How the forecasts held up.

Each projection is compared against the outturn for the same metric, geography, target period and measurement basis. A forecast published after its target period began is not counted, and a period with no outturn yet is simply absent.

10 scored forecasts2 resolved periods93 awaiting an outturnCompared against the latest outturn

Mean absolute error

3 institutions / %
Institution × forecast horizonDarker is a larger miss
Mean absolute error by institution and forecast horizon0-3 months3-6 months6-12 monthsFed Federal Reserve · 3-6 months · mean absolute error 0.20 over 4 forecasts 0.20 Federal Reserve · 6-12 months · mean absolute error 0.16 over 3 forecasts 0.16Goldman Sachs Goldman Sachs · 0-3 months · mean absolute error 0.39 over 1 forecast 0.39 Goldman Sachs · 6-12 months · mean absolute error 0.29 over 1 forecast 0.29Morgan Stanley Morgan Stanley · 3-6 months · mean absolute error 0.99 over 1 forecast 0.99

Every scored forecast

Largest miss first
InstitutionMetricPeriodForecastActualErrorHorizon
Morgan StanleyReal GDP growthQ4/Q42025 Q41.0%1.99%−0.99%3-6 months
FedReal GDP growthQ4/Q42025 Q41.4%1.99%−0.59%3-6 months
Goldman SachsReal GDP growthAnnual growth2025 full year2.5%2.11%+0.39%0-3 months
FedReal GDP growthQ4/Q42025 Q41.7%1.99%−0.29%6-12 months
Goldman SachsReal GDP growthQ4/Q42025 Q41.7%1.99%−0.29%6-12 months
FedHeadline PCE inflationQ4/Q42025 Q43.0%2.80%+0.20%3-6 months
FedHeadline PCE inflationQ4/Q42025 Q42.7%2.80%−0.10%6-12 months
FedUnemployment rateQ4 average2025 Q44.4%4.50%−0.10%6-12 months
FedUnemployment rateQ4 average2025 Q44.5%4.50%0.00%3-6 months
FedUnemployment rateQuarterly average2025 Q44.54.50%0.00%3-6 months

Error is the forecast minus the outturn. Actuals link to the underlying FRED series. Statistical agencies revise their own figures, so an outturn is dated from when this index first recorded it — for a period that resolved before ingestion began, the earliest recorded value and the latest are the same number.