| Title |
A Dynamic Regression Analysis of Interest Rate Sensitivity and Lag Effects in the Construction Order Market |
| Authors |
원규연(Won, Gyu-yeon) ; 임윤재(Lim, Yunjae) ; 한상원(Han, Sangwon) |
| DOI |
https://doi.org/10.5659/JAIK.2026.42.6.361 |
| Keywords |
Construction Order Forecasting; Dynamic Autoregressive Model; Cross-correlation Function (CCF); Industry Inertia; Out-of-Sample Prediction; Scenario Analysis |
| Abstract |
Construction order volumes are highly sensitive to macroeconomic shocks while also showing strong dependence on past order flows.
Previous studies have provided an important academic foundation by estimating the lagged effects of interest rates on the construction
economy, but dynamic approaches that explicitly account for endogenous inertia remain limited. This study uses annual data on total
construction orders in South Korea from 1994 to 2024 to examine how the lagged impact of interest rates changes when construction inertia
is controlled. By stepwise expansion of regression models, static lagged models are compared with a dynamic autoregressive model. The
results show that the previous year’s order volume strongly influences the current year’s market. When this inertia is incorporated, interest
rate changes that were previously interpreted as having long-term lag effects instead exert immediate downward pressure on current-year order
decisions. This finding reveals a statistical distortion caused by omitted variable bias. The robustness of the dynamic model is confirmed
through the Breusch?Godfrey test and expanding-window out-of-sample forecasting validation. Scenario simulations for the 2025 to 2027 order
market, including optimistic, base, and pessimistic cases, indicate that the path of market recovery or prolonged recession varies significantly
depending on the timing and magnitude of future interest rate adjustments. |