OFFICE OF FEDERAL HOUSING ENTERPRISE OVERSIGHT NEWS …€¦ · 401(h), Pub. L .101–73, 103 Stat. 357. 2. Section 360.7 is added to part 360 to read as follows: §360.7 Post-insolvency

OFFICE OF FEDERAL HOUSING ENTERPRISE OVERSIGHT

NEWS RELEASE

1700 G STREET N.W., WASHINGTON, D.C. 202.414.6922 FAX 202.414.3823

FOR IMMEDIATE RELEASE Contact: Stefanie MullinTuesday, December 11, 2001 202.414.6921

www.ofheo.gov

OFHEO ANNOUNCES PROPOSED IMPROVEMENTS TORISK-BASED CAPITAL RULE

(revised 12/18 to include actual language as printed in Federal Register)

WASHINGTON, D.C. -- Armando Falcon, Jr., Director of the Office of Federal Housing EnterpriseOversight (OFHEO), safety and soundness regulator of Fannie Mae and Freddie Mac (theEnterprises), today announced proposed improvements to OFHEO's final risk-based capital rulepublished September 13, 2001 in the Federal Register.

"These modifications will improve upon the ability of the rule to closely tie capital to risk. A dynamic,state-of-the-art rule such as this will necessarily undergo periodic revision to reflect changing riskprofiles and better information about the nature of risks," said Falcon. "We continuously research andanalyze risk characteristics of the Enterprises. That ongoing work has now resulted in the identificationof several areas where we can improve upon treatments currently in the rule. I assured Congress inSeptember that we would move expeditiously to implement such improvements."

Interested parties will have 30 days to comment from the date of publication in the Federal Register.Comments will be published on the OFHEO web site at www.ofheo.gov. The proposed rule wassubmitted to the Office of Management and Budget (OMB) Friday, Nov. 16 and cleared today,Tuesday, Dec. 11. OFHEO sent the proposed regulation to the Federal Register today (attached).

The proposed rule includes the following improvements:

Modification to Counterparty Haircuts

-The proposed rule would modify the calculation of counterparty haircuts to take loss severities intoaccount explicitly and would extend the phase-in period for investment grade counterparties from fiveyears to 10 years. These changes would tie haircuts more closely to the historical experience of bondissuers during stressful periods and phase them in evenly during the 10-year stress period.

-The proposed rule would reduce the differential in haircuts between AAA- and AA-ratedcounterparties to more closely reflect historical experience.

(more)

Multifamily Loans

-The proposal would make a number of changes to the multifamily default model, loss severity param-eters, and prepayment speeds that refine the measurement of risk in multifamily loans. The proposedchanges would reflect more accurately the differential risks between fixed rate and adjustable ratemortgages as well as the costs and recoveries associated with foreclosures. The proposal on prepay-ment speeds would more accurately reflect prepayment penalties.

Funding

-The proposed rule would refine funding rules to provide a more realistic picture of funding costs in astressful period by altering the target mix of long- and short-term debt to be maintained during the stressperiod from 50/50 to the Enterprise's actual mix of long- and short-term debt at the start of the stressperiod. The proposal would also add a 10-basis point premium to the cost of an Enterprise's newlyissued debt during the last nine years of the stress period, reflecting the impact of the stress on anEnterprise's cost of funds.

Technical and Clarifying Changes

-The proposal makes a number of other technical and clarifying changes that refine the measurement ofrisk in complex instruments, such as derivative contracts, foreign currency swaps, and instruments withcall options, and eliminates double counting and distortion in calculations.

In light of the proposed changes, OFHEO will now be using the First Quarter 2002 numbers tocalculate how the Enterprises will fare under the risk-based capital rule. OFHEO will publishthis information in June 2002.

###

Proposed Changes to the Risk-Based Capital Rule

• Modifications to Counterparty Haircuts

o Incorporates loss severities into haircut calculations for all counterparties by multiplying default rates by loss severity rates.

• Sets default rates for all counterparties equal to existing maximum haircuts for non-derivative counterparties, but lowers the default rate for AA-rated counterparties from 15 to 12.5 percent.

• Sets loss severity rates at 70 percent for non-derivative counterparties and 10 percent for derivative counterparties.

o Extends the phase-in period for counterparty haircuts from five to ten years.

o Provides for netting of swap payments and receipts at the counterparty level rather than the individual contract level. Provides an interim reduction in haircuts for derivative counterparties until netting is implemented.

o Permits a higher rating than BBB for unrated seller-servicers who participate in delegated underwriting and servicing programs when loss-sharing obligations are supported by a fully funded reserve deemed adequate by OFHEO.

• Modifications to Treatment of Multifamily Loans

o Makes technical adjustments to the treatment of ARM loans to reduce

excessive predicted losses on these loans in the up-rate scenario. o Modifies a variable in the model that increases defaults when property

cash flows are projected to be negative, moderating the model�s sensitivity to declines in cash flow.

o Makes a technical correction to the starting vacancy rate in the first month of the stress period.

o Re-estimates the loss severity rates applied to multifamily loans (which are currently based solely on Freddie Mac data from the 1980s) based upon a larger and more diverse sample of foreclosed properties from both Enterprises.

o Takes account of prepayment penalties by specifying no prepayments on loans where such penalties are effective. The rule provides for no prepayments in the up-rate scenario, but specifies that two percent of these loans prepay in the down-rate scenario.

• Additional Operational and Technical Changes

o Adds a ten-basis-point premium to the cost of an Enterprise�s newly issued debt during the last nine years of the stress period.

o Specifies that the stress test will issue new debt in amounts that maintain or move toward the mix of long- and short-term debt that exists at an Enterprise at the beginning of the stress test. The rule currently specifies a 50/50 mix.

o Corrects an overstatement of float income in the current rule. o Applies haircuts to foreign currency swaps. o Models American options more precisely. o Clarifies the appropriate House Price Index value to apply to loans

originated since the date of the most recent published Index. o Corrects a technical omission of a flag needed to distinguish loans without

prepayment penalties or yield maintenance provisions from those with in the calculation of the duration of prepayment penalties for loan groups.

1

TABLE 3–31—STRESS TEST MAXIMUM HAIRCUT BY RATINGS CLASSIFICATION

Ratings Classification

DerivativeContract

Counterpartiesprior to

Implementation ofNetting

DerivativeContract

Counterpartiesafter

Implementation ofNetting

Non-DerivativeContract

Counterparties orInstruments

Number ofPhase-inMonths

Cash 0 0% 0% N/A

AAA 0.3% 0.5% 3.5% 120

AA 0.75% 1.25% 8.75% 120

A 1.2% 2% 14% 120

BBB 2.4% 4% 28% 120

Below BBB andUnrated

100% 100% 100% 1

65146 Federal Register / Vol. 66, No. 243 / Tuesday, December 18, 2001 / Proposed Rules

due to changes in economic conditionsduring the life of the receivership.

Finally, the proposed rule providesthat post-insolvency interestdistributions would be calculated usinga simple interest method, rather than acompound interest method. The simpleinterest method is proposed because itappears to provide a reasonable amountof interest to compensate receivershipcreditors for the time value of moneyowed from the time the receivership isestablished until dividend payments arereceived.

III. Request for Public Comment

The FDIC hereby solicits commentson all aspects of the proposed rule, andspecifically whether post-insolvencyinterest should be paid according to theorder of priority described in thenational depositor preference statute oralternatively pro rata to all creditorsregardless of priority.

IV. Paperwork Reduction Act

The proposed rule will not involveany collection of information under thePaperwork Reduction Act (44 U.S.C.3501 et seq.). Consequently, noinformation has been submitted to theOffice of Management and Budget forreview.

V. Regulatory Flexibility Act

Pursuant to section 605(b) of theRegulatory Flexibility Act (5 U.S.C. 601et seq.) the FDIC certifies that theproposed rule will not have a significanteconomic impact on a substantialnumber of small entities. The proposedrule will only apply to FDIC-administered receiverships establishedafter the effective date of the rule, andit does not impose new reporting,recordkeeping or other compliancerequirements on receivership creditors.The proposed rule continues the FDIC’sexisting practice of making post-insolvency interest distributions tocreditors holding proven claims insurplus receiverships prior to makingdistributions to equityholders, based ontheir equity interests, in a failed insureddepository institution. In addition, theproposed rule will provide interestedparties, including small entities, withgreater certainty in future FDIC-administered receiverships byestablishing a single uniform interestrate and method for making post-insolvency interest distributions.Accordingly, the Act’s requirementsrelating to an initial regulatoryflexibility analysis are not applicable.

VI. The Treasury and GeneralGovernment Appropriations Act,1999—Assessment of FederalRegulations and Policies on Families

The FDIC has determined that theproposed rule will not affect familywell-being within the meaning ofsection 654 of the Treasury and GeneralGovernment Appropriations Act,enacted as part of the OmnibusConsolidated and EmergencySupplemental Appropriations Act of1999 (Public Law 105–277, 112 Stat.2681).

List of Subjects in 12 CFR Part 360Banks, banking, Savings associations.For the reasons set forth in the

preamble, the FDIC Board of Directorsproposes to amend 12 CFR part 360 asfollows:

PART 360—RESOLUTION ANDRECEIVERSHIP RULES

1. The authority for part 360 is revisedto read as follows:

Authority: 12 U.S.C. 1821(d)(1),1821(d)(10)(C), 1821(d)(11), 1821(e)(1),1821(e)(8)(D)(i), 1823(c)(4), 1823(e)(2); Sec.401(h), Pub. L .101–73, 103 Stat. 357.

2. Section 360.7 is added to part 360to read as follows:

§ 360.7 Post-insolvency interest.(a) Purpose and scope. This section

establishes rules governing thecalculation and distribution of post-insolvency interest to creditors withproven claims in all FDIC-administeredreceiverships established after [effectivedate of final rule].

(b) Definitions—(1) Equityholder. Theowner of an equity interest in a faileddepository institution, whether suchownership is represented by stock,membership in a mutual association, orotherwise.

(2) Post-insolvency interest. Interestcalculated from the date thereceivership is established on provencreditor claims in receiverships withsurplus funds.

(3) Post-insolvency interest rate. Forany calendar quarter, the couponequivalent yield of the average discountrate set on the three-month Treasury billat the last auction held by the UnitedStates Treasury Department during thepreceding calendar quarter, andadjusted each quarter thereafter.

(4) Principal amount. The provenclaim amount and any interest accruedthereon as of the date the receivershipis established.

(5) Proven claim. A claim that isallowed by a receiver or upon which afinal non-appealable judgment has beenentered in favor of a claimant against a

receivership by a court with jurisdictionto adjudicate the claim.

(c) Post-insolvency interestdistributions. (1) Post-insolvencyinterest shall only be distributedfollowing satisfaction by the receiver ofthe principal amount of all creditorclaims.

(2) The receiver shall distribute post-insolvency interest at the post-insolvency interest rate prior to makingany distribution to equityholders. Post-insolvency interest distributions shallbe made in the order of priority set forthin section 11(d)(11)(A) of the FederalDeposit Insurance Act, 12 U.S.C.1821(d)(11)(A).

(3) Post-insolvency interestdistributions shall be made at such timeas the receiver determines that suchdistributions are appropriate and only tothe extent of funds available in thereceivership estate. Post-insolvencyinterest shall be distributed on theoutstanding balance of a proven claim,as reduced from time to time by anyinterim dividend distributions, from thedate the receivership is established untilsuch time as the principal amount of aproven claim has been distributed butnot thereafter.

(4) Post-insolvency interest shall bedetermined using a simple interestmethod of calculation.

By order of the Board of Directors.Dated at Washington, DC this 10th day of

December, 2001.Federal Deposit Insurance Corporation.Robert E. Feldman,Executive Secretary.[FR Doc. 01–31162 Filed 12–17–01; 8:45 am]BILLING CODE 6714–01–P

DEPARTMENT OF HOUSING ANDURBAN DEVELOPMENT

Office of Federal Housing EnterpriseOversight

12 CFR Part 1750

RIN 2550–AA23

Risk-Based Capital

AGENCY: Office of Federal HousingEnterprise Oversight, HUD.ACTION: Proposed regulation.

SUMMARY: The Office of Federal HousingEnterprise Oversight (OFHEO) isproposing to amend Appendix A toSubpart B of 12 CFR Part 1750 Risk-Based Capital. The effect of theseamendments would be to modifyprovisions relating to counterpartyhaircuts, multifamily loans, andrefunding and to make several technical

VerDate 11<MAY>2000 19:13 Dec 17, 2001 Jkt 197001 PO 00000 Frm 00006 Fmt 4702 Sfmt 4702 E:\FR\FM\18DEP1.SGM pfrm04 PsN: 18DEP1

65147Federal Register / Vol. 66, No. 243 / Tuesday, December 18, 2001 / Proposed Rules

1 Risk-based Capital, 66 FR 47730 (September 13,2001).

2 For purposes of this proposal, Moody’s InvestorsService provided information on ‘‘LetterCumulative Default Rates (from 01/01/29 to 01/01/31)’’ on October 16, 2001. Data may be obtainedfrom Moody’s Investors Service by contacting Mr.Steve Liebling at Liebling@Moody’s.com.

3 W. Braddock Hickman, ‘‘Corporate BondQuality and Investor Experience,’’ 190 NationalBureau of Economic Research (1958).

adjustments and corrections. Theseamendments are intended to refine thestress test model to tie capital moreclosely to risk.DATES: Written comments must bereceived by January 17, 2002.ADDRESSES: Send written commentsconcerning the proposal to AlfredPollard, General Counsel, Office ofFederal Housing Enterprise Oversight,Fourth Floor, 1700 G Street, NW.,Washington, DC 20552. Writtencomments may also be sent to Mr.Pollard by electronic mail [email protected]. OFHEOrequests that written commentssubmitted in hard copy also beaccompanied by the electronic versionin MS Word or in portable documentformat (PDF) on 3.5″ disk.FOR FURTHER INFORMATION CONTACT:Edward J. Szymanoski, Acting AssociateDirector, Office of Risk Analysis andModel Development, telephone (202)414–3763 (not a toll-free number), orDavid Felt, Associate General Counsel,telephone (202) 414–3750 (not a toll-freenumber), Office of Federal HousingEnterprise Oversight, Fourth Floor, 1700G Street, NW., Washington, DC 20552.The telephone number for theTelecommunications Device for the Deafis (800) 877–8339.SUPPLEMENTARY INFORMATION:

Comments

The Office of Federal HousingEnterprise Oversight (OFHEO) invitescomments on the proposed regulationand will take all comments intoconsideration before issuing the finalregulation. Copies of all comments willbe posted on the OFHEO internet website at http://www.ofheo.gov. Inaddition, copies of all commentsreceived will be available forexamination by the public at the Officeof Federal Housing EnterpriseOversight, Fourth Floor, 1700 G Street,NW., Washington, DC 20552.

Background

On September 13, 2001, OFHEOpublished a final regulation setting fortha risk-based capital stress test, (Rule) 1

that is the basis for determining the risk-based capital requirement for theFederally sponsored housingenterprises—Federal National MortgageAssociation (Fannie Mae) and FederalHome Loan Mortgage Corporation(Freddie Mac) (collectively, theEnterprises). The risk-based capitalstress test set forth in the Rule simulatesthe performance of each Enterprise’s

assets, liabilities, and off-balance-sheetobligations under severe credit andinterest rate stress for a period of tenyears (stress period). The stress testprojects rates of default and prepaymentfor the mortgages guaranteed by theEnterprises, as well as cash flows fromthese and other assets, liabilities, andoff-balance-sheet obligations. Usingthese cash flows, the stress testproduces monthly balance sheets for the120 months of the stress period in orderto determine the amount of startingcapital that would be necessary tomaintain positive capital during the ten-year stress period. Thirty percent of theamount of capital so determined is thenadded to that amount to protect againstmanagement and operations risk.

OFHEO continuously seeks toimprove its measurements and formulasto tie capital more closely to risk andworks to ensure that the Rule supportsthe safety and soundness regime createdby Congress. In the preamble to theRule, OFHEO expressed its intention toreview, on an ongoing basis, theoperation of the stress test and itsvarious components and to evaluate theneed for revisions and improvements.Also, OFHEO committed to actexpeditiously to remedy any technicaland operational issues that arise duringthe one-year implementation periodfollowing promulgation. OFHEO is nowproposing to make refinements andtechnical adjustments and corrections tothe Rule to tie capital more closely torisk. Technical changes are included inthis proposal rather than issued as afinal regulation to provide acomprehensive package of changes.

A. Proposed Changes to CounterpartyHaircuts

The Rule gives the Enterprises creditfor cash payments that would bereceived during the stress period fromsecurities and various counterparties,such as mortgage insurance companiesand derivative counterparties. However,because Enterprise counterparties arethemselves likely to be adverselyaffected by the economic conditions ofthe stress period and to default on someor all of their obligations, the stress testdiscounts the value of cash paymentsreceived during the stress period by aspecified percentage, based on thepublic credit rating of the security orcounterparty. The amount by whichcash payments from a counterparty orsecurity are discounted in each monthof the stress period is the haircut. Thespecified haircut percentages increase asthe credit rating declines—the lowerthat rating, the more severe the haircut.In the Rule, the haircuts are phased inover the first five years of the stress

period, except for haircuts for below-investment-grade providers andinstruments, which are applied fully inthe first month of the stress period.

The Rule applies one set of haircutsfor non-derivative counterparties andsecurities, based on analysis ofhistorical bond default rates, and adifferent set of haircuts for derivativecounterparties, reflecting lowerexpected loss severities associated withthe use of strong collateral agreements.To further refine the Rule’s treatment ofhaircuts, OFHEO proposes to improveconsistency between haircuts forderivative counterparties and securitiesand non-derivative counterparties andsecurities by specifying default andseverity rates separately; to extend thephase-in period from five to ten years;to provide for netting of exposures tothe same derivative counterparty; and toprovide for an exception to the BBBhaircut for certain unrated seller/servicers as described in the proposedrule.

Default Rates. OFHEO proposes to usethe Rule’s haircut rates for non-derivative counterparties and securitiesas the cumulative default rates for allcounterparties and securities, but tolower slightly the default rate for AA-rated firms. After re-evaluating thehistorical data on differences inperformance of AA-rated and AAA-rated firms, including data that recentlyhas become available to OFHEO, theRule’s default ratio of three to one(based largely on the average exposureover the past 80 years) appears to bemore than is warranted for a period ofeconomic stress. Data were recentlymade available to OFHEO by Moody’sInvestors Service 2 for the worst annualcohorts of U.S. investment-grade issuerssince 1920, the cohorts formed at thebeginning of 1929, 1930, and 1931. Theaverage 10-year default rate for AA-ratedissuers (12.25 percent) was 2.6 times aslarge as the average default rate forAAA-rated issuers (4.72 percent), andthe ratio for the worst of those years wasonly 2.2. Furthermore, a study ofcorporate bond quality by W. BraddockHickman shows 12-year default rates forthe cohort formed at the beginning of1928 for AA-rated issuers (12.3 percent)to be 1.5 times as large as that for AAA-rated issuers (8.1 percent).3 More recentdata, in relatively favorable economic



4 ‘‘Default Recovery Rates of Corporate BondIssues: 2000,’’ 26 Moody’s Investor’s Service(February 2001).

5 ‘‘Ratings Performance 1997: Stability ofTransition,’’ 3 Standard and Poor’s (August 1998).

6 Hickman, at 460.

7 Hickman, at 119.8 ‘‘Historical Default Rates of Corporate Bond

Issuers, 1920–1996,’’ 12 Moody’s Investor Service(January 1997).

9 Moody’s (2001), at 24–25.

10 These percentages correspond to absolutechanges of 61 and 41 basis points, on average,during the period, but would be less than half asmuch at recent yield levels.

11 Loss severities of counterparty defaults aretypically expressed as percentages of derivativemarket value at the time of default. However, thestress test model reflects such losses as reductionsin net derivative cash flows. For example, in the up-rate stress scenario, after a 75 percent increase ininterest rates, a swap with a market value of zeroat the start of the stress test (i.e., a fixed-pay rateequal to the then-market rate) will have asignificantly increased market value during thestress period. Since short- and long-term rates arethe same in the last nine years of the stress periodin the up-rate scenario, net derivative cash flowsroughly equal the scenario-based change in long-term interest rates multiplied by the notional value,and the market value of the swap is the discountedpresent value of these cash flows. A ten percentreduction in those cash flows thus reflects theimpact on market value of a 7.5 percent change ininterest rates.

circumstances, also show greatersimilarity in the performance of issuersin these two rating categories. However,a partially offsetting factor is thatMoody’s data for both depressioncohorts and averages of all cohorts showthat defaults of AAA-rated issuers thatoccur within 10 years after the cohort isformed occur later in the 10-year periodthan those of AA-rated issuers.

The relationship between AA andAAA defaults is particularly relevantbecause most Enterprise counterpartyand security exposures are either AAA-or AA-rated. An excessive differentialbetween these ratings in the stress testcould create inappropriate businessincentives for the Enterprises. Afterweighing the above considerations,OFHEO proposes to lower thecumulative default rate for AA-ratedcounterparties and securities to 12.5percent (from 15 percent), which will be2.5 times the rate for AAA-ratedcounterparties and securities.

Severity Rates. To further refine riskmeasurement in the stress test, OFHEOproposes to take explicit account ofpotential recoveries in the event ofdefault by introducing a loss severityfactor. Before issuing the Rule, OFHEOreceived mixed comments regardingincorporation of recovery projections fornon-derivative security andcounterparty obligations after default.Such recoveries were not part of theproposed rule, however, and OFHEOdecided not to include them at thattime, pending further consideration.Historically, corporate bond recoverieshave averaged about 40 percent (i.e., a60 percent loss severity rate) over longperiods of time. A study of default andrecovery rates by Moody’s shows anaverage recovery rate of 39 percent overthe past 20 years.4 A study of defaultedbond recoveries by Standard and Poor’sshows an average recovery rate of 44percent from 1981 to 1997.5 TheHickman study shows an averagerecovery rate of 43 percent for largeissues from 1900 to 1943.6 Recoverieson Enterprise holdings of mortgage andother asset-backed securities and onmortgage insurance claims would likelybe substantial also, benefiting from assetvalues in the former case and premiumincome in the latter.

Data on recoveries in unusuallystressful times are less favorable.Hickman reported an average recoveryrate of 34 percent for large issues for

defaults in 1930 to 1943.7 Moody’s hasreported average recovery rate estimatesthat are substantially lower duringrecessions, and fall as low as 20 percentduring the 1930s.8 For 1930 to 1943,Moody’s average was 36 percent,despite higher rates during the latteryears of that period. A somewhat lowerprojection for the stress period used inthe rule is, therefore, appropriate.

All of the recovery studies show somedifferences in recovery rates dependingon the presence or absence of secured orsubordinated status. However, suchstatus is a factor used in determiningratings. Moody’s expressly states thatsecurities with different status may havesimilar probabilities of default, but berated differently in recognition of theeffect of security or subordination onlikely recoveries.9 Thus, a securedinstrument may have a somewhat higherprobability of default than average forits rating, but also have a somewhathigher expectation of recovery.Accordingly, OFHEO proposes tospecify a recovery rate of 30 percent (70percent loss severity rate) for all non-derivative counterparties and securitieswith investment-grade ratings.

OFHEO also proposes to maintain,with alteration, special treatment forderivative counterparty exposures.Current exposures are marked to marketat least weekly, and high qualitycollateral is posted against anysignificant exposures by counterpartieswith less than a AAA rating. TheEnterprises retain the right to requiresubstantial over-collateralization or totransfer the contract to a newcounterparty if a counterparty’s rating islowered to low investment-grade levelsor worse. Thus, the principal risk is thata relatively highly rated counterpartymay fail suddenly and that exposuresrise between the time a contract was lastcollateralized and the time theEnterprise takes action to transfer orreplace the contract. This period may beas much as ten business days.

The credit exposures on fixed-floatinginterest rate swaps and swaptions (thevast majority of Enterprise derivativecontracts) are closely tied to changes inmarket yields of securities withmaturities equal to those of the swap orswaptions. When interest rates rise, anEnterprise’s exposure rises on swaps forwhich it receives the floating-rate sideof the swap. When interest rates fall, anEnterprises’s exposure rises on swapsfor which it receives the fixed-rate side.

To develop loss severity rates fordefaulted derivative contracts, OFHEOexamined changes in Treasury securityinterest rates over periods of tenbusiness days during the past 25 years.For five-year Treasury securities,increases in yields of more than 7.5percent and decreases of more than 5.0percent, respectively, have occurredinfrequently-roughly 1 percent and 4percent, respectively, of the time.10

Thus, severity rates that reflect lossesassociated with yield changes of thesemagnitudes should be reasonablyconservative.

For application in the stress test’scash flow model, OFHEO must translatesuch changes into impacts on netderivative cash flows. During the stressperiod, net derivative cash flows arerelated to changes in the ten-yearTreasury yield-75 percent in the up-ratescenario and 50 percent in the down-rate scenario. For example, in the up-rate scenario, with its flat yield curve,the pay side of a ten-year pay-fixed/receive-floating swap implemented justbefore the start of the stress test wouldremain at its original rate and thereceive side would rise to 175 percentof the original pay-side rate. Thus, theswap would have net annual cash flowsfor the last nine years of the stress testroughly equal to 75 percent of the initialfixed rate used in the swap multipliedby the notional value. This is ten timesthe 7.5 percent market yield change thatmay be associated with losses on aderivative counterparty default in theup-rate scenario. Accordingly, OFHEOproposes to set severity rates forderivative exposures at ten percent.11

OFHEO recognizes that losses couldbe greater than ten percent if interestrates move exceptionally after a suddendefault, or if an Enterprise failed toreplace a contract with a defaultingcounterparty and market yields



12 NPR2 refers to the Second Notice of ProposedRulemaking issued by OFHEO before the Rule. 64FR 18084, 18159 (April 13, 1999).

continued to move unfavorably.However, OFHEO also recognizes thatyield changes near the time of a defaultcould easily be less unfavorable than the7.5 percent increase or 5 percentdecrease contemplated, and some

recoveries beyond the collateral alreadyheld might be available. Thus, OFHEOjudges that a ten percent severity rate forderivatives is adequate.

Haircuts. Under the proposal, haircutswould be determined by multiplying the

default rate for each rating category bythe severity rate. The resulting haircutsthat are proposed are set forth in Table1 below.

TABLE 1—STRESS TEST HAIRCUT BY RATINGS CLASSIFICATION

Ratings ClassificationDerivativeContract

Counterparties


Counterpartiesor Instruments

Cash 0% 0%

AAA 0.5% 3.5%

AA 1.25% 8.75%

A 2% 14%

BBB 4% 28%

Below BBB and Unrated 100% 100%

Phase-In. Under the Rule, haircuts forinvestment-grade counterparties andsecurities are phased-in over the firstfive years of the stress period, so thathaircuts are close to zero in the firstmonth of the stress period and rise totheir maximums in the 60th month,where they remain for the last fiveyears. In effect, all defaults occur withinthe first five years, and later haircuts tocash flows simply reflect theconsequences of previous defaults, asdefaulted counterparties are unable tomeet their obligations. This conservativeapproach takes into account that theinterest rate shocks and house priceshocks all occur in the first half of thestress period. Long-term averagehistorical data show more evenlydistributed defaults over time, butavailable data for especially stressfulperiods (e.g., the 1910s and 1930s) givelittle indication of timing. The recentlyobtained unpublished data fromMoody’s shows that for the worst cohort(starting in the beginning of 1930), only57 percent of ten-year investment-gradedefaults occurred during the first fiveyears. While the principal shocks mayoccur somewhat earlier in the stressperiod than they did for issuers in the1930s, a closer approximation of thehistorical patterns may better reflect theability of most highly rated firms tosurvive severe stresses for many years.Some of those that ultimately fail duringthe stress period may reasonably beexpected to fail during its final years.Accordingly, OFHEO proposes toextend the phase-in period from five

years to ten years for investment-gradecounterparties and securities. Thus, forcredit exposures to firms and securitiesrated BBB and higher, defaults willoccur evenly throughout the stressperiod.

Netting of derivative counterpartyexposures. The Enterprises regularlyenter into derivatives contracts,typically swaps, for debt and portfoliorisk management purposes. Thesecontracts expose the Enterprises to therisk of failure by a derivativecounterparty to perform its obligationsas anticipated by the terms of thecontract. The Enterprises, consistentwith accepted risk management andmarket practice, attempt to mitigatetheir derivative counterparty creditexposure through a number of methods,including the use of master nettingagreements. Master netting agreementsare used by the Enterprises when theyengage in multiple swap transactionswith the same counterparty. A masternetting agreement permits an Enterpriseto determine its aggregate total creditexposure to a particular counterparty bynetting the gains and losses across all ofthe contracts with that counterparty.This approach allows the Enterprises tonet their exposures at the counterpartylevel, rather than netting at theindividual contract level.

In NPR2, OFHEO proposed amethodology to recognize this practiceby modeling the terms of master nettingagreements and then applying specifiedhaircuts to the resulting net amountdue, if any, from each derivatives

counterparty.12 No comments werereceived on the proposal, and the Rule,reflecting OFHEO’s intent to modelmaster netting agreements, did notspecify a change from NPR2. However,due to a technical omission, OFHEO’sintent to model master nettingagreements was not operationalized inthe Rule. Recognition of master nettingagreements would result in a moreaccurate measurement of theEnterprises’ exposure to derivativecounterparties. Further, recognition ofmaster netting agreements is consistentwith OFHEO’s intent to modelEnterprise contracts according to theirrespective terms, and such recognitionallows OFHEO to tie capital to risk withgreater precision. The proposal wouldamend the Rule to model master nettingagreements explicitly, as originallycontemplated in NPR2.

OFHEO notes that this technicalcorrection will require animplementation period to allow fordevelopment and completion of thesoftware changes that will allow OFHEOto model master netting agreements.Therefore, during the implementation ofthe technical correction, OFHEO willrecognize the risk mitigation effects ofsuch agreements by reducing thehaircuts for derivatives contracts. Uponimplementation of the technicalcorrection, maximum haircuts forderivative contract counterparties willbe readjusted and netting bycounterparty will be implemented in thesoftware. The interim treatment willremain effective only for the period



13 The terms ‘‘benchmark region and period’’ referto the regional credit loss experience identified byOFHEO in compliance with the ‘‘Credit Loss’’parameters outlined in Title XIII of the Housing andCommunity Development Act of 1992, Pub. L. No.102–550, known as the Federal Housing EnterprisesFinancial Safety and Soundness Act of 1992 (1992Act), as described in additional detail in NPR2.

14 In the Rule’s single-family default andprepayment models, the level of borrower equity inthe property (property value less mortgage debt) isanalogous to multifamily DCR in that both measurescapture economic stress. The circumstance of asingle-family mortgage borrower having negativeequity is similar to that of a multifamily loan having

required to complete the technicalsoftware modifications necessary to

model master netting agreements. Theinterim and final haircuts for derivative

contract counterparties are as shown inthe Table 2 below:

TABLE 2—STRESS TEST HAIRCUTS FOR DERIVATIVE CONTRACT COUNTERPARTIES


Haircuts forDerivative

Counterpartiesprior to Imple-mentation of

Netting

Haircuts forDerivative

Counterpartiesupon Imple-mentation of

Netting


Cash 0% 0% N/A

AAA 0.3% 0.5% 120

AA 0.75% 1.25% 120

A 1.2% 2.0% 120

BBB 2.4% 4.0% 120

Below BBB and Unrated 100% 100%1

Unrated Seller/servicers. The Ruletreats unrated seller/servicers as BBB-rated counterparties. OFHEO recognizesthat certain unrated seller-servicers towhom underwriting and servicingauthority has been delegated enter intoloss-sharing agreements with theEnterprises and collateralize these loss-sharing obligations with fully fundedreserve accounts pledged to theEnterprise. OFHEO is proposing toamend the Rule to permit a higher ratingthan BBB for these seller-servicers if thefully funded reserve account is equal toor greater than an amount determinedby OFHEO to be adequate to support therisk borne by the seller-servicer underthe loss sharing agreement. Forexample, if the loss-sharing obligation ofa seller-servicer participating in FannieMae’s Delegated Underwriting andServicing (DUS) Program iscollateralized by a fully funded reserveaccount that is equal to or greater thanone percent of the seller-servicer’saggregate unpaid principal balancecovered by the loss-sharing agreement atthe start of the stress test, the rating ofthe issuer of the instrument backing thereserve account may be used, in lieu ofBBB, as the rating of the unrated seller-servicer, except that in no event will therating exceed AA. Determinations of therequired reserve amount and the ratingpermitted would be made on a program-by-program and Enterprise-by-Enterprise basis.

B. Proposed Changes to MultifamilyModel

OFHEO is proposing a number ofchanges to the multifamily defaultmodel, multifamily loss severityparameters, and multifamilyprepayment speeds specified in theRule. Proposed changes to the default

model include (1) a respecification ofexplanatory variables which has theeffects of reducing the model’ssensitivity to debt-service coverageratios (DCRs) falling below one andreducing predicted cumulative defaultrates on adjustable rate mortgages(ARMs) in the up-rate stress test, and (2)an increase to the initial vacancy rateused to update DCR during the stresstest making this rate consistent with thebenchmark region’s vacancy rate fromthe month prior to the start of thebenchmark period.13 OFHEO is alsoproposing changes for the multifamilyloss severity parameters that reflect thecosts, timing, and recoveries associatedwith a larger and more broad-based setof Enterprise foreclosures. The Rulereflects a decision not to model thecomplexities of prepayment premiumsthat may or may not be received by theEnterprises during stressful periodswithout further study. The proposedmultifamily prepayment speeds aremore consistent with that decision thanexisting pre-payment speeds. Eachproposed change is discussed in turn.

Underwater Debt Coverage Ratio flag(UWDCRF). In the Rule, the multifamilydefault model included an UnderwaterDebt Coverage Ratio Flag (UWDCRF),intended to cover the additional defaultrisk posed when the projected debtservice coverage ratio-net operatingincome (NOI) divided by mortgagepayment-falls below one during thestress test. A debt coverage ratio less

than one means that the NOI isinsufficient to cover the requiredmortgage payment, an occurrence thatsuggests a high probability of default.The stress test projects the DCR in eachmonth of the stress period from theprior month’s value by updating NOI,using rent growth rates and rentalvacancy rates that reflect the economicconditions of the benchmark region andperiod, and adjusting mortgagepayments monthly according to the noteterms and the stress test interest ratescenario. When this method is used toproject DCR, the types of loans forwhich the projected DCR falls belowone tend to be fixed rate mortgages(FRMs) that started the stress test witha low DCR and, in the up-rate scenario,most ARM loans, resulting incomparatively high cumulative defaultrates for these loans in the stress test.

OFHEO has found that the UWDCRFadds value to the multifamily defaultmodel by capturing the additional riskof default when NOI is insufficient tocover mortgage payments, but isconcerned that the sensitivity ofpredicted monthly defaults to projectedDCR falling below one may be too great,for two reasons. First, the UWDCRF isan indicator that is only turned on whenDCR is projected to be below one, andis turned off otherwise. There are nofiner gradations for this explanatoryvariable such as those that might becaptured if the projected DCR accountedfor individual property dispersionaround the mean.14 In the application of



a DCR below one because both are associated withincreased likelihood of default. However, in thesingle-family model, negative equity is captured asa probability and enters the model as categoricalvariable having eight possible values. These eightgradations for the probability of negative equityimprove the single-family model by avoiding abruptpredicted transitions from positive to negativeequity. OFHEO is able to calculate the probabilityof negative equity for single-family loans becauseprojected property value changes are based onOFHEO’s House Price Index and its associateddispersion parameters. No similar measures ofdispersion are currently available to projectmultifamily DCR or the probability of DCR fallingbelow one.

15 The Rule includes a New Book ARM flag (NAF)and a New Book Balloon flag (NBLF) as product-type offsets to the New Book flag (NBF), which isa categorical (or dummy) variable that distinguishesbetween ‘‘Old Book’’ loans that were made whenthe Enterprises first entered into the multifamilybusiness (before 1988 for Fannie Mae and before1993 for Freddie Mac) and ‘‘New Book’’ loans madeunder their more recent restructured programs.OFHEO’s research indicates that New Book loanshave shown lower defaults than Old Book loans ingeneral, although the amount of improvementvaries significantly among product types.Specifically, New Book fixed-rate balloon loansoutperformed Old Book fixed-rate balloon loans toa lesser degree than their fixed-rate fully amortizingcounterparts. ARM loan performance differentialswere even smaller. These differences are reflectedin the Rule in the NBLF and NAF offsets to theNBF.

16 This effect is captured in the Rule by the RatioUpdate Flag (RUF). Specifically, the RUF identifiesa subset of New Book loans—those for which theloan-to-value ratio (LTV) and debt-service coverageratio (DCR) have been calculated or delegated tohave been calculated by the Enterprises at loanorigination or for which the LTV and DCR havebeen recalculated or delegated to have been

Continued

the stress test, many multifamily loangroups will have DCRs projected to fallbelow one—some only slightly belowone, while others fall well below one.The additional risk of default may beoverstated for those loan groups withDCRs projected to fall only slightlybelow one by the abrupt transition ofthe UWDCRF variable. Second, evenwhen a multifamily property’s DCRdoes fall below one, only a fraction ofborrowers default, indicating that thosewho do not default may carry theirproperties with funds from othersources for a period of time while theytry to remedy the negative cash flowposition.

For these reasons, OFHEO decided tore-estimate the multifamily defaultmodel with a revised definition of theUWDCRF that turns the flag on onlywhen the DCR is projected to be wellbelow one. As a result of that re-estimation, OFHEO proposes to redefinethe UWDCRF to be equal to one (that is,to turn the flag on) when projected DCRis less than 0.98 (that is, when NOI ismore than two percentage points belowthe mortgage payment), rather thansetting the flag equal to one immediatelywhen the projected DCR falls below one.The re-estimated multifamily defaultmodel has a slightly lower coefficient onUWDCR, and the coefficients for theother explanatory variables do notchange materially. Simulations usingthe revised UWDCRF definition resultin lower predicted default rates forARMs in the up-rate scenario and forFRMs with low initial DCR in bothscenarios, making the model lesssensitive to the UWDCRF than theexisting model. The revised definitiondoes not substantially affect thepredicted default rates for most FRMs orfor ARMs in the down-rate scenario.OFHEO believes the respecified modelmore accurately captures the addedrisks associated with loans that havenegative cash flow in the stress test.

ARM Flags. OFHEO is concerned thatpredicted cumulative default rates forARM loans are excessive in the up-ratescenario. For example, a typical ARMpurchased by an Enterprise could have

a cumulative default rate of 95 percentin the up-rate scenario. These excessivedefault rates for ARMs in the up-ratestress test arise from two principalsources. First, the up-rate stress testprojects declining DCRs for ARMs, andtwo explanatory variables in the defaultmodel translate declining DCRs intohigher default rates: the DCR variable,itself, and the UWDCRF, whereapplicable. The second source is fromthe application of an ARM product-typeflag—New Book ARM Flag (NAF)—which further raises the predicted ARMdefault rates. OFHEO included the ARMproduct flag in the Rule because itobserved in the historical data from theEnterprises that ARM defaults appear tobe higher than those of otherwisecomparable FRMs even after controllingfor DCR changes due to interest ratechanges.

The stress test projects DCR in eachmonth of the stress period from theprior month’s value using rent growthrates and vacancy rates that reflect theeconomic conditions of the benchmarkregion and period along with monthlymortgage payment adjustmentsaccording to the note terms and thestress test interest rate scenarios. In theup-rate scenario, the mortgage paymentadjustments on ARMs cause theprojected DCR to fall much more thanthat of an otherwise comparable FRM.This more rapid decline in DCR causespredicted defaults on ARMs to be higherthan those of otherwise comparableFRMs, as one would expect, becausemortgage payments on an ARM maygrow to exceed net operating incomefrom the property. In addition, the NAFfurther raises new book ARM defaultsrelative to comparable new book FRMsto capture performance differences notrelated to projected changes in DCR.15

The theoretical justification for theinclusion of an ARM flag to account forperformance differences not related toARM payment changes is that ARMborrowers may possess higher creditrisk qualities than their fixed-rate

counterparts. Arguing against theinclusion of an ARM flag is theimprovement in the Enterprises’multifamily ARM underwriting inrecent years, which means that, overtime, differences in risk between loantypes due to differences in borrowercharacteristics will disappear. That is,the choice of ARM versus FRM in themultifamily mortgage market may bebecoming a strategic business decisionrelated to professional financialmanagement considerations and may, asa result, have a declining relationship toborrower credit quality.

OFHEO decided that the excessivepredicted default rates for ARM loans inthe up-rate stress test warrantedinvestigation of the default model’sspecification of ARM product type flags.OFHEO sought to determine if arespecification of the model couldmaintain a reasonable relationship tothe historical data while producingmore reasonable results in the stresstest. First, the estimation was performedwithout either of the two product typeflags, the NAF and the New BookBalloon Flag (NBLF). If the onlyadditional risk associated with ARMsrelative to FRMs resulted from theimpact of rate changes on mortgagepayments and DCR, then thisspecification for the default modelmight be appropriate. OFHEO found,however, that this model specificationcaused another explanatory variable, theRatio Update Flag (RUF) to be no longerstatistically significant. Next, OFHEOre-estimated the model without theRatio Update Flag. The result of thesecond re-estimation produced, asexpected, an averaging effect betweenNew Book ARM and FRM defaultrates—that is, the size of the coefficientfor New Book loans decreased (thecoefficient remained negative but had asmaller absolute value), reflecting thefact that the NBF was now averaging theproduct type differences that arecurrently separated out by the producttype flags in the Rule. This specificationalso reduced the sensitivity of defaultsto the distinction between New Bookand Old Book loans, holding otherfactors constant, because it no longerdistinguished between loans for whichloan-to-value ratio (LTV) and DCR ratiosare updated and those for which theyare not.16



recalculated by the Enterprises at Enterpriseacquisition according to current underwritingstandards. New Book loans for which originationand/or acquisition LTV and DCR are unknowncannot be considered to be ratio-updated.

17 Specifically, the twelfth root of month oversame month previous year rent indices minus one.

18 Reporting of vacancy rate data for MetropolitanStatistical Area located in the WSC Census divisionbegan in 1986. As a result, 1984 and 1985 rateswere estimated based on national rates using theratio of WSC Census division rates to U.S. rentalvacancy rates in 1986, a factor of 2.3. For 1983, alower factor of 1.8 is assumed because it predatesthe WSC Census division’s recession.

19 REO is real estate owned as a result of loandefault.

20 The ‘‘baseline’’ consists of a simple adding upof the cost components of the rate, withoutconsidering discounting, credit enhancements, orpassthrough interest on sold loans.

OFHEO rejected the above model re-specification, which eliminates theNAF, the NBLF, and the RUF, becauseit ignored two important factors thatOFHEO has observed in Enterprisehistorical data. First, OFHEO consideredthe evidence of higher Enterprise ARMdefault rates, compared with FRMdefault rates during historical periodswhen interest rates were flat todeclining. Since flat-to-declininginterest rates lead to stable or lowerARM payments and therefore stable orhigher DCRs, all else equal, OFHEOsuspected that factors unrelated tointerest-rate-related ARM paymentchanges (such as borrower creditquality) may still be underlying thehigher observed ARM default rates.Second, OFHEO found substantialdifferences in observed default rates forratio-updated versus not-ratio-updatedloans in Enterprise historical data.Ratio-updated loans appear to performbetter than those that are not, holdingother factors constant.

Therefore, OFHEO proposes to re-specify its multifamily default model asfollows. The proposed model has thesame explanatory variables as the modelin the Rule, except that NAF, NBLF, andRUF are removed, and a respecified flagis introduced that captures both thedistinction between ARMs and FRMsand the distinction between ratio-updated and not-ratio-updated loans.Specifically, the new variable OFHEO isproposing in its respecified defaultmodel is a Not-Ratio-updated ARM Flag(NRAF) which takes a value of one (thatis, it is turned on) if a loan is both anARM and not ratio-updated, and zerootherwise. Because nearly all of theARM loans in Enterprise historical dataare not ratio-updated, but nearly all ofthe FRMs are ratio-updated, OFHEOdetermined that it is statisticallydifficult to fully separate these effects asmeasures of historical performance. Theproposed model with the NRAF variablewould apply this new variablecoefficient during the stress testsimulation only to ARM loans that arenot ratio-updated, capturing thehistorical performance differences ofthese ARMs after controlling forpayment changes. ARM loans that haveundergone the ratio-update processwould not be subject to higher defaultrisk imposed by the NRAF, therebyreducing the differential between ARMand FRM defaults in the up-ratescenario for those loans.

OFHEO believes that a similardistinction between ratio-updated FRMsand not-ratio-updated FRMs shouldexist even though there are too few not-ratio-updated FRMs in the Enterprises’historical data to confirm thehypothesis. As a result, OFHEOproposes to multiply monthlyconditional default rates for not-ratio-updated FRMs by a factor of 1.2 timesthe rates for otherwise comparable ratio-updated FRMs to reflect the marginallyhigher risk expected with those loans.

OFHEO believes that, given theEnterprise data, the proposal handles avery complicated issue fairly and withstatistical soundness and goodjudgment. If, in the future, Enterprisedata show no differences between ARMand FRM risk other than the adverseeffect of rising interest rates on ARMpayments and ARM DCR, OFHEO mayrevisit this issue.

Initial Vacancy Rate. Estimated rentgrowth for the first month of the stresstest is based on the relative change in arent index from immediately prior to thestress test to month one of the stresstest.17 However, the estimated vacancyrate change in the first month of thestress test does not look back to thevalue of the vacancy rate immediatelyprior to the stress test, but rathercompares the vacancy rate in month oneof the stress test with a long-termnational historical average vacancy rate.To be consistent, the change in vacancyrates between the period immediatelyprior to the stress test and month one ofthe stress test should be based on thechange in the benchmark regionvacancy rate from the month prior to thebenchmark period to the first month ofthe benchmark period. OFHEO viewsthis change as a technical correction.

Specifically, the vacancy rate changein the Rule in the initial month of thestress test is from the Census Bureau’slong-term national historical average of6.23 percent to the West South Central(WSC) Census division’s estimatedJanuary, 1984, rate of 13.6 percent, withchanges thereafter based upon changesin rates through 1993 in that region.18

This specification has the effect ofimposing a greater percentage increasein vacancies than appears to haveoccurred during the benchmark lossexperience.

The proposed change is to set theinitial vacancy rate at ten percent,which is the estimated WSC Censusdivision vacancy rate in 1983. Thus, thevacancy rate change in the initial monthof the stress test would be from tenpercent to 13.6 percent.

Loss Severity. Loss severityparameters in the Rule were based uponthe experience of 705 Freddie Macmultifamily REO 19 properties from the1980s. OFHEO has now analyzed datareflecting the costs, timing, and recoveryrates associated with additional REOthat has been made available from bothEnterprises. Based upon that analysis,OFHEO is proposing to modify themultifamily severity parameters to takeinto consideration the performance ofFannie Mae REO in the 1980s and bothEnterprises’ more recent multifamilyREO. The multifamily loss severitycalculations that use the severityparameters in the Rule would notchange. Specifically, OFHEO proposesreducing net REO holding costs to sevenpercent from 13.33 percent andincreasing REO sales proceeds from58.88 percent to 63 percent of theunpaid principal balance as of thedefault date. Additionally, OFHEOproposes reducing the time from defaultto foreclosure completion from 18 to 9months while increasing the time fromREO acquisition to REO dispositionfrom 13 to 15 months. Changing theseseverity parameters yields a 44 percent‘‘baseline’’ severity rate, as compared tothe 55 percent ‘‘baseline’’ produced bythe model in the Rule. ‘‘Baseline’’severity is a simple way to compare oneset of severity parameters withanother.20

Prepayment Penalties. In the Rule, nocredit is given for cash flows fromprepayment penalties and yieldmaintenance provisions. Nevertheless,the Rule provides that two percent ofloans that are subject to such penaltiesor provisions prepay each year of thestress test in the down-rate scenario. Inthe preamble to the Rule, OFHEOexplained that the data indicated that asmall percentage of loans did prepaywhile subject to yield maintenanceprovisions and that OFHEO had no dataindicating to what extent prepaymentpenalties were actually paid byborrowers, as opposed to waived by theEnterprises or added to the balances ofrefinanced loans. Because it is likelythat some prepayment penalties are paidor other compensating consideration is



received by the Enterprises, OFHEOdecided to include some prepaymentson these loans in the down-ratescenario, but at a lower rate thanindicated by the data in order to takeprepayment penalties into account.

OFHEO is proposing to modify theRule to provide for no prepayments inthe down-rate scenario insideprepayment penalty or yieldmaintenance periods. This approach ismore consistent with OFHEO’spreference to model contractualinstruments according to their terms,but recognizes that modeling thesepenalties according to their terms wouldbe immensely complicated, becausethose terms vary greatly from loan toloan. The proposed approach is areasonable simplification becauseprepayment penalty provisions areactually liquidated damages clauses,which are intended to give the lenderthe benefit of full performance on theloan.

C. Proposed Changes to Yields onEnterprise Debt

The Rule does not impose a premiumupon an Enterprise’s cost of funds toreflect the reaction of the debt marketsto the financial stress imposed upon theEnterprise. However, the preamble tothe Rule suggested that a premiummight be appropriate and that thiswould likely be an area of futurechange. Upon further study, OFHEO hasfound that it is appropriate for the stresstest to recognize an increased cost ofdebt of ten basis points for an Enterprisein the stress test vis-a-vis otherborrowers in the debt markets.

OFHEO proposed in NPR2 to imposea 50-basis-point premium on newEnterprise debt for the last nine years ofthe stress period. The analysis thatOFHEO performed for NPR2 indicatedthat debt spreads to Treasury rates havewidened in times of financial stress forGovernment-sponsored enterprises(GSEs). NPR2 did not proposeadjustments to reflect unusual stress forany other interest rate series in thestress test.

In the final rule, OFHEO took note ofthe comments received in response toNPR2, some of which questioned theappropriateness of a premium on newEnterprise debt and the size of thatpremium. OFHEO conceded that dataupon which to base such a premiummay be too sparse to determinedefinitively whether other spreads toTreasuries would widen as much as theEnterprises’ spreads or to estimate howmuch the Enterprises’ spreads wouldwiden. The preamble to the final rulealso noted that some commenters feltthat no premium on new debt should be

charged because many of theEnterprises’ hedging instruments arebased upon rates other than Treasuries(e.g., LIBOR, COFI). The spreadsbetween these rates and Treasuriescould be expected to widen duringstressful conditions, thus mitigating theEnterprises’ risk. In light of thesecomments, OFHEO postponedimposition of any new debt premiumpending later refinements of the Rule.Nevertheless, OFHEO indicated that theimplicit assumption in the stress testthat the spreads of an Enterprise’s debtyields to other interest rates would beunaffected by the deterioratingcondition of the Enterprise ignored anarea of significant risk.

The risk of wider spreads in astressful period is important if assetlives, which are unusually long in theup-rate scenario, exceed terms-to-maturity of outstanding debt. In supportof this proposal, OFHEO notes thatsome funding strategies employed bythe Enterprises depend significantly ontheir ability to borrow in the future atrelatively favorable interest rates. Forexample, the Enterprises often fund aportion of their mortgage asset portfoliowith short-term debt accompanied byinterest rate swaps, in which they paya fixed rate and receive a floating rate.If the floating rate they pay on their ownshort-term debt is close to the floatingrate they receive on the swap, the neteffect is roughly the same as if they hadissued long-term fixed-rate debt at therate they pay on the swap. If, however,their cost of short-term funds risessignificantly, relative to the index onwhich the swap’s floating rate is based,their cost will be higher than if they hadissued long-term fixed-rate debt. Use offixed-pay swaptions to hedge against theeffect of rising interest rates on expectedasset lives creates a similar risk.Although the spreads to Treasury ratesof other interests rates may also widenin a stressful economic environment,the stress test is designed to beespecially stressful to the Enterprises.The stress test involves factors, such asa decline in housing prices, that mightnot affect the debt costs in other sectorsof the economy as much. OFHEO haschosen to propose a ten-basis-pointspread for the final nine years of thestress period, in part to reflect theserisks.

A ten-basis-point borrowing premiumincorporates these risks in a modestway. Firms in very stressfulcircumstances frequently face premiumsof several hundred basis points, if theyare able to borrow at all. GSEs, though,have always been able to borrow, evenwhen they are in very poor financialcondition, because of their perceived

special status. It is reasonable, therefore,to use a much smaller premium thanmight be appropriate for a non-GSE ina similar stress test. OFHEO alsoconsiders it appropriate to consider thatthe stresses affecting the Enterprises inthe stress test would also be affectingother borrowers in the market place. Toassume that they do not, as was the casein NPR2, which proposed a 50-basis-point premium, is inconsistent with thestress implied in the haircuts that thestress test applies to all counterpartiesof the Enterprises. An ideal stress testmight model different spreads fordifferent interest rate series, a complexapproach that OFHEO could notimplement in the foreseeable future.The ten-basis-point premium, therefore,can be viewed as a simplifyingassumption, which gives some effect tothe possibility that stress period marketconditions could impact an Enterprisemore adversely than the rest of themarket.

D. Proposed Changes to New Debt MixThe Rule provides for the funding of

all cash deficits by the issuance of newlong-or short-term debt, whichever is inshorter supply, until a 50/50 balance ofshort-to long-term debt is reached ineach Enterprise’s portfolio. Thereafter,long- and short-term debt are issued inwhatever ratio best contributes tomaintaining that balance. This approachwas chosen because OFHEO did notwish to include an assumption aboutany particular behavioral preference bythe Enterprises during the stress period.

On further consideration, however,OFHEO proposes to change the targetbalance embodied in this approach. A50/50 balance is generally unsuitable forfunding a portfolio of largely fixed-ratemortgage assets, and it could oftenresult in a substantial change in anEnterprise’s funding structure duringthe stress period. OFHEO proposes toreplace the 50/50 target with the actualratio of Enterprise debt obligations (asadjusted by interest rate swaps) at thestart of the stress period. Typically, theEnterprises have a long-term debt tototal debt ratio (swap adjusted) of 70percent to 90 percent. Use of such ratiosin the stress test will result in a morerealistic debt structure.

E. Miscellaneous Technical ChangesOperating Expenses. In the Rule, one

third of an Enterprise’s operatingexpenses at the start of the stress testremain fixed throughout the stressperiod, while the remainder decline inproportion to the decline in themortgage portfolio. The total of the fixedand variable components is thenreduced by one-third to recognize that a



21 1992 Act, section 1302(2) (12 U.S.C. 4501(2)).22 ‘‘Managing Risk in Housing Finance Markets:

Perspective from the Experience of the UnitedStates of America and Mexico,’’ Mortgage BankersAssociation of America (June 11, 1998).

cessation of new business would have asignificant impact upon operatingexpenses. The variable portion of theoperating expenses for a given month isdetermined by calculating theEnterprise’s mortgage portfolio at theend of each month of the stress periodas a percentage of the portfolio at thestart of the stress test. Starting-positionfixed-asset balances are held constantover the ten-year stress period, whilerelated depreciation is included in thebase on which operating expenses arecalculated for each month of the stressperiod. The implication of thistreatment is that fixed assets are beingregularly replaced throughout theperiod, which appears inconsistent withthe decline in financial assets asmortgages amortize and prepay.

To address this inconsistency,OFHEO is proposing to modify thestress test treatment of operatingexpenses by converting 75 percent ofstarting-position fixed-asset balances tocash over the ten-year stress period. Theproposal would retain 25 percent of thefixed assets on the Enterprise booksthroughout the stress period to reflectthe acquisition of some new fixedassets, such as computer equipment,which is likely even in a ‘‘wind-down’’scenario. The effect of this change is toreduce the Enterprises’ need for debt tocarry nonearning fixed assets.

Float Income. The Rule provides forthe modeling of float income associatedwith passthrough payments onsecurities issued by the Enterprises.Float income can be positive or negativedepending on whether the Enterpriseholds the funds for a period of timebefore remitting them to securityholders or remits funds to securityholders before they are actuallyreceived. When an Enterprise owns itsown passthrough securities, the timingof payment to itself is not relevant.However, the Rule includes thesesecurities in the calculation of floatincome, resulting in an overstatement offloat income. OFHEO proposes tocorrect this overstatement by reducingthe float income on passthroughsecurities issued by the reportingEnterprise by the percentage of theEnterprise’s ownership interest.However, when an Enterprise receivesprepayments and holds the funds for anumber of days during which investorsaccrue interest at the coupon rate of thesecurity, the difference between theyield the Enterprise can earn oninvested funds at that time of the stressperiod and the coupon rate willcontinue to be reflected for the relevantnumber of days.

Currency Swaps. As a simplifyingassumption in the Rule, OFHEO applied

no haircut to foreign currency swaps,but stated its intention to continue toexplore appropriate methodologies forapplying an appropriate haircut. Infurtherance of its commitment tocontinue to refine the stress test,OFHEO now proposes to eliminate thesimplifying assumption and applyhaircuts to foreign currency swapcounterparties. Because the stress testdoes not project foreign currency values,the haircut is applied by adjusting thepay (dollar-denominated) side of theswap upward by the amount of thehaircut percentage rather thanhaircutting the foreign-currency receiveside of the swap.

American Call Option. As asimplifying assumption in the Rule, anAmerican call option, which allows theissuer to exercise the option at any time,is treated as a Bermudan call option,which allows the issuer to exercise thecall only on a coupon date. However, inthe preamble to the Rule, OFHEOsignaled its intention to consider howAmerican call options might bemodeled more precisely. OFHEO is nowproposing to modify the stress test toevaluate American calls on the firstoption date in the exercise schedule andsubsequent monthly anniversaries of theinstrument’s first coupon date.

House Price Growth FactorClarification. The Rule requires the useof OFHEO’s most recent House PriceIndex as of the reporting date todetermine the house price growth factorused to calculate current loan-to-valueratios. The proposal expands theinstructions in Section 3.6 to clarify,consistent with Section 3.7, that whena loan was originated since thepublication of that report, a cumulativehouse price growth factor of one is used.

Technical Correction. The proposaladds a Prepayment Penalty Flag as anadditional classification variable formultifamily loan groups, to distinguishloans with active prepayment penaltiesor yield maintenance provisions fromthose without in the calculation ofprepayment penalty duration for loangroups.

Regulatory Impact

Executive Order 12866, RegulatoryPlanning and Review

The proposed amendment wouldamend a rule designated as a major ruleby the Office of Management andBudget (OMB). The proposedamendment is a refinement of that rulethat would tie the capital more closelyto risk. Although the impact of thatrefinement is not economicallysignificant, OMB has reviewed theproposed amendment to determine

whether the proposed changes may raisenovel policy issues. OFHEO is notrequired to provide the type ofregulatory impact analysis that isrequired for an economically significantrule. Nevertheless, in accordance withOMB’s guidance that all regulatoryactions should be consistent with theprinciples of E.O. 12866, OFHEO hasdetermined, after review by agencyeconomists, financial analysts, andattorneys, that the benefits of theproposed changes to the Rulesubstantially outweigh any economiccosts.

It is impossible to estimate preciselythe particular benefits and costsassociated with the risk-based capitalrequirement. While OFHEO believesthis group of enhancements andrefinements to the stress test will notgenerally increase or decrease theamount of required capital for anEnterprise to any substantial degree, theeffect in any particular quarter dependsupon how well that Enterprise is hedgedagainst the risks and conditionsspecified in the stress test. OFHEOcannot know whether or not hedges inplace at an Enterprise at the beginningof any quarter would have been in placein the absence of specific provisions ofthe risk-based capital rule or were putin place because of the test. Speculatingas to what the Enterprises would do inthe absence of specific provisions infuture quarters is even more difficult.Therefore, a detailed economic cost/benefit analysis is not practical.

Rather than trying to assess the costsand benefits of every change to thestress test, OFHEO looks to whether ornot the changes it is proposing make theRule better reflect the risks faced by theEnterprises. Improving the Rule in thismanner should reduce the potential forEnterprise insolvency by protectingbetter against interest rate, credit, andmanagement and operations risk. Byhelping to ensure the safety andsoundness of the Enterprises, theregulation allows them to continue tocarry out their public purposes, whichinclude providing stability in thesecondary market for residentialmortgages and providing access tomortgage credit in central cities, ruralareas, and underserved areas.21 Inaddition, the regulation helps ensurethat the Enterprises will continue toprovide benefits to the primarymortgage market, such as standardizingbusiness practices.22



23 Final Report of Standard & Poor’s to OFHEO,Contract No. HE09602C (February 3, 1997).

24 Contract No. HE09602C, at 10.

Adopting the proposed amendmentwill result in a capital requirement thatcorresponds more closely to capitallevels that the marketplace woulddemand in the absence of the benefitsafforded by the Governmentsponsorship of the Enterprises, leadingto gains in overall economic efficiency.By improving the Rule’s ability to reflectactual risks at the Enterprises, theamendment also may enhance investorconfidence in the ability of the stresstest to forewarn investors and regulatorsof financial weaknesses. This resultwould be consistent with a study byStandard & Poor’s (S&P) that providedrisk-to-the-government credit ratings forthe Enterprises.23 Although S&P hadrated Fannie Mae A- and Freddie MacA+ in 1991, the 1997 report upgradedthe ratings of both Enterprises to AA-.S&P cited increased governmentaloversight by OFHEO as an importantfactor in these higher ratings. It furthernoted that ‘‘OFHEO’s regulatoryoversight [of Freddie Mac] also givescomfort that appropriate interest raterisk mitigation steps would be taken asneeded.’’24

OFHEO can identify no significantcosts associated with implementing theproposed amendments. No new reportsare required, and net effects on requiredcapital likely will be very small. In sum,the benefits to the public, including theEnterprises and other private-sectorconcerns, of the proposed changes faroutweigh the already expended costs ofimplementing those changes.

Paperwork Reduction ActThis proposed regulation does not

contain any information collectionrequirements that require the approvalof the Office of Management and Budgetunder the Paperwork Reduction Act (44U.S.C. 3501 et seq.).

Regulatory Flexibility ActThe Regulatory Flexibility Act (5

U.S.C. 601 et seq.) requires that aregulation that has a significanteconomic impact on a substantialnumber of small entities, smallbusinesses, or small organizations mustinclude an initial regulatory flexibilityanalysis describing the regulation’s

impact on small entities. Such ananalysis need not be undertaken if theagency has certified that the regulationwill not have a significant economicimpact on a substantial number of smallentities. 5 U.S.C. 605(b). OFHEO hasconsidered the impact of the proposedregulation under the RegulatoryFlexibility Act. The General Counsel ofOFHEO certifies that the proposedregulation, if adopted, is not likely tohave a significant economic impact ona substantial number of small businessentities because the regulation isapplicable only to the Enterprises,which are not small entities forpurposes of the Regulatory FlexibilityAct.

List of Subjects in 12 CFR Part 1750

Capital classification, Mortgages,Risk-based capital.

Accordingly, for the reasons stated inthe preamble, OFHEO proposes toamend 12 CFR part 1750 as follows:

PART 1750—RISK-BASED CAPITAL

1. The authority citation for part 1750continues to read as follows:

Authority: 12 U.S.C. 4513, 4514, 4611,4612, 4614, 4618.

2. Amend Appendix A to subpart B ofpart 1750 as follows:

a. Revise Table 3–1 in paragraph3.1.1;

b. Revise Table 3–4 in paragraph3.1.2.1;

c. Revise paragraph 3.3.1 [b];d. Revise paragraph 3.3.3 [a] 3.c.;e. Add new paragraph 3.5.3 [a] 2.d.;f. Revise paragraph 3.5.3 [a] 3. and

Table 3–31;g. In sentence six of paragraph 3.6.1

[e], remove the comma after the words‘‘Credit Losses’’, add the word ‘‘and’’ inits place, and remove the words ‘‘andthe Float Income’’ after the words‘‘Guarantee Fee’’;

h. Revise paragraph 3.6.3.4.3.1 [a] 2.a.;i. Revise paragraph 3.6.3.5.1 [b];j. In paragraph 3.6.3.5.2, revise Table

3–38;k. Revise paragraph 3.6.3.5.3.1 [a] 2.;l. In paragraph 3.6.3.5.3.1 [a] 4,

remove the first equation: ‘‘UWDCRFm =1 if DCRm < 1 in month m’’ and add theequation ‘‘UWDCRFm = 1 if DCRm < 0.98in month m’’ in its place;

m. Revise paragraph 3.6.3.5.3.2 [a] 1.and Table 3–39;

n. Revise paragraph 3.6.3.5.3.2 [a]2.b.;

o. Revise paragraph 3.6.3.5.3.2 [a] 3.;p. Revise Table 3–44, in paragraph

3.6.3.6.3.2;q. In section 3.6.3.6.4.3, revise the

four paragraphs: [a] 1., [a] 3.b., [a] 4.b.and [a] 5.;

r. Revise paragraph 3.6.3.7.3 [a] 9.b.;s. Revise paragraph 3.7.3.1 [g] 1.;t. In paragraphs 3.7.3.2 [a] 5. and

3.7.3.3 [a] 3., add the words ‘‘, asappropriate’’ at the end of the sentencein each paragraph;

u. In paragraph 3.7.4 [a] removereference to ‘‘Table 3–55’’ and add‘‘Table 3–61’’ in its place;

v. Redesignate Tables 3–65 through 3–70 as Tables 3–66 through 3–71;

w. After paragraph 3.8.1 [e], add newparagraph 3.8.1 [f], new footnote 5, andnew Table 3–65;

x. In paragraphs 3.8.2 [a] and [b]remove references to ‘‘Table 3–65’’ andadd ‘‘Table 3–66’’ in their place;

y. Revise paragraph 3.8.3.1 [a] 3.a.;z. In paragraph 3.8.3.4 remove

reference to ‘‘Table 3–66’’ and add‘‘Table 3–67’’ in its place;

aa. In paragraphs 3.8.3.6.1 [e] 1. and[e] 2. remove both references to ‘‘Table3–67’’ and add ‘‘Table 3–68’’ in theirplace;

bb. In redesignated Table 3–69 inparagraph 3.8.3.9, remove bothreferences to ‘‘Table 3–65’’ and add‘‘Table 3–66’’ in their place;

cc. Revise paragraphs 3.8.3.10 [a], [b]and [c];

dd. In paragraph 3.9.2 removereference to ‘‘Table 3–69’’ and add‘‘Table 3–70’’ in its place;

ee. In paragraph 3.10.2 [a] removereference to ‘‘Table 3–70’’ and add‘‘Table 3–71’’ in its place;

ff. Revise paragraphs 3.10.3.1 [b] 2.and [b] 3.;

gg. Revise paragraph 3.10.3.6.2 [a] 5.;and

hh. Revise the definition of EnterpriseCost of Funds in paragraph 4.0 Glossary.

The revisions and additions read asfollows:

Appendix A to Subpart B of Part 1750—Risk-Based Capital Text Methodologyand Specifications

* * * * *3.1.2.1 * * *



TABLE 3–1-SOURCES OF STRESS TEST INPUT DATA

Section of this Appendix Table

Data Source(s)R = RBC ReportP = Public Data

F = Fixed Values

R P F Intermediate Outputs

3.1.3, Public Data 3–19, Stress Test Single Family QuarterlyHouse Price Growth Rates

F

3–20, Multifamily Monthly Rent Growth and Va-cancy Rates

F

3.2.2., Commitments Inputs Characteristics of securitized single family loansoriginated and delivered within 6 months priorto the Start of the Stress Test

R 3.3.4, Interest Rates Outputs

3.2.3., Commitments Procedures 3–25, Monthly Deliveries as a Percentage ofCommitments Outstanding (MDP)

F

3.3.2, Interest Rates Inputs 3–18, Interest Rate and Index Inputs P

3.3.3, Interest Rates Procedures 3–26, CMT Ratios to the Ten-Year CMT F

3.4.2., Property Valuation Inputs 3–28, Property Valuation Inputs 3.1.3, Public Data3.3.4, Interest Rates Outputs

3.5.3., Counterparty Defaults Procedures 3–30, Rating Agencies Mappings to OFHEORatings Categories

P

3–31, Stress Test Maximum Haircut by RatingsClassification

F

3.6.3.3.2, Mortgage Amortization Schedule In-puts

3–32, Loan Group Inputs for Mortgage Amorti-zation Calculation

3.3.4, Interest Rates Outputs

3.6.3.4.2, Single Family Default and PrepaymentInputs

3–34, Single Family Default and PrepaymentInputs

R F 3.6.3.3.4, Mortgage Amortization Schedule Out-puts

3.6.3.4.3.2, Prepayment and Default Rates andPerformance Fractions

3–35, Coefficients for Single Family Default andPrepayment Explanatory Variables

F

3.6.3.5.2, Multifamily Default and PrepaymentInputs

3–38, Loan Group Inputs for Multifamily Defaultand Prepayment Calculations

R F

3.6.3.5.3.2, Default and Prepayment Rates andPerformance Fractions

3–39, Explanatory Variable Coefficients for Mul-tifamily Default

F 3.6.3.3.4, Mortgage Amortization Schedule Out-puts

3.6.3.6.2.2, Single Family Gross Loss SeverityInputs

3–42, Loan Group Inputs for Gross Loss Sever-ity

F 3.3.4, Interest Rates Outputs3.6.3.3.4, Mortgage Amortization Schedule Out-

puts3.6.3.4.4, Single Family Default and Prepay-

ment Outputs

3.6.3.6.3.2, Multifamily Gross Loss Severity In-puts

3–44, Loan Group Inputs for Multifamily GrossLoss Severity

F 3.3.4, Interest Rates Outputs3.6.3.3.4, Mortgage Amortization Schedule Out-

puts

3.6.3.6.4.2, Mortgage Credit Enhancement In-puts

3–46, CE Inputs for each Loan Group R 3.6.3.3.4, Mortgage Amortization Schedule Out-puts

3.6.3.4.4, Single Family Default and Prepay-ment Outputs

3.6.3.5.4, Multifamily Default and PrepaymentOutputs

3.6.3.6.2.4, Single Family Gross Loss SeverityOutputs

3.6.3.6.3.4, Multifamily Gross Loss SeverityOutputs

3–47, Inputs for each Distinct CE Combination(DCC)

R

3.6.3.7.2, Stress Test Whole Loan Cash FlowInputs

3–51, Inputs for Final Calculation of Stress TestWhole Loan Cash Flows

R 3.3.4, Interest Rates Outputs3.6.3.3.4, Mortgage Amortization Schedule Out-

puts3.6.3.4.4, Single Family Default and Prepay-

ment Outputs3.6.3.5.4, Multifamily Default and Prepayment

Outputs3.6.3.6.5.2, Single Family and Multifamily Net

Loss Severity Outputs

3.6.3.8.2, Whole Loan Accounting Flows Inputs 3–54, Inputs for Whole Loan Accounting Flows R 3.6.3.7.4, Stress Test Whole Loan Cash FlowOutputs



TABLE 3–1-SOURCES OF STRESS TEST INPUT DATA—Continued

Section of this Appendix Table

Data Source(s)R = RBC ReportP = Public Data

F = Fixed Values

R P F Intermediate Outputs

3.7.2., Mortgage-Related Securities Inputs 3–56, RBC Report Inputs for Single Class MBSCash Flows

R

3–57, RBC Report Inputs for Multi-Class andDerivative MBS Cash Flows

R

3–58, RBC Report Inputs for MRBs and Deriva-tive MBS Cash Flows

R

3.8.2., Nonmortgage Instrument Inputs 3–66, Input Variables for Nonmortgage Instru-ment Cash flows

R

3.9.2., Alternative Modeling Treatments Inputs 3–70, Alternative Modeling Treatment Inputs R

3.10.2., Operations, Taxes, and Accounting In-puts

3–71, Operations, Taxes, and Accounting In-puts

R 3.3.4, Interest Rates Outputs3.6.3.7.4, Stress Test Whole Loan Cash Flow

Outputs3.7.4., Mortgage-Related Securities Outputs3.8.4., Nonmortgage Instrument Outputs

3.12.2., Risk-Based Capital Requirement Inputs R 3.3.4, Interest Rates Outputs3.9.4., Alternative Modeling Treatments Outputs3.10.4., Operations, Taxes, and Accounting

Outputs

* * * * * 3.1.2.1 * * *

TABLE 3–4.—ADDITIONAL MULTIFAMILY LOAN CLASSIFICATION VARIABLES

Variable Description Range

Multifamily Product Code Identifies the mortgage product types for multifamilyloans

Fixed Rate Fully AmortizingAdjustable Rate Fully Amortizing5 Year Fixed Rate Balloon7 Year Fixed Rate Balloon10 Year Fixed Rate Balloon15 Year Fixed Rate BalloonBalloon ARMOther

New Book Flag ’’New Book’’ is applied to Fannie Mae loans acquiredbeginning in 1988 and Freddie Mac loans acquiredbeginning in 1993, except for loans that were refi-nanced to avoid a default on a loan originated oracquired earlier.

New BookOld Book

Ratio Update Flag Indicates if the LTV and DCR were updated at origi-nation or at Enterprise acquisition

YesNo

Interest Only Flag Indicates if the loan is currently paying interest only.Loans that started as I/Os and are currently amor-tizing should be flagged as ‘N’.

YesNo

Current DCR Assigned classes for the Debt Service CoverageRatio based on the most recent annual operatingstatement

DCR < 1.001.00 <=DCR<1.101.10 <=DCR<1.201.20 <=DCR<1.301.30 <=DCR<1.401.40 <=DCR<1.501.50 <=DCR<1.601.60 <=DCR<1.701.70 <=DCR<1.801.80 <=DCR<1.901.90 <=DCR<2.002.00 <=DCR<2.502.50 <=DCR<4.00DCR >= 4.00

Prepayment Penalty Flag Indicates if prepayment of the loan is subject to ac-tive prepayment penalties or yield maintenanceprovisions

YesNo

* * * * * 3.3.1 * * *

[b] The process for determining interestrates is as follows: first, identify values forthe necessary Interest Rates at time zero;

second, project the ten-year CMT for eachmonth of the Stress Period as specified in the1992 Act; third, project the 1-month Treasuryyield, the 3-month, 6-month, 1-, 2-, 3-, 5-, 20-



and 30-year CMTs; fourth, project non-Treasury Interest Rates, including the FederalAgency Cost of Funds Index; and fifth,project the Enterprises Cost of Funds Index,which provides borrowing rates for theEnterprises during the Stress Period, byincreasing the Agency Cost of Funds Indexby 10 basis points for the last 108 months ofthe Stress Test.

* * * * *3.3.3 * * *

[a] * * *3. * * *

c. Enterprise Borrowing Rates. In the StressTest, the Federal Agency Cost of FundsIndex is the same as the Enterprise Costof Funds Index during the Stress Period,except that the Stress Test adds a 10basis-point credit spread to the FederalAgency Cost of Funds rates to projectEnterprise Cost of Funds rates for the last108 months of the Stress Period.

* * * * *3.5.3 * * *

[a] * * *2. * * *

d. The Stress Test will permit a higherrating to be used for an unrated seller-

servicer who participates in a delegatedunderwriting and servicing program thatrequires a loss-sharing agreement when:(1) The loss sharing agreement iscollateralized by a fully funded reserveaccount pledged to the Enterprise; and(2) the reserve account is in an amountthat is equal to or exceeds the amountthat OFHEO has determined to beadequate to support the seller-servicer’sloss-sharing obligation under theprogram. Determinations of the reserverequirement and of the rating that will bepermitted will be made on a program-by-program and Enterprise-by-Enterprisebasis by the Director.

3. Determine Maximum Haircuts. The StressTest specifies the Maximum Haircut (i.e.,the maximum reduction applied to cashflows during the Stress Test to reflect therisk of loss due to counterparty(including security) default) by ratingcategory and counterparty type as shownin Table 3–31.

a. The Maximum Haircut for a ratingcategory is the product of its default rateand its loss severity rate. For allcounterparties the default rates are 5percent for AAA, 12.5 percent for AA, 20percent for A, 40 percent for BBB and

100 percent for Below BBB and Unrated.For non-derivative counterparties, theloss severity rate is 70 percent; forderivative counterparties, it is 10percent. For all Below BBB and Unratedcounterparties, the loss severity rate is100 percent.

b. For periods prior to the implementationof netting, a separate set of MaximumHaircuts (set forth in Table 3–31) will beapplied to derivative contract cash flowsto approximate the impact of the netexposures to derivative contractcounterparties (see section 3.8.3,Nonmortgage Instrument Procedures).After the implementation of netting,exposures will be netted as described insection 3.8.3 before the haircut isapplied.

c. With the exception of haircuts for theBelow BBB and Unrated category,haircuts for all counterparty categoriesare phased-in linearly over the 120months of the Stress Period. TheMaximum Haircut is applied in month120 of the Stress Period. Haircuts for theBelow BBB and Unrated category areapplied fully starting in the first monthof the Stress Test.

TABLE 3–31.—STRESS TEST MAXIMUM HAIRCUT BY RATINGS CLASSIFICATION


DerivativeContract

Counterpartiesprior to Imple-mentation of

Netting

DerivativeContract

Counterpartiesafter Imple-mentation of

Netting


Counterpartiesor Instruments


Cash 0% 0% 0% N/A

AAA 0.3% 0.5% 3.5% 120

AA 0.75% 1.25% 8.75% 120

A 1.2% 2% 14% 120

BBB 2.4% 4% 28% 120

Below BBB and Unrated 100% 100% 100% 1

* * * * *3.6.3.4.3.1 * * *

[a] * * *2. * * *

a. LTVq is evaluated for a quarter q as:

LTVORIG ×

Ratio of current

Loan Group UPB

to Original UPB

Ratio of current property

value (based on HPI in

quarter q) to original

property value (based on

HPI at Origination)The HPI at Origination is updated to the

beginning of the Stress Test using actualhistorical experience as measured by theOFHEO HPI; and then updated within the

Stress Test using House Price Growth Factorsfrom the Benchmark region and time period:

LTV LTV

UPB

UPB

CHPGF HPGR

q ORIG

m q

ORIG

LGk

k

q

=

×

×

= −

=∑

3 3

01

exp

Where:UPBm=3q-3 = UPB for the month at the end of

the quarter prior to quarter qCHPGFoLG= 1.0 if the loan was originated in

the same quarter as or after the mostrecently available HPI as of the reportingdate

* * * * *3.6.3.5.1

[b] Explanatory Variables for Default Rates.Eight explanatory variables are used as

specified in the equations section 3.6.3.5.3.1,of this Appendix, to determine Default ratesfor multifamily loans: Mortgage Age,Mortgage Age Squared, New Book indicator,Not Ratio-updated ARM indicator, currentDebt-Service Coverage Ratio, UnderwaterCurrent Debt-Service Coverage indicator,Loan-To-Value Ratio at origination/acquisition, and a Balloon Maturity indicator.Regression coefficients (weights) areassociated with each variable. All of thisinformation is used to compute conditionalannual Default rates throughout the StressTest. The annualized Default rates areconverted to monthly conditional Defaultrates and are used together with monthlyconditional Prepayment rates to calculateStress Test Whole Loan Cash Flows. (Seesection 3.6.3.7, Stress Test Whole Loan CashFlows, of this appendix).

* * * * *

3.6.3.5.2



TABLE 3-38—LOAN GROUP INPUTS FOR MULTIFAMILY DEFAULT AND PREPAYMENT CALCULATIONS

Variable Description Source

Mortgage Product Type RBC Report

A0 Age immediately prior to start of Stress Test, in months (weighted average for Loan Group) RBC Report

NBF New Book Flag RBC Report

RUF Ratio Update Flag RBC Report

LTVORIG Loan-to-Value ratio at loan Origination RBC Report

DCR0 Debt Service Coverage Ratio at the start of the Stress Test RBC Report

PMT0 Amount of the mortgage Payment (principal and interest) prior to the start of the StressTest, or first Payment for new loans (aggregate for Loan Group)

RBC Report

PPEM Prepayment Penalty End Month number in the Stress Test (weighted average for LoanGroup)

RBC Report

RM Remaining term to Maturity in months (i.e., number of contractual payments due betweenthe start of the Stress Test and the contractual maturity date of the loan) (weighted aver-age for Loan Group)

RBC Report

RGRm Benchmark Rent Growth for months m = 1 120 of the Stress Test section 3.4.4, Property Valuation Outputs

RVRm Benchmark Vacancy Rates for months m = 1 120 of the Stress Test section 3.4.4, Property Valuation Outputs

PMTm Scheduled Payment for months m = 1 RM 3.6.3.3.4, Mortgage Amortization ScheduleOutputs

OE Operating expenses as a share of gross potential rents (0.472) fixed decimal from Benchmark region andtime period

RVRo Initial rental vacancy rate 0.10

* * * * *3.6.3.5.3.1 * * *

[a] * * *2. Assign product and ratio update flags

(NBF, NRAF). Note: these values do notchange over time for a given Loan Group.

a. New Book Flag (NBF):NBF = 1 for Fannie Mae loans acquired after

1987 and Freddie Mac loans acquired after1992, except for loans that were refinancedto avoid a Default on a loan originated oracquired earlier.

NBF = 0 otherwise.b. Not Ratio-updated Arm Flag (NRAF):

NRAF = 1 if both ARMF = 1 and RUF = 0,NRAF = 0 otherwise.Where:ARMF = 1 for ARMs (including Balloon

ARMs)ARMF = 0 otherwise, andRUF = 1 if the LTV and DCR were calculated

or delegated to have been calculated atorigination or recalculated or delegated tohave been recalculated at Enterpriseacquisition according to current Enterprisestandards.

RUF = 0 otherwise

* * * * *3.6.3.5.3.2 * * *

[a] * * *1. Compute the logits for multifamily Default

using inputs from Table 3–38 and

coefficients from Table 3–39. Forindexing purposes, the Default rate for aperiod m is the likelihood of missing themth payment; calculate its correspondinglogit (Xδm) based on Loan Groupcharacteristics as of the period prior tom, i.e. prior to making the mth payment.

X AY AY

NBF NRAF

DCR

UWDCRF

LTV

BMF

m AY m AY m

NBF NRAF

DCR m

UWDCRF m

LTV ORIG

BMF m

δ δ δ

δ δ

δ

δ

δ

δ δ

= +

+ +

+ ( )+

+ ( )+ +

− −

−

−

−

12

1

1

1 0

2 1

ln

ln

TABLE 3–39—EXPLANATORY VARIABLECOEFFICIENTS FOR MULTIFAMILY DE-FAULT

Explanatory Variable Default Weight (δv)

AY 0.5256

AY2 ¥0.0284

NBF ¥1.219

NRAF 0.4193

DCR ¥2.368

TABLE 3–39—EXPLANATORY VARIABLECOEFFICIENTS FOR MULTIFAMILY DE-FAULT—Continued

Explanatory Variable Default Weight (δv)

UWDCRF 1.220

LTV 0.8165

BMF 1.518

Intercept (δ0) ¥4.553

* * * * *2. * * *

b. For the down-rate scenario, APRm = 0percent during the Prepayment penaltyperiod (i.e., when m ≤ PPEM)

APRm = 25 percent after the Prepaymentpenalty period (i.e., when m > PPEM)

* * * * *3. Convert annual Prepayment and Default

rates to monthly rates (MPR and MDR)using the following formulas forsimultaneous processes:

MPRAPR

ADR APR

ADR APR

mm

m m

m m

=+

× − − −( )

1 1

112

If both ARMF = 0 and RUF = 0, then



MDRADR

ADR APR

ADR APR

mm

m m

m m

=+

× − − −( )

×

1 1

1 2

112

.otherwise,

MDRADR

ADR APR

ADR APR

mm

m m

m m

=+

× − − −( )

1 1

112

* * * * *

3.6.3.6.3.2 * * *

TABLE 3–44—LOAN GROUP INPUTS FOR MULTIFAMILY GROSS LOSS SEVERITY

Variable Description Value or Source

Government Flag RBC Report

DRm Discount Rate in month m (decimal per annum) 6-month Enterprise Cost of Funds from Sec-tion 3.3, Interest Rates

MQ Time during which delinquent loan interest is passed-through to MBS holders 4 for sold loans0 otherwise

PTRm Pass Through Rate applicable to payment due in month m (decimal per annum) section 3.6.3.3.4, Mortgage AmortizationSchedule Outputs

NYRm Net Yield Rate applicable to payment due in month m (decimal per annum) section 3.6.3.3.4, Mortgage AmortizationSchedule Outputs

RHC Net REO holding costs as a decimal fraction of Defaulted UPB 0.07

MF Time from Default to completion of foreclosure (REO acquisition) 9 months

MR Months from REO acquisition to REO disposition 15 months

RP REO proceeds as a decimal fraction of Defaulted UPB 0.63

* * * * *3.6.3.6.4.3 * * *

[a] * * *1. Determine Mortgage Insurance Payment

(MIm) for single family loans in the DCC,or Loss Sharing Payment (LSAm) formultifamily loans in the DCC, as apercentage of Defaulted UPB, applyingappropriate counterparty Haircuts fromsection 3.5, of this Appendix:

MI MIExp

C CLM

mMaxHct R

mDCC

mLG

MI DCCmMI LG

MI DCC

= −( )× ×

× − ′ × ( )

1

1120

, ,

,

LSA C CLM

mMaxHct R

mDCC LSA DCC

mLSA LG

LSA DCC

= ×

× − ′ × ( )

, ,

, 1120

Where:m′ = m, except for counterparties rated below

BBB, where m’ = 120

MIExp

UPB

UPB

MIExp

mLG

mLG

ORIGLG

mLG

=

×

<

=

=

1

0 78

0

if

LTV

otherwise

0.78 (78%) the LTV at which MI is cancelled if payments are

current

ORIG .

* * * * *3.* * *

b. Determine CE Payment in Dollars afterapplication of Haircuts:

PD PD

mMaxHct R

mDCC C H

mDCC C

DCC C

, , ,

,

1 1

1

120

=

× − ′ × ( )

1

Where:

m′ = m, except for counterparties rated belowBBB, where m′ = 120

* * * * *4.* * *

b. Determine CE Payment in Dollars afterapplication of Haircuts:

PD PD

mMaxHct R

mDCC C H

mDCC C

DCC C

, , ,

,

2 2

2

120

=

× − ′ × ( )

1

Where:

m′ = m, except for counterparties rated belowBBB, where m′ = 120

* * * * *5. Convert Aggregate Limit First and Second

Priority Contract receipts in Dollars foreach DCC in month m to a percentage ofDCC Defaulted UPB:

ALPDPD ELPI PD ELPI

DEF UPB PmDCC m

DCC C H DCC CmDCC C H DCC C

m mLG DCC=

×( ) + ×( )× ×−

, , , , , ,1 1 2 2

1



5 Ibid.

Where:ELPIDCC,C = 0 if ELPFDCC,C = Y (Yes,

indicating that Contract C is an EnterpriseLoss Position)

ELPIDCC,C = 1 otherwise

* * * * *3.6.3.7.3. * * *

[a] * * *9. * * *b. Float Income (FI) received in month m

FI SPR NIR GF

FDSPPR

FDP

FREP PIS

m m m m

m

m

= + −([ )([×

+ ×

× × −( )] −

FERm

365 365

1

Where:

Prepayment Interest Shortfall (PIS) in monthm is:

PIS UPB PREPTR

FER

PIS UPB PREPTR

FER

m m mm

m

m m mm

m

= × ×

− × ×

≥

= × ×

− × ×

≤ <

−

−

1

1

12

12

24

24

FREP PPR

if FDP 30

FREP PPR

if 15 FDP 30

m

m

* * * * *3.7.3.1 * * *

[g] * * *

1. Compute:

HctFacm

MaxHct Rm = ′ ×120

( )

Where:m’ = m, except for MBS credit rating below

BBB where m’=120R = MBS credit rating

* * * * *3.8.1 * * *

[f] In a currency swap, the Enterprisereceives payments that are denominated in aforeign currency and it makes payments inU.S. dollars. The main difference betweencurrency swaps and the type of swapsdiscussed above is that in a currency swapprincipal amounts are actually exchangedbetween the two counterparties. Currencyswaps are divided into two classes, as shownin Table 3–65 below.5

TABLE 3–65—CURRENCY SWAP CONTRACT CLASSIFICATION

Classification Description of Contract

Fixed-for-Fixed Currency Swap Enterprise receives fixed interest payments denominated in a foreign currency and makes fixed, US$-de-nominated payments

Fixed-for Floating Currency Swap Enterprise receives fixed interest payments denominated in a foreign currency and makes payments in US$based on a floating interest rate

* * * * *3.8.3.1 * * *

[a] * * *3. When applying the option exercise rule:

a. For zero coupon and discount securities,instruments with European options, andzero coupon swaps, evaluate optionexercise only on dates listed in theinstrument’s option exercise schedule.For Bermudan options, evaluate optionexercise on the first option date in theinstrument’s option exercise scheduleand subsequent coupon dates (coupondates on the fixed-rate leg for swaps). ForAmerican options, evaluate optionexercise on the first option date in theinstrument’s option exercise scheduleand subsequent monthly anniversaries ofthe instrument’s first coupon date.

* * * * *3.8.3.10 * * *

[a] Finally, the interest and principal cashflows received by the Enterprises for non-mortgage instruments other than swaps andforeign currency-related instruments areHaircut (i.e., reduced) by a percentage toaccount for the risk of counterpartyinsolvency, if a counterparty obligationexists. The amount of the Haircut iscalculated based on the public rating of thecounterparty and time during the stressperiod in which the cash flow occurs, asspecified in section 3.5, CounterpartyDefaults, of this Appendix.

[b] An Enterprise may issue debtdenominated in, or indexed to, foreigncurrencies, and eliminate the resultingforeign currency exposure by entering intocurrency swap agreements. The combination

of the debt and the swap creates syntheticdebt with principal and interest paymentsdenominated in U.S. dollars. The Haircuts forcurrency swaps are applied to the pay(dollar-denominated) side of the currencyswaps, or to the cash outflows of thesynthetic debt instrument. Therefore, thepayments made by the Enterprise on aforeign currency contract are increased by thehaircut amount. The Haircuts and the Phase-in periods for currency swaps are detailed inTable 3–31, under Derivative Contracts.

[c] Haircuts for swaps that are not foreigncurrency related are applied to the MonthlyInterest Accruals (as calculated in section3.8.3.8, of this Appendix) on the receive legminus the Monthly Interest Accruals on thepay leg when this difference is positive. Usethe maximum haircut from Table 3–31 forperiods before and after the implementationof netting, as appropriate. After theimplementation of netting, net the swapproceeds for each counterparty beforeapplying the haircuts. The following exampleapplies to an Enterprise having two swapswith the same counterparty. On the firstswap, the Enterprise pays fixed and receivesfloating and on the second swap it paysfloating and receives fixed. If thecounterparty is a net payer to the Enterprise,the haircuts will be applied to the sum of thetwo receive legs net of the sum of the two paylegs.

* * * * *3.10.3.1 * * *

[b] * * *2. In any month in which the cash position

is negative at the end of the month, theStress Test issues a mix of new short-termand long-term debt on the 15th day of that

month. New short-term debt issued is six-month discount notes with a discount rateat the six-month Enterprise Cost of Fundsas specified in section 3.3, Interest Rates,of this Appendix, with interest accruing ona 30/360 basis. New long-term debt issuedis five-year bonds not callable for the firstyear (‘‘five-year-no call-one’’) with anAmerican call at par after the end of thefirst year, semiannual coupons on a 30/360basis with principal paid at maturity orcall, and a coupon rate set at the five yearEnterprise Cost of Funds as specified insection 3.3, Interest Rates, of thisAppendix, plus a 50 basis point premiumfor the call option. An issuance cost of 2.5basis points is assessed on new short-termdebt at issue and an issuance cost of 20basis points is assessed on new long-termdebt at issue. New long-term debt is issuedto target a total debt mix of short to longterm debt that is the same as the short tolong term debt mix at the beginning of theStress Test. Issuance fees for new debt areamortized on a straight line basis to thematurity of the appropriate instrument.

3. Given the Net Cash Deficit (NCDm) inmonth m, use the following constantsand method to calculate the amount ofshort-term and long-term debt to issue inmonth m:

a. Set the Issuance Cost on new short-termdebt at issue (ISCOST):ISCOST = 0.00025

b. Set the Issuance Cost on new long-termdebt at issue (ILCOST):ILCOST = 0.002

c. Calculate Net Short-term DebtOutstanding (NSDO0) and Total DebtOutstanding (TDO0) at the start of the



Stress Test (m = 0) using the followingmethodology:

(1) For each month m and each debt andswap instrument i (each swap leg isconsidered a separate instrument),determine the Month of Next Repricing(MNRm) defined as the first monthgreater than m in which the instrumentmatures, an option is exercised, orrepricing can occur whether or not thecoupon rate actually changes. Set thePrincipal Balance (PBm) to be:

(a) the principal (or notional principal)outstanding if the instrument cash flowsare paid by the Enterprise,

(b) minus the principal (or notionalprincipal) outstanding if the instrumentcash flows are received by theEnterprise.

(2) Calculate NSDOm by summing PBm,i forall instruments where MNRm,i is lessthan or equal to m plus 12.

(3) Calculate TDOm by summing PBm,i forinstruments where MNRm,i is greaterthan m.

d. Set the Maximum Proportion of TotalDebt (MPD):

MPDTDO NSDO

TDO= −0 0

0

e. Calculate Discount Rate Factor (DRFm):

DRFCF

mm= +

1

12

6

Where: CFm = six month Enterprise Cost ofFunds for month m

f. Calculate the Adjustment Factor forShort-Term Debt Issuance Fees (AFSIFm):

AFSIFDRF

ISCOST DRFmm

m

=− ×1

g. Calculate the Adjustment Factor forLong-Term Debt Issuance Fees (AFLIFm):

AFLIFILCOSTm =

−1

1h. Calculate the Maximum Long-Term

Issuance (MLTIm):

MLTI NCD AFLIFm m m= ×i. Calculate Net Short-Term Debt

Outstanding (NSDOm) and Total DebtOutstanding (TDOm) for month m usingthe methodology described in section3.c. of this section. Note: Thiscalculation must reflect all newissuances, option exercises, andmaturities between the beginning of theStress Test and month m.

j. Calculate Interim Face Amount of Long-Term Debt to be issued this month(IFALDm):

IFALDMPD TDO NSDO MPD AFSIF NCD

MPD AFSIFMPD

AFLIF

mm m m m

mm

=−( ) ×( ) + + × ×( )

− + ×

1

1

k. Calculate Face Amount of Long-Term Debt to be issued (FALDm):

FALDm m= ( )( )min max MLTI IFALDm, ,0

l. Calculate Face Amount of Short-Term Debt to be issued (FASDm):

FASD AFSIFFALD

AFLIFm mm

m

= × −

NCDmmax 0,

* * * * *3.10.3.6.2 * * *

[a] * * *5. Fixed Assets. 25 percent of fixed assets

(net of accumulated depreciation) as ofthe beginning of the Stress Test remainconstant over the Stress Test. Theremaining 75 percent is converted tocash on a straight line basis over the ten-year Stress Period. Depreciation isincluded in the base on which operatingexpenses are calculated for each monthduring the Stress Period.

* * * * *4.0 * * *

Enterprise Cost of Funds: Cost of fundsused in computing the cost of new debt forthe Enterprises during the Stress Test, asspecified in section 3.3.3[a]3.c., of thisAppendix.

* * * * *Dated: December 11, 2001.

Armando Falcon, Jr.,Director, Office of Federal Housing EnterpriseOversight.

[FR Doc. 01–30898 Filed 12–17–01; 8:45 am]

BILLING CODE 4220–01–P

DEPARTMENT OF HOUSING ANDURBAN DEVELOPMENT

24 CFR Parts 5 and 202

[Docket No. FR–4681–C–02]

Uniform Financial ReportingStandardsFor HUD Housing Programs,Additional Entity Filing Requirements;Correction

AGENCY: Office of the General Counsel,HUD.ACTION: Proposed rule; correction.

SUMMARY: On November 30, 2001, HUDpublished a proposed rule entitled‘‘Uniform Financial ReportingStandards for HUD Housing Programs,Additional Entity Filing Requirements.’’The preamble to the rule (although notthe rule text) misstates the date bywhich the financial statements ofentities covered by the rule must submittheir financial statements electronically.This notice corrects the preamble.FOR FURTHER INFORMATION CONTACT: Forfurther information about the entitiescovered by the proposed rule and thiscorrection notice, Lynn Herbert, theOffice of Housing, U.S. Department of

Housing and Urban Development, 451Seventh Street, SW., Washington, DC20410, telephone 202–708–3976 (this isnot a toll-free number). For generalinformation about this notice and theproposed rule, Stacey Kniff, Real EstateAssessment Center, U.S. Department ofHousing and Urban Development, 1280Maryland Avenue, SW., Suite 800,Washington, DC 20024, telephoneTechnical Assistance Center, 1–888–245–4860 (this is a toll-free number).Persons with hearing or speechimpairments may access thesetelephone numbers via TTY by callingthe Federal Information Relay Service at(800) 877–8339. Additional informationis available from the REAC Web site athttp://www.hud.gov/reac/.

SUPPLEMENTARY INFORMATION: OnNovember 30, 2001, HUD published aproposed rule entitled ‘‘UniformFinancial Reporting Standards for HUDHousing Programs, Additional EntityFiling Requirements’’ at 66 FR 60132.The preamble to the proposed rule, inthe third column of that page,immediately above the ‘‘Findings andCertifications’’ section, states:


OFFICE OF FEDERAL HOUSING ENTERPRISE OVERSIGHT NEWS …€¦ · 401(h), Pub. L .101–73, 103 Stat. 357. 2. Section 360.7 is added to part 360 to read as follows: §360.7 Post-insolvency

Documents