Analytics & Credit Decisions

Introduction to our bloggers

Tracy Bremmer
Director, Product Marketing and Management for Decision Sciences

Tracy Bremmer is the Director of Product Marketing and Management in Experian’s Decision Sciences group. She leads the product team in new product initiatives, product development and management, and overall strategy execution. Previously, Bremmer managed a team of Strategy Consultants, advising clients within various industries on business intelligence benchmarking, custom model development, and strategies for risk-based pricing, line management, collections treatments, and account acquisition.

Bremmer joined Experian in 2001 as a Product Marketing Manager in the Fraud Solutions group, where she was responsible for developing and marketing innovative fraud prevention tools.
Bremmer holds a Bachelor of Arts in psychology from the University of California, Irvine and a Master of Business Administration from the University of California, Irvine.

Wendy Greenawalt
Senior Product Manager for Decision Sciences

Wendy Greenawalt is a Senior Product Manager who joined Experian’s Decision Sciences group in 2008. In her current role, she is responsible for the analytics product suite, including Premier AttributesSM, STAGG AttributesSM, Data IntelligenceSM, and the optimization product suite, which includes Marketswitch OptimizationSM, strategy tree optimization and custom optimization products. Greenawalt brings more than 12 years of consumer lender and credit reporting agency experience with extensive knowledge in automated decisioning and online product delivery.

Greenawalt earned a Bachelor of Science in business administration from California State University, Stanislaus, and is Six Sigma Certified.

Kelly Kent
Senior Product Manager/Business Consultant for Decision Sciences

Kelly Kent is a Senior Product Manager and Business Consultant for Experian Decision Analytics’ Decision Sciences. He manages strategy development, product design, and client and sales support for the Business Intelligence suite, including the Experian thought leadership publishing initiative with Oliver Wyman Consulting. In addition, Kent provides consulting services on custom Business Intelligence engagements.

Kent joined Experian in 2007 as a Program Manager, responsible for developing new products and opportunities for the Consumer Information Services business line. His previous experience includes Manager of Business Development for a large retail grocery chain and Implementation Consultant for a software company.

Kent earned an Honours Bachelor of Commerce from McMaster University in Hamilton, Canada, and a Master of Business Administration from the University of California at Los Angeles.

Kari Michel
Senior Product Manager for Decision Sciences

Kari Michel is a Senior Product Manager for Experian Decision Analytics’ Decision Sciences group. She is responsible for the scoring product suite, which includes generic and custom models, Fast StartSM, Performance InsightSM and Risk-based Pricing. Since joining Experian in 1995, Michel has gained extensive experience managing scoring-related products.

Prior to her product management role, Michel was a Business Manager for the analytics group, where she managed a team responsible for scoring and analytical products. Before joining Experian, she was a Senior Credit Risk Analyst with First Interstate Bank.

She earned a Bachelor of Business Administration from California State University, Northridge, and a Master of Business Administration from Chapman University.



 

--by Kari Michel

Most lenders use a credit scoring model in their decision process for opening new accounts; however, between 35 and 50 million adults in the US may be considered unscoreable with traditional credit scoring models. That is equivalent to 18-to-25 percent of the adult population. 

Due to recent market conditions and shrinking qualified candidates lenders have placed a renewed interest in assessing the risk of this under served population.  Unscoreable consumers could be a pocket of missed opportunity for many lenders. To assess these consumers, lenders must have the ability to better distinguish between consumers with a clear track record of unfavorable credit behaviors versus those that are just beginning to develop their credit history and credit risk models.

Unscoreable consumers can be split out into three populations:

• Infrequent credit users:  Consumers who have not been active on their accounts for the past six months, and who prefer to use non-traditional credit tools for their financial needs.

• New entrants:  Consumers who do not have at least one account with more than six months of activity; including young adults just entering the workforce,  recently divorced or widowed individuals with little or no credit history in their name, newly arrived immigrants, or people who avoid the traditional system by choice.

• Thin file consumers:  Consumers who have less than three accounts and rarely utilize traditional credit and likely prefer using alternative credit tools and credit score trends.

A study done by VantageScore® Solutions, LLC shows that a large percentage of the
unscoreable population can be scored with VantageScore* and a portion of these are credit-worthy (defined as the population of consumers who have a cumulative likelihood to become 90 days or more delinquent is less than 5 percent).  The following is a high-level summary of the findings for consumers who had at least one trade:

Lenders can review their credit decisioning process to determine if they have the tools in place to assess the risk of those unscoreable consumers.  As with this population there is an opportunity for portfolio expansion as demonstrated by the VantageScore study.

*VantageScore is a generic credit scoring model introduced to meet the market demands for a highly predictive consumer score. Developed as a joint venture among the three major credit reporting companies (CRCs) – Equifax, Experian and TransUnion.

-- By Kelly Kent

Source: Experian-Oliver Wyman Market Intelligence Reports

Analyzing recent trends from vintages published in the Experian-Oliver Wyman Market Intelligence Reports, there are numerous insights that can be gleaned from just a cursory review of the results.

Mortgage Trends

As noted in an earlier posting, recent mortgage vintages show a broad range of behaviors between more recent vintages and older, more established vintages that were originated before the significant run-up of housing prices seen in the middle of the decade. As below, the 30+ delinquency levels for mortgage vintages in 2005, 2006, and 2007 approach and in two cases exceed 10% of trades in the last 12 months of performance, and have spiked from historical trends, beginning almost immediately after origination. On the other end of the spectrum, the vintages from 2003 and 2002 have barely approached or exceeded 5% for the last 6 or 7 years.

Bankcard Trends

As one would expect, the 30+ delinquency trends demonstrated within bankcard vintages are vastly different from the trends of mortgage vintages. Firstly, card delinquencies show a clear seasonal trend, with a more consistent yearly pattern evident in all vintages, resulting from the revolving structure of the product. The most interesting trends within the card vintages do show that the more recent vintages, 2005 to 2008, display higher 30+ delinquency levels, especially the Q2 2007 vintage, which is far and away the underperformer of the group.Within Within each vintage pool, an analysis can extend into the risk distribution and details of the portfolio and further segment the pool by credit score, specifically VantageScore. In the chart below, an alarming trend is evident. This chart provides the 30+ delinquency levels for the subset of VantageScore A and VantageScore B consumers at the time of origination. In other words, the loans in this pool are only for the most creditworthy customers at the time of origination. The noticeable trend is that while these consumers were largely resistant to deteriorating economic conditions, each vintage segment has seen a spike in the most recent 9-12 months. Given that these consumers tend to have the highest limits and lowest utilization of any Vantage Score band, this trend encourages further account management consideration and raises flags about overall bankcard performance in coming months.

As shown, even a basic review of vintage pools and the subsequent analysis opportunities that result from this data can be extremely useful. This analysis can add a new perspective to risk management, supplementing more established analysis techniques, and further enhancing the ability to see the risk within the risk.

--by Wendy Greenawalt 

In the last installment of my three part series dispelling credit attribute myths, we’ll discuss the myth that the lift achieved by utilizing new attributes is minimal, so it is not worth the effort of evaluating and/or implementing new credit attributes. First, evaluating accuracy and efficiency of credit attributes is hard to measure. Experian data experts are some of the best in the business and, in this edition, we will discuss some of the methods Experian uses to evaluate attribute performance.

When considering any new attributes, the first method we use to validate statistical performance is to complete a statistical head-to-head comparison. This method incorporates the use of KS (Kolmogorov–Smirnov statistic), Gini coefficient, worst-scoring capture rate or odds ratio when comparing two samples. Once completed, we implement an established standard process to measure value from different outcomes in an automated and consistent format. While this process may be time and labor intensive, the reward can be found in the financial savings that can be obtained by identifying the right segments, including:

• Risk models that better identify “bad” accounts and minimizing losses
• Marketing models that improve targeting while maximizing campaign dollars spent
• Collections models that enhance identification of recoverable accounts leading to more recovered dollars with lower fixed costs

Credit attributes
Recently, Experian conducted a similar exercise and found that an improvement of 2-to-22 percent in risk prediction can be achieved through the implementation of new attributes. When these metrics are applied to a portfolio where several hundred bad accounts are now captured, the resulting savings can add up quickly (500 accounts with average loss rate of $3,000 = $1.5M potential savings). These savings over time more than justify the cost of evaluating and implementing new credit attributes.

 

-- by Wendy Greenawalt  

In the second installment of my three part series, dispelling credit attribute myths, we will discuss why attributes with similar descriptions are not always the same. The U.S. credit reporting bureaus are the most comprehensive in the world. Creating meaningful attributes requires extensive knowledge of the three credit bureaus’ data. Ensuring credit attributes are up-to-date and created by informed data experts.  Leveraging complete bureau data is also essential to obtaining long-term strategic success.

To illustrate why attributes with similar names may not be the same let’s discuss a basic attribute, such as “number of accounts paid satisfactory.” While the definition, may at first seem straight forward, once the analysis begins there are many variables that must be considered before finalizing the definition, including:

• Should the credit attributes include trades currently satisfactory or ever satisfactory?
• Do we include paid charge-offs, paid collections, etc.?
• Are there any date parameters for credit attributes?
• Are there any trades that should be excluded?
• Should accounts that have a final status of "paid” be included?

These types of questions and many others must be carefully identified and assessed to ensure the desired behavior is captured when creating credit attributes. Without careful attention to detail, a simple attribute definition could include behavior that was not intended.  This could negatively impact the risk level associated with an organization’s portfolio. Our recommendation is to complete a detailed analysis up-front and always validate the results to ensure the desired outcome is achieved. Incorporating this best practice will guarantee that credit attributes created are capturing the behavior intended.

 

--by Wendy Greenawalt

This blog kicks off a three part series exploring some common myths regarding credit attributes. Since Experian has relationships with thousands of organizations spanning multiple industries, we often get asked the same types of questions from clients of all sizes and industries. One of the questions we hear frequently from our clients is that they already have credit attributes in place, so there is little to no benefit in implementing a new attribute set.

Our response is that while existing credit attributes may continue to be predictive, changes to the type of data available from the credit bureaus can provide benefits when evaluating consumer behavior. To illustrate this point, let’s discuss a common problem that most lenders are facing today-- collections. Delinquency and charge-off continue to increase and many organizations are having difficulty trying to determine the appropriate action to take on an account because consumer behavior has drastically changed regarding credit attributes.

New codes and fields are now reported to the credit bureaus and can be effectively used to improve collection-related activities. Specifically, attributes can now be created to help identify consumers who are rebounding from previous account delinquencies. In addition, lenders can evaluate the number and outstanding balances of collection or other types of trades.  This can be achieved while considering the percentage of accounts that are delinquent and the specific type of accounts affected after assessing credit risk. The utilization of this type of data helps an organization to make collection decisions based on very granular account data.  This is done while considering new consumer trends such as strategic defaulters. Understanding all of the consumer variables will enable an organization to decide if the account should be allowed to self-cure.  If so, immediate action should be taken or modification of account terms should be contemplated. Incorporating new data sources and updating attributes on a regular basis allows lenders to react to market trends quickly by proactively managing strategies. 

 

-- by Kelly Kent

In a recent article, www.CNNMoney.com reported that Federal Reserve Chairman, Ben Bernanke, said that the pace of recovery in 2010 would be moderate and added that the unemployment rate would come down quite slowly, due to headwinds on ongoing credit problems and the effort by families to reduce household debt.’

While some media outlets promote an optimistic economic viewpoint, clearly there are signs that significant challenges lie ahead for lenders. As Bernanke forecasts, many issues that have plagued credit markets will sustain themselves in the coming years. Therefore lenders need to be equipped to monitor these continued credit problems if they wish to survive this protracted time of distress.

While banks and financial institutions are implementing increasingly sophisticated and thorough processes to monitor fluctuations in credit trends, they have little intelligence to compare their credit performance to that of their peers.  Lenders frequently cite that they are concerned about their lack of awareness or intelligence regarding the credit performance and status of their peers.  Marketing intelligence solutions are important for management of risk, loan portfolio monitoring and related decisioning strategies.

Currently, many vendors offer data on industry-wide trends, but few vendors provide the information needed to allow a lender to understand its position relative to a well-defined group of firms that it considers its peers. As a result, too many lenders are performing benchmarking using data sources that are biased, incomplete, inaccurate, or that lack the detail necessary to derive meaningful conclusions.

If you were going to measure yourself personally against a group to understand your comparative performance, why would you perform that comparison against people who had little or nothing in common with you? Does an elite runner measure himself against a weekend warrior to gauge his performance? No; he segments the runners by gender, age, and performance class to understand exactly how he stacks up.

Today’s lending environment is not forgiving enough for lenders to make broad industry comparisons if they want to ensure long-term success. Lenders cannot presume they are leading the pack, when, in fact, the race is closer than ever.

-- by Kari Michel

In August, consumer bankruptcy filings were up by 24 percent over the past year and are expected to increase to 1.4 million this year.  “Consumers continue to turn to bankruptcy as a shield from the sustained financial pressures of today’s economy,” said American Bankruptcy Institute’s Executive Director Samuel J. Gerdano.

What are lenders doing to protect themselves from bankruptcy losses? In my last blog, I talked about the differences and advantage of using both risk and bankruptcy scores. Many lenders are mitigating and managing bankruptcy losses by including bankruptcy scores into their standard account management programs. 

Here are some ways lenders are using bankruptcy scores:

• Incorporating them into existing internal segmentation schemes for enhanced separation and treatment assessment of high risk accounts;

• Developing improved strategies to act on high-bankruptcy-risk accounts
       • In order to manage at-risk consumers proactively and
       • Assessing low-risk customers for up-sell opportunities.

Implementation of a bankruptcy score is recommended given the economic conditions and expected rise in consumer bankruptcy. When conducting model validations/assessments, we recommend that you use the model that best rank orders bankruptcy or pushes more bankruptcies into the lowest scoring ranges.  In validating our Experian/Visa BankruptcyPredict score, results showed BankruptcyPredict was able to identify 18 to 30 percent more bankruptcy compared to other bankruptcy models.  It also identified 12 to 33 percent more bankruptcy compared to risk scores in the lowest five percent of the score range.  This supports the need to have distinct bankruptcy scores in addition to risk scores.

-- By Kelly Kent

Source: Experian-Oliver Wyman Market Intelligence Reports

In the most recent release of the Experian-Oliver Wyman Market Intelligence Reports, each product report contains a series of vintage data reports that shed light on the delinquency, charge-off, and prepayment trends discussed earlier in this series.

These examples of vintage pool curves are taken from the Q2 2009 release and pertain to the mortgage product.

Vintage performance - delinquency
The performance metrics of each vintage are the essence of the benchmarking process. Having properly weighed and balanced each vintage pool, a comparison can be made to the performances of each pool. In the chart shown here, “30+ delinquency rates as % of
trades,” each vintage pool is tracked based on the months on book since its origination. For instance, the longest trend line belongs to the oldest vintage, Q2 2002, and reflects the 30+ delinquency rates over the past 84 months. Conversely, the newest vintage, Q2 2008, is the shortest trend line and reflects only the performance for the past 12 months for those trades. In this chart, it can be easily observed that the delinquency levels for the vintages from 2005, 2006, and 2007 deviate significantly from the older vintages and have spiked for the past 12 to 18 months while older vintages have behaved more consistently.

Distribution of trades
As mentioned earlier, vintage pools are defined by the score at origination for each of the loans within the pool. This is significant in that the distribution of loans will impact the ability to correctly benchmark against each pool. For instance, the chart shown here displays the distributions in each vintage pool, by VantageScore band. 

Despite the clear advantages of using vintage analysis, a benchmarking exercise will require significant weighing and balancing to ensure that the risk profiles of the portfolios are comparable.

Vintage performance - prepayment
Less prominent to delinquency trends are the prepayment trends of each pool. From the moment of origination, each pool begins to change its composition as a result of prepayments/closures which need to be considered in any analysis in order to understand the changing composition of each pool. It is vital that a user understand the shifting risk profile of each vintage, over time. The risk profile, by VantageScore for instance, may skew away from the higher quality consumers over time as prepayment removes them from the pool, leaving only the lowest-scoring consumers in the pool.

These are just three examples of the data required in order to perform vintage analysis. For the sake of brevity, other aspects of these analyses, such as geographic footprint, have been excluded.  These would also add significant insight to the analysis results.


 

-- By Kelly Kent

Vintage analysis, specifically vintage pools, present numerous useful opportunities for any firm seeking to further understand the risks within specific portfolios. While most lenders have relatively strong reporting and metrics at hand  for their own loan portfolio monitoring...these to understand the specific performance characteristics of their own portfolios -- the ability to observe trends and benchmark against similar industry characteristics can enhance their insights significantly.

Assuming that a lender possesses the vintage data and vintage analysis capability necessary to perform benchmarking on its portfolio, the next step is defining the specific metrics upon which any comparisons will be made. As mentioned in a previous posting, three aspects of vintage performance are often used to define these points of comparison:

1. Vintage delinquency including charge-off curves, which allows for an understanding of the repayment trends within each pool. Specifically, standard delinquency measures (such as 30+ Days Past Due (DPD), 60+ DPD, 90+ DPD, and charge-off rates) provide measures of early and late stage delinquencies in each pool.

2. Payoff trends, which reflect the pace at which pools are being repaid. While planning for losses through delinquency benchmarking is a critical aspect of this process, so, too, is the ability to understand pre-repayment tendencies and trends. Pre-payment can significantly impact cash-flow modeling and can add insight to interest income estimates and loan duration calculations.

As part of the Experian-Oliver Wyman Market Intelligence Reports, these metrics are delivered each quarter, and provide a consistent, static pool base upon which vintage benchmarks can be conducted.

Clearly, this is a rather simplified perspective on what can be a very detailed analysis exercise. A properly conducted vintage analysis needs to consider aspects such as: lender portfolio mix at origination; lender portfolio footprint at origination; lender payoff trends and differences from benchmarked industry data in order to properly balance the benchmarked data against the lender portfolio.
 

-- By Kari Michel

Bankruptcies continue to rise and are expected to exceed 1.4 million by the end of this year, according to American Bankruptcy Institute Executive Director, Samuel J. Gerdano.  Although, the overall bankruptcy rates for a lender’s portfolio is small (about 1 percent), bankruptcies result in high dollar losses for lenders.  Bankruptcy losses as a percentage of total dollar losses are estimated to range from 45 percent for bankcard portfolios to 82 percent for credit unions.  Additionally, collection activity is restricted because of legislation around bankruptcy.  As a result, many lenders are using a bankruptcy score in conjunction with their new applicant risk score to make better acquisition decisions. This concept is a dual score strategy.  It is key in management of risk, to minimize fraud, and in managing the cost of credit.

Traditional risk scores are designed to predict risk (typically predicting 90 days past due or greater).  Although bankruptcies are included within this category, the actual count is relatively small.   For this reason the ability to distinguish characteristics typical of a “bankruptcy” are more difficult.  In addition, often times a consumer who filed bankruptcy was in “good standings” and not necessarily reflective of a typical risky consumer.   By separating out bankrupt consumers, you can more accurately identify characteristics specific to bankruptcy.  As mentioned previously, this is important because they account for a significant portion of the losses.
 
Bankruptcy scores provide added value when used with a risk score. A matrix approach is used to evaluate both scores to determine effective cutoff strategies.   Evaluating applicants with both a risk score and a bankruptcy score can identify more potentially profitable applicants and more high- risk accounts.

-- by Wendy Greenawalt

In my last blog post I discussed the value of leveraging optimization within your collections strategy. Next, I would like to discuss in detail the use of optimizing decisions within the account management of an existing portfolio. Account Management decisions vary from determining which consumers to target with cross-sell or up-sell campaigns to line management decisions where an organization is considering line increases or decreases.  Using optimization in your collections work stream is key.

Let’s first look at lines of credit and decisions related to credit line management. Uncollectible debt, delinquencies and charge-offs continue to rise across all line of credit products. In response, credit card and home equity lenders have begun aggressively reducing outstanding lines of credit.    One analyst predicts that the credit card industry will reduce credit limits by $2 trillion by 2010. If materialized, that would represent a 45 percent reduction in credit currently available to consumers. This estimate illustrates the immediate reaction many lenders have taken to minimize loss exposure. However, lenders should also consider the long-term impacts to customer retention, brand-loyalty and portfolio profitability before making any account management decision.

Optimization is a fundamental tool that can help lenders easily identify accounts that are high risk versus those that are profit drivers. In addition, optimization provides precise action that should be taken at the individual consumer level.

For example, optimization (and optimizing decisions) can provide recommendations for:

• when to contact a consumer;
• how to contact a consumer; and
• to what level a credit line could be reduced or increased...

…while considering organizational/business objectives such as:

• profits/revenue/bad debt;
• retention of desirable consumers; and
• product limitations (volume/regional).

In my next few blogs I will discuss each of these variables in detail and the complexities that optimization can consider.

-- By Kari Michel

This blog completes my discussion on monitoring new account decisions with a final focus: scorecard monitoring and performance.  It is imperative to validate acquisitions scorecards regularly to measure how well a model is able to distinguish good accounts from bad accounts. With a sufficient number of aged accounts, performance charts can be used to:

• Validate the predictive power of a credit scoring model;
• Determine if the model effectively ranks risk; and
• Identify the delinquency rate of recently booked accounts at various intervals above and below the primary cutoff score.

To summarize, successful lenders maximize their scoring investment by incorporating a number of best practices into their account acquisitions processes:

1. They keep a close watch on their scores, policies, and strategies to improve portfolio strength.
2. They create monthly reports to look at population stability, decision management, scoring models and scorecard performance.
3. They update their strategies to meet their organization’s profitability goals through sound acquisition strategies, scorecard monitoring and scorecard management.
 

-- By Wendy Greenawalt

The combined impact of rising unemployment, increasing consumer debt burdens and decreasing home values have caused lenders to shift resources away from prospecting and acquisitions to collection and recovery activities. As delinquencies and charge-off rates continue to increase, the likelihood of collecting on delinquent accounts decreases -- because outstanding debts mount for consumers and their ability to pay declines. Integrating optimized decisions into a collection strategy enables a lenders to assign appropriate collection treatments by assessing the level of risk associated with a consumer while considering a customer’s responsiveness to particular treatment options.  

Specifically, collections optimization uses mathematical algorithms to maximize organizational goals while applying constraints such as budget and call center capacity  -- providing explicit treatment strategies at the consumer level -- while producing the highest probability of collecting outstanding dollars. Optimization can be integrated into a real-time call center environment by targeting the right consumers for outbound calls and assigning resources to consumers most likely to pay.  It can also be integrated into traditional lettering campaigns to determine the number and frequency of letters, and the tone of each correspondence. The options for account treatment are virtually limitless and, unlike other techniques, optimization will determine the most profitable strategy while meeting operational and business constraints without simplification of the problem.

By incorporating optimization into a collection strategy that includes a predictive model or score and advanced segmentation, an organization can maximize collected dollars, minimize the costs of collection efforts, improve collections efficiency, and determine which accounts to sell off – all while maximizing organizational profits.

--  Kari Michel

This blog is a continuation of my previous discussion about monitoring your new account acquisition decisions with a focus on decision management. 

Decision management reports provide the insight to make more targeted decisions that are sound and profitable. These reports are used to identify: which lending decisions are consistent with scorecard recommendations; the effectiveness of overrides; and/or whether cutoffs should be adjusted.

Decision management reports include:

• Accept versus decline score distributions
• Override rates
• Override reason report
• Override by loan officer
• Decision by loan officer

Successful lending organizations review this type of information regularly to make better lending policy decisions.  Proactive monitoring provides feedback on existing strategies and helps evaluate if you are making the most effective use of your score(s). It helps to identify areas of opportunity to improve portfolio profitability. 

In my next blog, I will discuss the last set of monitoring reports, scorecard performance.

-- By Tracy Bremmer

In our last blog (July 30), we covered the first three stages of model development which are necessary whether developing a custom or generic model.  We will now discuss the next three stages, beginning with the “baking” stage:  scorecard development.
 
Scorecard development begins as segmentation analysis is taking place and any reject inference (if needed) is put into place. Considerations for scorecard development are whether the model will be binned (divides predictive attributes into intervals) or continuous (variable is modeled in its entirety), how to account for missing values (or “false zeros”), how to evaluate the validation sample (hold-out sample vs. an out-of-time sample), avoidance of over-fitting the model, and finally what statistics will be used to measure scorecard performance (KS, Gini coefficient, divergence, etc.).

Many times lenders assume that once the scorecard is developed, the work is done.   However, the remaining two steps are critical to development and application of a predictive model:  implementation/documentation and scorecard monitoring.   Neglecting these two steps is like baking a cake but never taking a bite to make sure it tastes good. 

Implementation and documentation is the last stage in developing a model that can be put to use for enhanced decisioning. Where the model will be implemented will determine the timeliness and complexity for when the models can be put into practice. Models can be developed in an in-house system, a third-party processor, a credit reporting agency, etc. Accurate documentation outlining the specifications of the model will be critical for successful implementation and model audits.

Scorecard monitoring will need to be put into place once the model is developed, implemented and put into use. Scorecard monitoring evaluates population stability, scorecard performance, and decision management to ensure that the model is performing as expected over the course of time. If at any time there are variations based on initial expectations, then scorecard monitoring allows for immediate modifications to strategies.

With all the right ingredients, the right approach, and the checks and balances in place, your model development process has the potential to come out “just right!”

-- By Wendy Greenawalt

When consulting with lenders, we are frequently asked what credit attributes are most predictive and valuable when developing models and scorecards. Because we receive this request often, we recently decided to perform the arduous analysis required to determine if there are material differences in the attribute make up of a credit risk model based on the portfolio on which it is applied.

The process we used to identify the most predictive attributes was a combination of art and sciences -- for which our data experts drew upon their extensive data bureau experience and knowledge obtained through engagements with clients from all types of industries. In addition, they applied an empirical process which provided statistical analysis and validation of the credit attributes included. Next, we built credit risk models for a variety of portfolios including bankcard, mortgage and auto and compared the credit attribute included in each.

What we found is that there are some attributes that are inherently predictive regardless for which portfolio the model was being developed. However, when we took the analysis one step further, we identified that there can be significant differences in the account-level data when comparing different portfolio models.

This discovery pointed to differences, not just in the behavior captured with the attributes, but in the mix of account designations included in the model. For example, in an auto risk model, we might see a mix of attributes from all trades, auto, installment and personal finance…as compared to a bankcard risk model which may be mainly comprised of bankcard, mortgage, student loan and all trades.  Additionally, the attribute granularity included in the models may be quite different, from specific derogatory and public record data to high level account balance or utilization characteristics.

What we concluded is that it is a valuable exercise to carefully analyze available data and consider all the possible credit attribute options in the model-building process – since substantial incremental lift in model performance can be gained from accounts and behavior that may not have been previously considered when assessing credit risk.

-- By Tracy Bremmer

Preheat the oven to 350 degrees. Grease the bottom of your pan. Mix all of your ingredients until combined. Pour mixture into pan and bake for 35 minutes. Cool before serving.

Model development, whether it is a custom or generic model, is much like baking. You need to conduct your preparatory stages (project design), collect all of your ingredients (data), mix appropriately (analysis), bake (development), prepare for consumption (implementation and documentation) and enjoy (monitor)!  

This blog will cover the first three steps in creating your model! 

Project design involves meetings with the business users and model developers to thoroughly investigate what kind of scoring system is needed for enhanced decision strategies. Is it a credit risk score, bankruptcy score, response score, etc.? Will the model be used for front-end acquisition, account management, collections or fraud?

Data collection and preparation evaluates what data sources are available and how best to incorporate these data elements within the model build process. Dependent variables (what you are trying to predict) and the type of independent variables (predictive attributes) to incorporate must be defined. Attribute standardization (leveling) and attribute auditing occur at this point. The final step before a model can be built is to define your sample selection.

Segmentation analysis provides the analytical basis to determine the optimal population splits for a suite of models to maximize the predictive power of the overall scoring system. Segmentation helps determine the degree to which multiple scores built on an individual population can provide lift over building just one single score.

Join us for our next blog where we will cover the next three stages of model development:  scorecard development; implementation/documentation; and scorecard monitoring. 
 

-- By Kari Michel

In my last blog I gave an overview of monitoring reports for new account acquisition decisions listing three main categories that reports typically fall into:  (1) population stability; (2) decision management; (3) scorecard performance.

Today, I want to focus on population stability.   Applicant pools may change over time as a result of new marketing strategies, changes in product mix, pricing updates, competition, economic changes or a combination of these. Population stability reports identify acquisition trends and the degree to which the applicant pool has shifted over time, including the scorecard components driving the shift in custom credit scoring models. 

Population stability reports include:

• Actual versus expected score distribution
• Actual versus expected scorecard characteristics distributions (available with custom models)
• Mean applicant scores
• Volumes, approval and booking rates

These types of reports provide information to help monitor trends over time, rather than spikes from month to month.  Understanding the trends allows one to be proactive in determining if the shifts warrant changes to lending policies or cut-off scores.

Population stability is only one area that needs to be monitored; in my next blog I will discuss decision management reports.

 

-- By Wendy Greenawalt

On any given day, US credit bureaus contain consumer trade data on approximately four billion trades. Interpreting data and defining how to categorize the accounts and build attributes, models and decisioning tools can and does change over time, due to the fact that the data reported to the bureaus by lenders and/or servicers also changes.

Over the last few years, new data elements have enabled organizations to create attributes to identify very specific consumer behavior. The challenge for organizations is identifying what reporting changes have occurred and the value that the new consumer data can bring to decisioning.

For example, a new reporting standard was introduced nearly a decade ago which enabled lenders to report if a trade was secured by money or real property. Before the change, lenders would report the accounts as secured trades making it nearly impossible to determine if the account was a home equity line of credit or a secured credit card. Since then, lender reporting practices have changed and, now, reports clearly state that home equity lines of credit are secured by property making it much easier to delineate the two types of accounts from one another.

By taking advantage of the most current credit bureau account data, lenders can create attributes to capture new account types.  They can also capture information (such as: past due amounts; utilization; closed accounts and derogatory information including foreclosure; charge-off and/or collection data) to make informed decisions across the customer life cycle.