articles Ratings /ratings/en/research/articles/221010-criteria-governments-u-s-public-finance-methodology-for-rating-u-s-public-finance-mortgage-revenue-bond-pr-12342112 content esgSubNav
In This List

Criteria | Governments | U.S. Public Finance: Methodology For Rating U.S. Public Finance Mortgage Revenue Bond Programs


Sustainability Insights: North American Wildfire Risks Could Spark Rating Pressure For Governments And Power Utilities, Absent Planning And Preparation


Table Of Contents: S&P Global Ratings Credit Rating Models


U.S. Privatized Student Housing Stabilizes As Recovery Continues And University Financial Support Dwindles


Credit Quality Improved At A Slower Pace In Third Quarter 2023

Criteria | Governments | U.S. Public Finance: Methodology For Rating U.S. Public Finance Mortgage Revenue Bond Programs


1. U.S. mortgage revenue bond (MRB) programs are backed by pools of mortgage loans on residential property. They are typically established and overseen by a state or local housing finance agency (HFA) for the purpose of improving housing affordability by financing mortgage lending to low- and moderate-income households or borrowers developing or preserving affordable rental housing.

2. MRBs are typically backed by pools of:

  • Single-family first and second mortgage whole loans;
  • Mortgage loans secured by rental housing that meet the qualifications of residential rental property as set forth in IRS Code Section 142(d);
  • Single-family and multifamily mortgage loans or MBS where the loans or MBS benefit from full credit enhancement via a guarantee, insurance, or credit enhancement instrument from U.S. federal agencies; and
  • Hybrid programs composed of a mix of the above.

3. We refer throughout this article to MRB programs issued by HFAs, but the criteria could be applied to MRB programs issued by other U.S. financing organizations, if the public policy mission, the legal structure, the degree of program management, and the underlying asset types are similar.

4. The criteria apply only to managed MRB programs where the issuing HFA (or other financing organization) has an active role in the general oversight of the program, as well as in the ongoing management of specific risks, such as asset-liability matching, debt profile and investment management, or liquidity and counterparty risks (see the Program Management And Operational Risk Analysis section for more details). The degree of such management (including both support and risks brought about by such involvement) is addressed directly in these criteria. MRB programs in scope of these criteria are generally active issuance vehicles used by the issuing HFA to fund its ongoing lending (although an MRB program would remain in scope if the HFA decides to amortize it down, for example, and fund its lending under a different program).

5. Conversely, the criteria don't apply to transactions that don't possess such program risk management characteristics (although the issuing organization may be involved in the servicing of securitized assets and the administrative operation of the transaction). In particular, non-managed transactions backed by static mortgage pools are generally rated under "U.S. Federally Enhanced Housing Bonds Rating Methodology," or "Methodology For Rating U.S. Public Finance Rental Housing Bonds," as applicable.

6. We also apply the "Credit Quality Of The Asset Pool" section of these criteria to determine loss assumptions used in our capital adequacy analysis under our criteria for rating HFAs ("Methodology And Assumptions: Housing Finance Agencies And Social Enterprise Lending Organizations").



Chart 1



Chart 2


7. The legal framework links the duties of the key transaction parties with the proper execution of the program. If typical legal provisions are not present in the transaction documents or the associated legal risks are not mitigated, we do not rate the transaction under these criteria. Typical legal provisions include security and collateral, flow of funds, events of default, acceleration, and redemptions, among others.

8. We also analyze bankruptcy and other legal risks that could adversely affect the ability to pay full and timely debt service. We address these concepts, to some extent, through a different lens, as compared with the approach we take in our analysis of U.S. structured finance transactions, because of the unique nature of U.S. public finance housing transactions (for example, more flexible, diverse, and dynamic structures that are often actively managed and affiliated with a U.S. municipal or quasi-municipal entity, and hence, there may not be formal separateness covenants or requirements for independent directors).

9. In some cases, events of default in the MRB program's transaction documents may include events linked to the creditworthiness of the HFA, such as the HFA's bankruptcy or inability to meet debt obligations outside of the MRB program. In order to assign an anchor that exceeds the HFA issuer credit rating (ICR), we assess whether the likelihood that such events could lead to an MRB program default is sufficiently remote.

10. Specifically, if the MRB program would face an event of default upon the HFA being unable to meet other debt obligations (without specific reference to bankruptcy), the MRB program's anchor is generally no higher than the HFA ICR. We may consider that the risk is remote if the HFA's debt obligations outside the MRB program are minimal in size, or less important, relative to the MRB program.

11. Furthermore, if the MRB program would face an event of default upon a bankruptcy of the HFA, we consider a range of analytical factors in order to determine whether the likelihood of an HFA bankruptcy is sufficiently remote to support an anchor higher than the HFA ICR. We generally expect that an HFA would meet the definition of a "municipality" under the U.S. Bankruptcy Code, and therefore would not be subject to an involuntary bankruptcy filing. In assessing the likelihood that an HFA would file for bankruptcy voluntarily, we consider whether the HFA is or could be authorized to file for bankruptcy under applicable state law. If so, we consider other factors that may disincentivize the HFA from filing for bankruptcy, such as the necessity of the service it provides to the local population, its continued need to access financial markets on favorable terms, and the size and importance of its debts outside of the MRB program, relative to the size and importance of the MRB program.

12. Where we deem appropriate to analyze whether legal risks are mitigated, we may request legal opinions that address one or more issues such as: automatic stay risk, preference risk, trust estate parameters, Chapter 9 status, non-consolidation, perfected security interest, enforceability of the transaction documents, or other applicable risks.


Chart 3


13. An HFA's robust active management and oversight can proactively mitigate risks over the life of a program. Conversely, weaknesses in program management may obscure or exacerbate credit risks in the program. As program management characteristics for MRB programs are typically strong, we consider strong program management as the base case, and do not apply any positive adjustment for management strengths. Our assessment of program management considers five factors, which we assess as neutral or negative. If any of the factors are assessed as negative, the anchor may be constrained.

14. Table 1 shows the rating category constraint that we generally apply, based on the number of factors we assess as negative. We consider the severity of each risk identified in the context of the specific program, in determining a rating cap in each category. For example, for a maximum achievable anchor in the 'aa' category according to table 1, we are more likely to cap the anchor at 'aa+' if only one negative attribute is identified, or if two negative attributes identified are both relatively immaterial. Conversely, we are more likely to cap the anchor at 'aa' or 'aa-' if two negative attributes are identified, or if a single negative attribute is more material or deteriorating further. In addition, if we assess a risk as severe, we may cap the anchor below the category indicated in table 1, including in the 'bb' or 'b' categories.

Table 1

Determining An MRB Program's Maximum Achievable Anchor Based On Program Management And Operational Risks
Number of factors assessed as negative Maximum achievable anchor
0 No constraint
1-2 'aa' category
3-4 'a' category
5 'bbb' category

15. Examples of attributes that may lead to a negative assessment for each of the five factors are shown in table 2. The presence of a specific attribute may not lead to a negative assessment if we view the identified risk as otherwise mitigated. The relevance of some attributes in table 2 varies for programs of different type. For example, an assessment of loan origination standards is less relevant in the context of a program backed by guaranteed MBS than for a program backed by a pool of single-family whole loans.

Table 2

Assessing Program Management And Operational Risks
Factor Examples of negative attributes
Program strategy & governance
Poorly defined strategy
Weak transparency/disclosure
Insufficient planning or internal controls; history of failing to set and achieve operational goals
Less sophisticated management team relative to that of peers, or frequent management turnover
Loan origination & monitoring
Origination standards are less robust relative to those of peers, as evidenced through new, riskier loan products or the targeting of riskier borrower profiles
Less robust monitoring process for servicers and lenders relative to that of peers
Weak track record of managing delinquencies/recoveries versus peers
Asset & liability management
Volatility or expected decline in parity ratio
Significant mismatch in asset/liability maturities (e.g. from bullet maturity bonds)
Significant presence of interest rate risk beyond what is captured in the program’s cash flows
Liquidity risk management
Reserve investment guidelines on credit quality/liquidity/maturity permitted are insufficient to support the program’s liquidity needs
Derivative contracts introduce liquidity risks to the program (e.g. collateral posting or termination payments)
Counterparty risk management  
Large exposures to individual, or any exposure to speculative grade, counterparties
Counterparties with no (or non-market standard) collateral arrangements
HFA termination rights are limited, particularly relative to speculative-grade counterparties
Lack of planning for expiring counterparty arrangements


Chart 4


16. We first calculate a projected loss assumption for the asset pool at each rating level, based on pool and loan characteristics. The projected loss assumption is an input in the next section of these criteria, the cash flow analysis, in which we determine the highest rating level at which the MRB program's available overcollateralization is sufficient to cover the projected loss assumption.

17. This section sets out our approach to determine projected loss assumptions for pools of single-family whole loans, multifamily mortgage loans, and single-family or multifamily loans or bonds backed by the full enhancement of a U.S. government entity (GE) or U.S. government-sponsored entity (GSE).

18. For hybrid programs (that is, programs with multiple asset types), the pool-level projected loss assumptions are the sum of projected loss assumptions for each asset type in the pool.

Projected loss assumptions for single-family whole loan pools

19. For single-family whole loan pools, two components are multiplied to generate the projected loss assumption: foreclosure frequency and loss severity. Foreclosure frequency represents the probability of a loan to enter foreclosure (default). Loss severity refers to the loss on foreclosure (i.e., the amount by which a loan balance, foreclosure costs, and lost interest exceed property sale proceeds). We generally base our analysis on cohort level data, with loans grouped on the basis of their ratio of current loan balance to property valuation at time of origination. We then determine a weighted average foreclosure frequency (WAFF) and weighted average loss severity (WALS) across the pool of single-family whole loans. Subsequently, we consider any recoveries from insurance or guarantees. Our projected loss assumption for the pool, before consideration of recoveries from insurance or guarantees, is represented by the following formula:

  • Equation 1: Projected loss assumption (at each rating level) = WAFF x WALS
Foreclosure frequency

20. We first define a base foreclosure frequency assumption at each rating level, shown in table 3. These base assumptions reflect risk characteristics of a typical pool of newly originated single-family whole loans. They are then adjusted based on specific loan and borrower characteristics of the actual pool, by applying adjustment multipliers above 1.0x for characteristics that we consider increase borrower default risk, and multipliers below 1.0x for characteristics that we consider decrease borrower default risk. The adjustment factors are summarized in table 4. Appendix I provides further detail on the adjustment factors, the definition of a typical pool, and the calibration of our analysis relative to our U.S. RMBS criteria.

Table 3

Base Foreclosure Frequency Assumptions For The Typical Single-Family Loan Pool
Rating Current loan pool balance* (%)
AAA 15.00
AA+ 13.00
AA 11.00
AA- 10.13
A+ 9.13
A 8.25
A- 7.13
BBB+ 6.13
BBB 5.00
BBB- 4.63
BB+ 4.38
BB 4.00
BB- 3.50
B+ 3.00
B 2.50
B- 2.00
*Before adjustments in table 4.

Table 4

Overview Of Foreclosure Frequency Adjustments For Single-Family Whole Loan Pools
Characteristic Adjustment factor (multipliers below 1.0x reflect characteristics that decrease risk; multipliers above 1.0x reflect characteristics that increase risk)
Adjustments reflecting loan characteristics
LTV 0.7x–2.1x for LTVs in the range 75%-95%, with an adjustment of approximately 1.0x for LTVs of 82%, increasing exponentially for higher LTVs (see chart 8)
Seasoning 0.5x–1.0x applied to performing loans with at least five years’ seasoning (see table 11)
Loan type 1.5x for loan types other than fixed-rate 30-year maturity amortizing loans. Can increase to 3.0x for balloon or negatively amortizing loans, if applicable (see table 10)
Property type 1.1x–2.0x for property types other than single-family residences (see table 12)
Delinquency status 2.5x–5.0x for loans that are at least 30 days delinquent, with 100% foreclosure frequency assumed for loans 90 days delinquent and above (see table 13)
Adjustments reflecting pool characteristics
Average borrower FICO 0.9x–2.5x for FICO scores ranging from greater than 725 to below 620 (see table 9)
Small pool adjustment factor Applies to pools of fewer than 250 loans (see chart 10)
Qualitative adjustment factor 1.0x–2.0x, applied where servicing or origination standards deviate significantly from the norm for HFA managed pools (see paragraph 60)
Loss severity

21. Our assumption for loss severity is calculated based on the following components, as detailed in Appendix I:

  • Market value decline (MVD) assumptions for the underlying properties;
  • Over/undervaluation adjustments (based on long-term price-to-income ratios) for current housing market conditions in the relevant state;
  • Foreclosure and preservation costs;
  • Lost interest, taxes, and insurance over the course of the liquidation timeline.
Minimum projected loss assumptions for single-family whole loans

22. To address idiosyncratic risks not otherwise captured in the adjustments in these criteria, we establish minimum projected loss assumptions at each rating level for single-family whole loan pools, as shown in table 5. We apply these minimum levels to the projected loss assumptions before consideration of any recoveries from insurance or guarantees.

Table 5

Minimum Projected Loss Assumption For Single-Family Whole Loan Pools
Rating Minimum projected losses (%)
AA+ 3.42
AA 2.83
AA- 2.58
A+ 2.28
A 2.03
A- 1.7
BBB+ 1.41
BBB 1.08
BBB- 0.97
BB+ 0.9
BB 0.79
BB- 0.64
B+ 0.5
B 0.35
B- 0.28
Projected loss assumptions for multifamily loan pools

23. For multifamily loan pools, we derive projected loss assumptions by setting a base credit loss at each rating level, to which we apply adjustment multipliers to capture concentration risks and borrower and pool characteristics. Subsequently, we consider any recoveries from insurance or guarantees. Table 6 shows the base credit loss assumptions commensurate with each rating level for multifamily loans.

Table 6

Base Credit Loss For A Typical Affordable Multifamily Loan Pool
Rating Current loan pool balance* (%)
AAA 10
AA+ 8.5
AA 7.5
AA- 6
A+ 5
A 4.25
A- 3.75
BBB+ 3
BBB 2.5
BBB- 2
BB+ 1.5
BB 1.25
BB- 1.1
B+ 0.9
B 0.75
B- 0.6
*Before adjustments described below.

24. As the base credit loss assumptions in table 6 reflect a diversified pool, we apply an upward adjustment to capture loan concentration risks in more concentrated pools. Specifically, we set a loan concentration threshold of 5% of the total multifamily loan pool balance. For each loan in a pool, we apply the base credit loss assumption on the balance of the loan that is up to this threshold. For any loan balance that exceeds the threshold, we apply an additional multiplier to the excess balance. The applicable multiplier varies with the creditworthiness of each loan, measured by the loan's debt service coverage (DSC) ratio, as shown in table 7. We then calculate the balance-weighted average credit loss assumption for the loan pool. An example of the application of this adjustment is included in Appendix II.

Table 7

Determining The Concentration Risk Multiplier
Loan DSC band* Concentration risk multiplier applied to the loan balance exceeding the concentration threshold
>= 2.0x 1.5x
[1.50 - 2.0x) 2.0x
[1.25x - 1.50x) 2.75x
[1.10 - 1.25x) 3.75x
[1.00 - 1.10x) 5.0x
Below 1.0x 10.0x
*The symbol "[" denotes the inclusion of the first data point in the range, and the symbol ")" denotes the exclusion of the last data point in the range.

25. After application of the loan concentration adjustment to the base credit loss assumptions in table 6, we apply an additional adjustment for pools exhibiting further credit risks or strengths not captured in our base credit loss assumptions. Specifically, we may adjust the base credit loss assumptions by applying a multiplier, generally within a range of 0.8x-1.5x.

26. Examples of negative characteristics that lead to an upward adjustment (multiplier above 1.0x) include:

  • Atypically weak performance of loans, as indicated, for example, by low or decreasing DSC or severely delinquent loans;
  • Presence of unenhanced construction loans or loans with unmitigated construction-related risks (early financing, acquisition, or bridge loans); and
  • Atypical loan provisions that we consider increase risk.

27. Examples of positive characteristics that lead to a downward adjustment (multiplier below 1.0x) to the base credit loss include:

  • Atypically strong performance of loans, as indicated for example by increasing DSC trends;
  • Additional loan level oversight that could increase stability and performance (for example, Low Income Housing Tax Credit);
  • Atypically strong underwriting requirements that we consider mitigate additional risk (for example, loan level reserve funding requirements); and
  • Atypical loan provisions that we consider mitigate risk.
Recoveries and credit enhancement on single-family and multifamily loans

28. Some loans in the program may benefit from full or partial credit enhancement in the form of mortgage insurance, government guarantees, or other guarantees, which can minimize the loss severity should the loans default. We determine projected loss assumptions for enhanced loans in the same way as for unenhanced loans, and then factor in recoveries from insurance or guarantees. We calculate such recoveries according to our "Methodology For Assessing Mortgage Insurance And Similar Guarantees And Supports In Structured And Public Sector Finance And Covered Bonds," and, if applicable, "U.S. Government Support In Structured Finance And Public Finance Ratings." The percentage or amount of the recovery is based on the specific insurance or guarantee parameters and takes into account the nature and rating of the insurance or guarantee.

Projected loss assumptions for loans or bonds backed by the full enhancement of a U.S. GE or GSE

29. For single-family or multifamily loans or bonds benefitting from full enhancement from a GE or GSE, we determine recoveries through the application of U.S. government support criteria (and if applicable, the criteria article "Incorporating Sovereign Risk In Rating Structured Finance Securities: Methodology And Assumptions"). In application of those criteria, we typically assume full recoveries (that is, no projected losses) at or below the rating on the GE/GSE, and partial recoveries at rating levels higher than the rating on the GE/GSE.

Small exposures to other asset types

30. MRB programs sometimes include small exposures to other asset types (for example, small business loans). If such exposures are below a threshold of approximately 10% of the total asset pool balance, we determine projected loss assumptions for these other assets as follows:

  • We use as a starting point the base credit loss assumptions for multifamily loan pools, as set out in table 6, which we adjust upward by a multiplier. For exposures consisting primarily of secured (or otherwise enhanced) loans, this is generally in the range of 1.5x to 3.0x;
  • If we assess that the HFA has a meaningful track record of prudent lending in the relevant asset type, and that the asset type is central to the HFA's public mission, we apply a multiplier of 1.5x;
  • If we assess that the HFA does not have a significant track record of prudent lending in the relevant asset type, or that the asset type is not central to its public mission, we apply a multiplier of 2.0x to the base credit loss;
  • If we assess that the HFA does not have a significant track record of prudent lending in the relevant asset type, AND that the asset type is not central to its public mission, we apply a multiplier of 3.0x to the base credit loss;
  • We may apply a further upward adjustment to the projected loss assumptions, for example, if most of the loans are unsecured or unenhanced, or if we observe significantly weaker credit performance of the sub-pool of other assets relative to the pool's main asset type(s).

31. We perform a further analysis of these assets if an MRB program includes an exposure to other asset types exceeding approximately 10% of the total asset pool balance. We generally use our securitization criteria for the relevant asset type as a starting point, and may make adjustments in some instances (for example, to apply the analysis based on stratified instead of loan-level data). In the absence of sufficient data to perform such an analysis, we apply a 100% projected loss assumption on the balance of such assets exceeding the 10% threshold.


Chart 5

Analysis of overcollateralization, cash flow scenarios and assumptions

32. In order to determine the highest rating level at which the MRB program's pledged assets are sufficient to cover the applicable credit loss and cash flow stress scenarios, we compare the program's available overcollateralization to the credit loss determined in the previous section.

33. In our analysis of available overcollateralization, we generally consider program assets, reserves, cash, and investments legally pledged to the bonds. We generally measure available overcollateralization as the minimum asset-to-liability parity ratio, simulated over the life of the bonds under stressed cash flow scenarios.

34. The cash flow scenarios relevant to an MRB program are described in Appendix IV. We typically receive cash flow scenarios run by the HFA or a third-party cash flow provider engaged by the organization, which has a track record of providing cash flows for MRB programs. We expect cash flows to model mortgage loan payment lag, timing mismatch between mortgage loan payments and bond payments, negative arbitrage, interest rate stresses (see "Methodology To Derive Stressed Interest Rates In Structured Finance"), prepayment speed assumptions, and reinvestment earnings assumptions.

35. If we assess that available overcollateralization is only sufficient to cover credit loss assumptions at a rating level of 'BB+' or lower, we may constrain the anchor in the 'B' category if we project that the DSC ratio could decline below 1.0x under 'BB' category cash flow stress scenarios. We do not apply this constraint if we assess that the decline in the projected DSC ratio is mitigated by offsetting factors, such as a line of credit or cash deposit. Specifically:

  • We limit the anchor to no higher than 'b+' if we project the DSC ratio to fall below 1.0x within four to 10 years;
  • We limit the anchor to no higher than 'b-' if we project the DSC ratio to fall below 1.0x within four years.
Programs with hedged variable-rate debt

36. Mostly, MRB programs are backed by fixed-rate loans, and issue fixed-rate debt. Some MRB programs issue a portion of variable-rate debt, and some of these programs use derivatives to hedge the resulting interest rate mismatch between assets and debt. Derivatives are typically secured against the program assets and rank equally with the most senior class of notes issued under the program. However, they require the HFA to meet ongoing collateral posting obligations. This is typically a general obligation (GO) of the HFA, and therefore, the HFA's creditworthiness in respect of this obligation is measured by the HFA ICR. A failure by the HFA to post collateral when due (that is, a default on its GO) may lead to a termination of the contract, which would introduce unhedged interest rate risk to the MRB program, and may materially affect the MRB program rating.

37. To mitigate the risk of a significant transition in the MRB program rating following a default of the HFA on its GO debt, we may constrain the MRB program anchor to a maximum elevation above the HFA ICR, as follows (and as further detailed in Appendix III):

  • We apply no constraint if the anchor is no higher than the ICR, or if the hedged exposure is immaterial and the HFA ICR is at least 'AA-';
  • We can assign an anchor of up to three notches above the ICR, if the HFA can demonstrate that it has sufficient liquidity to post collateral in a stress scenario;
  • In the absence of such a demonstration, or to assign an anchor higher than three notches above the HFA ICR, we consider cash flow scenarios assuming the unhedged risk materializes immediately, to assess the materiality of the hedging support in the MRB program rating. We constrain the elevation of the anchor above the HFA ICR based on this materiality assessment.

38. In rare cases, an MRB program's derivative documentation may contemplate that payments to counterparties are made by the HFA, rather than from the MRB program's asset cash flows. In these cases, we constrain the anchor at the HFA ICR, or assume that the risk is unhedged in the cash flow analysis.

Programs with multiple tranches of debt

39. For a program with a senior/subordinated structure, S&P Global Ratings determines if there is a bona fide distinction between the rankings of the security of the senior and subordinated liens. In the absence of a clear-cut determination, we may assign the same rating to both the senior and subordinated bonds. We address seven key components to substantiate a clear senior and junior position with respect to bondholders' liens: security pledge to bondholders, additional bond provisions, redemption provisions, flow of funds, default/cross-default, bondholder rights and approvals, and miscellaneous items.

40. We determine a specific anchor for each class. We calculate available overcollateralization for each class, based on the liabilities for the specific class and any prior or equal ranking class. We determine for each class the highest rating level at which projected credit losses are covered by available overcollateralization. We also consider the availability of liquidity for each class separately, under the program's legal provisions, and the degree to which the dependency on cash flow support from the HFA may differ by class, if applicable.

Programs with HFA GO pledge

41. Some MRB programs benefit from the supporting HFA's GO pledge, meaning that the HFA has pledged to cover the program's debt obligations if cash flows from program assets are insufficient. We capture this support in our determination of the anchor within these criteria, rather than through separate application of our multiple revenue streams criteria (see "Related Criteria"). Where all bonds issued under an MRB program benefit from a GO pledge, the MRB program anchor, SACP, and final rating can be no lower than the HFA ICR. Where all bonds within a specific tranche benefit from the GO pledge, but other tranches within the MRB program do not, the anchor, SACP, and final rating for the tranche benefitting from the GO pledge are also generally no lower than the HFA ICR. However, the anchor, SACP, or final rating can never exceed those of a more senior tranche within the MRB program.


Chart 6

Liquidity reserves modifier

42. This analysis focuses on an MRB program's ability to cover its debt obligations through short-term disruptions in asset cash flows, based on its available liquidity reserves. It may result in a negative modifier. We do not apply any negative modifier on the basis of liquidity reserves for pass-through programs, or programs entirely backed by fully enhanced MBS. For pass-through programs, we consider liquidity risk to be mitigated structurally. For programs backed by fully enhanced MBS, we consider liquidity risk to be mitigated both by guarantees on the bonds and by the inherent market liquidity of the underlying securities. For programs backed partly by a pool of whole loans, and partly by fully enhanced MBS, we compare the amount of liquidity reserves to the debt service associated with the whole loan pool. Absent information about the specific amount of debt associated with the whole loan pool, we approximate it by subtracting the balance of the MBS assets from total debt.

43. We generally consider the minimum reserve amount definition for reserves that are legally pledged to the program. If liquidity is provided by a third party, such as through a surety bond, we assess the relevant counterparty risk management in the program management and operational risk analysis. We include liquidity pledged to the program, but held outside of reserve accounts, only if all of the following apply:

  • The funds are available for debt service;
  • The funds are not designated for a different purpose (such as the acquisition of loans);
  • We expect the HFA to maintain this liquidity in the program in the future.

44. We apply no modifier to the anchor if the minimum pledged reserves definition is sufficient to cover at least 2% of the mortgage loan pool balance (excluding any loan acquisition fund from the pool balance) or the succeeding 12 months of debt service. Otherwise, we lower the anchor:

  • By one notch, if the minimum pledged reserves definition is sufficient to cover at least the succeeding six months of debt service;
  • By two notches, if the minimum pledged reserves definition is not sufficient to cover at least the succeeding six months of debt service.

45. For programs benefitting from the HFA's GO pledge, if pledged reserves are insufficient to fully mitigate liquidity risk, we further assess the availability of common reserves not legally pledged to the specific MRB program. However, the reliance on non-pledged reserves creates a dependency on the HFA ICR. We apply no liquidity risk modifier, if the anchor is no higher than three notches above the HFA ICR, and we assess that common reserves are managed such that the MRB program will have available liquidity to continuously cover the succeeding six months of debt service. If the anchor is higher than three notches above the HFA ICR, we apply the liquidity reserves modifier as described above (but if the anchor is four notches above the HFA ICR, we don't apply a two-notch adjustment).

46. Some MRB programs benefit from an additional external pledge (for example, from the state) to replenish reserves up to a certain amount. If pledged reserves are insufficient to fully mitigate liquidity risk, we assess whether this additional pledge mitigates the risk. The reliance on such replenishment creates a dependency on the pledgor applicable rating (for example, a state moral obligation rating). If we assess that the MRB program would be able to access, in a timely manner, sufficient replenishment funds to cover at least the succeeding six months of debt service, we apply no liquidity reserve modifier if the anchor is no higher than the pledgor rating. If the anchor exceeds the pledgor rating, we apply the modifier as described above (but this modifier cannot bring the anchor below the pledgor rating).

Market position modifier

47. A program's market position reflects characteristics of the area in which the loans are located, which is generally within a single state. These characteristics may affect the performance of a loan pool over the life of the bonds (e.g., likelihood of elevated delinquencies or foreclosures) or indicate potential challenges to operating in a certain state or metropolitan area (e.g., the supply and demand of real estate). Therefore, the market position modifier aims to capture factors not already considered in determining the anchor.

48. Table 8 presents examples of risk factors related to market position. We assess the presence and severity of risks at the relevant market level, relative to the national housing market. Depending on our view of the presence and severity of these risks for all three categories, we could lower the anchor by up to two notches. We may apply no notching modifier, despite the presence of market position risks, if we assess that these risks are mitigated by extraordinarily strong overcollateralization (generally, as demonstrated by an asset-liability parity ratio that exceeds 200%). We may also apply no modifier, if we assess that a risk is already captured through the constraint on the MRB program rating relative to the HFA ICR, or through loss assumptions applied in determining the anchor.

Table 8

Market Position Risk Assessment
Risk categories Examples of factors that generally indicate the presence of risks
Socioeconomic conditions -- Weak forecast population trends
-- Forecast significant increases in unemployment rate
Housing market conditions -- Forecast negative trends not captured in the loss coverage analysis.
-- Material changes to our market outlook assessment under U.S. residential mortgage-backed securities criteria
Other considerations -- Unmitigated (e.g., in the form of insufficient insurance), material geographic risks arising from potential environmental events that threaten properties’ physical condition
-- Local, state, or federal policy changes are likely to negatively affect the performance of housing loans or the entity’s lending ability
Holistic analysis

49. A holistic analysis is considered after applicable modifiers and caps to help capture a broader view of creditworthiness. The holistic analysis can have a one-notch impact up or down and is not limited by any credit specific caps or overrides. It can also result in no rating change at all. The analysis may be based on credit risk factors including our forward-looking view of operating and financial performance, if not already incorporated in the anchor. It may also reflect a comparable ratings analysis, or program-specific strengths or weaknesses that are not fully reflected through the application of the criteria.


Chart 7


50. When applicable, we factor in the influence of other criteria to arrive at the final rating. In particular, we apply additional criteria (see "Related Criteria") where relevant to analyze U.S. sovereign risk, construction risk on the underlying mortgage loans, or to assign ratings in the 'CCC' or 'CC' categories.

51. In cases where MRB programs benefit from a guarantee or other secondary pledge from a third party, (for example, a state government), our guarantee criteria or multiple revenue streams criteria apply (see "Related Criteria").

52. Our rating analysis for MRB programs under these criteria incorporates an assessment of environmental, social and governance (ESG) risks if we believe they have the potential to affect the securities' creditworthiness (see "Environmental, Social, And Governance Principles In Credit Ratings"). These factors are most likely to be reflected in changes to the analysis of the credit quality of the asset pool, or to the market position or holistic modifier.


53. S&P Global Ratings maintains approximately 74 public ratings on programs that are in scope of these criteria. We expect limited impact on ratings outstanding. Assuming the managed pools maintain their current credit characteristics, our testing suggests that the initial application of these criteria will affect fewer than 10% of ratings, which we would raise or lower by one notch. For some MRB programs that rely on derivatives to hedge variable-rate debt, we will review addition information regarding the HFA's ability to post collateral, as well as additional cash flow scenarios, to determine any constraint to the MRB program rating, relative to the HFA ICR (see "Cash Flow Analysis/Programs With Variable-Rate Debt"). Note that in our December 2021 RFC, we referenced 151 outstanding ratings, which included 76 ratings on issues that belong to two MRB programs, and have since been consolidated into two program ratings, and three programs that have since been fully redeemed. Our expectation of impact is unchanged from the RFC.

Appendix I: Detailed Analysis Of Credit Quality For Pools Of Single-Family Whole Loans

Base foreclosure frequency assumptions

54. These reflect the foreclosure frequency of a typical pool of newly originated single-family loans associated with an MRB program. The typical pool definition and base foreclosure frequency assumptions are benchmarked against our U.S. RMBS criteria "Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later." We have observed stronger credit performance of loans originated or purchased by HFAs, through periods of stress, relative to other U.S. mortgage loans to borrowers with similar characteristics. This stronger credit performance is driven by HFAs' robust borrower screening and third-party servicer monitoring, their focus on borrower education and preventative screening measures, and lower mortgage interest rates that place a lesser debt burden on borrowers. Therefore, we apply the same base foreclosure frequency assumptions as in our U.S. RMBS criteria, but to a typical pool definition that includes first-time homebuyers with higher loan-to-value (LTV) ratios at origination (and therefore, apply lower upward adjustments to foreclosure frequency for higher LTV ratios, than in our U.S. RMBS criteria). Based on the calibration of these criteria, the foreclosure frequency assumption for a typical pool of loans originated or purchased by HFAs is approximately 0.7x the foreclosure frequency assumption for a U.S. RMBS pool with otherwise similar borrower characteristics.

55. We define the typical pool to include:

  • At least 250 loans;
  • Loans originated or purchased by an HFA, with conforming documentation;
  • Newly originated loans that are current at the time of addition to the pool;
  • Loans for first-time or other qualified home buyers
  • First-lien mortgages secured by single-family, owner-occupied primary residences;
  • Properties of no more than four units;
  • Fully amortizing, fixed-rate loans with a 30-year maturity;
  • Loans have not been modified; and
  • Average borrower FICO score of 710-725 and average LTV ratio of 80%-85%.
Foreclosure frequency adjustments

56. The adjustment factors are set out in charts 8 and 9 and tables 9 to 13. If the value for a variable(s) listed in this section or any other variable used in our analysis is missing, we may assess other information that we believe is comparable so we can assess the relevant risks and, depending on the nature of the other information, we may make assumptions that in our view adequately address the risk in our analysis. For example, if FICO scores at origination are unavailable on a portion of legacy loans within a pool, we use the average FICO score for loans where it is available, and may adjust based on the origination standards and performance of those legacy loans relative to the overall pool. As a further example, if we received a different type of borrower credit score data rather than FICO, we would determine the relativity of borrower risk between the alternative score and FICO, and adjust the ranges and multipliers in table 9 accordingly.

57. The LTV ratio used in determining the adjustment in chart 8, for each cohort of loans, is the equal-weighted average of the loans':

  • Current LTV, that is, the current balance of the loan divided by the current valuation of the property; and
  • Original LTV ratio, that is, the balance of the loan at time of origination divided by the value of the property at the time of origination.

58. We generally analyze MRB program pools on the basis of loan cohorts grouped by their reported ratio of current loan balance to original property valuation. We then determine the current LTV for each cohort, by indexing property values based on the evolution of the Federal Housing Finance Agency (FHFA) house price index from the date of the original valuation. We reflect an adjustment of 50% of any upward movement and 100% of any downward movement in the index. We may make additional adjustments to the FHFA index value if we believe current changes in house prices are not fully captured by the index, given the lag in index reporting or rapidly shifting market conditions.

Chart 8


Chart 9


Table 9

FICO Adjustment Factor
Pool average borrower FICO* Adjustment factor
>725 0.9
(710;725] 1.0
(695;710] 1.2
(680;695] 1.4
(665;680] 1.6
(650;665] 1.8
(635;650] 2.0
(620;635] 2.2
<=620 2.5
*The symbol '(' denotes exclusion of the first data point in the range, and the symbol ']' denotes the inclusion of the last data point in the range.

Table 10

Loan Type Adjustment Factor
Loan type Adjustment factor
Fixed rate, 30-year maturity 1.0x
Other 1.5x
Balloon or negatively amortizing loans 3.0x

Table 11

Seasoning Adjustment Factor
Loan seasoning Adjustment factor*
0-5 years 1.0x
5-10 years 0.75x
> 10 years 0.5x
*Applies only to performing loans.

Table 12

Property Type Adjustment Factor
Property type Adjustment factor
Single-family residence, PUD 1.0x
Condominium, cooperative 1.1x
Two/three/four family 2.0x
Other* 2.0x
*Typically includes the following property types: manufactured housing; mixed-use; raw land; or other property types that, in our view, do not fall into above three categories. We also assume 100% loss severity for “Other” property types. PUD--Planned unit development.

Table 13

Delinquency Status Adjustment Factor
Delinquency status Adjustment factor
Presently current 1.0x
Presently 30 days delinquent 2.5x
Presently 60 days delinquent 5.0x
Presently 90+ days delinquent/real estate-owned/foreclosure 100% FF
FF--Foreclosure frequency.

Small pool adjustment factor

59. The primary purpose of the small-pool adjustment factor is to capture idiosyncratic risks and sample bias that may exist in small loan pools (less than or equal to 250 loans). When applicable, the small-pool adjustment factor follows a functional form (see equation 2 below) to reflect the greater risk from higher borrower concentration. The small-pool loss coverage adjustment factors are displayed in equation 2.


Chart 10


Qualitative adjustment factor

60. The calibration of these criteria considers better historical performance for loans originated or purchased by HFAs, relative to the broader U.S. mortgage market. In exceptional cases, we increase the pool's WAFF by a factor of up to 2.0x, if we assess that the overall historical performance of HFA loans is less indicative of performance expectations for a specific MRB program. For example, we would reach this conclusion if we assess that an MRB program's servicing and/or origination standards deviate significantly from the norm for HFA managed pools. A further example is an MRB program that includes a prevalence of modified loans. We would apply a multiplier of the order of 1.5x if we expect performance to be more comparable with the overall U.S. mortgage market. We would apply a multiplier of 2.0x if we expect performance to be more comparable with the more aggressive end of origination standards in the U.S. mortgage market.

Loss severity analysis

61. Similar to our foreclosure frequency analysis, we generally analyze MRB program pools on the basis of loan cohorts grouped by their reported current loan/original property valuation ratio. To estimate a pool's loss severity, we start from the current LTV for each cohort, determined in the previous section.

62. Next, we apply our MVD assumptions, and determine a rating-level repo MVD as set out in table 14. Our MVD assumptions include an assessment of the level of over/undervaluation in the prevailing property market at the state or territory level by comparing the long-term average of the ratio of house prices to income, to the current ratio. In the absence of specific state- or territory-level information, we may use national information to determine the level of over/undervaluation. To measure the degree of over/undervaluation in a particular property market, we may use various data sources that are available, including the FHFA, U.S. Bureau of Economic Analysis, U.S. Census Bureau, expert opinions, and independent research.

Table 14

Property Market Adjustments For Calculating Repossession MVD Modeling Assumptions
Rating category Fixed MVD (%) % of overvaluation added % of undervaluation deducted Forced-sale discount (%) Repo MVD, absent over/under valuation (%)
AAA 40 50 (20) 10 46
AA 36 43 (20) 11 43
A 28 36 (20) 12 37
BBB 23 30 (20) 13 33
BB 19 25 (20) 14 30
B 15 20 (20) 15 28
Note: Repo MVD = 1 - [1 - (Fixed MVD +/- percentage of over/undervaluation x over/under valuation)] x (1-FSD). MVD--Market value decline. FSD--Forced-sale discount.

63. The percentage added or deducted based on a property's assessed over/undervaluation status is detailed in columns 3 and 4 in table 14. The degree of over/undervaluation in a property market is only partly reflected in the repo MVD formula. This is to account for our view that property values during a housing cycle in part reflect the current valuation of the property market, but revert to the long-term average. The adjustment for overvaluation is linked to a particular rating category, and is highest for a 'AAA' rating. By contrast, we limit the reduction of the repo MVD in an undervalued property market and therefore use a constant percentage as an adjustment for undervaluation across all rating levels.

64. Fixed MVD is the fixed recessionary MVD shown in the second column of table 14. The forced sale discount (FSD) factor is detailed in the fifth column of table 14. Among other factors, foreclosed properties may sell at a discount due to the stigma of repossession. The FSD factor is larger at lower rating levels and smaller at higher rating levels because in a more severe recession, a greater proportion of all property transactions come from distressed sales.

65. Our loss severity assumptions are no lower than a floor, to address any potential idiosyncratic risk on loans with very low current LTVs that is not otherwise captured by our assumptions. The loss severity floors are 20% at the 'AAA' rating category, 18% at the 'AA' rating category, 16% at the 'A' rating category, 14% at the 'BBB' rating category, 12% at the 'BB' rating category, and 10% at the 'B' rating category.

Liquidation timelines, lost interest, and foreclosure costs

66. In our view, liquidation timelines (which include the assumed default, foreclosure, and real estate owned time periods) show variability across states based on the specific legal processes for foreclosure. Hence, for the loss severity calculation, we assume varying timelines across states and rating levels. We generally apply the same rating-level specific liquidation timeline assumptions set out in our U.S. RMBS criteria. In exceptional cases, we may lengthen or shorten these assumptions if we determine that a specific HFA may have different liquidation timelines relative to the overall mortgage market in its state of domicile. For example, we shorten the liquidation timeline assumption if we assess that an HFA's use of "claims without conveyance of title" may accelerate the agency's foreclosure process relative to the overall market, and that this difference would persist in times of economic stress. In contrast, we lengthen the liquidation timeline if we assess that an HFA will be slower to execute foreclosures in times of economic stress, relative to the overall mortgage market in its state of domicile.

67. In our analysis, variable liquidation costs (that is, lost interest, and taxes and insurance) accrue throughout each loan's liquidation timeline based on the state where the property securing the mortgage is located.

68. We apply the same lost interest, foreclosure cost, taxes and insurance assumptions as set out in our U.S. RMBS criteria, in our single-family whole loan pool analysis.

Loss severity for other property types

69. We typically assume 100% loss severity for loans secured by an "other" property type at all rating levels (see table 12) due to the low likelihood of any meaningful recoveries.

Loss severity for subordinate liens

70. If an MRB program includes subordinate-lien single-family mortgage loans, such as those made for down payment and closing cost assistance, we apply a loss severity of up to 100% of the balance of these loans, depending on the terms and structures of the subordinate liens. For example, we may apply a 100% loss severity for the percentage of loans in the pool that are fully forgivable as compared with loans that are amortizing or deferred (due on sale).

Appendix II: Example Of The Application Of The Concentration Risk Adjustment For Multifamily Loan Pools

71. The following example considers the calculation of the projected loss assumption commensurate with a 'AAA' rating level for a multifamily loan pool:

  • The base credit loss assumption is 10% (as per table 6). For all loans in a pool with a balance up to 5% of the loan pool balance, we apply the base credit loss assumption of 10%.
  • If a pool includes one loan representing 25% of the loan pool balance, we apply the base credit loss assumption of 10% to the balance of this loan that is below the 5% threshold.
  • For the excess amount (20% of the loan pool balance), we apply a multiplier that is dependent on the loan's DSC. If the loan's DSC is 1.4x, we apply a credit loss assumption of 2.75 x 10% = 27.5%.
  • Assuming all other loan balances are below the concentration threshold, we calculate the weighted-average projected loss for the pool as: 80% x 10% + 20% x 27.5% = 13.5% of the loan pool balance.
  • If no further adjustments apply, we assume a 13.5% projected loss assumption in our cash flow analysis to determine if the MRB program's available overcollateralization is sufficient to support a 'aaa' anchor.

Appendix III: MRB Programs With Variable-Rate Debt

72. As described in paragraphs 36 and 37, we may constrain the MRB program anchor if there is material reliance on the HFA to post collateral on derivative contracts, to support the hedging of variable-rate debt. We determine the applicability and level of any constraint as follows:

  • We don't constrain the anchor, if it does not exceed the HFA ICR, as we consider that the ICR captures the HFA's ability to post collateral;
  • We also don't constrain the anchor, if the HFA ICR is 'AA-' or higher, and the amount of hedged variable-rate debt is no greater than 2.0x net assets after subtracting our projected credit loss assumptions (at the applicable rating level);
  • We assign an anchor up to three notches above the ICR, if the HFA demonstrates that it has sufficient liquidity to post collateral. The amount of collateral that the HFA would need to post is typically dependent on the MRB program rating or the HFA ICR. The demonstration should consider the amount required under all derivative documents in the event of a two-notch downgrade of the relevant rating. It should also consider the increase in collateral posting that would result from a 150-basis point shift in the relevant interest rate;
  • Otherwise, we constrain the anchor based on the HFA ICR and additional cash flow scenarios indicating the hypothetical anchor if the hedging were not present.

73. Specifically, we determine the "hypothetical unhedged anchor" at which our cash flow analysis demonstrates that the program assets would generate sufficient cash flows to meet bond payments in stressed interest rate scenarios, in the absence of the applicable hedge. In assessing whether the MRB program's available overcollateralization is sufficient to cover projected losses at each rating level, we assess available overcollateralization based on the minimum parity ratio resulting from this unhedged cash flow analysis (which would generally be lower than the minimum parity ratio resulting from a cash flow analysis that includes the applicable hedge).

74. We then determine the maximum anchor, based on the HFA ICR and the hypothetical unhedged anchor. If either the HFA ICR or the hypothetical unhedged anchor are speculative-grade, the maximum supported anchor is the higher of the two. Otherwise, the maximum supported anchor may be higher, as shown in table 15.

Table 15

Maximum Potential Anchor Based On The Reliance On HFA Support To Maintain Hedging
Hypothetical unhedged anchor
HFA ICR aaa aa+ aa aa- a+ a a- bbb+ bbb bbb-
AAA aaa aaa aaa aaa aaa aaa aaa aaa aaa aaa
AA+ aaa aaa aaa aaa aaa aaa aaa aaa aaa aaa
AA aaa aaa aaa aaa aaa aaa aaa aaa aaa aa+
AA- aaa aaa aaa aaa aaa aaa aaa aaa aa+ aa
A+ aaa aaa aaa aaa aa+ aa+ aa+ aa+ aa aa-
A aaa aaa aaa aaa aa+ aa aa aa aa- a+
A- aaa aaa aaa aaa aa+ aa aa- aa- a+ a
BBB+ aaa aaa aaa aaa aa+ aa aa- a+ a a-
BBB aaa aaa aaa aa+ aa aa- a+ a a- bbb+
BBB- aaa aaa aa+ aa aa- a+ a a- bbb+ bbb

Appendix IV: Cash Flow Scenarios And Assumptions

75. S&P Global Ratings analyzes transaction cash flows to ensure that the transaction generally covers timing and sufficiency of payments, considering payment lags, reinvestment, and unscheduled prepayments. Single-family housing transactions typically experience prepayments over time because there are multiple mortgage loans as well the ability to refinance or prepay principal at any time. Multifamily housing transactions typically pay as scheduled and often will contain a prepayment lockout period, but will also typically allow for a variety of scenarios in which unscheduled prepayments may occur--including: non-origination (i.e. stress scenario in which non-origination of any loans results in assuming a full redemption of bonds on the date specified in the bond documents), recapitalization or refinancing, or default.

76. The assumptions used in the cash flow scenarios ("cash flow assumptions") are typically in line with applicable criteria, including, but not limited to, "Methodology To Derive Stressed Interest Rates In Structured Finance," ("stressed interest rate criteria"), and "Methodology And Assumptions For Stressed Reinvestment Rates For Fixed-Rate U.S. Debt Obligations," ("stressed reinvestment rate criteria"). The cash flow assumptions would also encompass variable-rate debt, as well as capture the effects of failed remarketings resulting in bank bonds, constraints regarding the availability of swaps, liquidity support, and market access. In addition, they may also include the effects of swap counterparty terminations and unhedged variable-rate bonds.

77. This appendix is organized in two primary sections beginning with Cash Flow Assumptions, in paragraphs 77-94, which identifies the structural variables underlying each transaction type that should be included in the cash flow analysis. The "Cash Flow Scenarios" section, in paragraphs 95-109, identifies the various scenarios that reflect a range of economic situations and structural features specific to particular transactions that should be reflected in cash flows.


78. S&P Global Ratings expects the following assumptions to be addressed in the cash flow scenarios:

  • Acquisition period
  • Fees, expenses, and rebates
  • Payment lags
  • Prepayments
  • Reinvestment earnings
  • Variable-rate assumptions

Acquisition period

79. Generally, we expect cash flows to assume a worst-case draw scenario if mortgage loans are not originated or MBS are not delivered contemporaneously with bond closing. This stresses the transaction, as it demonstrates the least amount of interest earnings until the mortgage loan(s) fully amortizes (that is, the maximum amount of negative arbitrage). For example, if the bond proceeds are held uninvested (or invested and earning a rate of return lower than the mortgage rate), the cash flows would be modeled to assume that the loans are acquired on the last possible date that is permitted under the transaction documents. Conversely, if the bond proceeds are invested and earning a rate higher than the mortgage rate, we expect the cash flows to be modeled to assume mortgage loan acquisition immediately after bond closing.

Fees, expenses, and rebates

80. In general, cash flows should demonstrate that there is sufficiency of assets and revenues to pay debt service and expenses under all scenarios on which the fees and expenses are calculated. All fees and expenses, including payments to the federal government for rebate, should be demonstrated, unless provided for outside of cash flows.

Payment lags

81. Cash flows should demonstrate the ability to withstand the payment lags described for each program. A payment lag occurs between the mortgage payment by the borrower and when the trustee receives the mortgage payment. We expect any additional payment lags to be incorporated on top of any standard payment lag assumptions, if applicable, to account for the fact that mortgages pay in arrears. A minimum 30-day lag (in addition to normal arrearage) in receipt of mortgage payments on newly originated and existing loans should be applied to the cash flow projections. S&P Global Ratings may request a lag greater than 30 days depending on historical delinquency levels or the program; this will be considered on a case-by-case basis.

GE/GSE payment lags

82. Nuances for each GE/GSE program should also be considered and modeled in the cash flows. To account for additional delays in payments such as weekends or holidays, S&P Global Ratings requires an extra lag of five days for loans enhanced by Ginnie Mae, Fannie Mae, or Freddie Mac.


83. For most single-family transactions, we review cash flows that use the Standard Prepayment Model for prepayment speed assumptions from the Securities Industry and Financial Markets Association. As this model is the standard used by the industry, we expect that transaction cash flows will use it as the basis for prepayment speed assumptions (PSA). The Standard Prepayment Model (100% PSA) assumes that the prepayment rate will increase by 0.2% each month for the first 30 months until it peaks at 6% in month 30 and then assumes a constant prepayment rate in the 30th month and beyond. Prepayment speeds for cash flow stress scenarios are quoted as multiples of the 100% PSA. For example, 0% PSA would equal 0% x 100% PSA, 50% PSA would equal half of the standard model, and 200% would equal twice.

Alternate prepayment curves

84. We may consider transactions that use an alternate prepayment speed curve on a case-by-case basis, depending on the historical information available (e.g. tax-exempt issues are required by the IRS to calculate cash flows for yield compliance by using 100% of FHA prepayment experience). If an alternate prepayment speed curve is proposed for a transaction, we would consider among other things, the length of time over which the prepayment curve has been established, how it compares to the industry standard, and the reasoning for using it as an alternative.

Prepayment penalties

85. No prepayment penalties should be assumed in cash flows, as payment of these penalties may not be enforceable.

Reinvestment earnings

86. The applicable reinvestment earnings assumptions applicable at each rating level are generally those set out in our stressed reinvestment criteria. These stressed, fixed reinvestment rate assumptions reflect a prolonged low interest rate environment.

87. For MRB programs with fixed rate debt, in the absence of an investment agreement or permitted investment with a stated rate, the assumptions set out in table 1 of our stressed reinvestment criteria should be used. These assumptions also apply for any period not covered by a guaranteed investment contract (GIC) or if the GIC terminates prior to mandatory bond redemption or maturity, unless there is a provision for permitted investments with a stated rate.

88. For MRB programs where the cash flow analysis models variable-rate debt through our interest rate stress scenarios (see "Variable-rate transactions" below), the cash flow analysis may assume that a portion of reinvestment earnings varies with the stressed interest rate, such that reinvestment earnings increase in an increasing interest rate scenario. Specifically, for a balance of reinvested funds no greater than the balance of debt modeled as variable rate, the cash flow analysis may assume a reinvestment earnings margin of 50 basis points below the stressed interest rate throughout the simulation, floored at the applicable stressed fixed reinvestment rate assumption.

Variable-rate transactions

89. For variable-rate debt transactions with unhedged interest rate exposure, we would assume stressed interest rates as per our criteria "Methodology To Derive Stressed Interest Rates In Structured Finance."

90. Hedged transactions can be modeled at the swap or cap rate associated with the hedge. Generally, all risks identified under hedge contracts should be incorporated into the cash flow modeling projections as expenses or additional interest due on bonds. Reserve funding or interest rate spread should be shown to cover any shortfalls produced as a result of the modeling.

Standby bond purchase agreements (SBPA)

91. For transactions involving SPBAs covering liquidity, we would analyze cash flow projections that incorporate nearing termination, expiration or substitution. Cash flows should reflect alternative plans, if any, that will address the potential of increased liquidity fees upon any of these events. Absent any alternative plans, we would assume that any SBPA expiring within six months will not be extended or replaced by the expiration date and the current SBPA provider will be required to purchase the bonds and hold as bank bonds unless evidence of a liquidity commitment can be provided. The assumptions for bank bonds described below will then apply.

92. For the expiration of SBPAs beyond six months, S&P Global Ratings will analyze cash flows based on current market conditions in relation to liquidity fees being charged by providers. Higher-rated bonds will be analyzed for the potential to withstand higher liquidity fees. By taking into consideration any increased cost of liquidity and adjusting the expense by rating category, we believe that transactions exposed to liquidity fees may mitigate credit concerns.

Bank bonds

93. S&P Global Ratings will analyze cash flow projections incorporating alternative plans the issuer may be contemplating that will address the potential effect of bank bonds on the transaction, if warranted. S&P Global Ratings will analyze cash flow projections initially assuming that any existing bank bonds will earn the bank rate specified under the SBPA for six months. For SBPAs with repayment rates based on a variable rate, S&P Global Ratings will analyze cash flows assuming its stressed interest rate assumptions. Absent any alternative plans, or if bank bonds remain outstanding six months after becoming held by the bank, cash flows will be analyzed assuming bank bonds for the entire term-out period as specified by the SBPA.

94. Depending on the cause of the bank bonds, we may also analyze cash flows assuming that any bonds with liquidity support from a bank for which remarketing agents have been unable to remarket a portion of the related debt will become bank bonds as well. The same assumptions described above will apply to these bonds.

Basis risk

95. S&P Global Ratings will assess the basis risk experienced on the variable-rate bonds outstanding of the issuer and determine any assumed adjustment based on the issuer's actual experience. Assumptions may differ for taxable bonds, or for bonds subject to the alternative minimum tax. Higher-rated bonds will be analyzed for the potential to withstand greater levels of basis risk.


96. Typical cash flow scenarios are shown in table 16, and further detailed below, with the minimum scenarios typically including:

  • Base-case scenario
  • Non-origination/non-delivery scenario, if applicable
  • Worst-case scenario, if applicable

97. If applicable to the transaction structure, we may require additional cash flow scenarios listed in table 16 and as further explained below (see Other prepayment and stress scenarios).

Base-case scenario

98. For all transactions, with the exception of some exempt pass-through structures, we generally expect a cash flow scenario showing a base-case scenario of prepayments (also referred to as 0% PSA scenario). Under the base-case scenario, we expect the underlying mortgage loan(s) pays according to the terms of the mortgage documents with no prepayment(s) or any other unscheduled receipts such as casualty or condemnation made at any time.

99. We may alternatively consider a base-case scenario at a minimum prepayment speed higher than 0% PSA (but no higher than 30% PSA), depending on the historical level of prepayments that an issuer can provide to substantiate such assumption. We typically look at 15 to 30 years of data to capture multiple economic cycles through the life of the loan portfolio, with a minimum data set reflecting the average life of the loan portfolio. We would need to validate this data set and assumption on an annual basis.

Non-origination/non-delivery scenario

100. For all new issue transactions, with the exception of direct pay credit enhancement instruments and pass-through transactions, we expect to analyze a stress scenario in which loans related with the bond issuance are not originated or MBS are not delivered, resulting in assuming a full redemption of bonds on the date specified in the bond documents. S&P Global Ratings may waive this requirement if the transaction structure is such that the mortgage loans are originated and delivered prior to or contemporaneously with the transaction's closing (that is, there is no origination or acquisition period), or if the transaction's structure is such that the scenario is not necessary.

Worst-case scenario(s)

101. Generally, we would look for multifamily transaction cash flows to demonstrate a default occurring at the worst possible time in the life of the bonds. This scenario may vary or may include multiple scenarios, depending on the complexity of the portfolio. We may take into account certain loan or pool characteristics, such as credit quality and loan payment status, when we determine the loss timing to apply in the cash flow analysis. For example, if a pool of assets is predominantly of lower credit quality or shorter maturity, we would generally expect that defaults will occur more quickly and to a larger degree, both in terms of dollar amount and percentage of the portfolio. In addition, we may look for the cash flows to demonstrate default occurring with the highest coupon loan(s) or strongest performing loan(s) first, such that less profitable or weaker performing loans would have to maintain a greater percentage of debt service. The timing of losses in the cash flow analysis may also account for economic cycles, or when we expect recapitalization/refinancing to occur (e.g. tax credit compliance period expiration).

Table 16

Typical USPF Housing Cash Flow Scenarios
Scenario Description Single-family transaction Multifamily transaction Pass-through transactions
Base-case scenario Full origination of loans/0% PSA prepayment experience, or higher if applicable (see "base case scenario" above) All structures All structures All structures
Non-origination/non-delivery scenario Non-origination of all loans assuming a full redemption of bonds on the date specified in the bond documents in the event full origination does not occur All structures, with exceptions* All structures, with exceptions* All structures, with exceptions*
100% PSA scenario Full origination of loans/100% PSA prepayment experience All structures N/A N/A
Three-year average life scenario Full origination of loans/three-year average life of the mortgage loans prepayment experience All structures N/A N/A
Rapid prepayment scenario PSA prepayment speed sufficient to retire all bonds within two years after origination; however, depending on the mortgage loan interest rate, the issuer, and whether or not the bonds are part of a parity program, this scenario may be run at slower prepayment speeds that retire all bonds within a greater number of years after origination (see table 17) Structures rated above the GE/GSE or sovereign rating N/A N/A
Planned amortization class (PAC) bond scenario Cash flows are run at the PSA prepayment percentage that the PAC bond is structured at, which is the level at which all prepayments first go toward calling the PAC bond (typically around 100% PSA), until the PAC bond is called in full, and then at 0% prepayments until bond maturity Certain structures§ with planned amortization class bond N/A N/A
Super-sinker stress scenario Cash flows are run at the three-year average life of the loans prepayment rate until the super-sinker priority term bond is called in full, and then at 0% prepayments until bond maturity. Certain structures† with super-sinker bond N/A N/A
Capital appreciation bond (CAB) remainder stress scenario Cash flows are run at the three-year average life prepayment rate until all current interest and other non-call-protected bonds are called in full, and then at 0% prepayments until bond maturity Certain structures including a capital appreciation bond‡ N/A N/A
Liquidity stress scenario Cash flows are run at the same speed as the PAC/super-sinker scenarios above, shutting off at the point of greatest decline in prepayment moneys received and remaining at 0% until bond maturity Certain structures with PAC or super-sinker bonds that are not called pro rata** N/A N/A
40-year mortgage scenario Cash flows are run with the 30-year loans prepaying at the appropriate rapid speed in accordance with the rating, assuming there are no prepayments on the 40-year loans Certain structures with 30-year and 40-year loans N/A N/A
Worst-case scenario(s) Default and/or prepayment occurring at the worst possible time in the life of the bonds N/A All structures, with exceptions§§ N/A
*When loan origination/delivery occurs prior to or contemporaneously with transaction closing. §When net interest rate on the PAC bond is among the lowest of all bonds in the structure. †When super-sinker bond is present in structure (typically seen in older series of bonds within a parity resolution). ‡When call-protected. **When serial bonds are not called on a pro rata basis with the PAC or super-sinker bonds. §§Exceptions include non-FHA, fully federally enhanced multifamily transactions, or programs HFA parity resolutions with strong excess cash flow, or where the bonds are supported by a general obligation pledge. N/A--Not applicable.

Pass-through transactions

102. In the case of pass-through transactions, mortgage loan principal and interest payments, and prepayments, are passed through to bondholders as they are received. These transactions are structured to mitigate the potential timing mismatches and negative arbitrage that can occur in structures that have monthly mortgage payments and semiannual debt service payments. These structures also reduce the risk of administrative error due to the close match of the timing of payments. Due to the monthly pass-through nature of these transactions, reinvestment risk is mitigated and does not need to be modeled into the cash flow runs. Pass-through transactions typically contain a minimum of the following structural features:

  • Bond denominations matched with the securities denomination;
  • Payment dates on the securities synced with or scheduled prior to the payment dates on the bonds;
  • Pass-through mortgage rate(s) greater than or equal to the highest bond rate(s);
  • Fees and expenses paid outside of the trust or they are expressed as a percentage of the securities outstanding and incorporated into the pass-through mortgage loan interest rate;
  • Lags, negative arbitrage and applicable reserves funded at closing, and sized according to the respective program; and
  • No mandatory or sinking fund redemption schedule.

103. For pass-through transactions, we typically expect to receive cash flow reports reflecting the initial transaction assumptions, including a non-origination/non-delivery scenario and worst-case scenario. There may be cases where the timing mechanics of the transaction result in the transaction creditworthiness being non-dependent on cash-flows, and as such allow us to waive our requirement for cash flow scenarios, such as if the mortgage loan(s) or MBS are delivered prior to or contemporaneously with bond closing. We typically do not receive ongoing cash flow reports for pass-through transactions but we expect transaction documents to require a cash flow certificate for any material changes in structure or portfolio composition.

Other prepayment and stress scenarios

Rapid prepayment scenario

104. Transactions rated above the sovereign or GE/GSE rating should include this stress scenario. Cash flows are prepared at a prepayment speed sufficient to retire all bonds within two years after origination; however, depending on the mortgage loan interest rate and the issuer, this scenario may be run at slower prepayment speeds that retire all bonds within a greater number of years after origination, as shown in table 17.

Table 17

Rapid Prepayment Scenario For Programs Rated Above The Sovereign Or GE/GSE Rating
Interest rate (%) Years until full redemption of bonds
6.50 or lower 5
6.51 to 7.00 4.5
7.01 to 7.50 4
7.51 to 8.00 3.5
8.01 to 8.50 3
8.51 to 9.00 2.5
9.01 and higher 2

105. Depending on historical prepayment speeds, we may request a rapid prepayment scenario for programs rated at or below the sovereign and/or GE/GSE rating.

Planned amortization class (PAC) bond stress scenario

106. If the bond structure includes a PAC bond, this stress run may be needed if the net interest rate on the PAC bond, factoring in any premium, is among the lowest of all bonds in the structure. Cash flows are run at the PSA prepayment percentage that the PAC bond is structured at, which is the level at which all prepayments first go toward calling the PAC bond (typically around 100% PSA), until the PAC bond is called in full, and then at 0% prepayments until bond maturity.

Liquidity stress scenario

107. If serial bonds are present in the structure when either a PAC or super-sinker bond is present and are not called on a pro rata basis with the PAC/super-sinker, a liquidity stress run can show the result of prepayments (run at the same speed as the PAC or three-year average life for a super-sinker) shutting off at the point of greatest decline in prepayment money received and remaining at 0% until bond maturity.

Forty-year mortgage scenario

108. Mortgage loans with terms longer than 30 years, usually generate less revenue on a semiannual basis than 30-year loans. If 30-year and 40-year loans are in the same indenture, S&P Global Ratings may request an additional cash flow with the 30-year loans prepaying at the appropriate PSA in accordance with the rating, assuming there are no prepayments on the 40-year loans. This would indicate whether the indenture could maintain debt service payments with the support of 40-year loans alone.

Recycling stress scenario

109. For MRB programs backed by single-family whole loans, we may look for additional cash flow runs if mortgage loan recycling is permitted under the legal documents. It is optimal that documents specify that new (recycled) mortgage loans are to be made only at the same rate and existing term as the original (prepaid) loan and such prepayment proceeds are to be held no longer than six months before being used to redeem bonds, or as tax law permits. Recycling can be done with terms other than the same mortgage rate, term of the loan, or with different holding periods of prepayment proceeds as long as the specific terms as outlined in the trust indenture and mortgage documents are properly modeled in the cash flows.

110. Recycling stress scenarios generally include, but are not necessarily limited to:

  • Full origination based on worst-case draw/three-year average life prepayment experience/hold prepayment proceeds for longest time stated in documents/recycle all loans on worst-case delivery/then 0% prepayments on recycled loans.
  • Full origination based on worst-case draw/three-year average life prepayment experience/hold prepayment proceeds for longest time stated in documents/then non-delivery of all prepayment proceeds.

Appendix V: Glossary

111.Asset-to-liability parity.   The ratio of total assets to total liabilities. Typically, total assets include mortgage loans, revenues, investments, reserves, and other fund balances, and total liabilities include the amount of debt outstanding in a given period and accrued interest. Asset-to-liability parity of over 100% indicates overcollateralization or net assets.

112.Debt service coverage (DSC).   A measure of the transaction's ability to cover bond debt service and associated fees from mortgage loan revenues and reinvestment earnings, if any. For purposes of these criteria, DSC equals the revenue fund balance plus current mortgage loan principal and interest, less servicing and guarantee fees, all divided by bond principal, interest, and trustee fees in the same period.

113.Government entities (GE).   U.S. government agencies such as the Federal Housing Administration or related entities such as Ginnie Mae.

114.Government-sponsored enterprise (GSE).   A type of financial services corporation created by the U.S. Congress. Well-known GSEs are Fannie Mae and Freddie Mac.

115.Key transaction parties.   Any party whose failure to perform as contracted poses a risk to the credit quality of a transaction, such as to adversely affect the rating on the transaction.

116.Negative arbitrage.   When the assets under the trust estate are earning interest at a lower coupon than the liabilities under the trust estate need to pay. Negative arbitrage is similar to reinvestment risk in that it could cause a cash flow shortfall to occur if no offsetting factors are present. Typically, negative arbitrage relates to the period of time from bond closing until the point where the mortgage loan(s) are originated and paying interest.

117.Payment lag.   A delay in payment on the mortgage loan that is in addition to the time period encompassed from the date of origination until the first scheduled mortgage loan payment date.

118.Prepayment speed assumption.   For most single-family transactions, we review cash flows that use the Standard Prepayment Model for prepayment speed curve assumptions from the Securities Industry and Financial Markets Association. We may consider transactions that use an alternate prepayment speed curve on a case-by-case basis, depending on the historical information available.

119.Pass-through program.   A program in which a mortgage loan or MBS pays principal and interest that is "passed through" on a monthly basis, and payments are used for debt service during the same period. Pass-through programs typically have monthly payments and are not subject to a lag in mortgage loan or MBS payments, reinvestment earnings, or timing mismatches between the mortgage loan or MBS payment and the bond payment.

120.Reinvestment risk.   When assets under the trust estate are reinvested at a rate that earns less than what is needed to pay debt service plus expenses. Reinvestment risk is similar to negative arbitrage in that a cash flow shortfall can occur if no offsetting factors are present. In general, it is associated with any unscheduled receipt of funds including other events such as casualty and condemnation receipts, insurance receipts, and other inflows from the mortgage loan(s).

121.Transaction documents.   Legal documents including, but not limited to, the trust indenture, loan agreement, mortgage, credit enhancement agreement, and other documents detailing the transaction's terms and provisions.

122.Whole loan pool.   Pool of mortgage loans that are not securitized into MBS.

Related Publications

Fully superseded criteria
Partially superseded criteria (as they will no longer apply to ratings in scope of the MRB program criteria)
Related Criteria
Related Sector And Industry Variables
Related Research

This report does not constitute a rating action.

This article is a Criteria article. Criteria are the published analytic framework for determining Credit Ratings. Criteria include fundamental factors, analytical principles, methodologies, and /or key assumptions that we use in the ratings process to produce our Credit Ratings. Criteria, like our Credit Ratings, are forward-looking in nature. Criteria are intended to help users of our Credit Ratings understand how S&P Global Ratings analysts generally approach the analysis of Issuers or Issues in a given sector. Criteria include those material methodological elements identified by S&P Global Ratings as being relevant to credit analysis. However, S&P Global Ratings recognizes that there are many unique factors / facts and circumstances that may potentially apply to the analysis of a given Issuer or Issue. Accordingly, S&P Global Ratings Criteria is not designed to provide an exhaustive list of all factors applied in our rating analyses. Analysts exercise analytic judgement in the application of Criteria through the Rating Committee process to arrive at rating determinations.

Analytical Contacts:Aulii T Limtiaco, San Francisco + 1 (415) 371 5023;
Marian Zucker, New York + 1 (212) 438 2150;
David Greenblatt, New York + 1 (212) 438 1383;
Daniel P Pulter, Centennial + 1 (303) 721 4646;
Methodology Contacts:Andrew O'Neill, CFA, London + 44 20 7176 3578;
Kenneth T Gacka, San Francisco + 1 (415) 371 5036;
Lapo Guadagnuolo, London + 44 20 7176 3507;

No content (including ratings, credit-related analyses and data, valuations, model, software, or other application or output therefrom) or any part thereof (Content) may be modified, reverse engineered, reproduced, or distributed in any form by any means, or stored in a database or retrieval system, without the prior written permission of Standard & Poor’s Financial Services LLC or its affiliates (collectively, S&P). The Content shall not be used for any unlawful or unauthorized purposes. S&P and any third-party providers, as well as their directors, officers, shareholders, employees, or agents (collectively S&P Parties) do not guarantee the accuracy, completeness, timeliness, or availability of the Content. S&P Parties are not responsible for any errors or omissions (negligent or otherwise), regardless of the cause, for the results obtained from the use of the Content, or for the security or maintenance of any data input by the user. The Content is provided on an “as is” basis. S&P PARTIES DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, ANY WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE OR USE, FREEDOM FROM BUGS, SOFTWARE ERRORS OR DEFECTS, THAT THE CONTENT’S FUNCTIONING WILL BE UNINTERRUPTED, OR THAT THE CONTENT WILL OPERATE WITH ANY SOFTWARE OR HARDWARE CONFIGURATION. In no event shall S&P Parties be liable to any party for any direct, indirect, incidental, exemplary, compensatory, punitive, special or consequential damages, costs, expenses, legal fees, or losses (including, without limitation, lost income or lost profits and opportunity costs or losses caused by negligence) in connection with any use of the Content even if advised of the possibility of such damages.

Credit-related and other analyses, including ratings, and statements in the Content are statements of opinion as of the date they are expressed and not statements of fact. S&P’s opinions, analyses, and rating acknowledgment decisions (described below) are not recommendations to purchase, hold, or sell any securities or to make any investment decisions, and do not address the suitability of any security. S&P assumes no obligation to update the Content following publication in any form or format. The Content should not be relied on and is not a substitute for the skill, judgment, and experience of the user, its management, employees, advisors, and/or clients when making investment and other business decisions. S&P does not act as a fiduciary or an investment advisor except where registered as such. While S&P has obtained information from sources it believes to be reliable, S&P does not perform an audit and undertakes no duty of due diligence or independent verification of any information it receives. Rating-related publications may be published for a variety of reasons that are not necessarily dependent on action by rating committees, including, but not limited to, the publication of a periodic update on a credit rating and related analyses.

To the extent that regulatory authorities allow a rating agency to acknowledge in one jurisdiction a rating issued in another jurisdiction for certain regulatory purposes, S&P reserves the right to assign, withdraw, or suspend such acknowledgement at any time and in its sole discretion. S&P Parties disclaim any duty whatsoever arising out of the assignment, withdrawal, or suspension of an acknowledgment as well as any liability for any damage alleged to have been suffered on account thereof.

S&P keeps certain activities of its business units separate from each other in order to preserve the independence and objectivity of their respective activities. As a result, certain business units of S&P may have information that is not available to other S&P business units. S&P has established policies and procedures to maintain the confidentiality of certain nonpublic information received in connection with each analytical process.

S&P may receive compensation for its ratings and certain analyses, normally from issuers or underwriters of securities or from obligors. S&P reserves the right to disseminate its opinions and analyses. S&P's public ratings and analyses are made available on its Web sites, (free of charge), and (subscription), and may be distributed through other means, including via S&P publications and third-party redistributors. Additional information about our ratings fees is available at

Register with S&P Global Ratings

Register now to access exclusive content, events, tools, and more.

Go Back