Comments on “Vintage and Credit Rating: What Matters in the ABX Data During the Credit Crunch” Shane Sherlund Federal Reserve Board* *The views expressed herein are mine alone and do not necessarily reflect the views of the Board of Governors, its members, or its staff.
Comments on “Vintage and Credit Rating: What Matters in the ABX Data During the Credit Crunch” Shane Sherlund Federal Reserve Board* *The views expressed.
Dec 13, 2015
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Comments on “Vintage and Credit Rating: What Matters in the ABX Data During the Credit Crunch”
Shane SherlundFederal Reserve Board*
*The views expressed herein are mine alone and do not necessarily reflect the views of the Board of Governors, its members, or its staff.
Discussion
Very nice, careful paper!
Decompose movements in ABX-HE prices
0
20
40
60
80
100
120
06Q1 06Q3 07Q1 07Q3 08Q1 08Q3
AB
X p
rice
Source: Markit.
Discussion
Common factors
0
20
40
60
80
100
120
-40
-20
0
20
40
06Q1 06Q3 07Q1 07Q3 08Q1 08Q3
2006-1 AAA2007-1 AAA
2006-1 BBB-2007-1 BBB-
HPA
AB
X p
rice
sH
PA
(percent)
Source: Markit, RadarLogic.
Discussion
Ratings factors
0
20
40
60
80
100
120
06Q1 06Q3 07Q1 07Q3 08Q1 08Q3
AAA AA A BBB BBB-
2006
-1 A
BX
pri
ces
Source: Markit.
Discussion
Vintage factors
20
30
40
50
60
70
80
90
100
110
06Q1 06Q3 07Q1 07Q3 08Q1 08Q3
2006-1 2006-2 2007-1 2007-2
AA
A A
BX
pri
ces
Source: Markit.
Cumulative Default CurvesSeptember 2008
Source: Calculations from First American LoanPerformance.
Discussion
Liquidity effects
0
20
40
60
80
100
120
-100
0
100
200
300
400
06Q1 06Q3 07Q1 07Q3 08Q1 08Q3
2006-1 AAA2007-1 AAA
2006-1 BBB-2007-1 BBB-
LIBOR-OIS
AB
X p
rice
s
LIBO
R-O
IS (basis points)
Source: Markit, Federal Reserve.
Discussion
Common factor Relatively small effect early on; low correlation Onset of financial turmoil, significant and
persistent increase Systematic, non-diversifiable risk High correlation across assets
Asset rating and vintage factors Effects decrease over time Distinguish between vintages and tranches Roles in asset volatility small
Discussion
Liquidity premium Influence increased at onset of financial turmoil Along with common factor, accounts for downward
pressure on ABX prices Across ratings Across vintages
Comments
Composition of each vintage Underwriting Geography
House Price Appreciation2000-2008
Source: FHFA House Price Index.
-30
-20
-10
0
10
20
30
2000 2002 2004 2006 2008
Per
cent
(an
nual
rat
e)
CA, FL, AZ, NV
OH, MI, IN
National
Percent of Subprime Mortgages with Negative Equity by State2005-2008
Source: Calculations from First American LoanPerformance, FHFA and S&P/Case-Shiller house price data.
0
10
20
30
40
50
60
70
2005 2006 2007 2008
CA, AZ, FL, NV
OH, MI, IN
Rest of U.S.
Per
cent
of
loan
s
Serious Delinquency Rates by State2005-2008
Source: First American LoanPerformance.
0
5
10
15
20
25
30
35
2005 2006 2007 2008
CA, AZ, FL, NV
OH, MI, INRest of U.S.
Per
cent
of
loan
s
Comments
Liquidity effect Investors? Borrowers? Both?
Subprime Mortgage Rate ResetsSeptember 2003
Source: First American LoanPerformance.
0
50000
100000
150000
200000
250000
300000
350000
2002 2004 2006 2008 2010 2012
Nu
mb
er
of l
oa
ns
Subprime Mortgage Rate ResetsSeptember 2004
Source: First American LoanPerformance.
0
50000
100000
150000
200000
250000
300000
350000
2002 2004 2006 2008 2010 2012
Nu
mb
er
of l
oa
ns
Subprime Mortgage Rate ResetsSeptember 2005
Source: First American LoanPerformance.
0
50000
100000
150000
200000
250000
300000
350000
2002 2004 2006 2008 2010 2012
Nu
mb
er
of l
oa
ns
Subprime Mortgage Rate ResetsSeptember 2006
Source: First American LoanPerformance.
0
50000
100000
150000
200000
250000
300000
350000
2002 2004 2006 2008 2010 2012
Nu
mb
er
of l
oa
ns
Subprime Mortgage Rate ResetsSeptember 2007
Source: First American LoanPerformance.
0
50000
100000
150000
200000
250000
300000
350000
2002 2004 2006 2008 2010 2012
Nu
mb
er
of l
oa
ns
Subprime Mortgage Rate ResetsSeptember 2008
Source: First American LoanPerformance.
0
50000
100000
150000
200000
250000
300000
350000
2002 2004 2006 2008 2010 2012
Nu
mb
er
of l
oa
ns
Cumulative Prepayment CurvesSeptember 2008
Source: Calculations from First American LoanPerformance.
Comments
Measure of fit? R-squared (pseudo R-squared) “Variance decomposition”
Summary
Ratings and vintage factors Effects decreased over time Role in asset volatility small
Common factor Significant and persistent effect once financial
turmoil started Systematic risk; high asset correlations
Liquidity premium Influence increases over time
Related Documents
