from the standalone essay published September 16.
The Federal Reserve raised its target range a quarter point to 3.75%-4.00% on September 16, its first increase since 2023, on a unanimous 12-0 vote. The decision itself was less informative than what came with it: a Summary of Economic Projections whose median dot now shows the funds rate ending this year at 4.1%, not 3.75%-4.00%, which means the median Committee member expects at least one more increase before December 31.
The new projections moved in one direction across the board. The funds rate path rose to 4.1% in 2026 and 2027, 3.9% in 2028, 3.6% in 2029 and 3.2% longer run, up from June’s 3.8%, 3.6%, 3.4% and 3.1%. Core PCE inflation is now projected at 3.4% this year before easing to 2.5%, 2.2% and 2.0% over the following three years. GDP growth is projected at 2.3% this year and 2.4% next, with unemployment expected to hold near 4.1% through 2028 before settling at 4.2% longer run. None of these are dramatic revisions individually. Together, they describe a Committee that raised rates and simultaneously told markets it isn’t finished, which is a more hawkish combination than either action would be alone.
The dot plot above shows the same distribution behind the median: two members consider September the last increase, twelve see one more move before year-end, and four see two. A Committee where fourteen of eighteen members expect at least one further hike is not a Committee signaling completion, whatever the headline vote count implied on the day. We take the Committee at its word and go a quarter point past its own median: at least two more quarter-point increases from here, one later this year and at least one more in 2027, with a terminal rate of 4.25%-4.50% or higher by the end of next year.
Every projection in that plot, and every word of our own call, carries the same caveat: it is ceteris paribus. Every dot assumes conditions the Committee cannot actually hold constant, and several of them have already moved since September 16. Brent crude has traded above $104, retail diesel above $6, a chokepoint in the Red Sea has drifted toward the control of a single armed faction, and a Pakistani mosque bombing has reopened a border conflict Islamabad and Kabul were already calling open war. None of these appear in the Committee’s own projections, because none of them are the Committee’s job to project. They are ours, and they are why our confidence in the near-term timing of the next move, October 27-28 or December 9, is lower than our confidence in the direction.
Full essay, “The Committee Signals One, We Expect Two”, published September 16, 2026, is available at piedmontcrescentcapital.com.
The Week Ahead
Monday: Chicago Fed National Activity Index.
Tuesday: Conference Board Consumer Confidence; flash S&P Global manufacturing and services PMIs.
Wednesday: New home sales.
Thursday: Initial jobless claims; durable goods orders; final estimate of second-quarter GDP.
Friday: August core PCE inflation, the Fed’s preferred gauge and the last one before the October 27-28 meeting; final University of Michigan consumer sentiment.
With the post-meeting blackout lifted, watch for the first round of Fed speeches since September 16 for how individual Committee members, Governor Waller in particular, characterize both the vote itself and the neutral-rate question that will shape whether October is genuinely a pause or a bridge to a larger December move. We expect a handful of regional Fed presidents to push for an October increase, which we read as more of a political endeavor than economic one.
The diplomatic calendar carries its own asymmetry this week. The UN General Assembly’s high-level week runs September 21-25 in New York, with President Trump addressing the Assembly Tuesday. Xi Jinping is skipping the Assembly itself and meeting Trump one-on-one at the White House on Thursday, September 24, with artificial intelligence, Iran-related sanctions, tariffs and Taiwan all on the agenda. Most coverage of the summit describes it as closer to a state visit than a negotiation, with limited concrete outcomes expected. We think that sets a low bar: with expectations already discounted, the asymmetry favors an upside surprise on trade or AI cooperation over a disappointment, the diplomatic mirror of a Fed decision the market had already priced to roughly 90%.
U.S. Economic & Financial Outlook
The table below carries our September 18 forecast vintage. The quarterly funds-rate path now carries 3.75%-4.00% through Q3-26, 4.00%-4.25% in Q4-26, consistent with the December 9 hike in our call above, and 4.25%-4.50% from Q1-27 through year-end 2027, consistent with the additional 2027 increase we call for in the text above. The two are now aligned; we are not carrying a further move beyond that path at this vintage.
U.S. Economic and Financial Outlook, as of September 18, 2026: annual
| Indicator | Units | 2024 Actual | 2025 Actual | 2026 Forecast | 2027 Forecast |
|---|---|---|---|---|---|
| Output | |||||
| Real GDP | % q/q SAAR | 2.8 | 2.1 | 2.3 | 2.7 |
| 4th Qtr-to-4th Qtr Change, Yr/Yr Chg for Qtly Data | % YoY | 2.4 | 2.0 | 2.6 | 2.7 |
| Final Sales to Private Domestic Purchasers | % q/q SAAR | 3.0 | 2.6 | 2.7 | 2.9 |
| Consumer Spending | % q/q SAAR | 2.9 | 2.6 | 2.3 | 2.0 |
| Nonresidential Fixed Investment | % q/q SAAR | 2.9 | 4.1 | 6.2 | 4.6 |
| Residential Investment | % q/q SAAR | 3.2 | -2.2 | -1.8 | 0.5 |
| Industrial Production, Manufacturing | % YoY | -0.9 | 1.0 | 2.0 | 3.9 |
| Light Vehicle Sales | Mil. units, SAAR | 15.9 | 16.2 | 16.1 | 16.4 |
| Labor Market | |||||
| Unemployment Rate | % | 4.0 | 4.3 | 4.2 | 4.0 |
| Nonfarm Payrolls | Avg. mo. chg., thous. | 150 | 62 | 82 | 95.25 |
| Employment Cost Index | % YoY | 3.9 | 3.5 | 3.4 | 3.5 |
| Housing Market | |||||
| Housing Starts | Thous. units, SAAR | 1,370 | 1,356 | 1,354 | 1,375 |
| New Home Sales | Thous. units, SAAR | 684 | 679 | 636 | 668 |
| Existing Home Sales | Thous. units, SAAR | 4,060 | 4,060 | 4,076 | 4,205 |
| Case-Shiller National Home Prices | % YoY | 5.1 | 2.2 | 1.4 | 2.5 |
| Inflation | |||||
| Consumer Price Index (CPI) | % YoY | 3.0 | 2.6 | 3.3 | 2.7 |
| Core CPI | % YoY | 3.4 | 2.8 | 2.6 | 2.8 |
| PCE Deflator | % YoY | 2.6 | 2.6 | 3.3 | 2.8 |
| Core PCE Deflator | % YoY | 2.9 | 2.8 | 3.1 | 2.7 |
| Interest Rates & Markets | |||||
| Fed Funds Target Range | Avg. / range, % EOP | 5.14 | 4.21 | 3.63 | 4.06 |
| 2-Year Treasury | %, period end | 4.37 | 3.81 | 3.92 | 4.06 |
| 10-Year Treasury | %, period end | 4.21 | 4.29 | 4.65 | 4.83 |
| 30-Year Mortgage Rate | %, period end | 6.72 | 6.60 | 6.66 | 6.83 |
| Brent Crude | $/bbl, period avg. | 79.91 | 68.32 | 92.05 | 84.50 |
| Trade-Weighted Dollar Index | Period average | 122.9 | 122.7 | 119.0 | 119.6 |
| Trade-Weighted Dollar Index | % chg., period | 1.7 | -0.1 | -3.0 | 0.5 |
Annual figures are full-year averages or year-over-year percent changes; the 2024-2025 columns are actual, 2026-2027 are forecasts.
U.S. Economic and Financial Outlook, as of September 18, 2026: quarterly
| Indicator | Q1-26 | Q2-26 | Q3-26 | Q4-26 | Q1-27 | Q2-27 | Q3-27 | Q4-27 |
|---|---|---|---|---|---|---|---|---|
| Output | ||||||||
| Real GDP | 2.1 | 1.5 | 3.6 | 3.1 | 2.5 | 2.7 | 2.8 | 2.7 |
| 4th Qtr-to-4th Qtr Change, Yr/Yr Chg for Qtly Data | 2.7 | 2.1 | 1.9 | 2.6 | 2.7 | 3.0 | 2.8 | 2.7 |
| Final Sales to Private Domestic Purchasers | 1.7 | 4.2 | 4.5 | 3.9 | 3.7 | 3.4 | 3.3 | 3.2 |
| Consumer Spending | 0.5 | 3.4 | 2.3 | 2.0 | 2.2 | 2.1 | 2.1 | 2.0 |
| Nonresidential Fixed Investment | 10.6 | 8.5 | 9.0 | 7.7 | 6.6 | 5.9 | 5.7 | 5.4 |
| Residential Investment | -7.8 | 0.0 | -3.8 | -1.3 | -0.5 | 0.4 | 0.8 | 1.2 |
| Industrial Production, Manufacturing | 1.0 | 1.6 | 1.6 | 3.6 | 4.0 | 3.6 | 3.7 | 3.7 |
| Light Vehicle Sales | 15.5 | 16.4 | 16.2 | 16.3 | 16.3 | 16.3 | 16.4 | 16.4 |
| Labor Market | ||||||||
| Unemployment Rate | 4.3 | 4.3 | 4.1 | 4.1 | 4.1 | 4.0 | 4.0 | 3.9 |
| Nonfarm Payrolls | 73 | 81 | 85 | 88 | 88 | 92 | 98 | 103 |
| Employment Cost Index | 3.4 | 3.4 | 3.4 | 3.4 | 3.5 | 3.4 | 3.5 | 3.5 |
| Housing Market | ||||||||
| Housing Starts | 1,418 | 1,337 | 1,320 | 1,340 | 1,360 | 1,380 | 1,380 | 1,380 |
| New Home Sales | 622 | 650 | 630 | 640 | 650 | 660 | 680 | 680 |
| Existing Home Sales | 4,043 | 4,120 | 4,040 | 4,100 | 4,160 | 4,190 | 4,220 | 4,250 |
| Case-Shiller National Home Prices | 0.8 | 1.2 | 1.5 | 1.9 | 2.3 | 2.5 | 2.6 | 2.7 |
| Inflation | ||||||||
| Consumer Price Index (CPI) | 2.7 | 3.8 | 3.4 | 3.2 | 3.0 | 2.3 | 2.6 | 2.6 |
| Core CPI | 2.5 | 2.7 | 2.7 | 2.6 | 2.7 | 2.8 | 2.8 | 2.8 |
| PCE Deflator | 3.1 | 3.9 | 3.2 | 3.0 | 2.9 | 2.5 | 2.7 | 2.7 |
| Core PCE Deflator | 3.1 | 3.4 | 3.1 | 2.8 | 2.7 | 2.4 | 2.6 | 2.6 |
| Interest Rates & Markets | ||||||||
| Fed Funds Target Range | 3.50-3.75 | 3.50-3.75 | 3.75-4.00 | 4.00-4.25 | 4.25-4.50 | 4.25-4.50 | 4.25-4.50 | 4.25-4.50 |
| 2-Year Treasury | 3.78 | 4.13 | 4.40 | 4.40 | 4.40 | 4.35 | 4.35 | 4.30 |
| 10-Year Treasury | 4.35 | 4.40 | 4.95 | 4.90 | 4.85 | 4.85 | 4.80 | 4.80 |
| 30-Year Mortgage Rate | 6.10 | 6.49 | 7.10 | 6.95 | 6.85 | 6.85 | 6.80 | 6.80 |
| Brent Crude | 78.00 | 100.20 | 98.00 | 92.00 | 90.00 | 88.00 | 82.00 | 78.00 |
| Trade-Weighted Dollar Index | 119.0 | 119.5 | 119.1 | 118.5 | 118.8 | 119.3 | 120.0 | 120.3 |
| Trade-Weighted Dollar Index | -1.4 | 0.4 | -0.4 | -0.5 | 0.3 | 0.4 | 0.5 | 0.2 |
Quarterly figures are seasonally adjusted annual rates, year-over-year changes where the units say so, or period averages and period-end levels for rates. The second-quarter 2026 column is the latest actual. Source: Piedmont Crescent Capital, BLS, BEA, Census, Federal Reserve, EIA.
Sources and Notes
Sources. Bureau of Labor Statistics (Consumer Price Index); U.S. Census Bureau (Advance Monthly Retail Trade, New Residential Construction); Federal Reserve Board (FOMC statement and Summary of Economic Projections, September 16, 2026; G.17 Industrial Production and Capacity Utilization); Federal Reserve Bank of Philadelphia (Manufacturing Business Outlook Survey); Federal Reserve Bank of New York (Empire State Manufacturing Survey); U.S. Department of Labor (Initial Jobless Claims); The Conference Board (Leading Economic Index, August 2026); Oxford Economics (core-goods and core PCE tracking, cited for its independent corroboration).
Geopolitical. Geo News and Arab News (Kohat attack and Pakistan’s statement on the Afghan Taliban); Al Jazeera (Pakistan-Afghanistan border conflict; Houthi consolidation of Yemen’s Red Sea coast); Bloomberg (Houthi shipping and Saudi oil risk; Ukraine’s strike on the Yaroslavl refinery; the US-Denmark Greenland framework); The Moscow Times, via Reuters (the September 20 strike on the Moscow refinery); Kyiv Independent and bne IntelliNews (Ukraine’s campaign against Russian refining capacity); NPR (US-Denmark Greenland agreement terms); Encyclopaedia Britannica (status of the 2026 Iran war); Arab News, via AFP (the September 24 Trump-Xi meeting and characterization of expectations for it); Scripps News (Trump’s UN General Assembly address); the United Nations (General Assembly high-level week dates); Piedmont Crescent Capital, “Trump’s Greenland Gambit: The Panama Canal Lesson America Forgot”, January 26, 2026 (our own prior Greenland argument).
Market data as of Friday, September 18, 2026 close: Yahoo Finance and TheStreet (equity indices); USAGold (precious metals); Fortune (Brent crude); U.S. Department of the Treasury (daily par yield curve); Freddie Mac (Primary Mortgage Market Survey); ICE BofA indices via FRED (investment-grade and high-yield option-adjusted spreads).
A View from the Piedmont is published weekly by Piedmont Crescent Capital for informational purposes only and does not constitute investment, legal, or tax advice. Views expressed are those of the author as of the date of publication and are subject to change. © 2026 Piedmont Crescent Capital, LLC. All rights reserved.
August Housing Starts: Single-Family Comes Back, Multifamily Falls Away
ECONOMIC INDICATOR REPORT · HOUSING STARTS, AUGUST 2026
Single-Family Comes Back, Multifamily Falls Away
August’s single-family rebound and multifamily pullback reversed the year-to-date pattern for one month, and builder sentiment fell to a one-year low the day before this report.
Early Signal
- Single-family starts posted their best month since March. Single-family starts rose 7.6% in August to 918,000 units at a seasonally adjusted annual rate, the fastest pace since March and enough to push the year-over-year comparison to plus 5.2%. Single-family permits held closer to flat, down 1.8% to 878,000, so the August jump reads more like builders working through a backlog than a fresh acceleration in demand. Year to date, however, single-family starts are still running 4.9% below the same eight months of 2025, so one strong month has not closed the gap.
- Multifamily construction fell off a cliff. Starts on buildings of five or more units dropped 22.5% to a 344,000-unit pace, reversing July’s spike and pulling the year-over-year comparison to minus 15.5%. Multifamily permits fell a smaller 3.1%, to 467,000, which argues August’s air pocket in starts is at least partly a timing story rather than a change in the pipeline. Year to date, multifamily starts are still running 10.0% above 2025’s pace, so August looks like a pause in a stronger multifamily year, not a new trend.
- The composite still slipped. Total starts fell 2.6% to 1,275,000, the third decline in the last four months, because the multifamily decline outweighed the single-family rebound. Total permits, the better leading indicator, fell 2.7% to 1,394,000 but held 3.5% above a year ago. Year to date, total starts are averaging 1,359,000 through August, down 1.0% from the same span of 2025, with single-family down 4.9% to 913,750 and multifamily up 10.0% to 430,875. Permits are averaging 1,416,000 through August, down 0.7%, with single-family down 3.4% to 890,000 and multifamily up 5.9% to 475,000, a less noisy version of the same split.
- Completions cratered. Total completions fell 11.9% to 1,128,000, down 27.1% from August 2025’s unusually high 1,548,000. The comparison flatters the size of the decline, because last year’s completions were still working off the record 2023-2024 multifamily construction wave, and that wave has largely cleared the pipeline.
- Builder sentiment matched its worst reading in a year. The NAHB/Wells Fargo Housing Market Index fell three points to 32 in September, tying the September 2025 low, with all three components, current sales, six-month expectations and buyer traffic, at or below their year-ago levels. Thirty-eight percent of builders cut prices in September, up from 35% in August, and 66% offered some form of sales incentive, the highest share since December.
- Mortgage rates moved further from the house forecast. The 30-year fixed rate reached 6.95% this week, up from 6.76% the week before and an August average near 6.67%, the highest level in more than a year and well above PCC’s own fourth-quarter workbook path of 6.60%, as Treasury yields absorbed this week’s Fed rate hike and elevated geopolitical risk. Higher financing costs were the reason builders cited most often for September’s drop in sentiment.
- Regulation and remodeling both point to a market with less room to maneuver. A new NAHB survey puts the regulatory cost of building a single-family home at $131,734, or 26.4% of the sale price, up more than 40% from $93,871 in the association’s 2021 study. Separately, Harvard’s Joint Center for Housing Studies now expects growth in home improvement spending to slow to just 0.5% year over year by the second quarter of 2027, down from 1.8% in the second quarter of this year.
- PCC’s own call remains on track, for now. Starts have averaged 1,359,000 through the first eight months of 2026, a bit ahead of our 1,336,000 full-year forecast despite August’s pullback. The single-family/multifamily mix, not the total, is the thing to watch, and the August starts moved against our standing expectation that multifamily gives up relatively less ground than single-family this cycle. Year-to-date permits are much closer to our expectations, averaging 1,416,000 through August, down just 0.7% from a year ago and with far smaller single-family/multifamily swings than starts have shown.
The Overview
The Census Bureau's New Residential Construction report for August, released this morning, and NAHB's Wells Fargo Housing Market Index for September, released yesterday, describe two different problems in the same market. The Census data show single-family builders finally getting relief, with starts and, to a lesser extent, permits both moving higher. The NAHB survey, fielded over roughly the same weeks, shows builders more pessimistic than at any point in the past year. Both can be true at once. Builders can be breaking ground on homes they already have permits and lots for while telling surveyors they expect the next few months to be harder, and that is the more likely reading of August than a genuine turn in either direction.
Total starts fell 2.6% to 1,275,000 at a seasonally adjusted annual rate, essentially unwinding July's 1,309,000 gain and leaving the level 1.2% below August 2025. Total permits fell a similar 2.7% to 1,394,000, still 3.5% above a year ago, which keeps the permits-to-starts relationship, the more reliable leading signal, pointed modestly higher on a trailing basis even as the single-month starts print disappointed.
The Mix: Single-Family vs. Multifamily
August's headline decline was entirely a multifamily story. Single-family starts rose 7.6% to 918,000, their best print since March, while starts on buildings of five or more units fell 22.5% to 344,000, erasing essentially all of the prior month’s surge. Permits tell a more consistent version of the story. Single-family permits slipped 1.8% to 878,000 and multifamily permits fell 3.1% to 467,000, both far smaller moves than the swings in starts. When starts move more than permits in the same direction, that is usually noise in the timing of groundbreakings rather than a change in what builders are approved to build, and this month's gap between the starts and permits moves, in both categories, argues for treating August as noisy rather than definitive. The year-to-date levels tell the more reliable story. Single-family starts are running 4.9% behind last year's pace while multifamily starts are running 10.0% ahead of it, the reverse of August's single-month print and the better read of where this cycle actually stands.
The multifamily air pocket is easier to explain than the single-family bounce. Builders delivered an outsized wave of apartments in 2023 and 2024, and the completions data below show that wave is still working through lease-up. A single soft month for multifamily starts, on top of that overhang, is the more parsimonious explanation than a sudden change in apartment demand.
The Regional Split
National single-family starts do not say where the improvement came from, and the answer matters for judging whether August's rebound is durable. Regional starts carry a wide margin of error and are unusable month to month on their own, so permits, the steadier series, are the better cross-check.
| Region | SF Starts, Aug (000s) | M/M | SF Permits, Aug (000s) | M/M |
|---|---|---|---|---|
| Northeast | 46 | -2.1% | 52 | -7.1% |
| Midwest | 143 | +23.3% | 130 | 0.0% |
| South | 531 | -0.9% | 527 | -0.6% |
| West | 198 | +28.6% | 169 | -5.1% |
Source: Census Bureau/HUD, New Residential Construction, Table 3a (starts) and Table 1a (permits), August 2026.
The West and Midwest produced essentially all of the national single-family gain, up 28.6% and 23.3% in a single month, while the South and Northeast were flat to lower. Permits in those same two regions moved the other way, flat in the Midwest and down 5.1% in the West. When starts move sharply in a region where permits do not confirm it, the more likely explanation is builders converting existing entitlements and already-developed lots into construction, not a fresh burst of new commitments.
That reading lines up with what builders themselves are describing. NAHB's September survey found 42% of builders rate lot availability as poor and 38% as fair, so the large majority are land-constrained rather than land-flush, and this month's trade coverage of the sector's cost structure describes finished-lot and land-development costs as now embedded in builders' cost basis in a way that pushes them to keep building through weak demand rather than around it. Our read, consistent with that commentary, is that a good share of August's single-family pickup reflects builders monetizing lots and development costs they cannot otherwise recover, not a change in buyer demand, which also fits a month in which builder sentiment fell to a one-year low.
The complication is that completed, unsold inventory is already near a cycle high, close to four times the February 2022 low, with total months' supply at 9.6, above the four-to-six-month range that has historically balanced the market. Building to work through sunk land costs is a sound explanation for why starts can rise even as sentiment falls, but it does not by itself resolve an already-elevated inventory of finished, unsold homes, and it argues for caution about reading August's single-family pickup as the start of a sustained upturn.
Completions and the Vacancy Overhang
Total completions fell 11.9% in August to 1,128,000, and the year-over-year comparison, down 27.1% from August 2025's 1,548,000, is the sharpest of any series in this release. Single-family completions fell 10.4% to 816,000 and multifamily completions fell to 302,000.
Harvard's Joint Center for Housing Studies put a number on the overhang in its June State of the Nation's Housing report. The rental vacancy rate has climbed to 7.3% as of the first quarter of 2026, up from a record low of 5.9% in 2022, and the for-sale vacancy rate has risen to roughly 1.1% from 0.81% in 2023. The Center's own read is that a multifamily construction wave layered on top of already-elevated single-family building pushed vacancy up from those historic lows, and that is precisely the supply now working its way through completions. Fewer completions from here is the expected, not the alarming, response to that much vacant inventory.
Builder Sentiment: The NAHB/Wells Fargo Survey
The HMI is a diffusion-style index. Each component is built from the share of builders rating conditions good, fair or poor, so a move in the index measures how many builders shifted their view, not how far conditions moved for any single builder.
The headline index fell three points in September to 32, tying the September 2025 reading as the weakest of the past year. All three components declined or held flat, as current sales conditions fell four points to 35, sales expectations for the next six months fell six points to 37, and prospective buyer traffic held at 23. Regionally, on NAHB's three-month moving average, the Midwest led at 44 (down one point), followed by the Northeast at 39 (down five), the South at 31 (down one) and the West at 28 (up one). No region is above the 50 breakeven line, so the gap between regions is a measure of relative resilience in a soft market rather than strength anywhere.
NAHB Chairman Bill Owens attributed the drop mainly to financing costs: “Buyer traffic weakened across much of the country largely due to rising mortgage rates. Builders face higher material costs, rising fuel prices, labor shortages, and immigration enforcement issues.” Chief Economist Robert Dietz added that “builder confidence is at its lowest level since September 2025 as tight lending conditions and elevated land, labor, and construction costs continue.” Builders are responding on price. Thirty-eight percent cut prices in September, up from 35% in August, with the average cut holding at 6% for a sixth straight month, and 66% offered a sales incentive, the highest share since December, up from 63%.
The Cost Side: Regulation and Remodeling
A new NAHB regulatory cost study, based on a March 2026 builder survey and January 2026 home price data, puts the total regulatory cost of a new single-family home at $131,734, or 26.4% of the final sale price. That is up more than 40% from the $93,871 the association measured in its 2021 study, split between $46,795 in lot development costs and $84,939 in costs tied directly to construction. Regulatory cost has grown faster than the home price it is measured against, which is the cleaner way to read the 40% increase: builders are not just paying more because homes cost more, they are paying a larger share of a home's price to regulation than they were five years ago.
Remodeling is decelerating alongside new construction rather than offsetting it. Harvard's Leading Indicator of Remodeling Activity, updated July 23, now projects annual home improvement and repair spending will slow to 0.5% year-over-year growth by the second quarter of 2027, down from 1.8% growth in the second quarter of 2026, as spending levels off near $519 billion. The Center points to flattening remodeling permits and building-material retail sales, alongside reduced housing starts, as the reasons homeowners are pulling back on renovation spending alongside builders pulling back on new construction.
Rates and the Road Ahead
The Federal Reserve raised the federal funds rate a quarter point to 3.75%-4.00% yesterday, its first increase since July 2023, on a unanimous 12-0 vote. PCC's call is that at least two more quarter-point increases are likely, one later in 2026 and at least one more in 2027, putting the terminal rate at 4.25%-4.50% or higher by the end of next year, a touch more hawkish than the Fed's own median projection. Every rate call of this kind is a ceteris paribus forecast. Each policymaker says what the Committee should do if nothing else changes, and something else is already changing. The 30-year fixed mortgage rate reached 6.95% this week, well above PCC's own fourth-quarter workbook path of 6.60%, and higher financing costs were the single most-cited reason builders gave for September's drop in sentiment.
A higher-for-longer funds rate argues for continued pressure on mortgage rates, and mortgage rates, not the funds rate directly, are what shows up in builder surveys and buyer traffic. If the Fed's hawks are right about a second 2026 hike, that pressure does not relent before the spring selling season.
Our Call
We are not changing our full-year forecast on the strength of one month. PCC still expects total housing starts to average 1,336,000 in 2026 and 1,353,000 in 2027. Starts have averaged 1,359,000 through August, running a little ahead of that call even after this month's pullback, so the level forecast needs no revision. What we are watching is the mix. Our standing view is that single-family production keeps giving up relative ground to multifamily through this cycle, and August moved the other way. Year to date, the mix still points our way. Single-family starts are down 4.9% from last year and multifamily is up 10.0%, so on the evidence so far August was the outlier, not the trend. Here is the test, with dates: if the September release, due in mid-October, shows single-family starts giving back most of August's gain while multifamily stays soft, that confirms August was noise. If single-family instead holds near 900,000 while multifamily stabilizes above 350,000, that would argue the mix call needs revisiting before year end.
President & Chief Economist, Piedmont Crescent Capital
Sources
U.S. Census Bureau and U.S. Department of Housing and Urban Development, New Residential Construction, August 2026, including Table 1a (regional permits) and Table 3a (regional starts) (released September 17, 2026). National Association of Home Builders and Wells Fargo, Housing Market Index, September 2026, including the lot-availability question (released September 16, 2026). National Association of Home Builders, Eye on Housing, regulatory cost survey (March 2026 survey; published June 2026). Harvard University, Joint Center for Housing Studies, The State of the Nation's Housing 2026 (released June 17, 2026) and the Leading Indicator of Remodeling Activity (released July 23, 2026). Trade press coverage of builder finished-lot and completed-inventory cost dynamics, September 2026. Freddie Mac, Primary Mortgage Market Survey, 30-year fixed rate, week of September 17, 2026. Board of Governors of the Federal Reserve System, FOMC statement and Summary of Economic Projections, September 16, 2026. Piedmont Crescent Capital internal forecast workbook.
This report reflects the analysis and opinions of Piedmont Crescent Capital as of the date shown and is provided for general informational purposes only. It does not constitute investment, legal, or tax advice, is not an offer or solicitation, and should not be relied upon as the basis for any investment decision. Forecasts and other forward-looking statements are inherently uncertain and actual outcomes may differ materially.
The Economy Was Stronger Than First Reported | A View from the Piedmont
A View from the Piedmont — Our Weekly Commentary on Money, Credit, Exchange Rates & Geopolitics
October 5, 2026 · Mark P. Vitner, Chief Economist · mark.vitner@piedmontcrescentcapital.com
The Economy Was Stronger Than First Reported
A government revision just told us the expansion was stronger, and more capital-intensive, than anyone had been measuring, even as a soft September jobs report drew the week’s attention. It did not change the character of the recovery. It confirmed it.
Editor’s note: Our latest issue covers the Bureau of Economic Analysis’s annual national accounts revision and third estimate of second-quarter GDP (September 30), the September employment report (October 2), second-quarter state GDP and personal income by state (September 30), Hormuz shipping and Iranian government statements, OPEC+’s November output decision and Monday’s oil-market action, Brazil’s October 4 first-round presidential vote, and the week’s data-center permitting news, through October 5.
Market Dashboard
(as of Friday, October 2 close, unless noted)
Note: Treasury yields are the Federal Reserve’s H.15 constant-maturity series for Thursday, October 1, the most recent full curve available at this writing. Equity, Brent, WTI and gold levels are Friday, October 2 close. Brent and WTI are shown against the contract months currently trading nearest the front of the curve.
Summary
A Bureau of Economic Analysis revision released alongside the third estimate of second-quarter GDP raised the growth rate for the first half of 2026 by the most in years, and roughly two-thirds of the upward move originated in categories tied to the AI buildout: data centers, power facilities and the equipment that they operate. Real GDP now stands at a 2.2 percent annual rate for the second quarter, versus 1.5 percent previously, and the first quarter was revised to 2.5 percent from 2.1 percent. Private data-center construction ran at an $85.0 billion annual rate in August, up 73.2 percent from a year earlier. The revision confirms what we have been describing all year: an economy spending readily on capital but sparingly on headcount.
The labor market told the other half of the story three days later. Nonfarm payrolls rose only 29,000 in September, the unemployment rate edged up to 4.2 percent. We feel it is better to view these numbers over time, however, as this past summer saw a great deal more volatility than past summers have. The three-month payroll average of 51,000 sits above our own breakeven estimate of zero to 40,000 a month but remains well south of what an economy growing at 2.5 percent-plus would ordinarily produce. Real consumer spending kept pace through August, rising 0.6 percent on the month, but income did not. The saving rate fell to 4.1 percent from a revised 4.6 percent, and the extra cushion still sits mostly with households living off pensions, interest and dividends, not paychecks. Re-emphasizing another key point we continue to make, that rising tide of retirees is responsible for much of the resiliency in the K-shaped economy. The revision also revealed that underlying inflation was slightly lower, with core PCE now 3.0 percent over the year rather than 3.3 percent, though diesel and a second-order list of refined products are still working their way toward the grocery aisle.
Outside the country, the shipping lanes are open but more expensive. An Iraqi state tanker completed an independent two-million-barrel transit past Hormuz on October 3, Middle East crude exports rose above prewar levels on four of the last seven days of September, and OPEC+ held its November output targets steady on October 4; war-risk premiums and rerouting costs are the tax now, not a blockade, and diesel, not crude, is the tight market.
Brazil’s October 4 first-round vote sends Lula into an October 25 runoff against Flávio Bolsonaro, who outpolled him 47.0 percent to 45.2 percent, and the U.S. midterms are four weeks out with the generic ballot and prediction markets both pointing toward divided government in Washington next year. Inside our own footprint, the same capital that is lifting the national accounts is running into power, water and local permitting limits faster than it is running into a shortage of money: Texas halted new data-center permits pending a grid audit, and Prince William County, Virginia cut its by-right development overlay by nearly two-thirds. We expect new policies to be enacted regarding data centers after the election, which will require hyperscalers to produce more tangible benefits to the local economy and provide assurances that they will not drive up energy costs.
We still look for at least two more quarter-point increases in the federal funds rate this cycle, one in December and at least one other in 2027. A reacceleration in core inflation or a further squeeze in diesel markets would put the October 28 meeting back in play, though we doubt a Committee that would rather not be seen meddling in politics a week before the midterms moves on it. A formal halt to the Hormuz traffic now moving would break our broader call on growth, which we discuss in Risks to our view, below.
This Week’s Argument
The books caught up to the buildout
Real GDP rose at a 2.2 percent annual rate in the second quarter, revised up 0.7 percentage point from the 1.5 percent pace in both the advance and second estimates. First-quarter growth was revised to 2.5 percent from 2.1 percent. The increase was led by consumer spending, investment and exports, while a larger rise in imports subtracted from the headline. Consensus expectations ahead of the release had anticipated no revision at all. Real final sales to private domestic purchasers, consumer spending plus gross private fixed investment, rose at a 4.6 percent annual rate, up from 4.2 percent and the strongest quarterly gain in that measure since the first quarter of 2023. Real gross domestic income rose 2.6 percent, revised up from 2.2 percent, leaving the average of GDP and GDI at 2.4 percent; in the first quarter, real GDI was revised to 2.5 percent from 1.2 percent. The income-side shortfall that had nagged the expansion was largely closed in the first half.
The annual figures carry the same story in smaller type. Real GDP growth in 2024 is now 3.0 percent, and 2025 growth is 2.3 percent, up from the 2.1 percent previously published. The 2023 estimate is unchanged at 2.9 percent. Compensation, drawn largely from new Quarterly Census of Employment and Wages data, led the first-quarter GDI revision, additional evidence that the income side had been understating activity rather than revealing a hidden slowdown.
The composition of the second-quarter revision fits the capital-heavy expansion we have been describing. BEA traced the upward revision in nonresidential structures to commercial and health-care construction, mainly data centers, using revised Census value-of-construction data for May and June. Private inventory investment was also revised higher, and equipment spending held at a double-digit pace. The revision shows up in the building, not in the servers. Census put private data-center construction at an $85.0 billion annual rate in August, up 73.2 percent from $49.1 billion a year earlier, with the pace rising from $61.9 billion in April. Data centers were about 63 percent of private office construction. General office construction fell 9.7 percent. Power construction, the companion line, was $165.9 billion, up 9.7 percent. Manufacturing construction, the offset, was $168.2 billion, down 19.8 percent, with computer and electrical plant down 45 percent. Construction of microchip plants had surged earlier in the decade, driven by the Chips and Science Act. An $85 billion construction pace is roughly three-tenths of one percent of GDP; the chips, the switchgear and most of the cooling equipment sit in equipment, not in structures.
Capital spending is accelerating well ahead of hiring. September payrolls rose only 29,000, and the unemployment rate edged up to 4.2 percent. August payrolls were revised to 133,000 net new jobs from 162,000 earlier, and July was revised to a loss of 10,000, leaving the three-month average at 51,000, above our own breakeven estimate of zero to 40,000 a month but short of the gain we had looked for after August’s initially reported increase. Average hourly earnings rose 3.0 percent over the year, the slowest pace since May 2021. That is how real GDP can run near a 2.5 percent pace while hiring stalls out. Real value added rose 2.5 percent among private services-producing industries and 2.3 percent among private goods-producing industries, while government was essentially flat; real estate and rental and leasing, information, durable-goods manufacturing and finance and insurance led the increase, while transportation and warehousing, retail trade and nondurable-goods manufacturing were the principal offsets.
Real consumer spending rose at a 3.8 percent annual rate in the second quarter, revised up from 3.4 percent, after a 0.7 percent pace in the first quarter, and August extended the pattern: current-dollar personal consumption expenditures rose $190.8 billion, or 0.9 percent, and real PCE rose 0.6 percent, the largest monthly gain since March 2025. Income did not keep up. Personal income rose 0.2 percent, disposable income rose 0.3 percent, and real disposable income was unchanged. The saving rate fell to 4.1 percent, from a revised 4.6 percent in July rather than the 3.0 percent previously published. That is a yellow flag, not a red one. The upper end of the K-shaped consumer still has equity gains, interest income and accumulated wealth to draw on; lower- and middle-income households have less room. We expect real spending to cool toward income growth over the next couple of months unless equity gains continue to fill the gap. A second soft month for real disposable income would turn the yellow flag red.
The price revisions moved the other way. First-quarter PCE inflation was revised to 4.2 percent from 4.6 percent, and core PCE to 3.9 percent from 4.4 percent. Second-quarter headline PCE inflation was revised to 5.0 percent from 5.3 percent, and core to 3.3 percent from 3.6 percent. August headline PCE rose 0.3 percent and 3.4 percent from a year earlier, down from 3.7 percent in July. Core PCE rose 0.2 percent and held at 3.0 percent from a year earlier, where it has sat since June. Some of that improvement is measurement: BEA changed its methods for several PCE components, including portfolio-management services, legal services and computer software and accessories, and the annual update folds in more complete source data. The broader message still holds. Core inflation shows modest improvement. Headline inflation is still being pushed around by energy and by events well outside the Federal Reserve’s reach. Policy can restrain domestic demand. It cannot reopen the Strait of Hormuz, restore lost refining capacity or bring down tanker insurance premiums.
Behind the Numbers
The maritime story has shifted: the question is no longer whether ships can get through, but at what cost. On October 3 the Iraqi Oil Tankers Company completed the transit of a chartered very large crude carrier carrying two million barrels past Hormuz, the first independent crude movement of that kind by the state shipper in decades, and Middle East crude exports ran above prewar levels on four of the last seven days of September. The next day Iran’s parliament speaker said the strait will not reopen until seven conditions in the June Islamabad memorandum are met, and OPEC+ held its November production targets unchanged, deferring any further adjustment into next year. A lane does not have to close to produce an economic shock; war-risk premiums, elevated tanker rates and rerouting are a tax on trade whether or not a barrel actually gets through. Venezuela’s crude exports fell almost 9 percent in September, to 1.08 million barrels a day, as traders pressed PDVSA for steeper discounts to cover freight, and an Aframax from Jose to the U.S. Gulf Coast was recently quoted near $3.5 million, about $5 a barrel, against roughly $1.35 million at the start of the year. Crude is responding to the easier flows: Brent eased to $101.82 and WTI to $89.61 in Monday trading, both down from Friday’s close, as better Middle East loading data and the prospect of the G7 reserve release outweighed OPEC+’s decision to hold supply flat. Diesel has not followed crude down, and that gap is this week’s more important story.
Diesel is the fuel that is actually biting, and the reason is refining chemistry, not crude supply. A barrel of crude is not a barrel of diesel: a refinery yields roughly 11 to 13 gallons of distillate per 42-gallon barrel, a ratio set by equipment that takes years and billions of dollars to change, and U.S. refineries are already running near 96 percent of capacity, with capacity itself down roughly 250,000 barrels a day from 2025. On the supply side, Russia has restricted product exports to protect its own refineries from Ukrainian strikes, Middle Eastern distillate exports are running thin on the same conflict taxing crude shipping, and China has suspended most fuel exports to rebuild its own stockpiles; U.S. refiners have answered by exporting record diesel volumes to capture the higher global price, which keeps the domestic cushion thin even as crude eases. Distillate inventories are below their five-year average and are on track to stay under 100 million barrels through 2027, the first time in more than two decades, which is why an ordinary refinery outage or a cold snap now reads as a price spike. The G7’s planned release of up to 100 million barrels of crude and diesel over the next four months works on the margin of that cushion. It does not rebuild it.
Two elections now sit on our calendar, and both bear on the policy backdrop we are forecasting into. Brazil’s October 4 first round sent Flávio Bolsonaro to an October 25 runoff against President Lula, 47.0 percent to 45.2 percent, on turnout the lowest since 1998; the Ibovespa rose 2.6 percent Monday and the real held steady near 5.16 as markets absorbed the result. Closer to home, the November 3 midterms are four weeks away. Our own framework reads a midterm as a referendum on the incumbent party’s economy, run through real after-tax income growth rather than the headline numbers, and this cycle’s income growth has been running on the soft side of that test, which is consistent with the signal in both the polling and the money. The RealClearPolitics generic-ballot average had Democrats ahead by 8.3 points as of October 1, and prediction markets price Democratic control of the House near 93 percent. A result anywhere close to that shifts the fiscal and regulatory debate in Washington before the new Congress is even seated.
In our own neighborhood, residents are arriving faster than income. In the 16 states we define as the South, real GDP rose at a 2.1 percent annual rate in the second quarter, against 2.3 percent elsewhere, and income per person rose 3.6 percent against 4.4 percent in the rest of the country as population grew faster than paychecks. Southern markets should keep generating volume, with less room to raise prices, rents and fees than the headline growth implies.
The binding constraint on the buildout has shifted from capital to power, water and local consent, which is the local version of the same story the revised national accounts just told. Governor Abbott directed Texas to halt data-center permits while agencies complete a power and water audit that ERCOT says will run at least into December. Prince William County ended by-right data-center development and cut its overlay from about 9,700 acres to 3,500. ERCOT has fielded some 474 gigawatts of connection requests, close to 90 percent of them from data centers. Applied Digital’s $3.2 billion Tuscaloosa campus works out to $32 million per permanent job. A data center is a tax-base strategy, not an employment strategy, and that is the regional expression of the capital-heavy, employment-light expansion the revised GDP accounts just confirmed nationally. The forces we have been tracking showed up. Hiring did not.
Our call runs ahead of a Street that just turned more dovish
The September 30 inflation revision and the soft September jobs report moved the policy debate, not just the headline numbers. Expectations for an October rate hike have been scaled back since the jobs report, as well as more cautious remarks from New York Fed President Williams. Some forecasters have not only dropped October from their forecast but see the FOMC holding steady in December as well. We have some sympathy for that view, given that a few IPOs have been put on hold, which is a sign of tightening financial conditions. The Fed clearly has more work to do, however, and it is too early to take a December hike off the table. Vice Chair Jefferson split the difference: he backed the September hike as an anchor for expectations, said future moves should follow the data and the balance of risks, and flagged that energy prices and AI-related building costs could still spill into broader inflation. We are the hawkish outlier, and the week’s data back it up. An economy just revised to be spending more on data centers, power and defense than any estimate showed a month ago has less slack than a reaction function built on the old numbers would assume. Every one of these calls, ours included, is ceteris paribus. Something keeps changing: since September 16 it has moved toward more inflationary pressure in energy and refined products.
The 10-year’s own history makes the point better than any single comparison: at 5.24 percent Thursday, and above 5.30 percent Monday, the 10-year is at its highest level since 2001, a 25-year high, not a one-quarter overshoot. Our current table puts the 10-year at 5.10 percent by year-end; one major Street rates desk, positioned for duration on French fiscal stress and a softening U.S. labor market, has it at 4.76 percent. The market is already running above both.
Oil shows a version of the same gap: published year-end Brent forecasts cluster in the mid-$80s to high-$80s against a $102.25 spot. Our current table looks for Brent crude to gradually fall back toward $86 a barrel by the end of next year.
Bottom Line
The revised national accounts strengthen the central theme: the economy is growing faster than the labor market suggests, investment, particularly the AI-related buildout in structures and equipment, is doing more of the lifting, and consumers are still spending readily even though income growth is no longer keeping pace. The principal threat to that domestic picture comes from outside the United States. The sea lanes are mostly functioning, but the hazards have multiplied and the cost of avoiding them is still rising. The Atlanta Fed’s GDPNow tracking estimate for the third quarter is 3.7 percent as of October 1, down from 5.0 percent on September 25, with a larger net-export subtraction doing most of the work. Our own forecast pegs Q3 growth about a percentage point slower at 2.7%, which would still mark a strong quarter.
We still look for at least two more quarter-point increases in the funds rate, one before year-end and at least one in 2027, which would put the rate at 4.25 to 4.50 percent or higher by the end of next year. The September employment report makes the case for waiting until December, not for stopping at one. We expect real growth to hold near a 2.5 percent pace into year-end if private domestic demand does not give back the second-quarter gain.
Scorecard tracks our own live, falsifiable calls, scored against the data every issue, not just when they are going our way.
Risks to our view
A reacceleration in core inflation. A hotter core PCE print, or a further squeeze in diesel and other refined products, would put the October 28 FOMC meeting back in play on the data alone, whatever the Committee’s appetite for acting on it.
A formal halt to Hormuz traffic. The lanes are open and more expensive, not closed. An actual halt to the traffic now moving would break our growth call, not just our rate call.
A sharper AI capital-spending slowdown. A pullback in hyperscaler spending would show up directly in the GDP accounts that just got revised higher on its account, and in the regional construction and power-demand pipelines we track across the South.
A second soft month for real disposable income. Income growth, not spending, is the weak link in the consumer; a second month like August would turn our yellow flag red and pull forward the slowdown we do not yet see in the data.
CFO and Treasurers Corner
Piedmont Perspective
The Piedmont Perspective: The Buildout Shows Up in the Books
The annual accounts revision confirmed, in official data, the capital-heavy, employment-light pattern we have argued since the start of the year.
The annual revision moved the official story toward ours. A single data point rarely does that. For most of this year we have described an economy that is more capital intensive than it used to be, more protein and less carbs, with investment in AI-related equipment, data centers and power doing a disproportionate share of the work while payrolls barely move. The revised national accounts now say that was understated, not overstated. Second-quarter real GDP is 2.2 percent, not 1.5 percent, and roughly two-thirds of the upward move landed in categories tied directly to the buildout.
Construction of buildings moved first in the data; the servers were always going to follow. Census put private data-center construction at an $85.0 billion annual rate in August, up 73.2 percent from a year earlier, and data centers now account for about 63 percent of all private office construction nationally, a category that barely existed as a line item a decade ago. Power construction rose alongside it, up 9.7 percent to $165.9 billion. Manufacturing construction, the prior cycle’s capital story, fell 19.8 percent over the same period, with computer and electrical plant down 45 percent, as investment shifted from building the factories that make equipment to building the places that run it.
Wall Street’s own capital-spending scenarios have not caught up to what has been booked. Hyperscaler capital spending for 2027 carries a consensus estimate near $920 billion, with scenarios built on cash flow and investment-grade borrowing capacity running as high as $1.4 trillion. Backlogs at the largest cloud providers more than doubled in a single quarter earlier this year, and analysts generally do not expect supply and demand in AI infrastructure to balance before the second half of 2027 at the earliest. Enterprise adoption, by contrast, is still early: the Census Bureau’s own survey work puts firm adoption of AI tools at under 20 percent currently, rising toward 23 percent within six months, mostly as a supplement to a narrow set of tasks rather than a wholesale change in how a business runs. The spending is real and arriving faster than the official estimates had captured. The payoff, in measured productivity and earnings, is still mostly ahead of us.
The revision has a regional reading as well as a national one. The same week the national accounts moved in our direction, Texas halted new data-center permits pending a power and water audit, which is the regional expression of the same capital-heavy, employment-light expansion the revised GDP accounts just confirmed nationally. See Behind the Numbers for the Texas, Virginia and ERCOT detail.
A slowdown in AI capital spending would not stay contained to technology stocks. An AI capital-spending slowdown, should one arrive, would not just dent a few technology stocks; it would show up directly in the GDP accounts that just got revised higher on its account, in the regional construction pipelines we track across the South, and in the power-demand forecasts utilities are already building their own capital plans around. We do not see that slowdown in this week’s data. We do see an expansion that is asking a shrinking pool of workers to do more with a rapidly growing pool of capital, which is the same story the labor market has been telling us all year from the other direction. The national accounts finally caught up to it.
The Week Ahead
Monday: ISM Services index for September.
Tuesday: August trade balance.
Wednesday: Minutes from the September 15-16 FOMC meeting, the first look at how broadly the Committee’s hawkish minority was shared; August consumer credit.
Thursday: Weekly initial jobless claims; August wholesale trade. Earnings from Delta Air Lines, the first major read on travel demand into year-end.
Friday: Preliminary University of Michigan consumer sentiment for October. Applied Digital, PepsiCo, Constellation Brands and Levi Strauss also report this week; Applied Digital’s results are the one to watch against this issue’s Tuscaloosa data-center discussion above.
What would move us: FOMC minutes showing more than a handful of participants already arguing for an October move would pull our second hike forward from December. A services ISM or trade report that reinforces the net-export drag behind this week’s drop in GDPNow would argue for trimming our third-quarter growth estimate rather than defending the 2.7 percent pace we are currently carrying.
U.S. Economic & Financial Outlook
The annual national accounts revision changed actual history more than it changed any single forecaster’s forward path. Real GDP growth is now 3.0 percent for 2024 and 2.3 percent for 2025, both revised up, with 2023 unchanged at 2.9 percent. Our full multi-year outlook, updated as of October 5, is below.
Sources and Notes
Sources. U.S. Bureau of Economic Analysis (annual national income and product accounts revision and third estimate of second-quarter GDP, September 30, 2026; second-quarter state GDP and personal income by state, September 30, 2026); U.S. Bureau of Labor Statistics (Employment Situation, September 2026, released October 2, 2026); U.S. Census Bureau (value of private construction put in place, including data-center and power categories, through August 2026); Federal Reserve Board (H.15 Selected Interest Rates, October 1, 2026); Federal Reserve Bank of Atlanta (GDPNow, October 1, 2026); Oxford Economics (U.S. weekly research, October 2, 2026); Office of the Texas Governor (data-center permit directive); Prince William County Board of Supervisors (data-center overlay rezoning); ERCOT (large-load interconnection queue); Applied Digital Corporation (Tuscaloosa, Alabama project disclosure); U.S. Energy Information Administration and refining-industry reporting on U.S. refinery utilization, capacity and distillate inventories, October 2026.
Geopolitical and political. Reuters and Al Jazeera (Iraqi Oil Tankers Company Hormuz transit, October 3, and Iranian parliament statement on strait conditions, October 4); OPEC+ (November production decision, October 4); shipping and tanker-market reporting on Middle East loading rates, Venezuelan crude exports and Jose-to-U.S. Gulf Coast freight rates, late September and early October; Brazil’s electoral authority and wire reporting on the October 4 first-round vote; RealClearPolitics (generic ballot average, October 1) and prediction-market pricing on 2026 House control.
Market data as of Friday, October 2, 2026 close, unless noted: Federal Reserve Board H.15 (Treasury yields, Thursday, October 1); major U.S. equity indices; ICE Brent and NYMEX WTI futures; COMEX gold. Monday, October 5 Brent, WTI and Brazilian market levels are intraday quotes, labeled as such where cited.
Mark P. Vitner
Chief Economist, Piedmont Crescent Capital
mark.vitner@piedmontcrescentcapital.com
A View from the Piedmont is published weekly by Piedmont Crescent Capital for informational purposes only and does not constitute investment, legal, or tax advice. Views expressed are those of the author as of the date of publication and are subject to change. © 2026 Piedmont Crescent Capital, LLC. All rights reserved.