Example: C&I SLOOS tightening (2-quarters ahead)

SLOOS C&I tightening from unemployment and the Chicago Fed NFCI

This model estimates the SLOOS net percentage of banks tightening C&I standards for large and middle-market firms two quarters ahead from the civilian unemployment rate and the Chicago Fed National Financial Conditions Index, both dated six months earlier. Coefficients are ordinary least squares on the first 60% of complete lagged pairs from 2004 and then frozen.

• UNRATE: civilian unemployment rate for the month six months before the quarter being forecasted
• NFCI: Chicago Fed National Financial Conditions Index for that same month

Our regression model is

SLOOS_estimated(t) = UNRATE * -5.265541038 + NFCI * 23.144768995 + 41.530059139
and yields R^2 = 55.4% as of April 2017 (n = 54 quarters) in-sample on paired actuals through 2017Q2.
On later paired actuals the same frozen coefficients yield R^2 = 6.4% as of April 2026 (n = 37 quarters) out of sample.

Coefficients are ordinary least squares on the first 60% of complete lagged pairs from 2004 (through 2017Q2). The remaining 40% of paired actuals, including the period shown as a forecast, use those frozen coefficients. On that leftover window the frozen fit beats predicting the holdout mean. The first R² above is in-sample through 2017Q2; the second is out-of-sample on later paired actuals.