CSW has created Home Price Forecasts to build upon its expertise in accurately tracking historical home prices. The HPFs blend CSW's home price trend and econometric analyses with the company's home price forecasting technology for residential real estate markets throughout the U.S. HPFs reflect an objective outlook for the coming year. The HPFs featured on this site use the same techniques as CSW's home price forecasts that are published by The Wall Street Journal. Click on any of the U.S. cities below for a sample Home Price Forecast.
Free Online Access (Selected HPFs)
Home Price Forecast Elements
Residential real estate market prices can be forecasted with greater reliability than prices in many other markets (for example, the stock market) because of significant market inefficiencies caused by the high transactions costs of buying and selling houses, the fact that most homes are purchased and sold for personal rather than investment reasons, and because of the high costs and inconveniences of owning homes as investments. Because of these factors, housing market information tends to work its way into current prices slowly. As this information gradually works its way into the housing market, prices will gradually move in a certain direction thereby making them predictable. Also, many markets exhibit seasonality that CSW can anticipate and accurately reflect within its forecasting models. The most challenging aspect of producing HPFs is CSW's integration of various leading indicators of future home price changes.
Home Price Forecast Performance
CSW has been publishing a sample of its Home Price Forecasts (for 23 metropolitan areas selected by The Wall Street Journal) more than five years. The majority of the time, these CSW Home Price Forecasts have been within two percentage points of the actual market change that unfolds for the forecasted period. This bar chart reflects performance scorecards for all CSW Home Price Forecasts published through March 2003.
In this scatter plot, the actual percent change is plotted against the forecast percent change for each forecast. The R2 statistic for a regression line fitted though all of the points equals 0.57, indicating that CSW's forecasting model predicts nearly 60% of the variation in actual one-year price changes. The dense clustering of points around the regression line shows that CSW's model produces consistently precise forecasts for all metro areas. (If the forecast percent changes exactly matched the actual percent changes, then all of the points on the chart would lie on a 45-degree line crossing though the origin and the R2 statistic would be equal to one.)
