Loan Level Mortgage Modeling

Transcription

1 Loan Level Mortgage Modeling Modeling and Data Challenges Shirish Chinchalkar October 2015

2 Agenda 1. The complexity of loan level modeling 2. Our approach for modeling mortgages 3. Data Challenges 4. Conclusion 2

3 The complexity of loan level modeling» Modeling complexity - Many (many) scenarios are required to capture the behavior of mortgages in different states of the world - Different types of loans behave very differently - Non-monotonic behavior with respect to several factors - Interaction terms needed to explain borrower behavior - Single period analysis cannot generally be used for path-dependent instruments like mortgages» Data complexity - A lot of data fields needed to fully describe the mortgage - Mortgages do not amortize quickly, so need to deal with legacy data - Need to incorporate future loan volume with loan level detail 3

4 Loan level modeling in different economies Many scenarios are required to capture the behavior of mortgages in different states of the world loan= 40 ; econ= 882 loan= 40 ; econ= 883 loan= 40 ; econ= 884 loan= 40 ; econ= LOAN # loan= 40 ; econ= loan= 98 ; econ= loan= 40 ; econ= 887 loan= 40 ; econ= Same loan in different economies exhibits different behavior and may be correlated differently loan= 98 ; econ= 883 loan= 98 ; econ= loan= 40 ; econ= loan= 98 ; econ= loan= 98 ; econ= loan= 98 ; econ= loan= 98 ; econ= loan= 98 ; econ= 889 LOAN #

5 Different baselines for different types of loans Different types of loans behave very differently 2/28 ARM 5/25 ARM /27 ARM FRM-30y

6 Non-monotonic behavior with respect to several factors 1.4 Prepayment Updated LTV Prepayment incentive is different for borrowers in different updated LTV buckets 6

7 Interaction terms needed to explain borrower behavior Low FICO High FICO Default Rate Updated LTV Default sensitivity is different for borrowers in different FICO buckets 7

8 Using Aggregate Pool Statistics Consider two pools drawn from this population: one homogeneous and one barbelled (but both with approximately the same mean CLTV and FICO) FICO SCORE Low <70 Combined LTV Medium [70,80) High [80,85) Very High >=85 Low < Medium [710,750) High [750,775) Very High >= FICO CLTV Def. rate Homogeneous Barbell Aggregate pool statistics may mask risk behavior. 8

9 Multi-period Simulation and path dependence» Home prices start at 100 and end, 10 years later, at Scenario 1: home price appreciation of 3% per year for 10 years - Scenario 2: home price depreciation of 20% over 3 years followed by a gain over the next 7 years Pool EL (Scenario 1) EL (Scenario 2) Multi-period simulation is valuable due to strong path dependency. 9

10 Why are Mortgages Complicated to Model?» A single loan can behave very differently in different economic scenarios.» Different loan types behave very differently in the same economic scenario.» Drivers of mortgage performance, including prepayment and default, are strongly path dependent.» Mortgages have many embedded options, including - the option to prepay (call) - the option to walk away from the loan (put).» The terms of these options do not generally average out analytically. Mortgages are one of the most difficult asset classes to model. 10

11 Our Approach» Our model is an analytic tool for assessing the credit risk of a portfolio of residential mortgages (RMBS & whole loans)» The model comprises loan-level econometric models for default, prepayment, and severity.» These models are integrated through common dependence on local macroeconomic factors, which are simulated at national and local (MSA) levels» This integration produces correlation in loan behaviors across the portfolio.» Because we use a multi-step Monte Carlo approach, the model can be combined with an external cash flow waterfall tool and used for simulation of RMBS transactions.» The models also use pool-level performance to update the output in realtime. 11

12 Our Approach» Panel data monthly observations for each loan through time» Survival models Different baselines or nominal hazard rates as a function of loan age for different types of mortgages» Competing risk framework Defaults and Prepayments are two types of loan exit.» Explanatory variables are observable macro-economic variables and loan and borrower characteristics» Implementation in a multi-period framework. 12

13 Applications» Stress testing of whole loan portfolios» Surveillance and monitoring of RMBS transactions» Selecting loans for securitization based on credit risk» Whole loan pricing and trading» Risk management of loan portfolios» Regulatory reporting for CCAR and DFAST submissions» ALLL and CECL calculation 13

14 Mortgage Modeling Overview FACTORS MODELS OUTPUT» Economic Data (Simulated or Scenario)» Default» Loan Level Pool Data (User Data)» Supplemental User data (Loan level Override, Pool Performance etc.)» Severity» Prepayment Loan Level E(L) Σ Pool Level E(L) 14

15 Mortgage Modeling Framework Economic Factors» The key economic processes that are simulated in the model are: - Interest rates (Yield curve & LIBOR rates) - Home Price Change (national, state, and MSA level) - Unemployment rates (national, state, and MSA level) - Loan market rates (Freddie Mac mortgage rate) 15

16 Typical delinquency pipeline Jumbo RMBS Delinquency and Cum Loss Rates, Vintage Jumbo Loss Tracker 16% 5.9% 5.9% Delinquencies/Current Balance 14% 12% 10% 8% 6% 4% 5.3% 4.7% 4.1% 3.5% 3.0% 2.4% 1.8% 1.2% 5.3% 4.7% 4.1% 3.5% 3.0% 2.4% 1.8% 1.2% 0.8% 0.7% 4.4% Cumulative Loss/Original Balance 2% 0.6% 0.6% 0% 0.0% May-05 Sep-05 Jan-06 May-06 Sep-06 Jan-07 May-07 Sep-07 Jan-08 May-08 Sep-08 Jan-09 May-09 Sep-09 Jan-10 May-10 Sep-10 Jan-11 May-11 Sep-11 Jan-12 May-12 Sep-12 Jan-13 May-13 Sep-13 Jan-14 May-14 Sep-14 Jan-15 May-15 Sep-15 Source: Moody s Investors Service REO% FC% 90+% 60-89% 30-59% Cum Loss % 0.0% Loss To Date Pipeline Loss Remaining Loss Delinquent loan pipeline makes up a key part of future losses. 16

17 Modeling Seasoned Mortgage Pools: Delinquent loans» We categorize delinquent loans into: 30 DPD and 60 DPD buckets» Default and prepayment hazard rates differ substantially between delinquent loans and current loans.» Each delinquency status has different default and prepayment behavior.» Can determine extra risk in delinquent loans as compared to equivalent current loans. Delinquent loans behave very differently from current loans. 17

18 Modeling Seasoned Loans: Incorporating pool-specific Realized Performance To-date» Realized performance can, on occasion, be very different than predicted due to unobservable differences in underwriting, servicing, borrower characteristics, etc.» It is important to incorporate individual components of the realized performance, namely default, prepayments, and severity, separately.» If past performance is available, it can be very valuable.» Need to back-test each individual component of the models against pool performance to calibrate the models. Pool-level idiosyncratic behavior can be useful in future projection. 18

19 A Mortgage Portfolio Loss Distribution Moody s MEDC Baseline Scenario: Loss = = 1.76% MEDC Moody s Total Protracted Collapse Slump Scenario: Scenario: Loss Loss = 13.05% = 13.05% Frequency Mean Loss = 4.15% Portfolio Loss In addition to generating the full loss distribution, we can estimate losses under Moody s or user-defined scenarios. 19

20 Tail Risk Contribution to VaR» TRC is the contribution a loan makes to the tail risk of a portfolio. EL 99.5% VaR Level 0% 5% 10% 15% 0% 5% 10% 15% Original portfolio 4.0% 12.6% With 100 highest EL loans removed With 100 highest contributors to VaR removed 2.9% 10.2% 3.1% 9.7% Tail risk of a loan is often different than its stand-alone risk. 20

21 Custom Scenarios» Custom scenarios are an integral part of stress testing» User can provide a view for one or more macro-economic variables such as interest rates, GDP, HPI, and Unemployment rate» Models can fill-out all the missing variables at the national, state, or MSA level and through time in a consistent manner» Can anchor a simulation around a custom scenario. We can then obtain a full loss distribution, Value at Risk, Tail Risk Contribution, and loan level detail. 21

22 Validation and Testing» How do we know that we have done a good job? - Calendar time plots» Segments modeled separately (for example, first liens and second liens)» Segments not part of the model (for example, low FICO, high LTV) Realized Predicted Default Rate Calendar time 22

23 Validation and Testing» Univariate plots - Model variables - Variables not included in the model Realized Predicted Default rate» Avoid over-fitting Updated LTV» Do not rely on statistical tests alone 23

24 Additional Calibration and Testing Some other things to look for» Do we need to extrapolate some functions? For example, unemployment rates outside the range of observable data.» Forecasts under different stress and baseline scenarios. How do projections over time compare with historical losses?» Sensitivities to key variables. Are the sensitivities in line with expectations?» Overrides or calibration for things not considered. Have some policies changed? 24

25 Modeling This Way Allows Us To Analyze portfolios of mortgages (and other asset classes) in a correlated fashion using an intuitive and consistent set of macro-economic variables» Generate losses for Moody s Analytics, Fed CCAR, and user defined scenarios.» Conduct scenario analysis using observable macro-economic factors.» Conduct validation using realized economies to-date.» Use the same framework to evaluate seasoned portfolios and new originations.» Determine a loss distribution and calculate VaR.» Calculate the tail risk contribution for each loan and manage the tail risk of a portfolio of mortgages.» Provide collateral loss distribution and cash flows that can be combined with a waterfall engine to produce tranche-level loss distributions. 25

26 The Role of Data Data Models Implementation Execution Data correctness and completeness is fundamental to the entire modeling effort. 26

27 Data Challenges Model Building Lack of good historical training data» Most banks do not have a long enough data history.» Post crisis data is usually of better quality, but pre-crisis data can have gaps.» Acquisitions and mergers create unique data problems.» Acquisitions and mergers can result in data of different underlying risks.» Policy decisions can result in unpredictable performance. Techniques for handling data inadequacies» Can use proxy data to derive relationships and calibrate to the bank s data.» Use segmentation to distinguish between different portfolios. For example, the portfolios of banks before and after an acquisition could be separately calibrated.» Each policy decision has to be addressed individually and carefully. 27

28 Data Challenges Model Execution Model Execution» Determine how a loan amortizes» Determine interest rate changes for ARM loans» Determine HELOC utilization» Delinquency information What information do we need» Loan terms» Underlying index, margin, caps, and floors for ARM loans» Model for HELOC utilization» May need to fill-in missing data since we may not be in a position to remove loans with bad data 28

29 Data Challenges Future Volume Future volume» Not all portfolios are in a run-off mode» Loans originated in the future will contribute to losses» Future volume depends on balance attrition as well as growth policies» Need to generate future originations with loan level detail» The loan level data should have the desired average values, distributions, and correlations How do we generate future volume with loan level detail?» Generate future volume by segments for example, first lien mortgages, second lien mortgages, fixed rate loans, ARM loans.» Examine a reference portfolio for each segment and infer distributions and correlations from it.» Allow override of any inferred distributions.» Generate loan level data by sampling from these distributions. 29

30 Conclusion» Modeling at the loan level offers significant detail in estimating losses.» Modeling each loan behavior (default, prepayment, and severity) separately provides substantial flexibility in calibration and specification.» The state of the local and national economy significantly impacts the performance of loans.» A loan level approach can be challenging from a data as well as a modeling point of view.» With good data collection and careful modeling, this approach can be used in different areas of the business. 30

31 2015 Moody s Corporation, Moody s Investors Service, Inc., Moody s Analytics, Inc. and/or their licensors and affiliates (collectively, MOODY S ). All rights reserved. CREDIT RATINGS ISSUED BY MOODY'S INVESTORS SERVICE, INC. AND ITS RATINGS AFFILIATES ( MIS ) ARE MOODY S CURRENT OPINIONS OF THE RELATIVE FUTURE CREDIT RISK OF ENTITIES, CREDIT COMMITMENTS, OR DEBT OR DEBT-LIKE SECURITIES, AND CREDIT RATINGS AND RESEARCH PUBLICATIONS PUBLISHED BY MOODY S ( MOODY S PUBLICATIONS ) MAY INCLUDE MOODY S CURRENT OPINIONS OF THE RELATIVE FUTURE CREDIT RISK OF ENTITIES, CREDIT COMMITMENTS, OR DEBT OR DEBT-LIKE SECURITIES. MOODY S DEFINES CREDIT RISK AS THE RISK THAT AN ENTITY MAY NOT MEET ITS CONTRACTUAL, FINANCIAL OBLIGATIONS AS THEY COME DUE AND ANY ESTIMATED FINANCIAL LOSS IN THE EVENT OF DEFAULT. CREDIT RATINGS DO NOT ADDRESS ANY OTHER RISK, INCLUDING BUT NOT LIMITED TO: LIQUIDITY RISK, MARKET VALUE RISK, OR PRICE VOLATILITY. CREDIT RATINGS AND MOODY S OPINIONS INCLUDED IN MOODY S PUBLICATIONS ARE NOT STATEMENTS OF CURRENT OR HISTORICAL FACT. MOODY S PUBLICATIONS MAY ALSO INCLUDE QUANTITATIVE MODEL-BASED ESTIMATES OF CREDIT RISK AND RELATED OPINIONS OR COMMENTARY PUBLISHED BY MOODY S ANALYTICS, INC. CREDIT RATINGS AND MOODY S PUBLICATIONS DO NOT CONSTITUTE OR PROVIDE INVESTMENT OR FINANCIAL ADVICE, AND CREDIT RATINGS AND MOODY S PUBLICATIONS ARE NOT AND DO NOT PROVIDE RECOMMENDATIONS TO PURCHASE, SELL, OR HOLD PARTICULAR SECURITIES. NEITHER CREDIT RATINGS NOR MOODY S PUBLICATIONS COMMENT ON THE SUITABILITY OF AN INVESTMENT FOR ANY PARTICULAR INVESTOR. MOODY S ISSUES ITS CREDIT RATINGS AND PUBLISHES MOODY S PUBLICATIONS WITH THE EXPECTATION AND UNDERSTANDING THAT EACH INVESTOR WILL, WITH DUE CARE, MAKE ITS OWN STUDY AND EVALUATION OF EACH SECURITY THAT IS UNDER CONSIDERATION FOR PURCHASE, HOLDING, OR SALE. MOODY S CREDIT RATINGS AND MOODY S PUBLICATIONS ARE NOT INTENDED FOR USE BY RETAIL INVESTORS AND IT WOULD BE RECKLESS FOR RETAIL INVESTORS TO CONSIDER MOODY S CREDIT RATINGS OR MOODY S PUBLICATIONS IN MAKING ANY INVESTMENT DECISION. IF IN DOUBT YOU SHOULD CONTACT YOUR FINANCIAL OR OTHER PROFESSIONAL ADVISER. ALL INFORMATION CONTAINED HEREIN IS PROTECTED BY LAW, INCLUDING BUT NOT LIMITED TO, COPYRIGHT LAW, AND NONE OF SUCH INFORMATION MAY BE COPIED OR OTHERWISE REPRODUCED, REPACKAGED, FURTHER TRANSMITTED, TRANSFERRED, DISSEMINATED, REDISTRIBUTED OR RESOLD, OR STORED FOR SUBSEQUENT USE FOR ANY SUCH PURPOSE, IN WHOLE OR IN PART, IN ANY FORM OR MANNER OR BY ANY MEANS WHATSOEVER, BY ANY PERSON WITHOUT MOODY S PRIOR WRITTEN CONSENT. All information contained herein is obtained by MOODY S from sources believed by it to be accurate and reliable. Because of the possibility of human or mechanical error as well as other factors, however, all information contained herein is provided AS IS without warranty of any kind. MOODY'S adopts all necessary measures so that the information it uses in assigning a credit rating is of sufficient quality and from sources MOODY'S considers to be reliable including, when appropriate, independent third-party sources. However, MOODY S is not an auditor and cannot in every instance independently verify or validate information received in the rating process or in preparing the Moody s Publications. To the extent permitted by law, MOODY S and its directors, officers, employees, agents, representatives, licensors and suppliers disclaim liability to any person or entity for any indirect, special, consequential, or incidental losses or damages whatsoever arising from or in connection with the information contained herein or the use of or inability to use any such information, even if MOODY S or any of its directors, officers, employees, agents, representatives, licensors or suppliers is advised in advance of the possibility of such losses or damages, including but not limited to: (a) any loss of present or prospective profits or (b) any loss or damage arising where the relevant financial instrument is not the subject of a particular credit rating assigned by MOODY S. To the extent permitted by law, MOODY S and its directors, officers, employees, agents, representatives, licensors and suppliers disclaim liability for any direct or compensatory losses or damages caused to any person or entity, including but not limited to by any negligence (but excluding fraud, willful misconduct or any other type of liability that, for the avoidance of doubt, by law cannot be excluded) on the part of, or any contingency within or beyond the control of, MOODY S or any of its directors, officers, employees, agents, representatives, licensors or suppliers, arising from or in connection with the information contained herein or the use of or inability to use any such information. NO WARRANTY, EXPRESS OR IMPLIED, AS TO THE ACCURACY, TIMELINESS, COMPLETENESS, MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE OF ANY SUCH RATING OR OTHER OPINION OR INFORMATION IS GIVEN OR MADE BY MOODY S IN ANY FORM OR MANNER WHATSOEVER. Moody s Investors Service, Inc., a wholly-owned credit rating agency subsidiary of Moody s Corporation ( MCO ), hereby discloses that most issuers of debt securities (including corporate and municipal bonds, debentures, notes and commercial paper) and preferred stock rated by Moody s Investors Service, Inc. have, prior to assignment of any rating, agreed to pay to Moody s Investors Service, Inc. for appraisal and rating services rendered by it fees ranging from $1,500 to approximately $2,500,000. MCO and MIS also maintain policies and procedures to address the independence of MIS s ratings and rating processes. Information regarding certain affiliations that may exist between directors of MCO and rated entities, and between entities who hold ratings from MIS and have also publicly reported to the SEC an ownership interest in MCO of more than 5%, is posted annually at under the heading Investor Relations Corporate Governance Director and Shareholder Affiliation Policy. For Australia only: Any publication into Australia of this document is pursuant to the Australian Financial Services License of MOODY S affiliate, Moody s Investors Service Pty Limited ABN AFSL and/or Moody s Analytics Australia Pty Ltd ABN AFSL (as applicable). This document is intended to be provided only to wholesale clients within the meaning of section 761G of the Corporations Act By continuing to access this document from within Australia, you represent to MOODY S that you are, or are accessing the document as a representative of, a wholesale client and that neither you nor the entity you represent will directly or indirectly disseminate this document or its contents to retail clients within the meaning of section 761G of the Corporations Act MOODY S credit rating is an opinion as to the creditworthiness of a debt obligation of the issuer, not on the equity securities of the issuer or any form of security that is available to retail clients. It would be dangerous for retail clients to make any investment decision based on MOODY S credit rating. If in doubt you should contact your financial or other professional adviser. For Japan only: Moody's Japan K.K. ( MJKK ) is a wholly-owned credit rating agency subsidiary of Moody's Group Japan G.K., which is wholly-owned by Moody s Overseas Holdings Inc., a wholly-owned subsidiary of MCO. Moody s SF Japan K.K. ( MSFJ ) is a wholly-owned credit rating agency subsidiary of MJKK. MSFJ is not a Nationally Recognized Statistical Rating Organization ( NRSRO ). Therefore, credit ratings assigned by MSFJ are Non-NRSRO Credit Ratings. Non-NRSRO Credit Ratings are assigned by an entity that is not a NRSRO and, consequently, the rated obligation will not qualify for certain types of treatment under U.S. laws. MJKK and MSFJ are credit rating agencies registered with the Japan Financial Services Agency and their registration numbers are FSA Commissioner (Ratings) No. 2 and 3 respectively. MJKK or MSFJ (as applicable) hereby disclose that most issuers of debt securities (including corporate and municipal bonds, debentures, notes and commercial paper) and preferred stock rated by MJKK or MSFJ (as applicable) have, prior to assignment of any rating, agreed to pay to MJKK or MSFJ (as applicable) for appraisal and rating services rendered by it fees ranging from JPY200,000 to approximately JPY350,000,000. MJKK and MSFJ also maintain policies and procedures to address Japanese regulatory requirements. 31