Methodology
Big data. Local insights.
Slider

Housing Market Vitality Indicators™ (HMVI) gauge the impact of prevailing economic conditions on future house prices. We employ multiple econometric and machine learning approaches to calculate the HMVI using data curated from public and proprietary sources.

Input data include:

  • National economic and financial indicators such as personal consumption, sales, inventories, interest rates, and yields
  • Local economic conditions as reflected in labor market metrics, business performance indicators, and future economic outlook
  • Housing data representing housing supply and new construction

The housing market responds to economic factors after a lag. Therefore, as necessary, we use lagged values of input data.  The optimal lag structure was selected based on a literature review and pragmatic concerns.

We compute national and Metropolitan Statistical Area (MSA) level indicators. In the case of larger MSAs we use the constituent Metropolitan Divisions.  Not all data are available at the metropolitan level and when necessary we impute based on state, regional, or county-level data using share-of-population-weighted averages. Currently, we cover 402 metros. Additionally, based on the metro markets comprising a state, we also compute state-level indicators.

The HMVI are fully revised with each monthly release. For non-monthly inputs, we interpolate monthly series.  The published series starts from Feb 2011.

Using historic data spanning over seven years, HMVI has been found to Granger-cause house price changes and improve the predictive accuracy of house price change forecasts by 40%.