Hidden Gems: Freddie Mac's Refinance Activity Reports

Of the myriad of public reports about the housing sector and mortgage industry, the quarterly statistics from Freddie Mac on refinance activity offer unique insights not only into the level of refinance lending but what that activity tells us about the housing sector.

And, if you know how to read them, the reports can offer strategic clues for the savvy lender.
First, the basics…
The McLean, Virginia-based GSE compiles data on loans it purchases which refinance loans within its portfolio. The data are used to produce two reports: the “Cash-out Refinance Report” and the “Refinance Product Transition Report.” Both are released quarterly.
The reports have a long and successful track record. The Cash-out Refinance Report dates back (annually) to 1986 and since Q1 1997 has been produced quarterly. The annual and quarterly cash-out volume series are covered from 1993 Q1 to the present. The Refinance Product Transition Report annual statistics are published for the period 1990 to the present. Quarterly statistics are presented for the past 20 quarters, e.g., 2006 Q4 through 2011 Q4.
The reports are published about 30 days after quarter’s end, making them quite timely and thus even more useful for strategic planning. While previously released data is revised, the revisions are usually small. (The revisions are due to Freddie Mac’s purchases of refinance loans originated in prior quarters. The additional data can cause the reported median and average statistics values to change.)
The reports provide key information estimating the dollar volume of equity extracted through refinancing which, among other things, affect Freddie’s own forecast of total prime conventional mortgage originations or refinance share changes.
The Cash-out Refinance report notes the percent of Freddie Mac-owned loans that:
were refinanced and resulted in new mortgages at least five percent higher in amount than the original mortgages,
the share that resulted in lower loan amounts,
the median ratio of the new loan interest rate to the old interest rate for fixed-rate mortgages,
the median age of the refinanced loan, and
the median amount of appreciation on the property since the previous loan was originated.
Those bits of data do more than describe the housing sector.
Loans refinanced into larger loans, for example, speak to a need to supplement lagging incomes – a consequence of the sharp reduction in jobs and hours worked in the Great Recession. As interest rates dropped, homeowners though could have drawn equity from their homes and maintained their previous payment.
Loans refinanced into smaller loans, to be sure allowed homeowners to reduce their monthly payments into more affordable payments, freeing cash for consumption. Personal consumption remains about 70 percent of the nation’s Gross Domestic Product (GDP).
Matching new and old interest rates of course offers insight into the impact of interest rate changes on the market.
Tracking the age of loans provides critical information in the construction of prepayment models.
Of course the median amount of appreciation provides further evidence of the trend in home values and prices. Freddie releases annual cash-out information for the whole U.S. and for the four Census regions. Quarterly data are published for the U.S. only.
Historical data from the reports show a distinct shift in refinance results. As real estate values continued to increase, refinances typically resulted in higher and higher loan balances, in spurts.
Indeed from 1985 through the 1990-91 recession, an average of 74 percent of refinance loans resulted in a loan balance at least 5 percent higher than the original loan. The increase in loan balances was offset by a drop in interest rates: the average median ratio of the new loan to the old loan, according to the reports was .89 which means the average rate of the new loan was about 11 percent lower. (Data from Freddie Mac’s product transition report for that period is not available to determine if borrowers shortened or extended terms of their loans.)
In the two years leading up to the Great Recession which began in December 2007, a whopping 85 percent of borrowers increased their outstanding loan balance through refis while accepting a slight increase in interest rates. At the same time, most borrowers increased loan terms to minimize payment shock.
By the end of 2008, just a year after the recession began, homeowners shifted dramatically to lowering their outstanding balance. From 2009, only 26 percent of refinance borrowers increased their outstanding loan balance, while 28 percent lowered their balance. The median of the change in interest rate showed rates came down about 19 percent and a significantly lower percentage of borrowers lengthened terms of their loans. The net impact was a noticeable improvement in homeowner balance sheets – and cash flow.
The quarterly report has changed over the years. When Freddie Mac first reported the interest-rate ratio, it was calculated as the original loan rate divided by the new loan rate. Starting with the 2006 Q3 report, Freddie flipped the ratio to report the new interest rate divided by the original rate, which is more intuitive: Values less than one indicate the borrower lowered his or her note rate; values greater than one indicate the rate increased.
The Cash-out Refinance Report also estimates another important macro-economic indicator: the amount of home equity borrowers are withdrawing from their home. Equity withdrawals through refinancing peaked at $83.7 billion in 2Q 2006 which, when combined with home equity loans and loans of credit produced equity withdrawals of $89.7 billion.
In the four years since the recession began, equity withdrawals through refinancing averaged $14.1 billion per quarter while the average median quarterly property appreciation was 1 percent; in the preceding four years the quarterly average withdrawal was $60.1 billion while median appreciation averaged 23 percent.
The Refinance Product Transition Report publishes the distribution of products chosen when borrowers refinanced their existing first mortgage. For each type of loan held initially, you can see what proportion kept the same product or opted for a different one. The report examines the ins and outs of six product categories:
1-year adjustable-rate mortgages (ARMs; this data includes some other types of loans for which rates reset at an equal frequency such as 3/3 and 5/5 ARMs),
hybrid ARMs (ARMs with the first rate reset period longer than subsequent rate reset periods, such as 3/1 ARMs and 5/1 ARMs),
balloon mortgages (loans that have one interest-rate adjustment, such as 5/25s),
15-year fixed-rate mortgages (FRMs; includes data on some shorter term loans),
20-25 year FRMs, and
30-year FRMs.
Trends for the transition report show a sharp move to shorter loan terms with borrowers moving from 20- and 30-year fixed rate loans to 15-year fixed rate loans, a trend that has accelerated since the beginning of 2010. From 2004 through 2009, the tendency had been to move to longer terms, from 15- and 20-year loans to 30-year loans.
Data from these reports can alert lenders as to the types of products borrowers prefer to adjust sales strategies, but at the same time can be used by macro-economists to understand consumer attitudes. In this case, it would appear consumers are increasingly reluctant to take on long term obligations, another indicator of lower levels of confidence about the economy.
The devil, of course, is in the details.