|
|
|
 Home
> Products & Services
> Products > Financial
Models > Default Model
Default Model
 Model Methodology:
| |
The Beyondbond Default Model is a dynamic econometric model that applies economic theory to analyze sub-prime mortgagor's default behaviors. The model does this first by fitting US sub-prime residential loans information data from 2000 to 2007, then projecting the future default behavior with required inputs.
In order to develop the model, essential loan or pool level characteristics such as loan-to-value (LTV), age, FICO score, debt-to-income ratio (DTI), weighted-average-coupon rate (WAC), and principal factor, as well as crucial economic variables, such as housing price appreciation (HPA), prevailing market mortgage rate, and conditional default curve information over time were selected to determine the default factors.
Three main factors differentiate our model from traditional models:
-
Dynamic HPA is built-in into our LTV to reflect housing market wealth effect during housing booming era;
- By fitting via market HPI and its derivatives, our model captures the expected value of housing market and successfully time the default;
- Systematically link delinquency rate and loss severity to default rates' time spectrum while delinquency rate is to aid our default forecast.
These factors will better serve our model to develop an accurate projection of default characteristic. The model employs spline functions using selected inputs to project the default behavior.
|
Model Factors and Economic Variables:
| | WAC | Loan Size | Issue Date |
| | Seasonality | IO period | Age |
| | FICO Credit | Issue year | Teaser period |
| | Housing price appreciation | Regional Housing price index | Regional factors |
| |
Tuning Parameters: CDR, FICO, LTV, Age, WAC
|
Model Documentation:
US Sub-Prime Performance Study (Draft Copy)
Model Inputs:
- CLTV (with built-in HPI);
- FICO scores;
- Ratio from original WAC;
- Seasoned age;
- DTI, debt-to-income ratio;
- IO period, original;
- Size in $1,000;
- Occupancy, [0/1/2/3/4]:[NA/Owner/2ndHome/Investor/OTHER];
- Purpose,[0/1/2/3/4/5/6]:[NA/Purchase/CashOut/NoCash/2ndMtge/
RefiUnknownCash/OTHER];
- Lien,[0/1/2/3]:[NA/1st/2nd/3rd];
- Loan Document, [0/1/2/3]:[NA/Full/Limit/NoDoc];
- Macro Data [HPA/UMP/CPI/Interest Rates/Mortgage Rates];
- Region Info, [ENC(East North Central):IL/IN/MI/WI/OH,
ESC(East South Central):KY/TN/MS/AL,
MAL(Middle Atlantic):NJ/NY/PA,
MOUN(Mountain):MT/ID/WY/NV/UT/CO/AZ/NM,
NED(New England):ME/VT/NH/MA/CT/RI,
PAC(Pacific):AK/CA/HI/OR/WA,
SAC(South Atlantic):WV/DC/NC/VA/SC/GA/FL/DE/MD,
WNC(West North Central):ND/SD/NE/KS/MN/IA/MO,
WSC(West South Central):OK/AR/LA/TX,
OTHER(Some Islands):VI/PR/GU/AP/N/H/F/S/L/W/P.];
- Property Type, [0/1/2/3/4/5/6/7/8/9]:[NA/SFR/CONDO/COOP/
2-4units/TOWNHOUSE/PUD/5+Units/MANUFACTURED
HOUSING/OTHER].
Model Flexibility:
| |
Beyondbond has created an easy to use C function API for users who wish to use the prepayment model. Model delivery can be either subroutine in C library or Microsoft Excel. As well, the model is customized in a user-friendly way such that users can customize and fine-tune it via any web browser.
|
Customer Flexibility:
| |
Customer support is provided as part of your software license agreement without cost or limitation. Training is also available for technical and system implementation. Unlike most models, Beyondbond offers a solution that can be completely integrated into your existing system's structure. We provide a library of deliverable analytics to be incorporated into your trading or valuation system. Company specific development requests are often directly incorporated into the system. Updates to system information
are provided to all clients on a monthly basis.
 Highly
trained and responsive staff
Complete
user documentation
Help
desk available
Rapid
development response cycle
Monthly
data update
|
*For live model projection, please Log-In
or Contact
Us for authorization.
|
