Tom:

In the article on Fannie’s considering a low-down-payment program, a representative of the Urban Institute is quoted as saying, “In an analysis this week, the Urban Institute compared loans with down payments of 3 percent to 5 percent with those with down payments of 5 percent to 10 percent. The loans with smaller down payments had lower default rates than the other mortgages when the credit scores of the borrowers were higher. ”

The key phrase here is, “… when the credit scores of the borrowers were higher.” Alas, the representative doesn’t tell us HOW MUCH higher. In doing so, we are left with the implication that down payment really isn’t that important. That is incorrect.

I worked for a major mortgage lender from 2004 to 2010 and was responsible for quantitative analysis of portfolio and operational performance. One of the more interesting findings we had in the aftermath of the collapse of the mortgage market was that down payment was a very important factor in default — and that it was not a linear relationship. Controlling for credit score and debt-to-income ratio, one sees relatively little effect as the down payment is reduced from 20% to 10%. However, as the payment was reduced BELOW 10%, there was a sharp increase in default.

If one thinks about it, this finding is not surprising. Consider two borrowers, both with equal credit scores and for whom the debt-to-income ratio on the loan is the same (and less than 30%). One has 10% to put down on the home; the other nothing. What is the difference between these two borrowers? Simple: one is living a lifestyle that allows some saving — i.e., he/she has a cushion. The other is living paycheck to paycheck. Hence, when adversity strikes, it is the second borrower that tends to default much more often than the first.

I wish we had recognized this when we decided to underwrite the second mortgage on some 80/20 loans in 2006 and 2007. It was the only part of our portfolio that suffered serious losses. The relatively strong credit scores of the borrowers did not protect us.

The lesson is that down payment DOES matter, though 10% is probably a better threshold than 20%. Maybe the reason the Urban Institute does not see this is that the two groups they were comparing (5%-to-10% vs 3%-to-5%) were both already under the threshold.

Friday Headlines: Fannie provides details on low down-payment plan; BofA takes forex charge

By Thomas K. Brown, November 7, 2014

New York Times: Fannie Mae Official Details Plans on Low-Down-Payment Mortgages

Bloomberg: Citigroup Said Poised to Get Bids for Japan Retail Bank

New York Times: Bank of America Takes $400 Million Legal Charge for Currency Inquiry

Chicago Tribune: Bank overdraft fees range from $5 to $45, survey says

CNBC: Weill tells Cramer: Banks aren’t the same anymore

Bloomberg: Pimco Offers Special Post-Gross Bonus to Retain Talent

Bloomberg: Fed Said to Seek More Time to Mull Actions on FX Traders

Washington Post: Why millennials prefer electronic gift cards

One Response to “Friday Headlines: Fannie provides details on low down-payment plan; BofA takes forex charge”

Randall Grossman November 8, 2014

Tom:

In the article on Fannie’s considering a low-down-payment program, a representative of the Urban Institute is quoted as saying, “In an analysis this week, the Urban Institute compared loans with down payments of 3 percent to 5 percent with those with down payments of 5 percent to 10 percent. The loans with smaller down payments had lower default rates than the other mortgages when the credit scores of the borrowers were higher. ”

The key phrase here is, “… when the credit scores of the borrowers were higher.” Alas, the representative doesn’t tell us HOW MUCH higher. In doing so, we are left with the implication that down payment really isn’t that important. That is incorrect.

I worked for a major mortgage lender from 2004 to 2010 and was responsible for quantitative analysis of portfolio and operational performance. One of the more interesting findings we had in the aftermath of the collapse of the mortgage market was that down payment was a very important factor in default — and that it was not a linear relationship. Controlling for credit score and debt-to-income ratio, one sees relatively little effect as the down payment is reduced from 20% to 10%. However, as the payment was reduced BELOW 10%, there was a sharp increase in default.

If one thinks about it, this finding is not surprising. Consider two borrowers, both with equal credit scores and for whom the debt-to-income ratio on the loan is the same (and less than 30%). One has 10% to put down on the home; the other nothing. What is the difference between these two borrowers? Simple: one is living a lifestyle that allows some saving — i.e., he/she has a cushion. The other is living paycheck to paycheck. Hence, when adversity strikes, it is the second borrower that tends to default much more often than the first.

I wish we had recognized this when we decided to underwrite the second mortgage on some 80/20 loans in 2006 and 2007. It was the only part of our portfolio that suffered serious losses. The relatively strong credit scores of the borrowers did not protect us.

The lesson is that down payment DOES matter, though 10% is probably a better threshold than 20%. Maybe the reason the Urban Institute does not see this is that the two groups they were comparing (5%-to-10% vs 3%-to-5%) were both already under the threshold.

Comments are closed.