Written by Steven Hansen
The March 2013 pending home sales index was released by the National Association of Realtors (NAR) today, and our analysis suggests:
- Econintersect‘s uses this March 2013 pending home sales data to forecast the April 2013 home sales – and our forecast for March is 440,000 (see details below);
- If this 440,000 historical correlation is correct, this would only be a 9.8% gain year-over-year in April existing home sales, and the 23rd month in a row of year-over-year gains.
The NAR reported the March pending home sales index up 1.5% month-over-month and up 7.0% year-over-year. The market was expecting growth of 0.0% to 0.1% (versus the 1.5% reported). Econintersect‘s evaluation shows the index down 4.6% month-over-month and up 5% year-over-year.
Econintersect is concerned with a developing trend of a cooling home market. The graph below puts the issue into perspective – where a “less good” home sales trend started in December 2012.
Unadjusted 3 Month Rolling Average of Year-over-Year Growth for Pending Home Sales (blue line) and Existing Home Sales (red line)
From Lawrence Yun , NAR chief economist:
The market appears to be leveling off. Contract activity has been in a narrow range in recent months, not from a pause in demand but because of limited supply. Little movement is expected in near-term sales closings, but they should edge up modestly as the year progresses. Job additions and rising household wealth will continue to support housing demand.
The PHSI in the Northeast was unchanged at 82.8 in March and is 6.3 percent higher than March 2012. In the Midwest the index increased 0.3 percent to 103.8 in March and is 13.7 percent above a year ago. Pending home sales in the South rose 2.7 percent to an index of 120.0 in March and are 10.4 percent higher than March 2012. In the West the index increased 1.5 percent in March to 102.9 but is 4.3 percent below a year ago.
Total existing-home sales are projected to increase 6.5 to 7 percent over 2012 to nearly 5 million sales this year, while the national median existing-home price is forecast to rise about 7.5 percent.
The National Association of Realtors (NAR) pending home sales index offers a window into predicting existing home sales. The actual home sale might appear in the month the contract was signed (cash buyers account for 30% of home sales in March according to the NAR), or in the following two months.
Econintersect evaluates by offsetting the index one month to project unadjusted existing home sales. Using this index offset one month suggests existing home sales of 440,000 in April 2013 (including a -20,000 fudge factor) for historical error of this methodology for the month of Aprils in years past. Note the graph below does not include fudge factors.
Using Pending Home Sales to Predict Existing Homes Sales – Unadjusted Existing Home Sales (blue line) & Predictive Forecast Using Pending Home Sales Index (red line)
Using this methodology, 355,000 (including a +30,000 fudge factor) existing home unadjusted sales were forecast for March 2013 sales vs the actual reported number of 386,000 (which is subject to further revision).
Unadjusted Year-over-Year Change in Existing Home Sales Volumes
As shown on the above graphic, since mid 2011 home sales have been positively growing year-over-year. However, the strong rate of growth seen since mid-2010 appears to have moderated to a lower growth channel as shown on the graph above.
Keeping things real – home sales volumes are only 2/3rds of previous levels.
Caveats on the Use of Pending Home Sales Index
According to the NAR:
NAR’s Pending Home Sales Index (PHSI) is released during the first week of each month. It is designed to be a leading indicator of housing activity.
The index measures housing contract activity. It is based on signed real estate contracts for existing single-family homes, condos and co-ops. A signed contract is not counted as a sale until the transaction closes. Modeling for the PHSI looks at the monthly relationship between existing-home sale contracts and transaction closings over the last four years.
…… When a seller accepts a sales contract on a property, it is recorded into a Multiple Listing Service (MLS) as a “pending home sale.” The majority of pending home sales become home sale transactions, typically one to two months later.
NAR now collects pending home sales data from MLSs and large brokers. Altogether, we receive data from over 100 MLSs & 60 large brokers, giving us a large sample size covering 50% of the EHS sample. This is equal to 20 percent of all transactions.
In other words, Pending Home Sales is an extrapolation of a sample equal to 20% of the whole. Econintersect uses Pending Home Index to forecast future existing home sales.
Econintersect reset the forecasting of existing home sales using the pending home sales index coincident with November 2011 Pending home sales analysis (see here) – as the NAR in November revised the historical existing home sales data.
The Econintersect forecasting methodology is influenced by the speed at which closings occur. When they slow down in a particular period – this method overestimates. The number of cash buyers are speeding up the process (cash buyers analysis here). A quick cash home sale process could begin and end in the same month. On the other hand, contracts for short sales can sometimes take months to close. Interpreting the pending home sales data is complicated by weighing offsetting effects in the current abnormal market.
Please note that Econintersect uses unadjusted data in its analysis.
Econintersect determines the month-over-month change by subtracting the current month’s year-over-year change from the previous month’s year-over-year change. This is the best of the bad options available to determine month-over-month trends – as the preferred methodology would be to use multi-year data (but the New Normal effects and the Great Recession distort historical data).
Analysis Blog articles on Housing
[iframe src=”http://econintersect.com/authors/author.htm?author=/home/aleta/public_html/authors/s_hansen.htm” width=”600″ height=”500″ frameborder=”0″ scrolling=”no”]