The November update on housing starts is scheduled for release today (8:30 am eastern) and a decline is widely expected vs. October’s 3.6% rise. The Capital Spectator’s average econometric forecast anticipates a 2.5% decrease. Another factor that implies that starts will fall in November is the relatively high level of new housing construction in October vs. the number of building permits issued in that month. The two series tend to track one another fairly closely through time and so any deviations between the pair don’t usually persist. In recent years, whenever starts exceeded permits, starts usually declined in the following month. In October, starts rose above the level of permits for the first time since April.
Here’s how the numbers stack up, followed by brief definitions for the methodologies behind The Capital Spectator’s projections.
VAR-4: A vector autoregression model that analyzes four economic series in search of interdependent relationships through history. The forecasts are run in R using the “vars” package and 40 years of historical data for the following indicators: housing starts, new home sales, newly issued permits for residential construction, and the monthly supply of homes for sale.
ARIMA: An autoregressive integrated moving average model that analyzes the 50-year-plus historical record of housing starts in R via the “forecast” package.
ES: An exponential smoothing model that analyzes the 50-year-plus historical record of housing starts in R via the “forecast” package.