Wednesday July 23, 2008

Home

Latest Headlines

Popular Stories

Bookmark Us

Reader Comments

SUBSCRIBE

SEARCH OUR SITE

New Study Puts Real Numbers On Mortgage Risk

RSS

COMMENTS(0)

LINK HERE

ADD NEWS TO YOUR WEBSITE

So how come your lender just quoted you 7.25 plus 2 points?

It's all about risk.

A lender wants a certain amount of assurance that, if he lends out 100,000 he is going to eventually get 100,000 back, plus interest. It might take 30 years to recoup all of that money although more and more people are selling houses or refinancing mortgages in a much shorter timeframe, but the idea is that the loan will eventually be repaid.

• the borrower's track record in paying back money
• the borrower's ability to continue to pay his obligations in the future
• the borrower's investment in the collateral
• and the underlying value of the collateral itself

In other words:

• credit history
• debt to income ratio
• amount of down payment
• and the appraised value of the car, jewelry, stocks, or real estate that secures the loan

It is a basic premise of lending and mortgage qualification that high risk loans have higher interest rates. A golden borrower, one with excellent credit, low debt to income ratio, and a substantial down payment is a borrower who presents little risk and thus usually gets an interest rate like those that are advertised in the paper and on-line. A borrower with a sketchy credit history (and it doesn't take much to turn excellent into sketchy when it comes to credit) a combination of high debt and insufficient income, and a low down payment is going to raise lender eyebrows and lenders are going to raise the interest rates and fees charged for the loan.

None of this should come as a surprise to anyone who has bought a house or a car. A new study, however, has identified other factors that contribute to a high interest rate and has actually quantified the impact of some of these factors on the rate.

The study "Mortgage Pricing is Based on Risk" was carried out by Richard F. DeMong and James E. Burroughs of the McIntire School of Commerce, University of Virginia on behalf of the law firm of Sirote & Permutt, P.C. of Birmingham, Alabama.

DeMong and Burroughs examined 961,344 mortgage loan applications made in 2004 trying to confirm that interest rate (measured in this case by APR) is tied to risk and which factors are considered as part of risk. They looked at FICO (credit scores), Loan to Value (LTV) and income, but they also weighed such factors as whether the home would be owner occupied, the type of lien (1st of 2nd mortgage) whether the application was full income verification or stated income mortgage loans and, the inclusion or omission of a prepayment fee in the final mortgage. Interestingly enough, they did not study the impact of debt or income to debt ratio except to the extent that it is factored into the FICO score.

The pair found that all of the risk factors they studied did have a positive correlation with APR. The strongest impact, as might be expected, was the type of lien. Even though borrowers in the study who were applying for 2nd mortgages had, overall higher FICO scores than those who were applying for 1st mortgages, the APR for second liens was found to be 253 basis points (2.5 percent) higher than for first liens due to the greater risk that full recovery might not be made on the collateral.

Where borrowers provided full income documentation to their lenders they received an APR that was, on average, 14 basis points lower than the borrowers who used the "stated income loans" option; that is did not provide any type of verification or documentation of the income they claimed to have.

Because of the proven impact of lien position and income verification, the study then focused only on 1st mortgages that had full income documentation in order to isolate the effect of FICO scores, LTV, income, prepayment fees and owner occupancy.

The study found that FICO scores had a beta of .010 which means that for each 10-point increase in the FICO score there was a corresponding decrease in borrowers' APR of 10 basis points with all other factors held constant. Since the lowest APR's went to those with FICO scores of 720 or higher, one can pretty easily calculate the effect that his own FICO score will have on his rate.

Likewise, both LTV and monthly income were significantly related to APR. The average LTV for the first liens studied was 78 percent. When LTV increased one percent the APR went up slightly less than one basis point - 0.6 basis points to be precise. This, however, was mitigated a bit by FICO scores. Lenders seemed willing make bigger loans with higher LTVs to borrowers with higher credit scores

The study estimated that APR declined 2.3 basis points for every thousand dollar increase in monthly income, but no baseline information was given.

The threat of an early payoff is considered a risk factor by many lenders. "Lenders are able, and do, offer loans with lower APRs to borrowers who are willing to accept a prepayment fee as a conditions of the loan…..because participants in the secondary market prefer loans with a more certain prepayment schedule and will pay a higher price for a package of mortgage loans with a prepayment fee as a result." Prepayment penalties for mortgage loans are illegal in the majority of states; nonetheless prepayment fees were included in 69 percent of the mortgages studied. Where such penalties were written into the loans, first mortgages had an average reduction in APR of 38 basis point and second mortgages 14 basis points. This reduction was found regardless of the borrower's FICO scores.

Not surprisingly, the owner-occupied status of a home had significant impact – first mortgage loans to borrowers for vacation homes or business investment had an APR 62 basis points higher than those which were to be owner occupied.

While the summary conclusion of the study "The greater the risk to the lender, the higher the APR charged for making the loan" seemed pretty obvious already, it is interesting to have some numbers attached to those risk factors.