Empirical Investigation of Temporal Association between Architecture Billings Index and Construction Spending Using Time-Series Methods

Have housing value indicators changed during COVID? Housing value prediction based on unemployment, construction spending, and housing consumer price index

“…The forecast indicates a positive relationship between housing value and PRCS: the higher the construction spending, the higher the housing value. Therefore, some of the indicators for construction spending, such as the ABI, as discovered by Abediniangerabi et al (2017), present attractive topics for future research.…”

“…Therefore, to facilitate the observation of a potential bidirectional causality relationship between the housing value and the indicators and to exclusively observe the relationship between the housing value and each of the three indicators, three models each studying the relationship between housing value vs unemployment, housing value vs private construction spending and housing value vs HCPI were constructed rather than a single VAR model that incorporates the home value time series and all three of its potential indicators (unemployment, private construction spending and HCPI) see Appendix. This study uses the standardized procedure for using the VAR model that is used in studies such as Abediniangerabi et al (2017), which includes the following steps: Investigating for stationarity. Identifying lead-lag relationships between each pair of variables using Granger causality tests. Investigating long-term relationships between the housing value and its potential indicators, identifying the optimal lag orders and preventing spurious regression. Validating the result by comparing the VAR model forecasts with the testing data. …”

“…Secondly, upon verification of stationarity, the Granger causality test was used to verify if the potential indicators are helpful in forecasting the housing value and vice versa (Granger, 1969). The test has been used in literature to study the temporal relationship between economic variables; for example, Abediniangerabi et al (2017) used the Granger causality test to verify that the residential and institutional Architecture Billing Index (ABI) are useful indicators for their respective corresponding construction spending while the ABI for non-residential construction is not. Ashuri et al (2012) also used the Granger causality test to identify the leading indicators of the Construction Cost Index in the USA.…”

“…Therefore, to facilitate the observation of a potential bidirectional causality relationship between the housing value and the indicators and to exclusively observe the relationship between the housing value and each of the three indicators, three models each studying the relationship between housing value vs unemployment, housing value vs private construction spending and housing value vs HCPI were constructed rather than a single VAR model that incorporates the home value time series and all three of its potential indicators (unemployment, private construction spending and HCPI) see Appendix. This study uses the standardized procedure for using the VAR model that is used in studies such as Abediniangerabi et al (2017), which includes the following steps:…”

“…Secondly, upon verification of stationarity, the Granger causality test was used to verify if the potential indicators are helpful in forecasting the housing value and vice versa (Granger, 1969). The test has been used in literature to study the temporal relationship between economic variables; for example, Abediniangerabi et al (2017) used the Granger causality test to verify that the residential and institutional Architecture Billing Index (ABI) are useful…”