The University of Michigan number that landed last week is genuinely difficult to reconcile with how the top decile of U.S. households is actually spending. That tension is the trade.
The University of Michigan’s consumer sentiment index fell to 51 in early August 2026, down from 55.2 in July and below market expectations of 54.5, ending two consecutive months of improvement. Both major components weakened, with the current conditions index declining to 51.8 and the expectations measure falling to 50.6. The deterioration was not narrow. The decline was broad-based across political and demographic groups, with particularly sharp falls among older, lower-income and less-educated consumers, who are more exposed to rising prices.
Inflation concerns increased slightly, with year-ahead expectations rising to 4.3% from 4.2%, while the long-term outlook remained unchanged at 3.3%. Only 8% of consumers now expect their incomes to grow faster than inflation, down from 18% in December 2024. That income-versus-inflation expectation is the most structurally damaging number in the release. When fewer than one in ten households expects real income growth, forward spending intentions deteriorate even if current balance sheets look fine.
The historical context matters here too. Consumer sentiment sits 12.4% below where it was a year ago.
The index bottomed at a record low of 44.8 in May 2026. June brought a modest uptick to 49.5, and July’s jump to 55.2 had some analysts cautiously optimistic that the worst had passed. August killed that thesis cleanly.
The Conference Board Lands Tuesday
The Conference Board’s Consumer Confidence index fell 1.4 points in July to 90.8, below the 92.4 expected. The decline was entirely in the Present Situation Index, which fell 3.6 points to 114.9, its lowest since February 2021, marking a third consecutive monthly decline on softer views of both current business conditions and the labor market. The Expectations Index remained below the 80 threshold that the Conference Board has flagged in the past as a potential recession signal.
The next Conference Board release is Tuesday, August 25, at 10 a.m. ET.
Given the Michigan miss and the geopolitical backdrop that has kept gasoline near $4 a gallon, another sub-91 reading is plausible. The question is whether the Present Situation component holds or accelerates lower, which would put real pressure on discretionary positioning.
XLY vs. XLP Into Friday’s PCE
As of August 24, XLY is trading around $118 and XLP is around $87. The divergence is notable: staples outperforming discretionary on a down-sentiment day is exactly the defensive rotation the data calls for.
The two-track consumer creates a two-track trade. Sentiment at 51 with 4.3% inflation expectations argues for XLP as the safer position into Friday. But actual card spending data, which Bank of America’s Consumer Checkpoint has shown at 5% year-over-year growth in July, suggests XLY is not as broken as the survey data implies. The Conference Board itself has said growth is expected to be driven by business investment in AI, while the higher cost of living may reduce consumer spending, particularly by lower- and middle-income households.
That income bifurcation is the answer. The households that drive XLY performance are not the ones showing up at the bottom of the Michigan survey. The trade is long XLP defensively into Tuesday’s Conference Board number, then reassess based on whether the Present Situation index confirms or contradicts the spending data before Friday’s PCE. A hot PCE reading flattens both positions and shifts focus entirely to Warsh’s response at Jackson Hole.
Watch the Conference Board at 10 a.m. Tuesday. That number, not Michigan, will tell you whether the sentiment collapse is spreading into the cohort that actually moves discretionary revenues.