STAR Performance Scorecard White Paper

Transcription

1 STAR Performance Scorecard White Paper March 2017

2 Table of Contents Table of Contents... 2 STAR Introduction... 3 What is STAR?... 3 Profiles and Relevant Metrics... 4 General Servicing Metric Definitions... 6 Solution Delivery Metric Definitions... 7 Timeline Management Metric Definitions... 8 Exceptions and Special Handling... 9 Comparable Pool Methodology Background Methodology Changes for 2017 Decision Trees Calculating Comparable Performance Evaluating Performance Calculating Relative Performance Determine Inference Based on Variance to Comp Resources Support Online Resources Appendix A Calculating Individual Metric Scores Appendix B Supplemental STAR Metric Definitions Credit Performance Metrics Capability Metrics Fannie Mae. Trademarks of Fannie Mae of 26

3 STAR Introduction What is STAR? The Total Achievement and Rewards (STAR ) Program is Fannie Mae s servicer performance management system. STAR is designed to: Provide consistent, specific, and measurable feedback to our servicers Align servicer performance with Fannie Mae s business goals Promote servicing knowledge and excellence across the housing industry and drive change Focus on actions that limit credit losses s in this program represent the majority of Fannie Mae s total credit risk exposure. Participants are evaluated based on their performance in the Capability Model (SCM) operational assessment and metrics and the STAR Performance Scorecard Fannie Mae. Trademarks of Fannie Mae of 26

4 Profiles and Relevant Metrics The STAR Reference Guide outlines three distinct business process areas in which servicers are evaluated through the STAR Program: General Servicing, Solution Delivery and Timeline Management. In each process area there are metrics that measure performance in terms of a servicer s ability to prevent credit losses for Fannie Mae (Credit Performance metrics) and those that measure their people, process and technology operations through the Capability Model (SCM). Collectively, these metrics represent the activities that are of strategic importance in achieving Fannie Mae s business objectives and demonstrate servicers proficiency in producing desired outcomes. The table below represents the STAR Scorecard metrics for each process area in Table 1: 2017 STAR Scorecard M etrics by Process Area Process Area Credit Performance Capability Model General Servicing Solution Delivery Timeline Management Transition to to Cure Retention Efficiency Liquidation Efficiency 6-Month Mod Performance Transition to Beyond Time Frame Multi-Occurrence Hard Reject Rate Multi-Occurrence Soft Reject Rate Shortage Percent Average Speed to Answer - Collections Abandonment Rate Collections Motions for Relief Referred Timely REOgrams Submitted within Timeline Title Issues Resolved within 45 Days Our servicers success is essential in achieving Fannie Mae s goal of preserving home ownership and reducing taxpayers exposure to credit losses. While every servicer participating in the STAR program manages a significant amount of risk for Fannie Mae in terms of the number of loans it services, not all STAR servicers manage enough delinquent loans for the scorecard to effectively measure relative performance in the Solution Delivery and Timeline Management process areas. Therefore, a servicer s level of participation in the STAR Program is tiered based on the number of loans it services that are relevant to each process area. The metrics a servicer receives in its monthly scorecard are representative of that profile, and its eligibility for recognition as STAR Performers Fannie Mae. Trademarks of Fannie Mae of 26

5 Figure 1: 2017 STAR Inclusion Hierarchy Inclusion in each process area based on the number of loans serviced in the following categories: Beyond Time Frame Timeline Management 60+ Days Delinquent Solution Delivery Total Loans Serviced General Servicing In addition to the metrics listed in Table 1, servicers will also receive supplemental STAR metrics relevant to each process area. These metrics measure activity that is considered foundational to achieving success in the STAR Scorecard and are used to refine Fannie Mae s evaluation of each process area. A full list of supplemental metrics measured in the 2017 STAR Scorecard is included in this document as Appendix B. Summary The STAR Scorecard measures servicer performance relative to other servicers and against established thresholds in three process areas related to loan servicing. s are selected for the STAR program based on the number of loans they service on behalf of Fannie Mae. All servicers are measured in the General Servicing category, and servicers that service a significant number of delinquent loans are also included in the Solution Delivery and/or Timeline Management categories. s monthly scorecards represent their performance in the relevant Scorecard metrics and includes supplemental metrics that measure key activities in each process area Fannie Mae. Trademarks of Fannie Mae of 26

6 General Servicing Metric Definitions Credit Performance Transition to 60+ Measures the number of loans that roll from a less than 60 day delinquent status to a 60+ day delinquent status or a loss liquidation status 1 over a three-month reporting period. Active trial modifications less than four months old are excluded from the metric. For example, if a servicer had 1,005 loans that were less than 60 days delinquent in January, and as of April, 20 of those loans were 60 or more days delinquent, but five had active trial modifications less than four months old, the 60+ transition rate for April would be 1.5%. Capability Model Multi-Occurrence Hard Reject Rate Measures the percentage of loans with multiple hard reject occurrences within the report period; for a full list of Reject codes, please refer to the Help Center section in the s Reconciliation Facility (SURF TM ). The numerator for the metric is based on the count of unique loans which had a payment hard reject in the report period and at least one or more times in the previous 5 months. The denominator is based on the total number of loans at the beginning of the processing period. If the loan is a bi-weekly loan and had a hard reject, it is included if the payment due date falls with the selected activity month, but it is only included once even if there are multiple payment due dates or hard reject occurrences within the selected activity month. Multi-Occurrence Soft Reject Rate Measures the percentage of loans with multiple soft reject occurrences within the report period; for a full list of reject codes, please refer to the Help Center section in the SURF. The numerator for the metric is based on the count of unique loans which had a payment soft reject in the report period and at least one or more times in the previous 5 months. The denominator is based on the total number of loans at the beginning of the processing period. If the loan is a bi-weekly loan and had a soft reject, it is included if the payment due date falls within the selected activity month, but it is only included once even if there are multiple payment due dates or soft reject occurrences within the selected activity month. Shortage Percent Measures the rate at which a servicer s monthly remittance is less than the total amount expected by Fannie Mae s Investor Reporting (SIR). The shortage percent is calculated as the total shortage amount divided by the total due, where total due is equal 1 Loss liquidation status includes short sales, Mortgage Release, foreclosure sales and third-party sales 2017 Fannie Mae. Trademarks of Fannie Mae of 26

7 to the total monthly remittance plus the total shortage amount less the total surplus amount. The total shortage amount is the shortage balance at the close of the cash reconciliation activity period for A/A, S/A, and S/S cash remittance. Solution Delivery Metric Definitions Credit Performance 60+ to Cure Measures the number of loans 60 or more days delinquent at the beginning of the month that are brought current, paid in full, or repurchased over a three-month reporting period. Active trial modifications less than four months old that are not converted during the reporting period are excluded from the metric. For example, if a servicer had 120 loans that were 60 or more days delinquent at the end of January, and as of April month end, 25 of these loans were current or paid in full and 20 had active trial modifications less than four months old (and not yet current), the 60 to Cure rate would be 25% for April. Retention Efficiency Measures the number of modifications initiated during the month as a percentage of loans 60 or more days delinquent 2 at the beginning of the month. Active trial modifications less than four months old are excluded from the metric. The metric ratio reported on the Performance Scorecard represents a three-month total of retention solutions in the numerator, divided by the three-month total of the monthly denominator population. Liquidation Efficiency Measures the number of completed Mortgage Releases, short sales, foreclosure sales and third-party sales during the month as a percentage of loans 60 or more days delinquent 2 at the beginning of the month. Active trial modifications less than four months old are excluded from the metric. The metric ratio reported on the Performance Scorecard represents a three-month total of liquidation solutions in the numerator, divided by the three-month total of monthly denominator population. 6-Month Modification Performance Measures the number of loan modifications completed 6 months prior that are now less than 30 days delinquent, paid in full, or repurchased, divided by the total number of loan modifications completed and not cancelled in the same time period. This metric uses a rolling three-month total in both the numerator and denominator. For example, the March metric denominator would be all modifications completed in July, August, and September 2 Retention and liquidation solutions delivered prior to the 60 th day of delinquency are included in the metric numerator and denominator and are treated as 60 days delinquent in the Comp Fannie Mae. Trademarks of Fannie Mae of 26

8 of 2016, and the metric numerator would be the sum of July 2016 modifications current as of January 2017, August 2016 modifications current as of February 2017, and September 2016 modifications current as of March Capability Model Average Speed to Answer Average number of seconds calls wait in the Collections/Loss Mitigation queues to be answered by an agent. Abandonment Rate Percentage of calls that enter the Collections/Loss Mitigation agent queue that are not answered by a live agent before the caller disconnects. This percentage represents all calls that enter the queue, with no exclusions for disconnects within a minimum threshold, blocked calls, or requested call backs. Timeline Management Metric Definitions Credit Performance Transition to Beyond Time Frame Measures the number of loans within 180 days of the state foreclosure time frame that transition to a beyond time frame status over a six-month reporting period. For example, if a servicer had 200 loans within 180 days of the state foreclosure time frame at the end of January, and as of July month end, 80 of these loans were still active and beyond the allowable foreclosure time frames, the Transition to Beyond Time Frame rate would be 40% for July. Capability Model Motions for Relief Referred Timely Measures the percentage of Motion for Relief referrals submitted to a bankruptcy attorney within Fannie Mae guidelines. Percentage of REOgrams Completed Timely Measures the percentage of REOgrams submitted within Fannie Mae timeline requirements Fannie Mae. Trademarks of Fannie Mae of 26

9 Percentage of Title Issues Resolved within 45 Days Measures the percentage of vesting-related title issues (e.g., HOA, deeds, and taxes) resolved within 45 days of the initial discovery or notification dates. Exceptions and Special Handling Loan products with unique characteristics or limited distribution among servicers are excluded from the Credit Performance metrics in STAR Scorecard this also applies to loans in certain geographic regions. As such, second liens, government insured loans, and loans securing properties in insular areas (e.g., Puerto Rico, Guam, and the U.S. Virgin Islands) are excluded from the metric and Comp calculations. Loans that transfer from one servicer to another are treated differently based on the metric being measured and the status of the loan at transfer. Table 2 represents the transfer exclusions applied for each Credit Performance Metric. Table 2: M etric Exclusions for Transferred Loans Metric Excluded from Transferor (Y/N) Months Excluded for Transferee Transition to 60+ Y to Cure Y 3 2 Retention Efficiency Y 2 Liquidation Efficiency Y 2 6-Month Mod Performance Y Not Measured Transition to Beyond Time Frame Y 2 3 Loans that transition to 60+ days delinquent or cure before the transfer date are included in the numerator and denominator for the transferor servicer 2017 Fannie Mae. Trademarks of Fannie Mae of 26

10 Comparable Pool Methodology Background In order to compare one servicer s performance to another, or to all other servicers collectively, it is first necessary to establish a methodology to account for differences in the credit characteristics of their respective portfolios. That way, the performance measured by the Credit Performance metrics represents the activities servicers do to impact credit losses, not simply how a servicer s individual loans perform. This section describes how Fannie Mae segments the portfolio based on key loan attributes, and how a servicer s performance is then measured against the performance of a similar population of loans. That process of comparing loans with like characteristics is referred to as Comping and the groupings that are generated are referred to as the Comp throughout this White Paper. The first step in comparing servicer performance on loans with similar characteristics is determining which loan attributes will be compared for a given metric. The selection of these attributes is based on statistical research performed by the analytics team at Fannie Mae using the historical loan performance of Fannie Mae's entire book of business. The attributes vary by metric, but common attributes include loan delinquency, status, and loan-to-value ratio (LTV). In previous versions of the STAR Scorecard the variables chosen for each metric were applied together, so that each attribute was represented in the final Comp bucket. This Comping methodology creates a highly refined set of loan characteristics at the individual bucket level; however, as the number of delinquent loans in the portfolio falls, it can become more difficult to ensure that there are enough observations in each bucket to make valid comparisons. Figure 2 illustrates how each additional layer of variables multiplies the number of buckets produced, and represents instances where there is a potential for limited observations. Figure 2: Comparable Pool Bucket M ethodology Collectibility 2 variables Delinquency x3 variables LTV x4 variables FC Time Frame x8 variables Thin Data 2017 Fannie Mae. Trademarks of Fannie Mae of 26

11 Methodology Changes for 2017 Decision Trees For 2017, the STAR Scorecard has adopted a conditional inference tree model to segment the loan attributes and create the Comp for each metric based on the historical performance of the Fannie Mae book of business. To do this, Fannie Mae leverages a publicly available recursive partitioning algorithm 4 to choose the loan characteristics that most significantly differentiate performance for each segment of loans. The recursive partitioning methodology makes a series of binary splits on a large set of control variables, ordering them by how important each variable is to the metric outcome. Once a control variable is selected, the algorithm then decides which split 5 best differentiates performance for that variable. As this process is repeated, the series of splits creates a tree-like structure referred to as a Decision Tree and each decision point in this tree is referred to as a node. Once the tree has extended to the point that there are no more variable splits that significantly differentiate performance, the branch terminates into its final, terminal node. It is the performance of each servicer s loans at the terminal node level that is then used to calculate servicers Comp for the metric. Figure 3 represents the decision tree model, where the nodes are created from the most relevant control variables for each segment of the tree. This allows the final metric tree 6 to evaluate many variables and still have fewer buckets or terminal nodes. Additionally, the trees can be limited so that the splitting ends if the number of loans falls below a certain threshold, ensuring that each terminal node contains a sufficient number of observations to make valid comparisons. Note that the example tree does not represent any particular metric in the STAR Scorecard; rather, it is a simple representation of the way the tree branches and nodes are created based on the attributes provided to the partitioning algorithm. To simplify the example, the outer branches of the sample tree have been limited so that just the first four levels are visible, and continuous variables like LTV have been rounded. Not all nodes are labeled in the image below but since the splits are binary, the alternate node represents the opposite of the inequality represented. 4 Torsten Hothorn, Kurt Hornik, Achim Zeileis (2006). party: A Laboratory for Recursive Partytioning 5 Splits can be either groupings of categorical variables like collectibility or thresholds for continuous variables like LTV 6 Trees for each metric are developed and calibrated using the entire Fannie Mae portfolio. The final tree for each metric is then applied to each servicer s population of loans Fannie Mae. Trademarks of Fannie Mae of 26

12 Figure 3: Comparable Pool Tree M ethodology All Loans LTV > 80% 1 8 LTV <= 80% Collectible 2 Uncollectible 7 9 LTV <= 70% 14 LTV > 70% Delinquency > Delinquency > Delinquency <= Delinquency > 6 Delinquency <= 6 Collectible FC Time Frame > 360 In this example the algorithm has determined that the most important segmentation for the entire metric is LTV and has first split the population based on those loans with a LTV greater than 80% or less than or equal to 80%. For loans with higher LTVs, the next segmentation is whether or not there are circumstances that would limit a servicer s ability to collect a payment (e.g., bankruptcy, forbearance, etc.). The collectible population is then further split by delinquency, while for loans that are deemed uncollectible the node terminates and there is no further segmentation for that population of loans. For loans with lower LTV ratios, the next node level further defines the population in terms of the same loan attribute (LTV). This is a valuable feature of the recursive partitioning methodology that allows attributes to be refined as they progress through the tree. The final splits are based on the delinquency of the loans and then the collectibility or foreclosure time frame. As mentioned previously, it is the performance at the final, terminal node level that is used to create the Comp for each credit performance metric. In the following section, we will select two terminal nodes from this tree that represent loans that might be considered more or less likely to achieve a curative event in the sample metric. We will then compare two servicers performance in those nodes to demonstrate how servicers portfolio characteristics impacts the construction of their Comp, and how servicer performance is evaluated relative to the Comp performance Fannie Mae. Trademarks of Fannie Mae of 26

13 Summary The attributes used to establish comparable portfolio characteristics for each metric are chosen based on analysis performed by Fannie Mae. The STAR Scorecard has modified its Comping methodology for 2017 to use decision trees to segment portfolios based on specific loan attributes. The attributes are chosen based on historical analysis performed by Fannie Mae, and the variables are then selected by a partitioning algorithm that segments the populations based on the how the attribute differentiates performance Fannie Mae. Trademarks of Fannie Mae of 26

14 Calculating Comparable Performance Once the terminal nodes have been created, a servicer's performance in each node is compared to the performance of all the other Fannie Mae loans for that node. This section will walk through the steps necessary to calculate the relative performance of a servicer s portfolio. Node 7 from Figure 3 represents a high LTV loan that is limited in terms of the solutions that can be solicited. Node 13 represents a low LTV loan that is no more than 4 months delinquent and potentially more curable. We will refer to these terminal nodes as having a high or low curative opportunity in the following example to demonstrate how the Comp is calculated and the way a servicer s portfolio stratification influences its performance. In Table 3, the performance of A is compared to the performance of the Comp, excluding A's loans, and against B who has a similar population of loans but with different attributes. Note that there are 500,000 loans in the Fannie Mae denominator, and A has 38,500 of these loans and B has 40,000. A has had 865 metric events and B has had 900 metric events ( Numerator column). A's Comp is the other 461,500 loans from the Fannie Mae denominator. B s comp, while a similar number of loans, is different from servicer A s based on the metric event opportunity of their portfolio. Table 3: Comparable Pool Construction Node Eve nt Opportunity Fannie Mae Numerator Fannie Mae Denominator Numerator Denominator Comp Numerator Comp Denominator A 7 Low 1, , ,500 1,380 91, High 10, , ,000 9, ,000 11, , ,500 10, ,500 B 7 Low 1, , ,000 1,250 84, High 10, , ,000 9, ,000 11, , ,000 10, ,000 Table 4 shows how a servicer s metric value is influenced by its portfolio makeup. Here, the denominator weight represents the portion of the servicer s metric population represented by that node. The product of the metric and the denominator weight yields that node s contribution to the servicer s overall metric value. This helps to illustrate the importance of segmenting the portfolio based on key loan attributes to account for the servicer s unique credit characteristics. Now we will take into account the key loan attributes and look at the performance in both the high and low opportunity nodes Fannie Mae. Trademarks of Fannie Mae of 26

15 Table 4: Understanding Node Weighting Based on s Portfolio Node Metric Event Opportunity Numerator Denominator Metric Denominator Weight Contribution to Metric Ratio A 7 Low 120 8, % 22.08% 0.31% 13 High , % 77.92% 1.94% , % % 2.25% B 7 Low , % 40.00% 0.63% 13 High , % 60.00% 1.63% , % % 2.25% A servicer s metric ratio is determined by dividing the metric numerator by the metric denominator. While metric performance can be expressed in aggregate for each servicer or for each node, the Comp is created based on how other loans with similar characteristics perform. Notice in Table 4 that the total number of loans (servicer denominator) for each servicer is similar, and servicers absolute performance is the same (2.25%). However, the grouping of the two populations by nodes based on the servicer portfolio composition is different, as shown in the Contribution to Metric Ratio column. Table 5: Calculating and Comp Performance Node Metric Event Opportunity Numerator Denominator Metric Comp Numerator Comp Denominator Comp Ratio A 7 Low 120 8, % 1,380 91, % 13 High , % 9, , % , % B 7 Low , % 1,250 84, % 13 High , % 9, , % , % The same ratio calculation is made against the servicer s Comp pool of loans to establish the performance of all loans in the Fannie Mae portfolio with those characteristics. Once the Comp ratio is calculated (Table 5), you can determine how many loans would have had a metric event if the servicer performed at Comp by multiplying the Comp ratio by the servicer s denominator for both nodes as in Table 6 below. The servicer s Comp for the metric is then calculated as the sum of the individual node Comp values. Table 6: Determining M etric Comp Values Node Metric Event Opportunity Comp Ratio Denominator Comp Value A 7 Low 1.51% 8, High 2.50% 30, , B 7 Low 1.49% 16, High 2.49% 24, , Fannie Mae. Trademarks of Fannie Mae of 26

16 Summary For STAR metrics that measure relative performance, we segment the loan population for each metric based on loan attributes that impact servicer performance in the metric. Once the loan populations are segmented into their final, terminal nodes, the metric ratio for all other Fannie Mae servicers with similar loans is calculated. That ratio is then applied to the servicer s loan volume to establish the Comp for that node. The results are then aggregated for all nodes to establish the servicer s metric Comp which represents expected metric performance Fannie Mae. Trademarks of Fannie Mae of 26

17 Evaluating Performance Calculating Relative Performance Once the Comp value has been calculated, we can see to what degree the rates for each servicer are higher or lower than the Comp and find the percentage their performance is different from the Comp values. That aggregated difference will become the basis of how the STAR Performance Scorecard measures servicer s relative performance. The following table reflects how the Comp values we calculated are used to calculate the percent difference to Comp for each servicer. To do this, start by subtracting the Comp value from the servicer s numerator value to establish the difference between those numbers. Then divide that total by the Comp value to establish the percent at which it varies. For metrics where a lower value is more desirable, like Transition to 60+, the result is adjusted by a factor of -1 so that larger values are always better. Table 7: Calculating Variance to Comp Node Metric Event Opportunity Numerator Comp Value Variance to Comp Pct. Variance to Comp A 7 Low % 13 High % % B 7 Low % 13 High % % Recall from Table 5 that the absolute performance of these two servicers was nearly equal. However, the variance to comp is different between the two servicers as a result of the difference of portfolio composition of the two servicers. To illustrate how a servicer s performance compared to its Comp relates to other servicers for the same metric, the STAR Scorecard will include metric scores based on the servicer s variance to Comp for each. An explanation of the scoring methodology and the calculations used to produce the individual matric scores is provided in Appendix A. Determine Inference Based on Variance to Comp Once the servicer s metric performance and its relative performance to the Comp is calculated, an inference is made to determine, within a reasonable degree of statistical certainty, if the servicer performed above Comp or below Comp. To do this, STAR evaluates the variance to Comp for each servicer to first determine whether the servicer s measured variance is statistically different from zero at a 99% confidence level. If the servicer s variance to Comp does not meet the threshold of being statistically 2017 Fannie Mae. Trademarks of Fannie Mae of 26

18 different, the servicer s performance is deemed to be at Comp for the evaluation period. If the servicer s variance is statistically different from zero with a 99% degree of certainty, the inference will be determined as above Comp if the variance is positive or below Comp if the variance is negative. Figure 4: Inference Test Summary Once a servicer s Comp is established based on the servicer s performance in each node of the decision tree, we then compare the servicer s results to that of its Comp to determine relative performance. That performance is referred to as the servicer s variance to Comp and a servicer s relative performance to other servicers measured is represented by a metric score. An inference test is applied to determine whether or not there is a statistically significant difference between the servicer s performance and that of its Comp Fannie Mae. Trademarks of Fannie Mae of 26

19 Resources Support For servicer support or additional information related to the STAR Program: Contact your Fannie Mae Portfolio Manager us at Call our Support Center at FANNIE Online Resources Online resources can be accessed at The site contains a Program Overview, the STAR Reference Guide, Frequently Asked Questions (FAQs), and other training resources. For additional details regarding Investor Reporting and Investor Reporting Scorecard metrics, please visit Fannie Mae. Trademarks of Fannie Mae of 26

20 Appendix A Calculating Individual Metric Scores To provide servicers with a representation of their performance relative to other servicers in a given metric, the STAR Scorecard provides a numerical score. This multi-step calculation takes into account the range of outcomes for the servicers peer group and then re-scales the range from 5 to 95 in the following steps based on the results calculated for B: Step Description Formula Example Result 1 Calculate variance to Comp for servicer (Actual - Comp) / Comp ( ) / % 2 Calculate adjusted variance to Comp Variance to Comp * -1 (only for metrics w here low er score is better) 6.33% x % 3 Determine the highest variance to Comp for the month Maximum variance to the Comp for the servicer's peer group for the month 12.18% 12.18% 4 Determine the lowest variance to Comp for the month Minimum variance to the Comp for the servicer's peer group for the month % % 5 Calculate the range of differences Maximum - Minimum 12.18% - (-15.20%) 27.38% 6 Calculate the relative position for the servicer within the range (Adjusted Variance to Comp Minimum of Range) / Range (6.33% - (-15.20%)) / 27.38% 78.66% 7 Convert relative position to score (Relative Position of * (95-5)) + 5 ( x (95-5)) Fannie Mae. Trademarks of Fannie Mae of 26

21 Appendix B Supplemental STAR Metric Definitions Credit Performance Metrics Process Area General Servicing Solution Delivery Sub-Process Area Metric Definition Formula Threshold Collections Current to and Loss Worse Mitigation Collections and Loss Mitigation The number of loans in a current status at the beginning of the month that transitioned to a delinquent status at the end of the month divided by the beginning current loan count. 30 to Worse The number of loans in a 30 day delinquent status at the beginning of the month that transitioned to a 60+ day delinquency status at the end of the month divided by the beginning 30 day delinquent loan count. 60 to Worse The number of loans in a 60 day delinquent status at the beginning of the month that transitioned to a 90+ day delinquency status at the end of the month divided by the beginning 60 day delinquent loan count. 90+ to Worse The number of loans in a 90+ day delinquent status at the beginning of the month that transitioned to a more delinquent status at the end of the month divided by the beginning 90+ day Retention Solution Timing Liquidation Solution Timing delinquent loan count. The percentage of trials initiated by the servicer before the 210th day of delinquency. This metric uses a rolling three month total in both the numerator and denominator. For short sales and Mortgage Release, the percentage of cases completed by the servicer before the 210th day of delinquency. This metric uses a rolling three month total in both the numerator and denominator. Loans that transition to 30+ or loss Loans current at beginning of month Loans that transition to 60+ or loss Loans 30 days delinquent at beginning of month Loans that transition to 90+ or loss Loans 60 days delinquent at beginning of month Loans that transition to 120+ or loss Loans 90+ days delinquent at beginning of month Trial mods started < 210 days delinquent All trial mod starts Preferred liquidations < 210 days delinquent All preferred liquidations Inference Test Inference Test Inference Test Inference Test Inference Test Inference Test

22 Process Area Solution Delivery Sub-Process Area Metric Definition Formula Threshold Collections Mod The percentage of trials converted based and Loss Conversion on trials started 6 months prior Mitigation Converted trial mods Trial mods started 6 month's prior Inference Test Timeline Management Foreclosure 12-Month Mod Performance Beyond Time Fram e Resolution The number of loan modifications completed 12 months prior that are now less than 30 days delinquent, paid in full, or repurchased, divided by the total number of loan modifications completed and not cancelled in the same time period. This metric uses a rolling three month total in both the numerator and denominator. Measures loan dispositions and delinquency cures as a percentage of loans that began the month beyond the allow able foreclosure time frame. Mods that are < 30 days delinquent Mods converted 12 month's prior Current Loans and Liquidations Loans beyond time frame at beginning of month Inference Test Inference Test 2017 Fannie Mae. Trademarks of Fannie Mae of 26

23 Capability Metrics Process Area General Servicing Sub-Process Area Metric Definition Formula Threshold Investor Ending Hard Percent of ending hard rejects to total Reporting Reject Rate loans in portfolio Hard rejects in current period Loans in portfolio MIN: % MAX: % Aged Recurring Hard Reject Rate Percent of hard rejects that have occurred on individual loans over the span of five consecutive reporting periods, including the current reporting period Recurring hard rejects Loans in portfolio MIN: % MAX: % Aged Recurring Soft Reject Rate Percent of soft rejects that have occurred on individual loans over the span of five consecutive reporting periods, including the current reporting period Recurring soft rejects Loans in portfolio MIN: % MAX: % Surplus Percent Percent of total month end cash surplus amount to total due for the period Month end cash surplus in the period Total amount due for the period MIN: % MAX: % Percent of Loans not Reported Percentage of unique active loans in the portfolio w hich w ere not reported (missing LAR) in a timely manner as per Fannie Mae's policies and guidelines Number of loans not reported Loans in portfolio TBD Average Days Reporting Liquidations Average days for a servicer to report a liquidation LAR in an accepted state. Business days (Action Date - Accepted Date) Number of Liquidated Loans TBD LAR 83 Discrepancy Rate Percentage of unique active ARM loans in portfolio w ith outstanding note rate and/or payment change discrepancies (those due for a scheduled change but missing LAR 83 and those w ith LAR 83 exceptions) at any point in a given activity month. LAR 83 Discrepancy ARM Loan Count ARM Projections for the activity month TBD 2017 Fannie Mae. Trademarks of Fannie Mae of 26

24 Process Area Solution Delivery Sub-Process Area Metric Definition Formula Threshold Collections Outbound and Loss Efficiency Mitigation N/A Total right party contact (RPC) (borrow er or authorized party) spoken to divided by total outbound connects (dialing connects w ith a "live" phone number) Outbound RPCs Outbound live connects Penetration Ratio Percent of delinquent loans that received a manned outbound attempt for the reported month Loans that receive an outbound attempt Eligible delinquent loan population 100% Collections Calls Offered Total number of calls offered to collection queues N/A N/A Collections Calls Answered Total number of calls received through a collections queue answ ered by an agent N/A N/A Repayment Plan Success Rate Percentage of borrow er payments made in accordance w ith a repayment plan Loans that made a plan payment Loans on active payment plan at start of month N/A Timeline Management Bankruptcy Proofs of Claim Filed Timely Percentage of Proofs of Claim filed on or before to Proof of Claim deadline/bar date (excluding Amended Claims), for each month Number of Proofs of Claim filed timely Proofs of Claim filed 80.0% Bankruptcy No LPI Movement to Bankruptcy Inventory Percentage of bankruptcy loans w ith no payment(s) received Bankruptcy loans w ithout LPI movement Total bankruptcy inventory 3.5% Foreclosure 180+ Loans Not in a Foreclosure Status Percentage of loans that are 180+ days delinquent but not reported as in foreclosure Loans w ithout a foreclosure status code 180+ day delinquent loans 2.0% 2017 Fannie Mae. Trademarks of Fannie Mae of 26

25 Process Area Timeline Management Sub-Process Area Metric Definition Formula Threshold Foreclosure Pull-Through Rate Total foreclosure sales held divided by all scheduled foreclosure sales for the month, excluding Mortgage Release and Reverse Mortgages Total sales held All scheduled sales for the month 70.0% Scheduled Sales to Foreclosure Inventory The ratio of loans w ith a scheduled foreclosure sales to the active foreclosure inventory at the end of the month Loans w ith a scheduled foreclosure sale Loans in foreclosure 5.0% Foreclosure Holds to Inventory Percentage of loans in foreclosure w ith a foreclosure hold Loans w ith a foreclosure hold Loans in foreclosure < 33.0% REO Fulfillment REOgr ams Submitted Number of total REOgrams submitted for the period N/A N/A Elimination Exceptions Percentage Percentage of servicer elimination/rescission notification exceptions relative to all eliminations reported Elimination or rescission exceptions Total REO eliminations N/A Title Exception Percentage Percentage of title exceptions for the reporting period Title exceptions reported for the period Total REO properties conveyed to Fannie Mae N/A Tax title Issues Resolved within 45 Days Percentage tax title issues resolved w ithin 45 days of initial discovery or notification date Tax issues resolved w ithin 45 days Tax issues resolved 72.0% Foreclosure Deeds Resolved within 45 Days Percentage of completed foreclosure deed vesting issues resolved w ithin 45 days of initial discovery or notification date Foreclosure deed issues resolved w /in 45 days Foreclosure deed issues resolved 72.0% 2017 Fannie Mae. Trademarks of Fannie Mae of 26

26 Process Area Timeline Management Sub-Process Area Metric Definition Formula Threshold REO Number of homeow ner association (HOA) Fulfillment title issues for the reporting period HOA Title Issues Resolved within 45 Days HOA issues resolved w /in 45 days HOA issues resolved 72.0% Seller Not in Title Issues Resolved within 45 Days Percentage of completed seller not in title issues resolved w ithin 45 days of initial discovery or notification date Seller not in title issues resolved w /in 45 days Seller not in title deed issues resolved 72.0% Property Preservation Maintenance Score For a sample SDQ population, the percentage of evaluated properties that meet Fannie Mae standards. Properties that meet expectation Properties evaluated TBD 2017 Fannie Mae. Trademarks of Fannie Mae of 26