Collateral Analytics, a provider of comprehensive automated valuation solutions and real estate analytic products for large vendors and the financial services industry, is extending its Home Price Forecasting Models to include shorter term market-based indicators of home prices, along with the longer term economically-based drivers previously used.
Long-term home price forecasts can be explained well by fundamentals such as employment, incomes and interest rates, the company said in a news release. However, the new model leads to significant improvement in the predictive power of the shorter term forecasts as well.
The model has two features. One is that the model is estimated at the county level, offering the opportunity to take a more geographically look at the drivers of house prices than one based on metropolitan areas, states or the nation.
The second is that the model now is estimated with a set of variables that includes both long-run and short-run drivers of house prices. The long-run drivers include local area employment, wages and mortgage interests. Over the long run, increases in local employment and income and decreases in the mortgage interest rate are positively correlated with house price growth.
The short-run drivers or local market conditions include a number of measures of activity in the sales of single family properties such as the number of listings, months of remaining inventory, and others. In aggregate form, the CA market Conditions Index (MCIndex) reflects several highly correlated market conditions in a single variable.
“We have found that these shorter-term market-based indicators do a much better job of identifying turning points in real estate price cycles,” Collateral Analytics CEO and President Michael Sklarz said in the release. “In addition, these indicators did a much better job of predicting the magnitude of both the price bubble and subsequent crash which occurred in the 2004 to 2009 period, compared to more traditional models.”