Federal Reserve Bank of New York

Federal Reserve Bank of New York is an alternative data provider in the Credit & Lending category^[2], listed in the open provider register. Its coverage is focused on United States^[2]. It has been operating since 1999^[1].

Household Debt and Credit report: mortgage, auto, student and card debt balances, delinquency and origination flows.^[1]

Overview

Federal Reserve Bank of New York sells to institutional buyers — hedge funds, asset managers and quant teams — looking for credit & lending signals with a track record they can backtest. The register lists its delivery channels as Bulk^[2]. Its listings sit in the Free price band^[2].

Most alt-data engagements follow the same arc: a free sample, a historical backtest, a paid pilot and — if the signal survives — an enterprise license. The sections below describe what that process looks like for this kind of data, and what separates a usable product from an expensive story.

The signal

Aggregated credit and lending data tracking consumer debt health: originations, balances, utilization and delinquency across mortgages, cards, autos and personal loans. The signal turns loan-book dynamics into an early read on the consumer credit cycle and lender fundamentals.

Cohort-level series record originations by product and channel, outstanding balances, credit utilization and delinquency status, segmented by score band and geography. Vintage views follow each origination quarter's performance curve, while roll-rate matrices capture transitions between current, late and default states. Public benchmarks such as the New York Fed's household debt series provide quarterly aggregates built from anonymized credit reports.

Delinquency inflections precede charge-offs, reserve builds and bank guidance changes by quarters, so lenders' equity reacts to cohort deterioration before management concedes it. Origination-mix shifts expose fintech growth or underwriting tightening ahead of disclosures, and aggregate consumer credit momentum informs rate-cut and consumption forecasts. Structured-finance desks use loan-level vintages to price securitizations against deal assumptions. Buyers rarely use a single alt-data source in isolation: this kind of signal is typically combined with fundamental estimates or other datasets to build a composite edge.

Collection, delivery and evaluation

Inputs combine credit-bureau partnerships covering anonymized credit-report panels, direct data-sharing agreements with banks and fintech lenders, and loan-level files from securitization trustees. Vendors anonymize records into cohorts with minimum cell sizes, then publish stock-and-flow series with vintage cuts that preserve point-in-time performance. Quarterly public series from central-bank research units serve both as calibration references and as products in their own right.

According to the register, Federal Reserve Bank of New York makes its data available via Bulk^[2]; delivery ergonomics matter, and buyers typically start with an API sample and move to bulk delivery (S3, Snowflake or Parquet) once a signal is validated. Before licensing data from Federal Reserve Bank of New York, a fund's data-sourcing team will typically check: history depth and survivorship, point-in-time correctness, coverage (the register lists United States)^[2], entity resolution to tickers or companies, and the compliance story behind collection. A practical sequence: request a free sample with a data dictionary, reconstruct a known historical window, and only then discuss licensing terms.

Panel composition shifts silently and moves aggregates without any underlying borrower change. Regulatory reporting lags and restatements delay signals, and servicing transfers can masquerade as delinquencies. Balance definitions drift as products evolve, complicating long-history comparisons.

Who uses it

Bank, card and fintech equity analysts read origination and delinquency turns early; macro teams track household leverage as a consumption and policy input. ABS and credit investors benchmark pool performance against panel cohorts.

Questions to ask Federal Reserve Bank of New York

How stable is cohort composition, and how do you flag composition breaks? What share comes from bureaus versus direct lender feeds versus trustee reports? Is history shipped point-in-time with vintage integrity? How do you handle regulatory reporting lags and revisions? What aggregation thresholds prevent re-identification of thin cohorts?

History and landscape

The 2008 financial crisis turned subprime monitoring into a standing discipline and elevated household-leverage statistics into market-moving releases. The post-2015 fintech lending boom layered high-frequency origination panels on top of quarterly bureau aggregates. Consumer-credit turns now circulate through trading floors within days of panel publication rather than waiting for bank earnings.

Within that lineage, Federal Reserve Bank of New York is one of 9 providers listed in the Credit & Lending category of the register; comparing their coverage, history depth and delivery is the fastest way to map the competitive landscape.

Compliance and legal considerations

Among the most regulated categories alongside card data: GLBA and FCRA govern bureau-derived information, GDPR applies in Europe, and fair-lending rules constrain analysis split by protected attributes. Aggregation must prevent re-identification of small or unusual cohorts.

Complementary signals

Buyers of this signal typically combine it with these adjacent categories — cross-coverage lowers single-source risk and widens the alpha surface.

Further reading