Borrowing plans do not confirm the New York Fed survey's application-rate high

By Cumulant Research | 20 July 2026

The New York Fed's nationally representative survey recorded its highest trailing application share since October 2021. But compared with February 2026, respondents' average likelihood of applying during the coming year fell somewhat for a new credit card, a higher card limit, an auto loan and mortgage refinancingmortgage refinancingReplacing an existing mortgage with a new home loan to change its rate, repayment period or balance. [New York Fed questionnaire](https://www.newyorkfed.org/medialibrary/interactives/sce/sce/downloads/data/2015-sce-credit-access-survey-questionaire.pdf). The mortgage measure rose slightly. [New York Fed release](https://www.newyorkfed.org/microeconomics/sce/credit-access.html) [New York Fed FAQ](https://www.newyorkfed.org/microeconomics/sce/sce-faq)

Those forward-looking measures do not confirm continued momentum. Most of the products highlighted in the release moved in the opposite direction from the broad trailing measure. That is a descriptive divergence, not a forecast that applications will fall and not proof that June marked a peak.

The application rate is a rear-view mirror. Borrowing plans are a shopping list. Neither tells us with certainty where the car goes next.

The headline measure counts the weighted shareweighted shareA survey percentage adjusted so the respondent sample better represents the population instead of counting every answer equally. [New York Fed FAQ](https://www.newyorkfed.org/microeconomics/sce/sce-faq) of respondents who reported at least one covered credit request. It does not count every application, add up requested dollars or measure how intensively each applicant searched for credit. Think of a station turnstile that records whether a person entered, not how many trains that person took. [New York Fed FAQ](https://www.newyorkfed.org/microeconomics/sce/sce-faq) [New York Fed chart guide](https://www.newyorkfed.org/medialibrary/interactives/sce/sce/downloads/glossary/sce_ca_chartglossary.pdf)

A high level is also different from acceleration. Acceleration would mean the pace of increase itself was rising. The New York Fed's summary establishes only that the trailing application share increased to its highest level since October 2021; the page does not print the exact June application rateapplication rateThe weighted share of respondents who reported at least one covered credit request during the previous 12 months. [New York Fed FAQ](https://www.newyorkfed.org/microeconomics/sce/sce-faq). [New York Fed release](https://www.newyorkfed.org/microeconomics/sce/credit-access.html)

The two clocks do not match. The application question covers the preceding 12 months, while the product-level likelihood questions ask for a percentage chance of applying during the next 12 months. Because the Credit Access SurveyCredit Access SurveyAn SCE module that asks people about recent credit applications, their outcomes and expected applications during the next year. [New York Fed Credit Access Survey](https://www.newyorkfed.org/microeconomics/sce/credit-access.html) is fielded every four months, consecutive backward-looking waves cover roughly eight of the same calendar months. [New York Fed FAQ](https://www.newyorkfed.org/microeconomics/sce/sce-faq) [New York Fed questionnaire](https://www.newyorkfed.org/medialibrary/interactives/sce/sce/downloads/data/2015-sce-credit-access-survey-questionaire.pdf)

The broad application measure covers seven request types: credit cards, card-limit increases, mortgages or home-based loans, auto loans, increases in an existing loan's limit, mortgage refinancing and student loans. The release's product-level forward summary names five measures. Four declines among those five therefore are not a complete counterpart to the seven-category headline. [New York Fed chart guide](https://www.newyorkfed.org/medialibrary/interactives/sce/sce/downloads/glossary/sce_ca_chartglossary.pdf) [New York Fed release](https://www.newyorkfed.org/microeconomics/sce/credit-access.html)

Magnitude matters. The New York Fed says the four likelihoods declined somewhat and the mortgage measure rose slightly, but its summary page does not print the February and June estimates for those measures. The safe result is therefore directional: most highlighted product-level plans weakened, but the published summary alone does not show a large reversal. [New York Fed release](https://www.newyorkfed.org/microeconomics/sce/credit-access.html)

The Federal Reserve's April 2026 bank survey asked about changes during the previous three months, generally corresponding to the first quarter. For credit cards, auto loans and other consumer loans, weaker-demand responses exceeded stronger-demand responses. [Federal Reserve April 2026 SLOOSSLOOSThe Federal Reserve's Senior Loan Officer Opinion Survey, which asks banks how lending standards, terms and demand have changed. [Federal Reserve SLOOS](https://www.federalreserve.gov/data/sloos/sloos-202604.htm)](https://www.federalreserve.gov/data/sloos/sloos-202604.htm)

The full response mix is less dramatic than the net figures alone. About the same was the largest answer in each category: 62.0% for credit cards, 74.5% for auto loans and 80.0% for other consumer loans. The corresponding Federal Reserve net-demand readings, calculated as stronger minus weaker, were -6.0, -9.9 and -12.8 percentage points. [Federal Reserve SLOOS table](https://www.federalreserve.gov/data/sloos/sloos-202604-table-1.htm) [Federal Reserve SLOOS definition](https://www.federalreserve.gov/data/sloos/sloos-202604.htm)

This is supporting context, not a replication of the household result. SLOOS records banks' assessments of loan demandloan demandThe amount of interest that households or businesses show in borrowing, measured in SLOOS through banks' reports of whether demand became stronger, weaker or stayed about the same. [Federal Reserve SLOOS table](https://www.federalreserve.gov/data/sloos/sloos-202604-table-1.htm) during a three-month period. The Credit Access Survey records households' reported requests over a year and their subjective probabilities for the next year. The respondents, questions and time windows are different. [Federal Reserve SLOOS](https://www.federalreserve.gov/data/sloos/sloos-202604.htm) [New York Fed chart guide](https://www.newyorkfed.org/medialibrary/interactives/sce/sce/downloads/glossary/sce_ca_chartglossary.pdf)

The Federal Reserve's G.19G.19A Federal Reserve statistical release measuring consumer credit outstanding, excluding loans secured by real estate. [Federal Reserve G.19 methodology](https://www.federalreserve.gov/releases/g19/about.htm) release shows that the seasonally adjusted annual growth rate of revolving creditrevolving creditCredit that can be borrowed again after repayment, with credit-card loans making up most of the Federal Reserve's G.19 measure. [Federal Reserve G.19 methodology](https://www.federalreserve.gov/releases/g19/about.htm) was 9.7% in March and 10.4% in April before turning to -4.7% in May. Total consumer credit was unchanged in May, while nonrevolving creditnonrevolving creditClosed-end credit repaid on a schedule, with motor-vehicle and education loans making up most of the G.19 category. [Federal Reserve G.19 methodology](https://www.federalreserve.gov/releases/g19/about.htm) grew at a 1.6% annual rate. March and April are revisedrevisedA previously published estimate that has been updated after the Federal Reserve received or incorporated additional information. [Federal Reserve G.19](https://www.federalreserve.gov/releases/g19/current/) estimates; May is preliminarypreliminaryAn initial estimate that may be changed in a later data release. [Federal Reserve G.19](https://www.federalreserve.gov/releases/g19/current/). [Federal Reserve G.19](https://www.federalreserve.gov/releases/g19/current/)

Revolving credit is mostly credit-card borrowing, while nonrevolving credit chiefly includes motor-vehicle and education loans. G.19 excludes mortgages and other loans secured by real estate. [Federal Reserve G.19 methodology](https://www.federalreserve.gov/releases/g19/about.htm)

Balances are not applications. They change when people borrow, repay or receive adjustments, and the G.19 table does not identify which of those channels explains a monthly movement. The May reversal therefore supplies context for softer card borrowing but cannot confirm what happened to application counts or what households will do next. [Federal Reserve G.19 methodology](https://www.federalreserve.gov/releases/g19/about.htm)

The overall rejection rate rose slightly in June 2026 to 16.1%, according to the New York Fed, but remained below the 23.1% reading from June 2025. The measure is calculated only among respondents who reported applying, and a person is counted as rejected if at least one covered request was fully rejected. [New York Fed release](https://www.newyorkfed.org/microeconomics/sce/credit-access.html) [New York Fed chart guide](https://www.newyorkfed.org/medialibrary/interactives/sce/sce/downloads/glossary/sce_ca_chartglossary.pdf)

The 7.0-percentage-point year-over-year decline does not isolate a cause. Approval conditions may have changed, but so may the mix of people and products entering the applicant pool. The published aggregate cannot separate those possibilities.

• Confirmed: the trailing application share increased to its highest level since October 2021. [New York Fed release](https://www.newyorkfed.org/microeconomics/sce/credit-access.html)
• Confirmed: four of five highlighted product-level year-ahead likelihoods fell somewhat, while the mortgage measure rose slightly. [New York Fed release](https://www.newyorkfed.org/microeconomics/sce/credit-access.html)
• Supported as context: banks most often reported stable demand, although weaker answers exceeded stronger answers in three consumer-loan categories. [Federal Reserve SLOOS table](https://www.federalreserve.gov/data/sloos/sloos-202604-table-1.htm)
• Not established: that application growth is accelerating, that June was a peak, or that applications will decline during the next year.
• Not tested: any financial-market reaction to the release or any change in real economic activity caused by it.

For a general reader, the message is simple: more people said they had sought at least one covered form of credit, but their plans did not become more ambitious across most of the highlighted products. For a finance professional, the interpretation is narrower: the survey's forward product signals fail to confirm the broad trailing high, while bank-demand and balance data provide mixed rather than decisive confirmation.