Servicers are overwhelmed by volumes and often do not have the resources they need, both in terms of personnel and technology (due to the increasing costs of servicing).  For example, they lack technologies (such as loan portfolio management) to determine the best option for each loan in distress, whether that be loss mitigation, a short sale, a third party sale or foreclosure.  They have yet to adjust to their new role as “life coaches” (not just loan counselors focused on completing a task) in working with so many distressed borrowers.   Regulations, such as the Home Affordable Modification Plan (HAMP) in the US, have been challenging to implement and execute.

So why should the residential mortgage industry use loan portfolio management – now?

Loan portfolio management has been shown to reduce re-default and significantly improve average unpaid-principal-balance increase in net present value (NPV) from modifications using loan portfolio management (relative to nonoptimized loan modifications using general risk scores) from vendors such as Response Analytics (Distressed Portfolio Management and others).  Segmentation and clustering analysis supports optimal treatment strategies as well.  For example, First American CoreLogic (WillCap) provides cluster-driven treatment strategies that segments borrowers into more than 20 segments in order to reduce default rates, decrease losses and accomplish socio-political goals (for example, keeping borrowers in their homes and reducing foreclosures).

And preventing a customer from falling behind on payments may be the best course of all.  I spoke with someone in the industry today that said something like 20%-30% of borrowers who are current on their mortgage at the time of modification later end up defaulting.  If a borrower has fallen behind on payments, the rate is more like 70%-80%.  Pre-delinquency management is also a core capability of loan portfolio management.

For example, credit card issuers are reducing limits on consumers whose risk profile is changing, potentially due to unemployment or other factors.  They are monitoring individual consumer spending patterns and can tell when a consumer begins using their credit card in a way that signals distress – for example, if they start using their credit card to pay their mortgage.  In this case, the financial institution may shut down the card altogether.  Or in some cases, depending on the risk profile of the customer, they may increase the spending limit.

Case studies show that card issuers using loan portfolio management technology that supports pattern-based strategy and seek-model-adapt have reduced bad debt by 25% – 50% and increased revenues by 25% – 30% (by optimizing authorization rates, credit lines and marketing campaigns).

Credit card issuers are already moving this direction as in the examples above – other lenders would benefit from following.

• “Introducing Pattern-Based Strategy“
• “Balance Investment in Four Categories to Support Pattern-Based Strategy”
• “Five Eras of IT Business Value Add: From Automation to Pattern-Based Strategy”
• “CEOs and Chief Strategy Officers: Balance Investments With Pattern-Based Strategy”

Pattern-based strategies enable business leaders to actively seek, amplify, examine and exploit patterns (see “Introducing Pattern-Based Strategy“). Seeking patterns will require new disciplines and technologies that identify patterns of change to indicate opportunity or risk. Exploiting patterns will require new disciplines and technologies that enable an organization to establish a consistent and repeatable response (focused on results) to patterns of change.

Loan portfolio management should immediately become a critical element of a lender’s corporate pattern-based strategy because it uses technology to expose signals that may lead to a pattern that will have a positive or negative impact on operations.  It can be used to significantly the credit losses forecasted for the global financial services industry.

Regulatory change, falling balances, record delinquencies and general recessionary stress require a new approach to credit cards.  Loan portfolio management can help issuers rebuild portfolio profitability by factoring risk and profitability analytics into credit line, authorizations and marketing decisions.

Credit card issuers are already moving this direction, and other lenders would benefit from following.

• Distressed portfolio management – ability to determine the optimal treatment strategy:
  • Loan modification
  • Short sale
  • Third party sale
  • Foreclosure
• Loan modification risk – ability to analyze the risk of re-default on modified loans and determine the optimal course of action that minimizes lender, investor and insurer loss
• Pricing analytics for securities and insurance – value-based pricing for securities and insurance that includes risk factors such as default risk, interest rate exposure and prepayment risk
• Pricing analytics for real estate owned (REO) property – value-based pricing for properties in REO
• Mortgage/loan asset management – acquisition, due diligence, service and surveillance, loss mitigation, asset disposition, portfolio performance analysis
• Lending portfolio stress testing – analysis of loan portfolio and securities as macroeconomic parameters change

Loan portfolio management also supports other functions (beyond reducing credit loss) within the lending lifecycle:

• Credit scoring and risk evaluation – the risk of default for loans in the origination process
• Prepayment risk – individual loan-level propensity for prepayment risk and exposure estimates
• Fraud modeling – predictive modeling to identify fraud
• Reengineering credit card profitability – regulatory change, falling balances (-11% as of Q12009 (source:  FICO)), record delinquencies and general recessionary stress require a new approach to credit cards; loan portfolio management can help issuers rebuild portfolio profitability
• Customer service support – real-time decisioning and support integrated with collections systems (for example, workout support decisioning).

I’m sure there are more use cases I’m missing.  What others have you seen?

Credit models are typically based around “who” – who is likely to go delinquent.  Once delinquent, treatment strategies are then selected (loss mitigation, short sale, third party sale, foreclosure).  In the US, regulations such as the Home Affordable Modification Plan (HAMP) are pushing modifications.  But a big problem is emerging.  From what I can see, modifications are not really preventing default – they are delaying default.  Fitch Ratings estimates that 55% to 65% of overall loan modifications, and 75% of sub-prime loans, will become at least 60 days delinquent in the next 12 months.  This focus on “who” has been (and will continue to be) very costly for mortgage servicers and investors.

This is where they “why” comes in.  Default behaviors have been changing, and therefore borrower segmentation is important.  For example, borrowers with negative equity in their home tend to look like good credit.  However, their LTV value ends up making more sense to “strategically default” on their mortgage rather than remain current.  In talking to Experian, they said this population of borrowers in 2004 -2009 went from 3% to 18% of defaulters.  The behavior and motivations of different segments of borrowers is important to take into account in terms of identifying optimal treatment strategies.

So, “who” is important, but “why” helps show the way forward when it comes to optimizing treatment strategies.

For one thing, not all homeowners qualify (Freddie Mac or Frannie Mae guarantee $5.5 trillion of the $10.9 trillion of residential mortgages in the US).  This program has struggled as well – only 360,165 (or 12%) of 2.9 million eligible distressed homeowners have started trial modifications under the Home Affordable Modification Program (HAMP) (as of August 2009).  Some of the largest US banks, such as Bank of America, have only done HAMP modifications for 7% of eligible loans.  In addition, earlier this year Fitch Ratings estimated that 55% to 65% of overall loan modifications, and 75% of sub-prime loans, will become at least 60 days delinquent in the next 12 months.

Loan portfolio management has been shown to improve average unpaid-principal-balance increase in net present value (NPV) from modifications using loan portfolio management (relative to nonoptimized loan modifications using general risk scores) by 25%.  With such a big impact like this (sometimes up to 35% of NPV), I think banks and servicers in the US mortgage industry will turn back to loan portfolio management once banks either decouple from the Treasury by paying off TARP funds, determine that HAMP modifications are not sustainable (low bank adoption rates would suggest resistance already), or both.

Source:  FDIC, Gartner (September 2009)

Financial institutions must find a better way to reduce pending loan losses or risk failure.

Financial institutions are largely fighting their way through the financial crisis by leveraging default management and loss mitigation solutions.  However, the primary function of loss mitigation is not to proactively identify distress, but rather to deal with a default on payment after it has occurred.  To prevent loss, lenders must have a mechanism to prevent default in the first place.  They must have a means to reduce the risk of re-default on modified loans and price distressed assets.

Loan portfolio management applications leverage analytics and modeling capabilities to price distressed assets and proactively identify distress – which, once identified, can then be handled by loss mitigation applications.

• New loan application risk: the risk of default for loan mortgage applications
• Prepayment risk: individual loan-level propensity for prepayment risk and exposure estimates
• Pricing analytics: value-based pricing for new loans and securities that includes risk factors such as default risk, interest rate exposure and prepayment risk
• Portfolio stress testing: analysis of loan portfolio and securities as macroeconomic parameters change
• Fraud modeling: predictive modeling to identify fraud
• Optimization: mathematical formulas and science-based decision modeling applied against each individual loan to determine its most-appropriate value in the face of uncertain conditions and risk factors
• Behavior modeling: adaptive recovery rate modeling (to predict the success of a loan modification) that changes over time as market conditions fluctuate (similar to predicting the risk of prepayment)
• Customer service support: real-time decisioning and workout support integrated with collection systems.

 End-users of loan portfolio management solutions include banks and servicers, distressed asset investors, mortgage insurers, brokerdealers, money managers, and insurance firms.

 The technology vendor community has only recently begun providing portfolio management solutions for lending, and to date most solutions have been focused on mainly mortgage or commercial lending.  However, the financial crisis has accelerated interest in portfolio management solutions in order to reduce loss and execute loan modification more proactively, particularly in the area of mortgage loans (and advanced mortgage analytics) given they have been in the most distress to date.  We suggest evaluating your existing analytics capabilities against loan portfolio management solutions – clearly, it’s never been more important to address business opportunities and challenges quickly in order to maintain liquidity and drive new revenue