12 Jul Is there a better way to measure first-time homebuyers?
In any discussion on affordable homes the focus is generally on the first-time buyer. As the first-time buyer looks to move on from renting and transition to home ownership finding the down-payment can be a huge hurdle. This is particularly the case for those living in areas where their rent takes up a large portion of their total household income. Unlike the first-time buyer, a repeat buyer already on the housing ladder typically relies on the equity in their current home to provide a down payment, allowing them to trade up. Their track record of payments on their existing mortgage benefits their credit score and improvements here also act to make them a more attractive proposition to lenders.
There is little reliable data on first-time homebuyers, despite a stated government policy focus.
This post looks at how to provide a better measure of the characteristics of this group and how the dynamics of first-time buyers have changed over the last 17 year. It also examines how sustainable first-time home-ownership is. At the present time the only data available on first-time buyers is available from only one official source and a survey. In 1992 the Federal Housing Enterprises Financial Safety and Soundness Act was passed. This Act defined a first-time buyer as anyone who had not had a mortgage for the preceding 3 years. This meant that the Universal Residential Loan Application (section VIIm) now includes a question to establish whether someone was buying for the first time. This data is used by the American Enterprise Institute (AEI) as the measure for first-time buyers. Due to the way the question is worded and the limits of the 3-year window, the number of first-time buyers is over estimated, this skews the data because it includes previous homeowners who left property ownership only to choose to return later.
The National Association of Realtors (NAR) produces an annual survey which is the second measure of first-time home ownership but also includes recent sales and purchases of homes. The main advantage of this data is that the measurement is not limited to those who need to obtain a mortgage to buy. In 2017 the NAR data indicated that 96% of first-time homeowners were reliant on a mortgage. The main shortcoming of the NAR survey is that their sample is relatively small making it questionable as to whether it is truly representative of the overall housing market.
How do we gain a more accurate measure of first-time buyers?
A key method would be to use the Consumer Credit Panel (CCP) of the Federal Reserve Bank of New York. This takes a random sample of 5% of US households and looks at credit files obtained via Equifax. The data allows us to follow the credit files of the individual members of a household over a period of time. The quarterly collection of data began in 1999 and individual mortgage liens such as the Veteran’s Administration (VA), and the Federal Housing Administration (FHA), can be identified using narrative codes. The data also allows for identification of GSE’s (government-supported enterprises) and any other privately secured bank or portfolio finance. Quarterly data showing the status of the loan, payments, and balance are shown by zip code with the buyer’s address hidden to protect privacy.
The CCP data gives a much clearer picture of who the first-time buyers are by looking at the household credit history as far back as 1999, combined with information which shows the age of the mortgage, and whether the mortgage is open or closed. All these metrics are provided in the report and makes it easy to see the age of each mortgage. The definition of a first-time buyer using this measure is that they appear on the report as an active mortgage after 1999 with no history of any previously closed mortgage. The CCP starts the analysis from 2000 and ends in 2016 focusing on the cohort of first-time buyers per year. The CCP data has the advantage of having a sample that can be considered representative of the national picture; based on mortgages within a credit file it also prevents skewing of data by removing cash purchases.
Once we have ascertained a view of the numbers of first-time homeowners, we can look at how important they are over time. By excluding refinance data, we can focus on the percentage of new mortgages taken out by first-time buyers.
In 2003 there was a spike in purchased mortgages at over 7 million even though the peak in house prices did not occur until 2007. Mortgage purchases declined from 2004 and through the financial crises, and by 2011 had reached 2.4 million purchases. This represents only a third of peak volume. Since 2011 new mortgage lending has slowly increased reaching over 3.5 million by 2015 and 2016.
Although house prices soared at the beginning of the 2000’s, the CCP data shows that the number of first-time buyers only fell slightly. In 2001 the share of purchase mortgages owned by first-time buyers was 44% and this dropped by only 4% to 40% in 2005. As part of the boom and bust housing cycle, the first-time ownership share rose to over 50% by 2010 then began a downward trend until it reached the mid-40’s in 2013. After that a rise began, by 2016 the share reached 46%.
Using the official data for the same three years, it consistently shows first-time buyers as having a higher share, thus overstating their importance in the market.
The NAR results are close to that provided by the CCP between 2001 to 2010, at which point the NAR first-time buyer share fell below CCP. By 2016 the gap between the two was 11%. The NAR data could generate a concern regarding the accessibility of credit for first-time home ownership, however this is not shown in the data produced by the CCP.
The next area of scrutiny is where first-time homeowners buy their mortgage using statistics from the FHA, VA, GSEs, and others. GSEs were the largest provider of first-time mortgages in the early 2000’s, however as the boom took hold more first-time mortgages were securitized by the “others” group. This group mainly consists of the private-label security market who began to draw the first-time buyers away from the FHA resulting in the FHA seeing a decline to below 10% market share of first-time owners in 2006. As the housing market started its downturn the private-label security market also collapsed leading the FHA to increase its first-time buyer share to as much as 49% in 2010. More recently, the GSE share has increased seeing a reduction in the FHA share to 30%.
A look at the FHA insurance program and its focus on first-time buyers.
Of the four groups the FHA has the largest share of purchase mortgages for first time home buyers. The FHA had lost market share back in the early 2000’s as the private-label security market saw an upsurge. At this stage first-time home owners accounted for 75% of FHA mortgage lending. During the financial crisis this share dropped to around 65%. Over the last few years the FHA has again increased its share, bringing it up to around 60%. The FHA share of this market has never recovered to the level it achieved in the early 2000’s, despite selling most mortgages to first time buyers across the groups.
To conclude, the analysis of the dynamics of first time home owners over the last 17 years and where they obtain mortgage funding provides interesting data if we use the new measure. The housing market has had its booms and busts over the last 17 years, yet the first-time buyer share is very similar when looking at the data in 2016 and comparing it to the early 2000’s.