Frequently I am asked either what historical lending has been, or what total lending will be next year. Unlike the ebb and flows of the ocean tide, residential lending volume changes have been more akin to the wildest roller coaster ride in the world. And even more amazing is the rapid rate that future forecasts can change—up and down.
There are three primary sources that forecast housing markets, lending volumes and related metrics. Fannie Mae, Freddie Mac and the Mortgage Bankers Association (MBA) each release monthly forecasts. When you compare their forecasts, you quickly realize there is no plagiarism amongst the three.
To see these forecasts:
Fannie Mae — http://www.fanniemae.com/portal/research-and-analysis/emma.html Then Click the Housing Forecast link which has the month name in front.
Freddie Mac — http://www.freddiemac.com/news/finance/tab_outlook.html Click the link (preceded by the year) U.S. Economic and Housing Market Outlook.
MBA — http://www.mortgagebankers.org/ResearchAndForecasts/ForecastsAndCommentary Click the link Mortgage Finance Forecast and on the next screen click the box Related Document(s).
So what are the latest forecasts from these three oracles of housing and financial data? The following table shows the breakdown in residential purchase and refinance transactions for 2011, 2012, 2013, and from Freddie Mac and the MBA, 2014. Freddie Mac’s data are based on their December 2012 forecast while the Fannie Mae and MBA numbers are from their November forecasts (December not yet released).
While each of the three are forecasting double-digit percentage declines for 2013 in total residential lending, that does not equate to a similar decline in title industry premiums. Because of reissue credits, a title premium on a refinance transaction is less than the title premium on the purchase. On average, and this varies from state to state, refinance title premium revenues are 40 percent less than purchase premium revenues. Note that in each of the three forecasts, 2013 purchase lending volumes are greater while the reduction in total lending results from reduced refinance volumes.
To compensate for the reduced revenues from refinance title premiums, I calculate what I call Effective Lending. Effective Lending is calculated by adding purchase lending plus 60 percent of refinance lending.
The following graph illustrates total U.S. title industry statutory title premiums (these are just title premiums and do not include other costs such as escrow fees) and compares them to historical Effective Lending volumes. In addition, the dark bars are the latest forecast from Fannie Mae for 2012 and 2013.
How well does Effective Lending forecast U.S. total statutory title premiums? Extremely well. The Pearson Product Moment Correlation Coefficient is 0.968, and when squared, 0.937. Hence, by just knowing total U.S. Effective Lending from 1990 through 2011, 93.7 percent of the variability in statutory title premium revenues can be explained.
Do note that for 2012 and 2013, Effective Lending likely underestimates U.S. statutory title premiums since the number of all-cash transactions (not having a loan) are double the normal level, running at 30 percent of all U.S. homes sales versus a typical 12 to 14 percent.
These forecasts are not set in concrete and can change dramatically, up and down. Examine the following table and realize that for 2012 lending volumes, from October 2011 to November 2012, Fannie Mae has almost doubled their forecast. Do not criticize Fannie Mae for this almost 100 percent adjustment in 13 months, but rather appreciate how rapidly economic conditions can evolve.
So if you are looking for residential lending historical data and forecasts, head to these sources. And appreciate how quickly these data can change.