Shock response simulation

How does the modeled economy respond after borrowing rates rise? Each line compares one assumed starting state with its own no-shock baseline. Higher positive shortfall means more damage; a negative value means activity has moved above that baseline.

Peak modeled shortfall: 0.087–0.144 % output, across four assumed states. Each is compared with its own no-shock baseline.

Conditional on assumed states: retained 2025Q4 debt constructions and a 2023 wage-share mapping. This figure is a dated experiment, not a current US measurement or recession probability. The four lines are separate assumptions, not confidence-band edges. Starting employment uses November–December 2025 only because October is missing; it is not a complete-quarter observation.

0.1 means output is 0.1% below the same model’s no-shock output at that date. For example, baseline output of 100 and shocked output of 99.9 give 0.1%. Formula: 100 × (baseline − shocked output) / baseline. It is neither a 10% loss nor a recession probability. Zero = no difference; positive = below baseline; negative = above baseline. This shows the difference at each time, not accumulated loss.

Calendar placement: the assumed shock starts 1 January 2026, immediately after the 2025Q4 reference quarter, and the experiment ends 1 January 2029. These are illustrative dates assigned to the unchanged model path, not a forecast issued in 2025 or a claim that a shock happened in 2026. Within each model year, dates are interpolated between calendar anniversaries; the exact elapsed time is retained in the table. For nonzero shock choices, the shaded first year represents the chosen rate increase. The vertical scale is fixed across shock choices within each response.

Reading the four lines: D_A and D_B are different debt constructions; R_LOW and R_HIGH are lower and higher assumed borrowing-rate mappings. They have equal status, not statistical weights.

Estimate the response to the same standard shock from each past quarter’s starting conditions, then plot the resulting peak shortfall against that quarter’s date. That requires consistent debt, income and borrowing-rate definitions and explicit data vintages. No such calendar-quarter series is published in this chart. A daily website refresh does not create daily balance-sheet observations.

D_A and D_B are different debt constructions; R_LOW and R_HIGH are alternative assumed starting rates. No state is preferred. Rebounds remain visible in these paths; they do not cancel the separately reported accumulated-positive-shortfall measure.