It comes as no surprise to those in the real estate industry that there is a seasonal pattern to home sales data. This concept is also easily understood by those outside of the industry who can picture families with school-age children preferring to purchase homes and move in the summer flooding the market during those months so as to avoid disruption to education. It is also easy to picture singles, young couples, or empty nesters remaining in the market regardless of the season. While there are undoubtedly other potential explanations behind the seasonal pattern in home sales, one does not need to understand all of the possible drivers of a seasonal pattern to understand that it exists.
For this reason, we adjust our data to account for these normal seasonal patterns, taking out some of the noise of the boom of activity in the summer and dearth in the winter. (See other articles on seasonality here). Normally, these adjustments are fairly standard from year to year. But as you probably know, the housing markets have been anything but normal these past few years. This is why when NAR revised seasonal adjustment factors in conjunction with the February EHS release as we do every year, the revisions may have seemed slightly larger than normal.
In the graph below, we see the percentage of annual sales that occur in each month. This is created using data that has not been seasonally adjusted. Depicting the seasonal pattern this way allows us to make year to year comparisons easily. The thick blue line shows the pattern of the average of the years from 2000 to 2011—we’ll call this the “typical year.” As you see from the cluster of lines around the blue line, most years are not typical, but they’re pretty close.
Looking at the graph, you also notice a handful of other thicker lines—these depict the years 2007 to 2011. Both 2009 and 2010 show seasonal patterns that are noticeably different from that of the typical year, and 2007 and 2008 were odd in the later part of the year, too. The revision made in February’s EHS release uses 2011 data to revise factors from 2009 to 2011 and forecast them for 2012. The seasonal adjustment factors done in February 2011 used data through the end of 2010 to revise and forecast. Because 2011 was so much closer to a typical year than years in the recent past, the forecast adjustment factors fit the data less well than they usually do, and the factors were revised by a more noticeable amount than usual.
Even in very abnormal market years, the seasonal pattern in home sales is evident. As the housing market continues to heal, we expect the pattern to resume a more typical trajectory, and expect more typical small adjustments to seasonal factors.