The Delinquency of Subprime Mortgages Michelle A. Danis U.S - - PowerPoint PPT Presentation

The Delinquency of Subprime Mortgages

Michelle A. Danis U.S Securities and Exchange Commission & Anthony Pennington-Cross Marquette University The views expressed in this research are those of the authors and do not represent policies or positions of the U.S Securities and Exchange Commission, other officers, agencies, or instrumentalities of the United States Government, as well as any other author affiliations.

2

Subprime Literature

• Where and when is subprime used?

– High-risk locations

• Low-income, minority census tracts
• High foreclosure rate census tracts
• High proportion of individuals with “very low

credit score”

• Lower house price appreciation areas

– Less knowledgeable and high transition rates

• Performance

– Higher default for third party originations – Less responsive to interest rates – Higher loss severity – Hybrids terminate early

3

Option Theory Literature

• Traditional mortgage model

– Default, prepay, current (continue or alive) – Ruthless default

• Value of delay -- option pricing

framework

– Kau and Keenan, REE 1994 – Ambrose, Buttimer, and Capone, JMCB1997

• Expected and realized value of property matters
• Delay of foreclosure assumed exogenous
• Longer delay and higher LTV -- leads to default

– Short delay and more variance -- leads to default – Long delay and very high variance ↔ Prob(default) ↓

4

30-years of Delinquency Research

• Examine one or two states of delinquency

in isolation

– Assume Independence – Ignore competing risks

• Termination (default or prepayment)

– Remarkable consistent results

• LTV at origination
• Income
• Other financial assets
• Contemporaneous economic conditions

– Housing and labor market

• Credit scores
• Lender behavior/characteristics

5

Nested Logit Model

terminate delinquent prepay current 30 late 60 late 90 late default Loan Performance “branches” indexed by l

• utcomes in

“nests” indexed by j current terminate delinquent prepay current 30 late 60 late 90 late default Loan Performance “branches” indexed by l

• utcomes in

“nests” indexed by j current

6

90+ Days Delinquent

90 Late Frequency

0.5 1 1.5 2 2.5 3 3.5 4 1 Q 1 9 9 9 2 Q 1 9 9 9 3 Q 1 9 9 9 4 Q 1 9 9 9 1 Q 2 2 Q 2 3 Q 2 4 Q 2 1 Q 2 1 2 Q 2 1 3 Q 2 1 4 Q 2 1 1 Q 2 2 2 Q 2 2 3 Q 2 2 4 Q 2 2 1 Q 2 3 2 Q 2 3 Quarter Frequency (percentage) Sample MBA

7

90+ Days Delinquent and Age of Loan

0.0% 0.1% 0.2% 0.3% 0.4% 0.5% 0.6% 0.7% 0.8% 0.9% 1.0% 10 20 30 40 50 60 Age of Loan in Months Probability

• - -- -- -- Nested Logit

___________ Multinomial Logit

8

90+ Days Delinquent and FICO

0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 430 480 530 580 630 680 730 780 830 FICO Probability

• - -- -- -- Nested Logit

___________ Multinomial Logit

9

90+ Days Delinquent and House Prices

0.0% 0.2% 0.4% 0.6% 0.8% 1.0% 1.2% 1.4% 1.6% 1.8% 2.0%

• 3%

7% 17% 27% 37% 47% 57% 67% 77% Change in House Price Index Probability

• - -- -- -- Nested Logit

___________ Multinomial Logit

10

90+ Days Delinquent & House Price Volatility

0% 1% 2% 3% 4% 5% 6% 7% 8% 0.015 0.035 0.055 0.075 0.095 0.115 0.135 0.155 0.175 0.195 Standard Error of HPI Probability

• - -- -- -- Nested Logit

___________ Multinomial Logit

11

Current 30-Late 90+-Late Default Prepaid Prepay Penalty 0.7% 7.0% 2.5% 2.6%

• 27.6%

Low Doc

• 1.9%

87.6% 87.9% 19.8% 1.7% No Doc

• 5.8%

270.7% 336.8% 24.3% 5.8%

Elasticity Estimates The Recipe for Delinquency

Young loans Moderately volatile and flat or declining house prices Low credit scores Low & no documentation