Data sources

Know where every important reading begins.

VisionBoard combines official economic releases with third-party market observations. These sources have different schedules, revision policies, and failure modes; “latest” does not mean every value was measured at the same instant.

Primary economic series

Core macro observations are retrieved through the Federal Reserve Bank of St. Louis FRED service. FRED republishes data from the originating statistical agencies; the source and release notes on each series page remain authoritative.

FRED series requested by current core and alternative-data workflows
SeriesReadingTypical frequency
A191RL1Q225SBEAReal GDP growthQuarterly
T10Y2Y10-year minus 2-year Treasury spreadDaily
DFFEffective federal-funds rateDaily
CPIAUCSLConsumer Price IndexMonthly
UMCSENTUniversity of Michigan consumer sentimentMonthly
INDPROIndustrial production indexMonthly
MEDCPIM159SFRBCLECleveland Fed median CPI, year-over-year changeMonthly
UNRATE / U6RATEU-3 unemployment / broader U-6 labor underutilizationMonthly
AHETPIAverage hourly earnings, private production and nonsupervisory employeesMonthly
CES0500000003Average hourly earnings input used by the legacy comparison workflowMonthly
REVOLSLRevolving consumer credit inputMonthly
DRCCLACBS / CORCCACBSCredit-card delinquency / charge-off rates at commercial banksQuarterly
MSPUSMedian sales price of houses sold in the United StatesQuarterly
MORTGAGE30US30-year fixed mortgage-rate inputWeekly
MEHOINUSA646NMedian household income inputAnnual

Additional release-calendar, liquidity, housing, credit, inflation, and labor series can appear in feature-specific views. The series identifier shown with a reading is the clearest way to identify the exact definition.

Reported comparisons and model estimates

The source-backed comparison workflow uses published series alongside derived estimates. Its metadata identifies source_comparison_v3 and fred_official_csv, with a source identifier and observation period for each input. An older field name such as “adjusted inflation” does not, by itself, describe the current measure.

How to read source-backed comparisons
ComparisonWhat the values mean
InflationHeadline CPI growth (CPIAUCSL) and Cleveland Fed median CPI. These are different inflation measures, not an official figure versus a VisionBoard measurement.
LaborU-3 unemployment and broader U-6 labor underutilization. Both are published measures; U-6 is not a multiplier estimate in this workflow.
WagesNominal earnings growth based on AHETPI, alongside the feed’s model-adjusted growth estimate. Its public metadata does not specify the adjustment formula. Do not treat the adjusted value as a separately observed wage series; check the earnings and inflation observation periods together.
CreditCredit-card delinquency and charge-off rates. They measure different outcomes; their difference is not a correction to either rate.
HousingA 30% affordability benchmark and a modeled payment share. The benchmark is not an observed national affordability rate. The source list includes home prices and mortgage rates; the result also depends on modeling assumptions.

Legacy and fallback calculations differ. The older comparison workflow uses fixed multipliers for inflation and broad unemployment, fixed housing assumptions, and revolving-credit growth with a delinquency benchmark. Where that workflow or an estimated fallback is shown, treat its outputs as model estimates—not retrieved median CPI, U-6, or independent measurements.

Read the definition, not just the gap

Compare like units and observation periods. A percentage-point difference between two published rates does not establish which one is “right,” and gaps across inflation, wages, credit, and housing are not interchangeable. Source identifiers alone do not disclose every assumption in a derived estimate.

Market and company observations

Depending on the feature and upstream availability, price, performance, company, options, and news observations may come from Yahoo Finance endpoints, Stooq, Financial Modeling Prep, Alpha Vantage, Polygon, or a stored prior observation. VisionBoard can switch to a fallback when a primary endpoint is delayed, restricted, incomplete, or unavailable.

A market quote displayed by the product is informational and may be delayed. It is not an executable quote, official exchange record, or substitute for a broker’s order-entry data. Where a feature exposes a source or fallback badge, use it together with the observation timestamp.

Four dates that matter

Observation date

The period the value describes—for example, a month or quarter.

Release date

When the source first published or revised the observation.

Retrieved or cached at

When VisionBoard last obtained and stored the value.

Page updated

When the surrounding analysis or page was materially changed.

These dates should not be treated as interchangeable. A recently rendered page can contain an older monthly or quarterly observation because that is the newest release available.

Revisions and fallback states

Economic agencies routinely revise seasonal adjustments and prior observations. VisionBoard generally reflects the source value available when the cache refreshes, so historical dashboard readings may differ from later vintages unless a feature explicitly preserves vintages.

When an upstream request fails, a feature may show a cached value, alternate provider, restricted view, or unavailable state. It should not silently present a fabricated real-time observation. If a source label and the displayed value appear inconsistent, please report it.

How to cite or verify a reading

  1. Record the VisionBoard page URL, displayed value, source label, and timestamp.
  2. For economic data, open the linked FRED series and compare the same observation period and units.
  3. Check the source notes for seasonal adjustment, units, frequency, and revisions.
  4. For market data, verify the same symbol, session, adjustment convention, and timestamp with an authoritative market source.