TL;DR: Through June, American households kept earning steadily, but their financial cushion thinned to its lowest level in two years. Income held as the anchor while spending adapted and credit cooled. The strain is concentrated at the bottom, where the households with the least room are running their credit hardest. That divergence, not the calm-looking average, is the signal risk teams should watch into Q3.
Steady income, thinner cushions
The Q2 2026 Atlas x Pave Consumer Financial Health Index, built on real-time behavioral and cashflow data across more than 10 million accounts, reads stable on the surface and more fragile underneath. Income stayed the steadiest of the three pillars we track, spending kept adapting rather than retreating, and credit use cooled in aggregate. But the margin for error narrowed, and household liquidity fell for a sixth straight quarter to its weakest reading in the data.
The pillars told three different stories. Inflation pressure scored 53.6, down 8.8 from Q1. Income stability held at 77.8, essentially flat. Liquidity and credit slipped to 40.0, the lowest we have measured. Read together, they describe a consumer earning steadily but with less room to absorb the next shock.
Inflation came back at the pump
After a first-quarter improvement, inflation pressure reversed, and gas did almost all of the work. Gas rose to 16.1% of card spend, up 2.1 points from Q1, as pump prices climbed through May before easing sharply in June. The rest of the essentials basket stayed calm, and utility bills fell as a share of spend. In practical terms, of every $1,000 a household put on its card this quarter, about $21 more went to gas than last quarter.
Consumers kept trading down to defend the rest of the budget. Walmart posted the largest grocery-share decline of any merchant in our data, while ALDI posted the largest gain. When the value leader loses share to harder discounters, price sensitivity is rising, not easing.
Income held as the anchor
Income remained the steadiest pillar and the reason Q2 still reads as stable. The share of households receiving payroll deposits sat about 6.4 points above its January 2025 baseline in April, keeping income in the healthy band for a sixth straight quarter. May brought the first notable dip in the series, worth watching but not yet a trend.
Gig work stayed a supplement rather than a primary income source, with one exception. The lowest earners were the only group gigging more than a year ago, a small but telling sign of where the pressure sits. When a paycheck stopped, the labor market still absorbed the churn: of the roughly 2.2% of users who missed a paycheck in a given month, nearly two-thirds were paid again within six months.
Liquidity hit a two-year low
The cushion behind that steady income kept thinning. The liquidity and credit pillar fell for a sixth consecutive quarter to its weakest reading in the data. Fewer households opened new credit in Q2, which on its own looks healthy. The strain shows up underneath the aggregate.
The fragile floor: where stress shows first
The clearest signal in the report is the widening gap between households with room and households without. Among the lowest cashflow segment, the share of users running above 90% credit utilization rose 17.6 points over the year. Among the highest cashflow segment, it rose just 4.4 points. Same behavior, split by cashflow, diverging in real time.
We call this lower cashflow group the fragile floor. It is a behavioral segment, not a demographic one: the households that react first to every shift in gas prices, grocery costs, or paycheck timing. Their behavior today previews what the broader middle may experience in six months if conditions deteriorate. Right now they are using credit as a bridge between paychecks and bills, and running that bridge with far less room for error. This is a timing and liquidity signal, not yet a delinquency one, but it is the place broader stress would surface first.
A signal we are watching: gambling
New this quarter, we began tracking gambling activity as an exploratory signal. It appears in roughly 6% to 10% of active users in a given month and correlates with liquidity stress: gambling months are more likely to include a cash advance and other credit use than non-gambling months. This is an association, not a cause, and the most recent months are still settling. We are including it as a candidate for future editions rather than a finding, but it sits squarely inside the fragile-floor story.
What to watch into Q3
Three questions will shape the next read. First, whether June's sharp reversal in gas prices holds through peak summer driving season, or whether pressure returns. Second, whether rising utilization at the bottom turns into payback stress. Third, whether May's payroll dip repeats or proves to be a one-month wobble. Income is still the floor under everything else, so that last one matters most.
What the Q2 read means for underwriting
For a risk team, the lesson of Q2 is that the aggregate hides the exposure. A portfolio can look stable on averaged metrics while a meaningful share of it runs closer to the edge every month, and cashflow data makes that share visible at decision time.
At the point of underwriting, the fragile-floor pattern argues for reading three things a bureau file is slow to show:
- Liquidity trajectory: whether balances and utilization are trending up or down, not just where they sit today.
- Income stability and timing: whether pay is regular, and how a missed or delayed paycheck would land against upcoming obligations.
- Return and NSF risk: whether recent account behavior points to a timing problem rather than an ability-to-pay problem.
Read this way, cashflow signals sharpen the credit box instead of simply tightening it. The same view that flags a stretched low cashflow borrower can also surface a thin-file applicant who is managing cashflow well, which is where cashflow underwriting can expand approvals rather than shrinking them. The logic holds across consumer products, cash advance, small dollar, personal loans, and cards, and across SMB products like charge cards and merchant cash advance, anywhere timing risk sits between a paycheck or a deposit and a bill.
What it means for monitoring
Underwriting is a snapshot. The Q2 data is a reminder that stress builds between snapshots. Utilization at the bottom did not spike in a single month, it built steadily over the past year while the aggregate looked calm.
That is the case for monitoring the book on cashflow signals between bureau pulls:
- Watch the early-warning behaviors: rising utilization in lower cashflow segments, slower recovery after a missed paycheck, and stacking of short-term credit.
- Segment by cashflow, not just by score, so the fragile floor shows up as its own cohort instead of being averaged away.
- Act on timing: align repayment and outreach with when income actually arrives, which can turn a likely missed payment into a made one.
The payoff is lead time. When the signal is behavioral and current, a risk team can adjust lines, pricing, and collections before stress becomes delinquency, rather than reading about it in next quarter's bureau data.
The full Q2 2026 Consumer Financial Health Index goes deeper on all three pillars, the fragile-floor divergence, and the methodology behind the index. Download the full report.
FAQ
What does the Consumer Financial Health Index measure?
It scores three dimensions of consumer behavior, inflation pressure, income stability, and liquidity and credit, on a 0 to 100 scale against two years of baseline data, where higher is healthier.
Who is in the data?
Aggregated, anonymized activity from consumers who use cashflow-based products across the Atlas and Pave network, more than 10 million accounts. It skews toward mainstream, fintech-adopting households and is best read as a leading indicator rather than a nationally representative panel.
How does it differ from CPI or bureau data?
CPI and bureau reports are lagging and periodic. This index reads real-time behavior, so shifts in spending, income, and credit show up here before they appear in monthly or quarterly prints.
