oversight

Evaluation of Fannie Mae's Servicer Reimbursement Operations for Delinquency Expenses

Published by the Federal Housing Finance Agency, Office of Inspector General on 2013-09-18.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

             Federal Housing Finance Agency
                 Office of Inspector General




 Evaluation of Fannie Mae’s Servicer
   Reimbursement Operations for
       Delinquency Expenses




Evaluation Report  EVL-2013-012  September 18, 2013
                Evaluation of Fannie Mae’s Servicer Reimbursement
                Operations for Delinquency Expenses
                Why OIG Did This Report
                The Federal National Mortgage Association (Fannie Mae) relies on mortgage
                servicers to make various payments on behalf of delinquent borrowers. Generally,
                these payments are for property maintenance, insurance, taxes, and loan liquidation
Synopsis        costs. Ultimately, Fannie Mae reimburses the servicers for payments that comply with
                the Fannie Mae Single Family Servicing Guide. These servicer reimbursements are
   ———          the subject of this evaluation.

September 18,   To collect reimbursement, servicers submit a form electronically to Fannie Mae.
    2013        Eighty percent of these forms require manual processing – a task that Fannie Mae
                outsources to Accenture LLP (Accenture). Accenture analysts review the submissions
                and decide to pay, curtail, or deny the reimbursements. Accenture relays these
                decisions to Fannie Mae, which reimburses servicers in the amount approved by the
                Accenture analysts.
                In 2012, Fannie Mae reimbursed servicers $2.9 billion as a result of Accenture’s
                manual processing decisions. Occasionally, however, Accenture analysts erroneously
                approve or deny servicer reimbursements. As a result of these errors, Fannie Mae
                overpays or underpays servicers. Thus, operational oversight is critical.
                In 2010, Fannie Mae’s Internal Audit Division began finding serious deficiencies in
                the controls and processes Fannie Mae used to ensure servicers were reimbursed in the
                correct amount. The Federal Housing Finance Agency (FHFA or Agency) Office of
                Inspector General (OIG) commenced this evaluation to assess FHFA’s oversight of
                Fannie Mae’s servicer reimbursement operations.

                What OIG Found
                Prior to 2013, FHFA conducted minimal oversight of Fannie Mae’s servicer
                reimbursement operations. Recently, however, it has bolstered its monitoring and is
                currently conducting a targeted examination of the area.
                Fannie Mae oversees Accenture principally by measuring its contractual performance.
                Although Fannie Mae collects useful data through sampling, it does not effectively
                utilize these data to minimize processing errors. For example, Fannie Mae does not
                aggregate the extent to which it overpays servicers as a result of Accenture’s
                processing errors. Without attempting to quantify the problem, it may be difficult for
                Fannie Mae managers to make fully informed business decisions or measure their
                corrective actions.
                Moreover, Fannie Mae has yet to implement a red flag system for servicer
                reimbursements, despite repeated urgings. In this context, a red flag system would
                identify reimbursement trends and data anomalies that adversely impact Fannie Mae
                financially.
                To quantify the effectiveness of Fannie Mae’s controls, OIG analyzed Fannie Mae’s
                sampling data. OIG estimates that Accenture analysts incorrectly approved 3.1% of
                servicer reimbursements in 2012. These processing errors prompted Fannie Mae to
                pay servicers $89 million (i.e., “overpayments”).

                What OIG Recommends
                OIG recommends that FHFA:

Synopsis           1. Ensure Fannie Mae takes the actions necessary to reduce Accenture
                      processing errors. These actions should include utilizing its process accuracy
   ———                data in a more effective manner and implementing a red flag system.
                   2. Require Fannie Mae to:
September 18,
    2013                   a. Quantify and aggregate its overpayments to servicers regularly;
                           b. Implement a plan to reduce these overpayments by (i) identifying
                              their root causes, (ii) creating reduction targets, and (iii) holding
                              managers accountable; and
                           c. Report its findings and progress to FHFA periodically.
                   3. Publish Fannie Mae’s reduction targets and overpayment findings.
TABLE OF CONTENTS ................................................................

ABBREVIATIONS .........................................................................................................................6

PREFACE ........................................................................................................................................7

CONTEXT .......................................................................................................................................8
      Fannie Mae Loan Liquidation ..................................................................................................8
      Property Management During Delinquency and REO .............................................................8
              Property Management During Delinquency .....................................................................9
              Property Management During REO .................................................................................9
      Fannie Mae’s Servicer Reimbursement Operations ...............................................................10
              Form 571 .........................................................................................................................10
              Manual Processing ..........................................................................................................11
              Fannie Mae Oversight of Accenture’s Manual Processing ............................................13
              Internal Recognition of Deficiencies in Fannie Mae’s Oversight ..................................15

FINDINGS .....................................................................................................................................17
      1. Prior to 2013, FHFA Oversight of Fannie Mae’s Servicer Reimbursement
      Operations Was Minimal; Recently, FHFA Has Bolstered Its Monitoring ............................17
              FHFA Oversight Prior to 2013 .......................................................................................17
              FHFA Oversight in 2013 ................................................................................................18
      2. Fannie Mae’s Oversight of Accenture’s Claim Processing Focuses on
      Contractual Compliance Rather than Minimizing Overpayments to Servicers......................18
              Data Collection and Analysis .........................................................................................19
              Red Flag System .............................................................................................................19
      3. Accenture Processing Errors Prompted Fannie Mae to Overpay Servicers
      $89 Million in 2012 ................................................................................................................20

CONCLUSIONS............................................................................................................................22

RECOMMENDATIONS ...............................................................................................................23
      FHFA Response to Recommendations ...................................................................................23



                                         OIG  EVL-2013-012  September 18, 2013                                                                4
OBJECTIVE, SCOPE, AND METHODOLOGY .........................................................................25
      Estimation Methodology ........................................................................................................25
             Technical Errors ..............................................................................................................25
             Claim Review Methodology ...........................................................................................26
      Improper Payment Analysis ...................................................................................................28
      Authority .................................................................................................................................29

APPENDIX A ................................................................................................................................30
      FHFA’s Comments on OIG’s Findings and Recommendations ............................................30

APPENDIX B ................................................................................................................................32
      OIG’s Response to FHFA’s Comments .................................................................................32
             Methodology Concerns ...................................................................................................32
             OIG’s Third Recommendation .......................................................................................33

ADDITIONAL INFORMATION AND COPIES .........................................................................34




                                         OIG  EVL-2013-012  September 18, 2013                                                                 5
ABBREVIATIONS .......................................................................

Accenture             Accenture LLP

CPM-Ops               Credit Portfolio Management Operations

Enterprises           Fannie Mae and Freddie Mac

Fannie Mae            Federal National Mortgage Association

FHFA or Agency        Federal Housing Finance Agency

Form 571              Expense Reimbursement Submission Request

Freddie Mac           Federal Home Loan Mortgage Corporation

HERA                  Housing and Economic Recovery Act of 2008

IPERA                 Improper Payments Elimination and Recovery Act of 2010

Navigant              Navigant Consulting, Inc.

OIG                   Federal Housing Finance Agency Office of Inspector General

REO                   Real Estate Owned

Servicing Guide       Fannie Mae Single Family Servicing Guide

SLAs                  Service Level Agreements




                        OIG  EVL-2013-012  September 18, 2013                     6
PREFACE ...................................................................................

FHFA was established by the Housing and Economic Recovery Act of 2008 (HERA),1
which amended the Inspector General Act of 1978.2 HERA requires FHFA to oversee the
prudential operations of Fannie Mae and the Federal Home Loan Mortgage Corporation
(Freddie Mac) (collectively the Enterprises). Additionally, in September 2008, FHFA placed
the Enterprises into conservatorships. The goals of the conservatorships are to preserve the
Enterprises’ assets and minimize taxpayer losses.

This report evaluates Fannie Mae’s servicer reimbursement operations and FHFA’s oversight
of those operations. OIG appreciates the cooperation of all those who contributed to this
evaluation, which was led by Bruce McWilliams, Senior Investigative Evaluator; Brian
Harris, Investigative Counsel; Alexa Strear, Investigative Counsel; Omolola Anderson,
Statistician; Christine Eldarrat, Senior Policy Advisor; and Desiree Yang, Financial Analyst.
It has been distributed to Congress, the Office of Management and Budget, and others and
will be posted on OIG’s website, www.fhfaoig.gov.




George Grob
Deputy Inspector General for Evaluations




1
    Pub. L. No. 110-289.
2
    Pub. L. No. 95-452.




                           OIG  EVL-2013-012  September 18, 2013                              7
CONTEXT ..................................................................................

Fannie Mae Loan Liquidation

When Fannie Mae purchases or guarantees a mortgage, it acquires a lien on the home that
secures the mortgage. This lien enables Fannie Mae to liquidate the loan and sell the home if
the borrower defaults. Accordingly, Fannie Mae has a financial interest in each of these
homes.

The primary goal of loan liquidation is to sell the home to a third party. The loan can be
liquidated in one of four ways: (1) short sale, (2) third-party sale, (3) deed-in-lieu of
foreclosure, or (4) foreclosure. The former two liquidation options result in a direct sale of the
home to a third party, thus ending Fannie Mae’s affiliation with the property. In contrast, the
latter two liquidation options result in Fannie Mae temporarily becoming the legal owner of
the home. During this temporary ownership period, the property is classified as real estate
owned (REO). Over time, REO properties are sold to third parties on the real estate market.
Ultimately, loan liquidation and the subsequent sale of the home serve to mitigate the credit
losses Fannie Mae incurs from a borrower default.

                     FIGURE 1. FANNIE MAE’S FOUR LOAN LIQUIDATION METHODS



                                                          Property
                               Short Sale
                                                         Disposition

                                                          Property
                            Third-Party Sale
                                                         Disposition
  Loan Liquidation
                             Deed-in-Lieu of                                        Property
                                                            REO
                              Foreclosure                                          Disposition

                                                                                    Property
                              Foreclosure                   REO
                                                                                   Disposition




Property Management During Delinquency and REO

Fannie Mae has measures in place to safeguard the value of its security interests when
borrowers begin to miss payments. Depending on a loan’s status, either the servicer or
contractors working for Fannie Mae will manage the property and, if necessary, sell it.




                             OIG  EVL-2013-012  September 18, 2013                                 8
   Property Management During Delinquency

Delinquency begins when a borrower misses a
mortgage payment and ends after (1) the borrower            Servicer: The entity responsible
cures the delinquency or (2) one of the four loan           for collecting principal and
liquidation events occurs. During this delinquency          interest payments from the
                                                            borrower. The servicer may be
period, the loan’s servicer is contractually obligated
                                                            the loan’s originating bank.
to protect Fannie Mae’s beneficial interest in the home
and, when appropriate, liquidate the loan. The former
obligation entails making various payments on the borrower’s behalf; generally, these
payments are for property maintenance, insurance, taxes, and loan liquidation costs.
Ultimately, Fannie Mae will reimburse servicers for all payments that comply with the Fannie
Mae Single Family Servicing Guide (Servicing Guide). These servicer reimbursements are the
subject of this evaluation.

   Property Management During REO

When a home becomes REO (i.e., following foreclosure or deed-in-lieu of foreclosure), the
servicer is relieved of its duty to protect Fannie Mae’s beneficial interest in the property. To
ensure home value and facilitate property disposition during this stage, Fannie Mae contracts
with REO vendors. REO vendors are independent contractors that perform many of the same
tasks that servicers perform during delinquency. Like servicers, Fannie Mae reimburses REO
vendors for their expenses if they comply with Fannie Mae guidelines. REO vendor
reimbursements are not the subject of this evaluation. However, this contextual information is
critical to understanding OIG’s findings.

 FIGURE 2. MAINTENANCE OF FANNIE MAE’S SECURITY INTERESTS DURING DIFFERENT LOAN STAGES


      Performing                        Delinquency                              REO

                                          Fannie Mae contracts
                                                                             Fannie Mae contracts
       Fannie Mae relies on                  with servicers to
                                                                             with REO vendors to
      borrowers to maintain               maintain home value
                                                                             maintain home value
           home value                       and facilitate loan
                                                                               and facilitate sale
                                                liquidation




                              OIG  EVL-2013-012  September 18, 2013                                9
Fannie Mae’s Servicer Reimbursement Operations

Fannie Mae reimbursed servicers approximately $3.8 billion in 2012 for their delinquency
expenses.3 Generally, to obtain reimbursement for these expenses, servicers access Fannie
Mae’s Asset Management Network and electronically submit an Expense Reimbursement
Submission Request (Form 571).

    Form 571

Form 571 is comprised of 13 broad, billable categories. These categories are further refined
into a total of 79 unique line items. Figure 3 provides various examples of these categories
and the line items within them.

Each Form 571 submitted to Fannie                  FIGURE 3. EXAMPLES OF CATEGORIES AND LINE ITEMS
Mae is referred to as a claim.4 Claim                               ON FORM 571
complexity and submission timing
                                                         Category                         Line Item
vary. The average servicer claim
                                                                                Landscaping
contains approximately seven line                Property Preservation
                                                 Expenses                       Trash Removal
items, but they can contain as few as                                           Locksmith
one.5 Generally, servicers submit their                                         Hazard Premium
claims near the end of a borrower’s              Insurance                      Mortgage Insurance Premium
delinquency period or immediately                                               Title Insurance
after loan liquidation. Thus, it is              Taxes
                                                                                State Taxes
possible for servicers to make                                                  Property Taxes
payments for years prior to requesting           Foreclosure Costs and          Eviction Costs
reimbursement from Fannie Mae.                   Expenses                       Sherriff’s Fees and Costs

Fannie Mae processes claims using three mutually exclusive methods:

    1. Scripted. Scripted claims refer to those that require approval from an internal Fannie
       Mae business unit.

    2. Automatically. If a claim’s reimbursement request falls within specific, predetermined
       parameters, then Fannie Mae automatically reimburses the claim.

3
 This number has not been audited or validated by Fannie Mae for financial reporting purposes or compliance
with the Sarbanes-Oxley Act of 2002.
4
 Fannie Mae reimburses both servicers and REO vendors through the same means – Fannie Mae’s Asset
Management Network. When a distinction is necessary, this evaluation refers to servicer reimbursement
submissions as “servicer claims” and REO vendor reimbursement submissions as “REO vendor claims.”
5
 OIG calculated this average from Fannie Mae’s process accuracy sample (discussed below). OIG found that
50% of the claims in the sample contain three or fewer line items.




                                OIG  EVL-2013-012  September 18, 2013                                       10
    3. Manually. If a claim does not fall within the parameters referenced above, then it is
       processed manually. This entails an analyst reviewing each line item on the claim for
       compliance with the Servicing Guide.

Figure 4 illustrates that a majority
                                                              FIGURE 4. SERVICER CLAIM VOLUME BY
of servicer claims require manual
                                                                  PROCESSING METHOD IN 20126
processing. Accordingly, OIG’s
evaluation focuses on manual                           1,400,000
processing operations.6                                1,200,000
                                                       1,000,000
                                                         800,000
    Manual Processing                                    600,000
                                                         400,000
Prior to 2011, Fannie Mae internally                     200,000
reviewed all claims that required manual                       0
processing in its Dallas, Texas, satellite             Automatically   Scripted     Manually
                                                           (< 1%)       (19%)         (80%)
office. In 2011, however, Fannie Mae
decided to outsource this task to Accenture, a management consulting, technology, and
outsourcing company. By July 2011, Accenture was processing all claims that required
manual review – a practice that continues today. Accenture’s claim processing staff consists
of approximately 84 employees who work at an Accenture delivery center in San Antonio,
Texas.

Fannie Mae prescribes specific servicer claim review
procedures that Accenture analysts are required to                  Curtail: Reimburse the
follow when reviewing each claim. These procedures                  servicer a lesser amount than
are intended to ensure Fannie Mae only reimburses                   requested.
servicers for claims that comply with the Servicing
Guide. After reviewing each line item on the claim, the analyst (1) approves reimbursement
of the claim in full, (2) curtails the claim, or (3) rejects it.7 This decision is relayed to Fannie
Mae, which distributes the approved amount of money to the servicer. In 2012, Accenture
manually processed approximately 1.3 million claims,8 approving a total of $2.9 billion in
servicer reimbursements.



6
 These numbers have not been audited or validated by Fannie Mae for financial reporting purposes or
compliance with the Sarbanes-Oxley Act of 2002.
7
 If a servicer believes its claim was unjustly curtailed or denied, it can contact Fannie Mae for an explanation.
Additionally, the servicer can repeatedly resubmit a curtailed or denied claim without penalty.
8
 This number has not been audited or validated by Fannie Mae for financial reporting purposes or compliance
with the Sarbanes-Oxley Act of 2002.




                                  OIG  EVL-2013-012  September 18, 2013                                           11
                FIGURE 5. CHRONOLOGY OF SERVICER REIMBURSEMENT PROCESS FOR
                                MANUALLY PROCESSED CLAIMS




Ideally, Accenture analysts would only approve and deny claims in compliance with the
Servicing Guide. However, errors do arise because of the (1) inconsistent application of
guidelines by different reviewers, (2) limited resources devoted to avoiding such
inconsistencies, (3) complex subject matter, and (4) large volume of servicer claims. When an
Accenture analyst erroneously approves or denies a line item in a claim, it results in Fannie
Mae making an overpayment or underpayment, respectively, to a servicer. Figure 6 provides
four illustrative examples of errors that Accenture analysts are prone to make and their
consequences.

    FIGURE 6. EXAMPLES OF PROCESSING ERRORS THAT LEAD TO AN INCORRECT REIMBURSEMENT

                              Accenture Analyst Error                                   Consequence
 An analyst reviews a servicer claim that contains a reimbursement request for         Fannie Mae
 hazard insurance. The hazard insurance request is for a period of time that exceeds   overpays the
 the maximum allowed by the Servicing Guide. The analyst erroneously approves          servicer.
 the amount requested.
 An analyst reviews a duplicate servicer claim. The claim is for a service that was    Fannie Mae
 already processed and approved by another analyst. Unaware that it is a duplicate,    overpays the
 the analyst erroneously approves the amount requested.                                servicer.
 An analyst reviews a servicer claim that contains a reimbursement request for         Fannie Mae
 condominium fees. The servicer, however, did not provide the documentation            overpays the
 required by the Servicing Guide. The analyst erroneously approves the amount          servicer.
 requested.
 An analyst reviews a servicer claim that contains a reimbursement request for         Fannie Mae
 attorney fees. The reimbursement request is within the Servicing Guide’s              underpays the
 acceptable boundaries. The analyst erroneously rejects the request as excessive.      servicer.



                              OIG  EVL-2013-012  September 18, 2013                                  12
Presently, neither Fannie Mae nor Accenture estimates how much money Fannie Mae
overpays or underpays servicers. Further, operational factors tend to favor the correction of
underpayments as opposed to overpayments. Notably, the burden falls on Fannie Mae to
identify and recover its own overpayments. Currently however, Fannie Mae has no remedial
measures to recover overpayments on 99% of claims because it only samples 1% of claims for
accuracy (discussed below). Alternatively, when a servicer is underpaid, it often contacts
Fannie Mae to discuss the underpayment and resubmits the claim at no additional cost.
Because of this asymmetry, netting overpayments and underpayments does not accurately
represent the financial loss that Fannie Mae incurs as a result of Accenture errors.

    Fannie Mae Oversight of Accenture’s Manual Processing

Fannie Mae conducts oversight of Accenture’s manual processing operations though its Credit
Portfolio Management Operations (CPM-Ops) business group.9 CPM-Ops conducts oversight
primarily by (1) quantifying a variety of claim processing metrics that are specified in the
Fannie Mae-Accenture contract and (2) identifying data trends.

    a. Quantifying Claim Processing Metrics

The Fannie Mae-Accenture contract specifies minimum claim processing standards known as
service level agreements (SLAs). If Accenture fails to satisfy any of these SLAs, then the
contract requires Accenture to remedy the failure operationally or by paying a penalty. The
SLA most pertinent to this evaluation is the process accuracy rate.

The process accuracy rate measures the degree to which Accenture analysts correctly
followed the claim review procedures when reviewing claims. To measure the process
accuracy rate, Fannie Mae draws a weekly sample of about one percent of claims that have
been manually processed by Accenture analysts. Then, a Fannie Mae reviewer examines each
claim in the sample and answers the following five, weighted questions affirmatively or
negatively:
    1. Did the Accenture analyst make the correct reimbursement decision?                            45 points
    2. If applicable, did the Accenture analyst obtain the necessary approvals?                      20 points
    3. If applicable, did the Accenture analyst review proper documentation?                         15 points
    4. If applicable, did the Accenture analyst properly address any payments                        15 points
       that exceeded the reimbursement threshold for a specific line item?

9
  The exact name and scope of the business unit that conducts Accenture oversight varies and has changed
multiple times during this evaluation. In June 2013, these oversight duties were transferred to Credit Business
Operation Solutions. For historical consistency, however, this report will refer to CPM-Ops as the business unit
tasked with Accenture oversight.




                                 OIG  EVL-2013-012  September 18, 2013                                           13
   5. Were the Accenture analyst’s internal comments sufficient to support             5 points
      and explain the disposition of the claim?
The reviewer records the answers electronically and makes an annotation in a comment box
each time an Accenture analyst makes an error. If the Fannie Mae reviewer discovers an
overpayment, the reviewer records the dollar amount and refers the claim to another business
unit, which attempts to retrieve the overpayment. However, the primary goal of measuring the
process accuracy rate is to grade Accenture’s contractual performance, not to recapture
overpayments to servicers. Additionally, Fannie Mae samples only about 1% of claims for
process accuracy – it has no remedial means to recapture overpayments in the other 99% of
claims.

After the reviewer answers the five process questions, Fannie Mae’s software tracks the total
process points awarded and the total process points possible for each claim. Then, on a
monthly basis, Fannie Mae computes the process accuracy rate by dividing the sum of the
total process points awarded on all reviewed claims in that month by the sum of the total
process points possible in that month.

                                                 Process Points Awarded on All Claims)
        Monthly Process Accuracy Rate =
                                                  Process Points Possible on All Claims)


                       FIGURE 7. MEASURING THE PROCESS ACCURACY RATE

  Fannie Mae draws a sample of manually processed claims


         Fannie Mae reviewers score the five process accuracy questions

                Fannie Mae computes the process accuracy rate monthly to measure
                Accenture's contractual compliance



This contextual information is critical to OIG’s second and third findings, below.

   b. Identifying Data Trends

Additionally, CPM-Ops is responsible for designing and implementing a red flag system for
servicer claims. Yet, no such system exists today.

In this context, red flags should identify reimbursement trends and data anomalies that
adversely impact Fannie Mae financially. They signal that something unusual may need to be
further investigated and can provide warning signs of fraud. They also can be indicative of an


                            OIG  EVL-2013-012  September 18, 2013                               14
unclear Fannie Mae procedure or a lack of training in a particular area. Figure 8 provides
various examples that could raise red flags.

                         FIGURE 8. EXAMPLES OF TRENDS AND DATA ANOMALIES

       Trend or Anomaly                                            Description
                                   Submission of more than one claim for the same service on the same
 Duplicate Payment
                                   loan
                                   Submission outside of expected service months or seasons (e.g., pool
 Odd Timing
                                   service in winter months in northern states)
                                   Excessive claim submission or claim submission occurring more than
 Unusual Frequency                 expected (i.e., more than one submission for a service that is generally
                                   performed once per the life of a loan)
 Overpayment Request               Frequent claim submission for payment amounts above tolerances


     Internal Recognition of Deficiencies in Fannie Mae’s Oversight

Fannie Mae’s Internal Audit Division (Internal Audit) evaluates the effectiveness of controls
implemented by various internal business units. Over the past three years, Internal Audit has
identified numerous deficiencies within Fannie Mae’s servicer reimbursement operations.
These deficiencies were documented in the three most recent audits of CPM-Ops.10 The
summaries of these audits below are not exhaustive; they are intended to illustrate the
seriousness of Internal Audit’s findings.

In 2010, Internal Audit rated the controls surrounding Fannie Mae’s servicer reimbursement
operations “unsatisfactory” – the lowest possible rating. At that time, Fannie Mae employees
performed manual claim processing in-house. Notably, the audit found that Fannie Mae had
no red flag system and lacked the controls necessary to prevent duplicate payments of the
same claim.

The 2011 audit rated Fannie Mae’s servicer reimbursement operations “needs improvement.”
The audit was performed while Fannie Mae was transitioning manual processing
responsibilities to Accenture. The audit reported that, although Fannie Mae was aware of
overpayments to servicers, it had performed no analysis to quantify their potential impact.
Moreover, the audit noted that Accenture was not satisfying its contractual SLAs.


10
  The audits were conducted in October 2010, December 2011, and November 2012. The exact scope of these
audits is broader than Fannie Mae’s CPM-Ops business unit. For example, the scope of the 2012 audit included
two operational groups – CPM-Ops and Single Family REO Accounting. However, a significant amount of
each internal audit focuses on the deficiencies found within CPM-Ops processes and controls. Additionally,
several of Internal Audit’s findings were self-identified by Fannie Mae management.




                                OIG  EVL-2013-012  September 18, 2013                                        15
In 2012, Internal Audit downgraded Fannie Mae’s servicer reimbursement operations rating
to “unsatisfactory.” It reported that controls “related to processes performed by CPM-Ops are
unsatisfactory and do not provide reasonable assurance that risks are being managed
effectively.” Similar to 2010, Internal Audit concluded that Fannie Mae had yet to establish a
recurring and sustainable process to identify red flags, and there was no formal process to
identify the source of duplicate payments. Additionally, the audit noted that the delegations of
authority between Fannie Mae business units were insufficient.




                            OIG  EVL-2013-012  September 18, 2013                                16
FINDINGS .................................................................................

1. Prior to 2013, FHFA Oversight of Fannie Mae’s Servicer Reimbursement Operations
   Was Minimal; Recently, FHFA Has Bolstered Its Monitoring

As Fannie Mae’s regulator, FHFA is responsible for ensuring that it operates in a safe and
sound manner. To ensure the safety and soundness of Fannie Mae’s operations, FHFA
monitors the quality of its risk management systems through three supervisory activities:
(1) targeted examinations, (2) continuous supervision, and (3) supervisory analyses. A
targeted examination is an in-depth evaluation of a specific risk or risk management system.
It is FHFA’s most thorough risk assessment product. After the completion of an exam, FHFA
drafts a conclusion letter that communicates and supports the exam’s findings. Continuous
supervision encompasses various activities that are designed to monitor and analyze trends or
emerging risks. Supervisory analyses involve research efforts to enhance FHFA’s general
understanding of a specific risk.

       FHFA Oversight Prior to 2013

Prior to 2013, FHFA oversight of Fannie Mae’s servicer reimbursement operations was
limited. Despite Fannie Mae Internal Audit’s findings, no targeted examinations were
performed. There were only two instances in which FHFA independently analyzed the risks
associated with Fannie Mae’s claim reimbursement operations. First, FHFA engaged
Navigant Consulting, Inc. (Navigant), an independent contractor, to perform an operational
risk assessment of Fannie Mae’s REO vendor reimbursement operations. Subsequent to the
Navigant report, FHFA performed continuous supervision of both the servicer and REO
vendor reimbursement operations.

       a. Navigant Report

In January 2012, Navigant completed an analysis of the operational risk associated with
Fannie Mae’s REO vendor reimbursement operations. However, its scope did not include
Fannie Mae’s servicer reimbursement operations; thus, it does not constitute direct oversight
of those operations.

Nevertheless, the report warrants consideration. It identified several key deficiencies in
Fannie Mae’s REO vendor reimbursement operations.11 Therefore, at a minimum, FHFA

11
     The following were among Navigant’s findings:
          System limitations constrain Fannie Mae’s ability to provide adequate REO vendor oversight;
          Preventative fraud controls need improvement in order to fully mitigate the risk of erroneous
           or fraudulent payments; and



                                   OIG  EVL-2013-012  September 18, 2013                                 17
knew that Fannie Mae’s REO vendor reimbursement operations – which are similar to its
servicer reimbursement operations – contained deficiencies that exposed it to serious
operational risk. Yet, at that time, FHFA did not engage a consultant nor did its examiners
conduct a targeted examination of the servicer reimbursement operations. Rather, FHFA
proceeded with limited continuous supervision.

   b. FHFA’s Continuous Supervision

According to FHFA’s Division of Enterprise Regulation, FHFA examiners were performing
continuous supervision of Fannie Mae’s claim reimbursement operations in 2012. However,
FHFA provided OIG with little evidence of this continuous supervision. The only document
OIG received in connection with FHFA’s oversight activities was a summary report of Fannie
Mae internal documentation that was never finalized. It was drafted in May 2012 by two
examiners who had visited Fannie Mae’s claim reimbursement operations in Dallas, Texas.
The draft’s primary purpose was to further develop FHFA’s understanding of Fannie Mae’s
claim reimbursement operations. The draft, however, contains minimal substantive
information regarding servicer reimbursement operations; instead, it is a high-level, nine-page
review of several topics and largely consists of charts summarizing Fannie Mae’s internal
statistics. OIG concludes that this draft is more akin to a supervisory analysis than continuous
supervision. Regardless, these actions do not constitute rigorous oversight of a multi-billion
dollar operation that consistently received poor ratings from Internal Audit and presents a risk
of overpaying servicers tens of millions of dollars annually.

   FHFA Oversight in 2013

Subsequent to the high-level review noted above, FHFA left Fannie Mae’s servicer
reimbursement operations largely unchecked until March 2013 – approximately ten months
after OIG initiated this evaluation – when it assigned a recently hired examiner to Fannie
Mae’s claim reimbursement unit. On April 8, 2013, the examiner conducted an initial
interview with the business unit and added it to the continuous supervision portfolio. In May
2013, FHFA initiated a targeted examination of Fannie Mae’s claim processing operations,
which remains open as of the date of this report.

2. Fannie Mae’s Oversight of Accenture’s Claim Processing Focuses on Contractual
   Compliance Rather than Minimizing Overpayments to Servicers

Fannie Mae can help reduce manual processing errors by utilizing the data it collects more
effectively and implementing a red flag system.

      The lack of a centralized management system to monitor complaints against REO vendors
       inhibits Fannie Mae’s ability to identify questionable business practices.



                              OIG  EVL-2013-012  September 18, 2013                              18
   Data Collection and Analysis

As explained above, Fannie Mae reviewers sample Accenture’s work to measure it against the
contractual SLAs. When assessing the SLAs, the reviewers collect a wealth of information
about Accenture’s processing. For example, when computing the process accuracy rate,
Fannie Mae reviewers record which of the five process elements an Accenture analyst failed
(if any), annotate the reason why, and document overpayments and underpayments. However,
Fannie Mae management does not effectively utilize these data to identify error trends.
Rather, the data are used solely to grade Accenture’s processing performance.

Moreover, Fannie Mae reviewers attempt to record every overpayment that they find within
the sampled claims. Yet, Fannie Mae has no systematic method of aggregating these amounts
or projecting them onto the entire claim population. Without quantifying the problem, Fannie
Mae managers cannot make fully informed business decisions or measure their corrective
actions.

Additionally, communication between Fannie Mae business units has been a consistent
impediment to operational improvement. When a Fannie Mae reviewer records an
overpayment, the claim is transferred to another business unit that attempts to recapture it.
However, communication between the two business units is minimal. To date, Fannie Mae
has not performed a cost-benefit analysis to determine whether investing additional money
into or expanding its current overpayment recovery system would be beneficial. Indeed, such
an analysis would be difficult to conduct without improved communications between the two
units.

   Red Flag System

Beginning in 2010, Internal Audit recommended that Fannie Mae implement a red flag
system for servicer reimbursements. To date, Fannie Mae has not established a recurring and
sustainable process to identify red flags for servicer claims. Thus, Fannie Mae is unable to
identify reimbursement trends and data anomalies. This increases Fannie Mae’s susceptibility
to overpayments and potential servicer fraud.

For example, without a red flag system, Fannie Mae is susceptible to duplicate payments. A
duplicate payment occurs when Accenture analysts approve more than one claim for the same
service on the same loan. This results in the servicer being reimbursed twice for one service.
Fannie Mae occasionally receives returned checks from servicers for duplicate payments that
have gone unnoticed. Moreover, in 2012, Fannie Mae reported three operational incidents that
resulted in a total of $20 million in duplicate payments. Although Fannie Mae eventually
recovered the overpayments, these operational incidents highlight the importance of a red flag
system that should, among other things, identify duplicate payments.



                           OIG  EVL-2013-012  September 18, 2013                               19
Notably, Fannie Mae has implemented a red flag system for REO vendor reimbursements.
The REO vendor red flag system entails Fannie Mae reviewers examining a sample of claims.
The reviewers either validate each claim or designate it as a finding. A finding indicates that
at least a portion of the claim should not have been reimbursed. The system generates
monthly reports that include the amount of findings, trends within the findings, and training
opportunities. For example, the October 2012 red flag report observed that 70% of REO
maintenance vendor claims erroneously included maintenance charges for the first month of
services. These charges, which aggregate to $310,800, are non-reimbursable.

By searching for observable trends (such as the one above) within servicer reimbursements,
Fannie Mae may be able to reduce overpayments. Fannie Mae, however, chose to focus its
resources on REO vendors first because their claims are not as complicated as servicer claims.
Presently, Fannie Mae states that it is actively researching and building an analogous red flag
system for servicer reimbursements.

3. Accenture Processing Errors Prompted Fannie Mae to Overpay Servicers
   $89 Million in 2012

Currently, neither FHFA nor Fannie Mae quantifies the magnitude of overpayments or
underpayments to servicers.12 OIG estimated these numbers using Fannie Mae’s 2012 process
accuracy rate data.13 This involved OIG (1) aggregating all overpayments and underpayments
resulting from Accenture errors, (2) interpreting Fannie Mae reviewers’ annotations, and
(3) crosschecking the findings. A thorough description of the methodology can be found in
the Objective, Scope, and Methodology section below.14

OIG estimates that Accenture analysts incorrectly approved 3.1% of servicer reimbursements
in 2012, which prompted Fannie Mae to overpay servicers $89 million. Although these

12
  As used in this report, an “overpayment” is any servicer reimbursement approved by an Accenture analyst
for an ineligible service or fee, a duplicate payment, or services not performed. It also includes instances when
the servicer did not submit proper documentation or correctly follow Fannie Mae procedures, but was
reimbursed nonetheless. Conversely, an “underpayment” is any payment withheld by Fannie Mae that should
have been approved at the time of processing. In either case, the actual amount of the servicer’s claim that is
classified as “over paid” or “under paid” is the portion of the line item that a Fannie Mae reviewer identified as
an overpayment or underpayment. (In contrast to FHFA’s characterization of OIG’s methodology, OIG did not
count “the whole amount” of any claim as an overpayment simply because one line item within the claim was
reimbursed erroneously. See Appendix A.)
13
  As noted above, Fannie Mae collected these data solely to measure Accenture’s processing performance.
The data were not collected to estimate overpayments and underpayments. Further details regarding how
OIG supplemented the process accuracy data to improve the accuracy of its estimation can be found in the
Objective, Scope, and Methodology section below.
14
  Additionally, see Objective, Scope, and Methodology for an improper payment analysis that uses the
Improper Payments Elimination and Recovery Act of 2010 for guidance.




                                  OIG  EVL-2013-012  September 18, 2013                                            20
overpayments may not equate directly to financial harm against Fannie Mae, they represent
a fundamental problem that undermines the reliability and integrity of Fannie Mae’s servicer
reimbursement operations. Given the absence of available data, OIG has been unable to
estimate the amount of financial harm caused by these overpayments (see Appendix A and
Appendix B for a more in depth discussion of financial harm).15

Additionally, OIG estimates Accenture analysts incorrectly denied 0.9% of servicer
reimbursements in 2012, which prompted Fannie Mae to withhold $27 million from servicers.
Simply netting or aggregating overpayments and underpayments is not appropriate due to the
different treatment accorded to each error. For example, when a servicer is underpaid, it can
resubmit another claim to recover any suspected shortfalls. Fannie Mae, however, does not
currently possess the means to discover or recapture overpayments for 99% of
reimbursements that fall outside of its process accuracy sample.




15
   In FHFA’s comments to this report, the Agency asserts that some overpayments may not cause Fannie Mae
financial harm. Although there may be some merit to FHFA’s assertion, the objective of OIG’s evaluation was
to determine whether Fannie Mae – under the conservatorship of FHFA – had a significant amount of
erroneous servicer reimbursements and, if so, whether effective controls were in place to identify and mitigate
such erroneous reimbursements. The objective was not to calculate the overall financial impact of all erroneous
reimbursements (which neither the Agency nor OIG is in a position to do).




                                 OIG  EVL-2013-012  September 18, 2013                                          21
CONCLUSIONS ..........................................................................

Prior to 2013, FHFA conducted limited oversight of Fannie Mae’s servicer reimbursement
operations. Recently, however, it has bolstered its monitoring, and is currently conducting an
examination of the area.

Fannie Mae’s oversight of Accenture’s manual claim processing focuses on measuring
Accenture’s contractual performance rather than minimizing overpayments to servicers.
Although Fannie Mae collects useful data in its sampling process, it does not utilize these data
effectively to reduce payment errors. Additionally, Fannie Mae has yet to implement a red
flag system for servicer reimbursements.

Neither FHFA nor Fannie Mae aggregates the amount of overpayments to servicers that result
from Accenture processing errors. Without attempting to quantify the problem, it is difficult
for Fannie Mae managers to make fully informed business decisions or measure their
corrective actions. OIG estimates that Accenture processing errors prompted Fannie Mae to
overpay servicers $89 million in 2012.




                            OIG  EVL-2013-012  September 18, 2013                                22
RECOMMENDATIONS ...............................................................

OIG recommends that FHFA:

     1. Ensure Fannie Mae takes the actions necessary to reduce Accenture processing errors.
        These actions should include utilizing its process accuracy data in a more effective
        manner and implementing a red flag system.

     2. Require Fannie Mae to:

             a. Quantify and aggregate its overpayments to servicers regularly;

             b. Implement a plan to reduce these overpayments by (i) identifying their root
                causes, (ii) creating reduction targets, and (iii) holding managers accountable;
                and

             c. Report its findings and progress to FHFA periodically.

     3. Publish Fannie Mae’s reduction targets and overpayment findings.

Based on its data analysis, OIG projects that Fannie Mae – if it continues to follow its current
procedures without remediation – will overpay servicers $312 million between July 1, 2013,
and December 31, 2016 – the date when Accenture’s contract expires.16 However, by
following the above recommendations, FHFA and Fannie Mae can reduce overpayments to
servicers and, thereby, reduce waste.

FHFA Response to Recommendations

After this evaluation was drafted, FHFA had the opportunity to review the report and its
recommendations. FHFA’s formal response can be found in Appendix A. The Agency agreed
to implement the first two recommendations. OIG appreciates FHFA’s support for these
recommendations and its plans to implement them.

However, the Agency expressed concerns over OIG’s methodology and disagreed with our
third recommendation – the publication of reduction targets and overpayment findings.



16
  OIG calculated this figure by assuming $89 million in annual overpayments and multiplying it by three and
one-half years (representing the time between July 1, 2013, to December 31, 2016). The figure does not
account for other fluctuating variables (e.g., the housing market, volume of servicer claims, improved
oversight, and improved manual processing procedures).




                                OIG  EVL-2013-012  September 18, 2013                                       23
Prior to finalizing this report, OIG concluded that the methodology used in this evaluation –
which is fully detailed below – is sound and objective.

Additionally, OIG’s third recommendation is intended to improve transparency for U.S.
taxpayers. Because Fannie Mae’s quarterly net worth directly impacts Treasury, OIG believes
that such transparency is appropriate.

For a more thorough response to FHFA’s concerns, see Appendix B.




                            OIG  EVL-2013-012  September 18, 2013                             24
OBJECTIVE, SCOPE, AND METHODOLOGY .................................

The objective of this report was to evaluate Fannie Mae’s servicer reimbursement operations
for delinquency expenses and FHFA’s oversight of those operations. To achieve our
objectives, OIG interviewed various Fannie Mae and Accenture personnel to gather
information about the servicer reimbursement operations. OIG also interviewed FHFA
personnel to assess the extent of its oversight. Additionally, OIG requested and reviewed
numerous documents from Fannie Mae and FHFA.

When possible, OIG attempts to identify and estimate any “funds [that can be] put to better
use.”17 This term includes any reduction in outlays that results from our recommendations. By
following the recommendations outlined above, FHFA will ensure that Fannie Mae reduces
the overpayments made to servicers as a result of Accenture errors.

Estimation Methodology

After conducting our field work, OIG developed the methodology below to estimate the
figures in Finding Three. OIG confined its analysis to Accenture’s manual processing
operations; it did not assess claims that were processed by Fannie Mae’s scripted or automatic
methods. In addition, OIG did not count technical errors (discussed below) in calculating
overpayments and underpayments.

       Technical Errors

OIG recognized that an Accenture analyst can produce two distinct types of errors. The first
occurs when an Accenture analyst approves or denies a claim in the wrong amount. OIG
terms these types of errors overpayments and underpayments, and they are the type of errors
reported in Finding Three. The second type of error occurs when an Accenture analyst
approves or denies a claim in the correct amount, but fails to comply with Fannie Mae’s claim
review procedures. OIG terms these types of errors technical overpayments and technical
underpayments because Fannie Mae reimbursed the servicer in the correct amount, but the
Accenture analyst made a procedural mistake during the claim review. Such technical errors
were not included in determining the $89 million in overpayments or the $29 million in
underpayments discussed in Finding Three.




17
     Inspector General Act of 1978, 5 U.S.C. App. 3 § 5(a)(6).




                                   OIG  EVL-2013-012  September 18, 2013                       25
             FIGURE 9. EXAMPLE OF OVERPAYMENT VERSUS TECHNICAL OVERPAYMENT

                  Error Description                         Reimbursement             Consequence
 An analyst reviews a servicer claim that contains a   Accenture analyst             Overpayment
 reimbursement request for hazard insurance. The       approved the wrong
 hazard insurance request is for an amount that        reimbursement amount.
 exceeds the maximum allowed by the Servicing Guide.
 The analyst erroneously approves the request.
 An Accenture analyst did not record internal          Accenture analyst             Technical
 comments that were sufficient to support and          made the correct              Overpayment
 explain the disposition of the claim.                 reimbursement decision
                                                       but failed to follow Fannie
                                                       Mae’s claim review
                                                       procedures.

   Claim Review Methodology

OIG’s claim review relied on the data Fannie Mae routinely collects to calculate Accenture’s
process accuracy rate. To collect this data, Fannie Mae selects and reviews a random sample
of manually processed claims. Fannie Mae determines the necessary sample size to achieve a
99% statistical confidence level. Then, the Fannie Mae reviewer (1) answers the five process
accuracy questions found on page 13, (2) records an error amount (if the reviewer found any),
and (3) briefly explains the findings in a comment box. In 2012, Fannie Mae reviewed 13,917
servicer claims for process accuracy out of the 1.3 million that were processed.

The process accuracy data, however, were not collected for the purpose of quantifying and
aggregating how much Fannie Mae overpaid servicers as a result of manual processing errors.
Consequently, an isolated examination of these data made estimation difficult, particularly
because the data are maintained at the claim level. To supplement Fannie Mae’s process
accuracy data, OIG requested and received individual line item data from Fannie Mae that
further detailed each claim within the sample.

Of the 13,917 servicer claims that were reviewed for process accuracy, Fannie Mae found that
approximately ten percent – 1,331 – failed at least one of the five process accuracy questions.
To obtain a more granular understanding of the data and exclude technical errors, OIG
reviewed the 1,331 failed claims and their corresponding line item data. In doing so, the OIG
reviewer primarily read and interpreted the Fannie Mae reviewer’s comments. This is where
the most instructive information was found. Then, the OIG reviewer examined various claim
and line item data to help interpret the comments. For example, if the Fannie Mae reviewer
enumerated an error amount, then the OIG reviewer used that number for guidance in
identifying any errors.



                             OIG  EVL-2013-012  September 18, 2013                                26
To differentiate the errors that resulted in an incorrect reimbursement amount (overpayments
and underpayments) from the technical errors (technical overpayments and technical
underpayments), OIG used the Fannie Mae reviewer’s answers to the five process accuracy
questions. If the Fannie Mae reviewer determined that the Accenture analyst did not make the
correct reimbursement decision (Question 1), then OIG presumed that the error was an
overpayment or underpayment, unless the data indicated otherwise. If the Fannie Mae
reviewer determined that the Accenture analyst made the correct reimbursement decision but
failed one of the remaining four process questions, then OIG presumed that the error was a
technical overpayment or technical underpayment.

After a full examination of the claim and line item data, the OIG reviewer attempted to
categorize and quantify an Accenture error into one of four categories: (1) overpayment,
(2) underpayment, (3) technical overpayment, and (4) technical underpayment. If the OIG
reviewer determined there was not enough data to do so, then the claim was categorized as
ambiguous. OIG did not include any errors categorized as ambiguous in its projections. The
following flow chart demonstrates OIG’s claim review process.

                          FIGURE 10. OIG CLAIM REVIEW FLOW CHART



                                    Interpret the Fannie Mae
                                       reviewer's comment


                                    Use relevant data to help
                                    interpret the Fannie Mae
                                       reviewer's comment



                                                     If the data do not contain enough
                          Categorize and                  instructive information to
                         quantify the error            categorize or quantify an error,
                           (if possible)                   then identify the error as
                                                                  Ambiguous



                                      Technical         Technical
 Overpayment      Underpayment                                            Ambiguous
                                     Overpayment      Underpayment




                           OIG  EVL-2013-012  September 18, 2013                             27
After an OIG reviewer finished examining one month’s data, a second OIG reviewer
examined the first reviewer’s decisions for accuracy. If there was a difference of opinion
about how to categorize an error or the quantity of an error, the two reviewers discussed the
issue and resolved the difference.

Finally, as discussed in Finding Three and     FIGURE 11. OIG PROJECTION OF OVERPAYMENTS
illustrated in Figure 11, OIG projected its             AND UNDERPAYMENTS IN 2012
findings onto the 2012 claim population.
In that year, Fannie Mae made a total of                     Overpayments    Underpayments
$2.9 billion in servicer reimbursements      Projection          $89 million     $27 million
based on Accenture’s manual processing.      Error Rate                3.1%           0.9%
OIG determined that 3.1% of these
reimbursements, $89 million, were overpayments and 0.9%, $27 million, were
underpayments.

Improper Payment Analysis

OIG also conducted an improper payment analysis using guidance from the Improper
Payments Elimination and Recovery Act of 2010 (IPERA).18 Although Fannie Mae is not
obligated to comply with IPERA, OIG used the statute as a guide for best practices.

IPERA directs federal agencies within the executive branch to estimate the annual amount of
gross improper payments for every program.19 The Office of Management and Budget, the
agency charged with issuing IPERA guidance, defines an improper payment as any payment
that should not have been made or that was made in an incorrect amount under contractual
requirements.20 Incorrect amounts are overpayments or underpayments, including
inappropriate approvals and denials – even if the amount reimbursed or withheld was correct.
Thus, an improper payment includes technical overpayments and technical underpayments,
which were not included in Finding Three.

Using the IPERA methodology, OIG estimates that Accenture processing errors produced
$140 million in improper payments in 2012. This number is the sum of the projected
18
     Pub. L. No. 111-204.
19
  See Office of Management and Budget Memorandum M-11-16, Issuance of Revised Parts I and II to
Appendix C of OMB Circular A-123, at 5 (Apr. 14, 2011). A program includes any credit program. Id. at 4.
20
   Additionally, the Office of Management and Budget’s guidance states that improper overpayments and
improper underpayments should be summed in an agency’s estimation of improper payments. A program’s
susceptibility is considered significant if its improper payments (1) exceed $10,000,000 and are disbursed at an
improper payment rate greater than 1.5% or (2) exceed $100,000,000. If an agency estimates that a program’s
susceptibility to improper payments is significant, then the agency must (1) obtain a statistically valid projection
of the program’s improper payments and (2) implement a plan to reduce them.




                                  OIG  EVL-2013-012  September 18, 2013                                              28
overpayments ($89 million), underpayments ($27 million), technical overpayments ($10
million), and technical underpayments ($14 million) – the four categories noted in Figure 12;
$140 million equates to an improper payment rate of 4.8% based on the $2.9 billion in
manually processed servicer reimbursements in 2012.

                         FIGURE 12. OIG PROJECTION OF IMPROPER PAYMENTS IN 2012
                                                                                                         Total
                                                              Technical          Technical             Improper
                      Overpayments       Underpayments      Overpayments       Underpayments           Payments
 Projection               $89 million        $27 million        $10 million         $14 million        $140 million
 Error Rate                      3.1%                0.9%              0.4%                0.5%              4.8%21


OIG concludes that IPERA would classify this level of susceptibility to improper payments as
significant because Fannie Mae’s gross improper payments for manually processed servicer
reimbursements exceed $10 million and a rate of 1.5%.22

Authority

This evaluation was conducted under the authority of the Inspector General Act of 1978, as
amended, and is in accordance with the Quality Standards for Inspection and Evaluation
(January 2012), which was promulgated by the Council of the Inspectors General on Integrity
and Efficiency. These standards require OIG to plan and perform an evaluation that obtains
evidence sufficient to provide reasonable bases to support the findings and recommendations
made herein. OIG believes that the findings and recommendations discussed in this report
meet these standards.

The performance period for this evaluation was from May 2012 to May 2013.




21
     Figures do not sum precisely due to rounding.
22
  Additionally, this level of improper payments qualifies as significant under IPERA’s other criteria: a gross
improper payment amount in excess of $100 million.




                                   OIG  EVL-2013-012  September 18, 2013                                        29
APPENDIX A .............................................................................

FHFA’s Comments on OIG’s Findings and Recommendations




                         OIG  EVL-2013-012  September 18, 2013                       30
OIG  EVL-2013-012  September 18, 2013   31
APPENDIX B..............................................................................

OIG’s Response to FHFA’s Comments

After this evaluation was drafted, FHFA had the opportunity to review the report and its
recommendations. The Agency (1) expressed concerns over OIG’s methodology and
(2) disagreed with OIG’s third recommendation.

   Methodology Concerns

Although OIG appreciates FHFA’s perspective and thoughtfulness, we stand behind this
report’s findings and recommendations.

In its official comments, the Agency stated that “FHFA-OIG counted the whole amount of
the payment and not just the portion representing the overpayment.” This statement is not
accurate. As noted in the Objective, Scope, and Methodology section, OIG carefully
conducted our analysis at the line item level. Furthermore, OIG only counted the dollar
amount within each line item that the Fannie Mae reviewer identified as an overpayment. OIG
did not count “the whole amount” of any claim as an overpayment simply because one line
item within the claim was approved erroneously.

Additionally, FHFA commented that our “analysis focuses on the gross payment amounts
processed with some type of technical [emphasis added] inaccuracy.” The $89 million of
overpayments noted in Finding Three, however, are not merely the result of a “technical”
problem. They are the direct result of Accenture analysts erroneously approving servicer
claims that are not in compliance with the Servicing Guide, many of which result in Fannie
Mae paying for services that are non-reimbursable.

Finally, FHFA stated that Fannie Mae’s actual financial harm is “substantially smaller” than
$89 million. The following scenario illustrates this concept:

       A servicer submits a claim for reimbursement to Fannie Mae without the
       documentation required by the Servicing Guide. The Accenture analyst does
       not notice and erroneously approves the claim, which results in Fannie Mae
       reimbursing the servicer. Although the analyst approved a claim that was not
       in compliance with the Servicing Guide, it is conceivable that, if asked, the
       servicer could provide documentation for the claim. Thus, in this scenario,
       Fannie Mae would have suffered no actual financial harm even though the
       Accenture analyst erroneously approved the claim.




                           OIG  EVL-2013-012  September 18, 2013                             32
OIG acknowledges that there may be some merit to this viewpoint. However, OIG decided
not to incorporate it for the following reasons.

First, Fannie Mae currently has no formal system to track which overpayments cause financial
harm and which do not. Therefore, even if OIG agreed with this alternative viewpoint, we
would have no reliable means of measuring which Accenture errors ultimately cause financial
harm to Fannie Mae. This point further emphasizes the importance of OIG’s first and second
recommendations.23

More importantly, Accenture’s inability to accurately process claims represents a serious
problem that undermines the reliability and integrity of Fannie Mae’s servicer reimbursement
operations. In the hypothetical above, the Accenture analyst approved a claim that should
have been denied at the time of processing.

Finally, OIG believes that FHFA’s and Fannie Mae’s remedial efforts should be focused on
risk management processes that lead to an improvement in Accenture’s front-end performance
– before the Accenture analyst makes an error. This approach will improve Accenture’s
processing accuracy for the entire population of servicer claims. Conversely, FHFA’s
alternative perspective focuses on back-end corrections (i.e., performing additional analysis
to determine the percentage of Accenture errors that theoretically would have no financial
impact after a Fannie Mae overpayment). This approach centers on how much of an
overpayment Fannie Mae should recover after an error occurs. As noted, however, any
recovery of overpayments identified on the back-end would be limited to the 1% sample and
have no effect on the 99% of claims not sampled.

       OIG’s Third Recommendation

FHFA disagreed with OIG’s third recommendation: to publish Fannie Mae’s reduction targets
and overpayment findings. As noted above, OIG recognizes that FHFA has no legal
obligation to publish the payment error rate or any action taken to reduce it. OIG made this
recommendation to improve transparency for U.S. taxpayers as Fannie Mae continues to be in
a conservatorship overseen by FHFA.




23
     FHFA agreed to OIG’s first and second recommendations.




                                 OIG  EVL-2013-012  September 18, 2013                        33
ADDITIONAL INFORMATION AND COPIES .................................


For additional copies of this report:

      Call: 202–730–0880
      Fax: 202–318–0239
      Visit: www.fhfaoig.gov



To report potential fraud, waste, abuse, mismanagement, or any other kind of criminal or
noncriminal misconduct relative to FHFA’s programs or operations:

      Call: 1–800–793–7724
      Fax: 202–318–0358
      Visit: www.fhfaoig.gov/ReportFraud
      Write:
                FHFA Office of Inspector General
                Attn: Office of Investigation – Hotline
                400 Seventh Street, S.W.
                Washington, DC 20024




                             OIG  EVL-2013-012  September 18, 2013                       34