Example: CRE loan delinquency (2-quarters ahead)

FDIC CRE delinquency from its own two-quarter lag and the Chicago Fed NFCI

This model estimates the delinquency rate on commercial real-estate loans (excluding farmland) two quarters ahead from the last observed CRE delinquency print and the Chicago Fed National Financial Conditions Index, both dated six months earlier. CRE delinquency is highly persistent; NFCI is the public financial-conditions series used here. Office vacancy and remote-work indexes are not on FRED. Coefficients are fit through December 2023 and then frozen.

• CREDEL_LAG: CRE loan delinquency rate for the quarter six months before the quarter being forecasted
• NFCI: Chicago Fed National Financial Conditions Index for that same month

Our regression model is

CREDEL_estimated(t) = CREDEL_LAG * 0.852903160 + NFCI * 1.035595582 + 0.719760904
and yields R^2 = 97.5% as of October 2023 (n = 80 quarters) in-sample on paired actuals through 2023Q4.
On later paired actuals the same frozen coefficients yield R^2 = 47.7% as of April 2026 (n = 10 quarters) out of sample.

Coefficients are ordinary least squares on complete lagged pairs from 2004 through 2023Q4. Later dates, including the period shown as a forecast, use those frozen coefficients. On the leftover paired actuals that frozen fit beats predicting the holdout mean. The first R² above is in-sample through 2023Q4; the second is out-of-sample on later paired actuals.