Federal Agricultural Mortgage Corporation Governance, etc.:

Federal Register: April 27, 2011 (Volume 76, Number 81)

Rules and Regulations

Page 23459-23469

From the Federal Register Online via GPO Access [wais.access.gpo.gov]

DOCID:fr27ap11-2

FARM CREDIT ADMINISTRATION 12 CFR Parts 651 and 652

RIN 3052-AC51

Federal Agricultural Mortgage Corporation Governance and Federal

Agricultural Mortgage Corporation Funding and Fiscal Affairs; Risk-

Based Capital Requirements

AGENCY: Farm Credit Administration.

ACTION: Final rule.

SUMMARY: The Farm Credit Administration (FCA, Agency, us, or we) issues this final rule amending our regulations on the Risk-Based Capital

Stress Test (RBCST or model) used by the Federal Agricultural Mortgage

Corporation (Farmer Mac). This rulemaking updates the model to ensure that it continues to appropriately reflect risk in a manner consistent with statutory requirements for calculating Farmer Mac's regulatory minimum capital level under a risk-based capital stress test. This rule updates the model to estimate the capital requirements associated with

Farmer Mac's statutory authority to finance rural utility loans and to revise the treatment of certain secured general obligations held by

Farmer Mac as program investments. This rule also revises the treatment of counterparty risk on non-program investments in the model by adjusting the haircuts applied to those investments to keep the model internally consistent with revisions made to stressed historical corporate bond default and recovery rates.

DATES: Effective date: This regulation will be effective 30 days after publication in the Federal Register during which either or both Houses of Congress are in session. We will publish a notice of the effective date in the Federal Register.

Compliance date: Compliance with the changes to the model must be achieved by the first day of the fiscal quarter following the effective date of the rule. All other provisions require compliance on the effective date of this rule.

FOR FURTHER INFORMATION CONTACT:

Joseph T. Connor, Associate Director for Policy and Analysis, Office of

Secondary Market Oversight, Farm Credit Administration, McLean, VA 22102-5090, (703) 883-4280, TTY (703) 883-4434; or

Laura McFarland, Senior Counsel, Office of the General Counsel, Farm

Credit Administration, McLean, VA 22102-5090, (703) 883-4020, TTY (703) 883-4020.

SUPPLEMENTARY INFORMATION:

  1. Objective

    The objective of this final rule is to ensure that the RBCST for

    Farmer Mac continues to determine regulatory capital requirements in a manner consistent with statutory requirements.

  2. Background

    The FCA is an independent agency in the executive branch of the

    Federal Government that, in part, serves as the safety and soundness regulator of Farmer Mac. The FCA regulates Farmer Mac through the

    Office of Secondary Market Oversight (OSMO). Farmer Mac is a stockholder-owned instrumentality of the United States, chartered by

    Congress to establish a secondary market for agricultural real estate, rural housing mortgage loans, and rural utilities loans. Farmer Mac also facilitates the capital markets funding for USDA-guaranteed farm program and rural development loans. Section 5406 of the Food,

    Conservation and Energy Act of 2008 (2008 Farm Bill) \1\ amended the definition of ``qualified loan'' in Title VIII of the Farm Credit Act of 1971, as amended, (Act) \2\ to include rural utility loans. This change gave Farmer Mac the authority to purchase and guarantee securities backed by loans to rural electric and telephone utility cooperatives as program business. The

    Page 23460

    2008 Farm Bill further directed FCA to estimate the credit risk on the portfolio covered by this new authority at a rate of default and severity reasonably related to the risks in rural electric and telephone facility loans. The existing RBCST (Version 3.0) for Farmer

    Mac is contained in part 652, subpart B, and is used to determine the minimum level of regulatory capital Farmer Mac must hold to maintain positive capital during a 10-year period, as characterized by stressful credit and interest rate conditions. Version 3.0 of the RBCST was developed according to the provisions of section 8.32 of the Act before

    Farmer Mac was given rural utility authority and thus lacks a component to directly recognize the credit risk on such loans.\3\ The updated version of the RBCST will be identified as Version 4.0.

    \1\ Public Law 110-246, 122 Stat. 1651 (June 18, 2008)

    (repealing and replacing Pub. L. 110-234).

    \2\ Public Law 92 181, 85 Stat. 583 (December 10, 1971).

    \3\ FCA currently treats Farmer Mac's portfolio of investments in rural utility loans as non-program investments.

    On January 22, 2010, we published a proposed rule (75 FR 3647) to enhance the RBCST for Farmer Mac and to add a component addressing

    Farmer Mac's recently acquired authority to purchase and guarantee securities backed by loans to rural electric and telephone utility cooperatives. The comment period closed on April 22, 2010.\4\ This rulemaking finalizes policies proposed prior to the passage of the

    Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010

    (Dodd-Frank Act).\5\ Section 939A of the Dodd-Frank Act requires federal agencies to review all regulatory references to Nationally

    Recognized Statistical Ratings Organization (NRSRO) credit ratings by

    July 21, 2011, and, as a result of this review, to remove those references. While this rule maintains existing reliance on NRSRO credit ratings, the Agency intends to begin a rulemaking initiative immediately following this one to address the requirements of the Dodd-

    Frank Act.

    \4\ 75 FR 13682 (March 23, 2010).

    \5\ Public Law 111-203, 124 Stat. 1376, (H.R. 4173), July 21, 2010.

  3. Comments and Our Response

    We received several comments on the proposed rule from Farmer Mac and one comment letter from the Farm Credit Council (FCC), acting for its membership and each of the five Farm Credit banks. The FCC expressed support for using a more conservative approach to loss rate estimation in the AgVantage portfolio. It also noted its belief that capital standards for Farmer Mac should be equivalent to those of Farm

    Credit System (FCS or System) lenders. The FCC was also generally supportive of the proposed characterization of credit risk in the rural utility portfolio, but noted that the approach requires vigilant oversight of Farmer Mac's guarantee fee-pricing procedures.

    While we appreciate the FCC's comment, the Act provides for a different treatment of capital than that of the other System institutions. As such, the FCC's suggestion to make the capital standards equivalent to those of other FCS lenders is outside the scope of this rulemaking. Farmer Mac submitted comments on three aspects of the proposed rule--the method of characterizing credit losses on rural utility loans, the stress factor applied to the general obligation adjustment (GOA) to estimated losses in the AgVantage portfolio, and the concentration risk adjustment to the GOA factors. Farmer Mac stated that the proposed method of characterizing losses in the rural utility loans is not consistent across different market environments because it was too high relative to both the historical loss experience in that sector as well as levels that could be reasonably applied to agricultural mortgages. Farmer Mac also commented that the multiplier selected to stress GOA factors was too high, and the concentration risk adjustment to the GOA factors was unwarranted and duplicative to the use of credit ratings in the base GOA factors. Farmer Mac asked that the concentration risk be reversed in its impact to reflect a reduction in Farmer Mac's risk exposure in light of the counterparty's relative portfolio diversification.

    We discuss the comments specific to our proposed rule and our responses below. For purposes of responding to the comments made regarding GOA factors, we will be using the following terms to distinguish between the existing ``base GOA'' factors to refer to those set forth in Version 3.0, which are based solely on historical corporate bond default and recovery rates, and ``stressed GOA'' factors to refer Version 4.0 where base GOA factors are increased by a multiple of 3. Those areas of the proposed rule not receiving comment are finalized as proposed unless otherwise discussed in this preamble.

    1. Credit Loss Estimation on Rural Utility Loans [Sec. Sec. 652.50 and 652.65(b); Appendix A to Part 652] 1. Guarantee Fee

      We proposed amending Sec. 652.50 by adding a definition for guarantee fees charged on rural utility loans to distinguish treatment of these fees from those assessed against all other loans guaranteed by

      Farmer Mac. We explained ``rural utility guarantee fee,'' as it pertains to funded volume, means the gross spread over cost of funds, not a subset of that spread. Farmer Mac requested that we clarify whether or not the definition of ``rural utility guarantee fee'' is meant to reflect a subset of the term ``pricing spread.''

      We apply the term ``rural utility guarantee fee'' as a standalone term and not as a subset of pricing spread, and therefore, no component of the pricing spread should be netted. The rule defines ``rural utility guarantee fee'' as the actual guarantee fee charged for off- balance sheet volume and the earnings spread over Farmer Mac's funding costs for on-balance sheet volume on rural utility loans.\6\ As explained in the proposed rulemaking, we use the phrase ``earnings spread'' in the guarantee fee definition to represent the incoming cashflow rate minus Farmer Mac's total funding rate associated with that volume. We expect Farmer Mac to maintain records of these spreads when they are established for each transaction. We do not consider this an overly burdensome expectation given Farmer Mac's current practice of documenting such approvals of such spreads. Thus, the guarantee fee is the gross spread over cost of funds, not a subset of that spread. We are finalizing the definition as proposed. As a conforming technical change, we finalize amendments to sections 1.0.a., 4.1.b., 4.2.b.(2), and 4.2.b.(3) of the model in Appendix A of part 652 to add rural utility guarantee fees.

      \6\ For purposes of the mechanics within the spreadsheets of

      RBCST Version 4.0, on-balance sheet volume will, if necessary, be divided into those with AgVantage Plus-type structures and those that are outright loan purchases similar in structure to Farmer

      Mac's cash window for agricultural mortgages.

      1. Credit Risk

      We proposed amending the model in Appendix A of part 652 to include rural utility program volume by using a stylized approach to characterizing credit risk for rural utility program volume by multiplying the dollar-weighted average rural utility guarantee fee by a factor of two to characterize stressed annual loss rates.\7\ We also proposed clarifying the applicability of individual sections of the model to the

      Page 23461

      rural utility portfolio and adding new sections 2.6, 4.1.e., and 4.3.e. to calculate losses for rural utility loans.

      \7\ In the proposed rule, in this context, we used the phrase

      ``average annual loss rates.'' We believe the phrase ``stressed annual loss rates'' is clearer. What we intend to convey is that while agricultural lifetime loss rates are calculated by the model and then distributed on a front-loaded basis, we characterize rural utility loss rates as equal annual loss rates, or what could be referred to as average loss rates over a period of worst case stress.

      Farmer Mac objected to the proposed approach on the grounds that it results in projected stressed credit losses on rural utility loans that are inconsistent across different market environments and exceed both the historical experience in the rural utility sector and levels that could be reasonably applied to agricultural mortgages. Farmer Mac explained that the stressed credit loss characterizations on rural utility loans will be inconsistent across different market environments because it would be subject to inaccuracy due to potential volatility in the pricing by Farmer Mac of similar exposures under varying market conditions through time. In other words, investor risk tolerances vary with changes in perceived levels of overall risk in the market, and such changes could enable Farmer Mac to charge higher rates on rural utility loans despite no change in the underlying fundamentals of the sector or the specific loans it guarantees. We disagree with the suggestion that the stressed credit loss characterizations on rural utility loans will be inconsistent across different market environments. We used a multiple of the Farmer Mac rural utility guarantee fee as a proxy for stressed loss rates because the data on historical losses are not suitable for the development of a more statistically reliable estimate. We elected not to decompose the guarantee fee and earnings spreads into their component parts

      (including required versus ``excess'' spread) as that approach would have: (1) Required significant assumptions regarding what portion might be attributable to Farmer Mac's perception of market conditions versus credit risk; and (2) added a level of calculation complexity that is disproportionate to the coarse level of precision achievable given the data limitations. In other words, we take the view that the market clearing price reflects the market consensus of risk at a point in time.

      Farmer Mac asserts that the proposed approach is also incongruous because it characterizes losses of on- and off-balance sheet rural utility volume identically, though the rural utility guarantee fee would be inherently different. Farmer Mac suggests that the earnings spread on on-balance sheet volume might be larger than the guarantee fee on off-balance sheet volume. Farmer Mac clarified this comment by explaining that the return on equity component of the earnings spread would be larger for on-balance sheet volume ``[i]f the return on equity pricing is determined using current statutory minimum capital requirements (or any other capital requirements set using a differential approach to capital allocation).'' The comment references the statutory minimum requirements for on-balance sheet exposure (2.75 percent) and off-balance sheet exposure (0.75 percent) of outstanding principal. We understand the comment to indicate that program investment decisions, i.e., capital allocations, might be made on the basis of some required equity return margin over the associated statutory minimum capital requirements rather than on the basis of the risk and expense characteristics of the investments. We disagree with this premise. We are aware of no reason to base return on equity requirements on fixed statutory minimum capital requirements or to use such minimum capital requirements as a proxy for capital allocated to specific program investments. We reject the suggestion that such fixed minimums could be appropriately used as a basis to justify differential return on equity requirements on investments that have otherwise exactly the same risk and expense characteristics.

      Farmer Mac also commented that a multiple of two times the rural utility guarantee fee would not be consistent with FCA's stated position that the agriculture sector is generally more risky than the rural utility sector. Farmer Mac used a hypothetical example to demonstrate its comment. In this example, the cumulative annual loss rate characterization on rural utility volume over the 10 years of the modeling horizon slightly exceeded the estimated lifetime loss rate on newly originated, agricultural loans underwritten according to Farmer

      Mac's minimum standards. Farmer Mac modified the example to create a situation where the two sets of loans were equally seasoned and concluded that the cumulative loss rate for electrical loans in such cases would always exceed that of the agricultural real estate loans.

      Farmer Mac explained that the example demonstrated that the rule's approach would not be consistent with the statute's authorizing language requiring modeled loss rates to be ``reasonably related to risks'' in rural electric and telephone facility loans. Farmer Mac instead suggests that cumulative loss rates should, at the very least, be no greater than those for comparably sized agricultural mortgage loans. While Farmer Mac noted that the multiplier of two could be reduced, it instead asked FCA to adopt a credit risk estimate supported by historical loss and recovery rate trends.

      We disagree with the commenter's use of FCA Bookletter BL-053,

      ``Revised Regulatory Capital Treatment for Certain Electric Cooperative

      Assets,'' to support the contention that the proposed treatment is inconsistent with the bookletter's conclusion that the electric cooperative sector has a lower risk profile than the agricultural sector.\8\ While under normal conditions an average dollar of exposure to a rural electric cooperative is viewed as a lower credit risk than an average dollar of agricultural real estate mortgage exposure, the purpose of the RBCST is to represent a worst-case loss scenario for program-related assets. We view the concept of ``worst case'' in the rural utility cooperative sector as fundamentally different from the agriculture sector. The rule's approach inherently reflects our expectation that worst-case losses in the rural utility sector will occur far less frequently than worst-case losses in the agriculture sector--but when they occur, can be far more severe. While the average annual loss rate over the long term may be viewed as likely to be lower in the rural utility sector due to the infrequent occurrence of loss events, in a scenario where worst-case losses do occur, they will involve much greater loss rates than worst-case losses in agriculture.

      Further, the relationship between the two cumulative 10-year loss rates

      (agricultural versus rural utility) is not instructive, as the sector with the higher cumulative rate will vary depending on rural utility guarantee fee rates and the credit risk characteristics of the agriculture portfolio at any given time. Thus, in attempting to characterize both sectors' worst-case scenarios in the RBCST over a 10- year modeling horizon, having 10 years of loss rates that do not always sum to lower cumulative rate in the rural utility portfolio is not inconsistent with the general tenet that the electric cooperative sector typically has a lower risk profile.

      \8\ While BL-053 pertains to Farm Credit System banks and associations, and not to Farmer Mac, we believe the general tenets set forth in it apply to those same certain loan types in Farmer

      Mac's portfolio.

      Notwithstanding our position on this comment, using the suggested approach, it would be more appropriate to compare cumulative loss rates only to the modeling year at which the model indicates capital would approach its limit of zero (the zero-year) because losses recognized by the model in subsequent modeling years do not impact the calculation of the minimum capital requirement. Expanding on

      Page 23462

      Farmer Mac's example, if the zero-year occurred at year three, cumulative losses over those 3 years in agriculture portfolio would be 9.87 percent versus 4.2 percent in the rural utility portfolio.

      Seasoning could further affect the relative impacts of credit risk in the model. Given our stated view of the fundamentally different concepts of ``worst-case'' in the two sectors, this fact does not contradict the Agency's stated position.

      Farmer Mac's comment goes on to suggest various approaches to achieve the ``result'' recommended (that cumulative losses projected in the RBCST for rural utilities loans should be, on a relative basis, no greater than those for comparably sized agricultural mortgage loans).

      Farmer Mac notes that this result could be achieved by reducing the multiplier of two, but suggests instead that we abandon the proposed approach of applying a multiplier to Farmer Mac pricing factors in favor of an approach that references historical loss trends. In the proposed rule's preamble, we discussed in detail the insufficiency of historical lost trend data, as well as other alternatives to the proposed approach that were considered and why they were rejected.

      Farmer Mac also stated that the proposed approach was inconsistent with historical loss trends. We disagree because the comment is based on the premise that appropriate historical loss trend information is available. As discussed in the proposed rulemaking, we determined that a data set suitable to build a reliable default probability loss function is not available due to the fact that historical losses in the electric cooperative sub-sector of the utilities industry have been extremely rare and dissimilar.\9\ We also note that historical instances of default appear largely unrelated to specific underwriting decisions. Further, even among the few historical instances of non- performing loans in the data we obtained, restructured credit defaults have in many instances become more profitable than the original loan in terms of interest income, while others were never fully resolved despite exceptionally long periods of time since initial default. For those reasons, an empirical frequency-based analog for estimating credit risk, as was used to arrive at the model's approach to estimating agricultural loan risks, was not feasible for rural utilities. Instead, the rule characterizes credit risk on rural utility loans using the stylized approach of multiplying the dollar-weighted average rural utility guarantee fee by a factor of two to characterize stressed annual loss rates.

      \9\ In evaluating the suitability of empirical data sources, we examined historical loan performance data of the U.S. Department of

      Agriculture's (USDA) loan programs and interviewed market participants including the National Rural Utility Cooperative

      Financing Corporation, CoBank, and USDA's Rural Utility Service.

      Finally, Farmer Mac commented that the proposed approach to characterizing credit losses in the rural utility portfolio is inconsistent with the Act. We disagree with this assessment because the

      Act does not require us to use any particular statistical methodology.

      The Act, at section 8.32(a)(1)(B), requires us to estimate credit loss risk ``at a rate of default and severity reasonably related to risks in electric and telephone facility loans * * * as determined by the

      Director [of OSMO].'' The proposed rulemaking explained in some detail the reason behind selecting the method of identifying rural utilities credit loss risk, and Farmer Mac has offered no evidence to demonstrate that our method does not reasonably relate to actual risks in the rural utilities sector.

      We selected a method that relies directly on the notion that the assessment of relative risk would be reflected in differences in priced guarantee fees charged by Farmer Mac. These fees represent Farmer Mac's estimate of likely long-term average annual losses on an investment, in addition to fee loads to cover operating costs and return-on-equity requirements. We selected the combination of the total earnings spread with a lower stress multiple because the total spread also represents agreement on the value of the transaction between at least two parties:

      Farmer Mac and its counterparty (i.e., a market clearing price).

      For these reasons, we finalize this section and the conforming changes as proposed to reflect the treatment of the rural utility authority. As we gain more experience and data in this sector, the

      Agency may revisit this approach.

    2. Modification of the Treatment of Loans Backed by an Obligation of the Counterparty and Loans for Which Pledged Loan Collateral Volume

      Exceeds Farmer Mac-Guaranteed Volume [Sec. Sec. 652.50 and 652.65(d);

      Appendix A to Part 652]

      We are amending sections 2.4.b.3, 2.4.b.4, 4.1.f., and 4.2.b. of the model in Appendix A of part 652 to increase the GOA factors, address counterparty concentration risks, and ensure AgVantage Plus volume maturities are recognized in the model. 1. GOA Factors--Treatment of Loan Volume

      We proposed revising the GOA factors by stressing the historical corporate bond loss rates to levels intended to represent stressed conditions instead of average conditions. We accomplish this in the model by modifying the GOA factors through the application of increases

      (or ``haircuts'') to the estimated historical loss rates by whole- letter credit rating category using a multiple of three.

      Farmer Mac commented that our selection of three as the multiplier appeared to be much too high based on data in reports issued by Moody's

      Investor Services. Farmer Mac explained that the multiple and its implied assumption of a coefficient of variation (CV) equal to one lacked empirical support or theoretical justification. Farmer Mac askedthat the implied underlying CV ratio be much lower than one and that separate multipliers, scaled by whole-letter credit rating, be applied based on the historical variability over time of each whole- letter credit rating. Farmer Mac based this request on Moody's data on the standard deviations for 10-year cumulative default rates. Farmer

      Mac recommends these data be used to derive empirically based multiples of GOA factors to represent stress on issuer counterparties.

      We disagree with the recommendation as we believe it to be based on a mistaken reliance on CVs of average default rates within credit rating categories over time, rather than cross-sectional CVs of the individual issuer defaults within each period.\10\ The long-term average rate of the annual average default rate combined with the standard deviation of those average default rates do not convey a reasonable measure of ``worst-case'' default risk, but rather, as identified in the Moody's report, are primarily related to sample size used in construction of the estimated average loss rates. We believe our approach places the adjusted corporate bond loss estimate in a range that provides a meaningfully stressful representation, given limited data, and reflects generally accepted statistical principles and relationships. We selected the multiplier of three on the basis that it was a reasonable policy position given that the most accurate alternative to the selected multiple using statistical theory to establish the limits on probability from the sample variance (i.e.,

      Chebychev's theorem as discussed in

      Page 23463

      the proposed rule) would have yielded a proposed multiple many times higher than three. We continue to believe that use of the limit of probability established through limited sample information to require too extreme a multiple, and instead maintain our more moderate treatment through the use of our proposed value of three.

      \10\ In the proposed rule, we used a CV of one in an example to demonstrate a point and not as a factual premise of this rulemaking.

      We further disagree that one can accurately infer individual variability directly from the variance of a set of pooled experiences

      (aggregate annual default rates) through time. The primary purpose of the cited report, as explained by Moody's in the report, appears fundamentally different from its use in the comment letter. Moody's report explicitly states its purpose is to present confidence intervals around historical average cumulative default rates and, as warning against interpretation as a cross-sectional variance, the report indicates that standard errors around estimated long-run average default rates ``should not be confused with the much greater bands of uncertainty associated with the expected performance of particular cohorts of issuers formed at specific points in time (cross section).''

      \11\

      \11\ Cantor, R; Hamilton, D.; Tennant, J. ``Confidence Intervals for Corporate Default Rates'', Moody's Investor Services, Global

      Credit Research: Special Comment, April 2007; p. 1-2.

      We finalize this provision as proposed. 2. GOA Factors--Concentration Ratios

      We proposed modifying GOA factors to recognize the risk associated with a counterparty's (also referred to as the AgVantage Plus issuer) loan portfolio concentration in the industry sector used in an

      AgVantage Plus issuance. We also proposed modifying section 2.4.b.3.A. of Appendix A to allow the Director of OSMO to make final determinations of concentration ratios on a case-by-case basis by using publicly reported data on counterparty portfolios, non-public data submitted and certified by Farmer Mac as part of its RBCST submissions, and generally recognizing two rural utility sectors--rural electric cooperatives and rural telephone cooperatives.

      Farmer Mac objected to the GOA modifications because it believes the change creates redundancy in two ways: (1) The level of an issuer's loan portfolio concentration is already captured in the NRSRO's credit rating and therefore already captured in the level of the base GOA factor (prior to the proposed concentration risk adjustment), and (2) base GOA factors already capture stress associated with ``tail'' events according to the newly proposed stressed corporate bond loss-rate multiple. Farmer Mac suggests instead that the new GOA factors be adjusted to reflect a reduction in risk due to the level of diversification of the issuer, not an increase in risk due to the issuer's portfolio concentration.

      Farmer Mac further commented that the proposed methodology is vague and might oversimplify industry concentration. Farmer Mac asked that at least two sub-sectors of rural electric utilities be recognized in the concentration adjustment: Distribution cooperatives and generation and transmission (G&T) cooperatives. Farmer Mac explained that the magnitude of the concentration risk-adjusted GOA (CRAGOA) factors are driven more by the concentration risk adjustment than by the stressed historical corporate bond default and recovery rates (stressed GOA factors). Farmer Mac states that this is counterintuitive to the concept of the GOA because it associates more of the final effect of the CRAGOA adjustment with the issuer's portfolio structure than is warranted. Farmer Mac illustrates this point using the example of a sovereign issuer without credit risk. In this scenario, the CRAGOA factor would equal the concentration ratio, due to the mathematical relationship between the stressed GOA (pre-concentration risk adjustment) and the CRAGOA (i.e., 1-(1-GOA) (1-concentration ratio), where GOA = 0)). If that concentration ratio were one, then no risk- mitigation would be recognized in the general obligation of the sovereign issuer even if the issuer were rated AAA. Farmer Mac views this as placing an overly heavy emphasis on the issuer's portfolio concentration.

      Farmer Mac contends that our approach is inherently deficient because, in the example, the percentage increase in the GOA factor after adjustment for concentration risk is much greater for the AAA issuer (1,800 percent) than it is for the BBB issuer (300 percent), though the magnitudes of change stated in percentage terms are actually artifacts of the scale of remaining credit risk within each whole- letter rating category, as we discuss in depth below. Farmer Mac commented that the concentration risk adjustment should, if it has any impact at all, reduce risk rather than increase risk. Farmer Mac suggested replacing the mathematical relationship we had proposed with a multiplicative relationship--i.e., because the concentration ratio will frequently be less than one, that the stressed GOA factor should be reduced for any level of issuer portfolio diversification, rather than increased for any level of portfolio concentration. Farmer Mac suggests the following formula: CRAGOA = stressed GOA * CR.

      We appreciate Farmer Mac's concern that the two sub-sectors of rural electric utilities be recognized. However, we believe the rule provides for recognition of those sub-sectors and others on a case-by- case basis. We recognize Farmer Mac's authority to finance four industry sectors: Agriculture (including farms and agribusiness), rural electric distribution cooperatives, rural electric G&T cooperatives, and rural telephone cooperatives. The modifications to section 2.4.b.3.A. of Appendix A will allow the Director of OSMO (Director) to make final determinations of concentration ratios, including recognizing two rural utility sectors--rural electric cooperatives and rural telephone cooperatives. However, we disagree that the GOA factors contain redundancy. While NRSRO's may consider the extent of diversification of assets generally in their credit ratings, they do not do so in a worst-case context. Nor would the NRSRO's consideration of diversification always specifically include the impact of the issuer's relative exposure to industry sectors that Farmer Mac is authorized to finance. Agriculture and rural utility cooperative exposures are often combined with other sector exposures in publicly reported documents--including sectors that Farmer Mac is not authorized to finance. While it's possible that an NRSRO might require the issuer to disaggregate that information, its rating determination would not specifically focus on the degree of exposure to the Farmer Mac- authorized sectors. Hence, credit ratings do not provide the level of granularity of information needed. Nor does an NRSRO rating necessarily consider the issuer's exposure to the specific industry sector involved in the specific AgVantage Plus pool being modeled as this approach does. We do not believe that consideration of these specific risk components to the modeling of AgVantage Plus volume is sufficiently reflected in credit ratings to use them as suggested. For example, an

      NRSRO rating on a 100-percent concentrated issuer (e.g., a single- sector lender) says little or nothing about its ability to guarantee the credit on loan volume that it would pledge to Farmer Mac. In a worst-case loss scenario in that single sector, the issuer's ability to liquidate its unpledged assets to fulfill its general obligation to

      Farmer Mac at a price near the outstanding principal would be

      Page 23464

      severely reduced. This rule effectively evaluates the degree of that reduced ability at 100 percent. In other words, we do not believe it to be plausible that an issuer whose unpledged assets are experiencing worst-case losses would be able to continue as a going concern if it were forced to liquidate a significant volume of those unpledged, but highly impaired assets in order to fulfill its general obligation to

      Farmer Mac.

      Farmer Mac asked that we define the sectors but did not suggest any definition with the request. We decline to do so because we believe the general understanding of what these sectors include is sufficient for setting a parameter but flexible enough to allow the Director to use his discretion in a manner appropriate to each case presented. In addition, we do not view the fact that the concentration risk adjustment has a significant impact on the CRAGOA as counterintuitive.

      We believe it is logically consistent to view the concentration ratio as potentially a more significant driver of the value of the issuer's general obligation than the estimated corporate bond loss rate. We view the concentration risk adjustment as a critical component of the CRAGOA because it reflects the ability of the specific counterparty to augment the more generalized component derived from stressed corporate bond default rates by whole-letter credit rating.

      Farmer Mac's comment included an example of a sovereign (credit- risk-free) issuer and AgVantage Plus counterparty. We believe this example is too extreme to be applicable even for illustrative purposes.

      As a risk-free issuer, the hypothetical sovereign issuer in the example would be guaranteeing the credit risk on the subject loan volume, thus making the transaction more akin to the Farmer Mac II program than to the AgVantage Plus product.\12\ The RBCST already contains an approach on this type of transaction, i.e., it does not recognize credit risk and therefore would it not be appropriate to model this volume using the treatment for AgVantage Plus. Such transactions would result in a gross loss estimate of zero to which the CRAGOA (equal to the concentration ratio as previously discussed) would be applied for a net loss estimate of zero. However, to the more general point outside of this extreme case, i.e., a single-sector AAA issuer, we believe it reasonably and logically consistent for the single sector characteristic to weigh most heavily in the CRAGOA. The discussion and tables below further describe these relationships.

      \12\ Farmer Mac's program investments in loans that are guaranteed by the USDA as described in section 8.0(9)(B) of the Act, and which are securitized by Farmer Mac, are known as the ``Farmer

      Mac II'' program.

      Farmer Mac argued that our approach is inherently deficient due to the fact that the CRAGOA factor increases (relative to the stressed

      GOA) so much more for the AAA issuer (18 times) than it does for the

      BBB issuer (three times). We disagree and use the following tables to illustrate the ultimate effects of the CRA across a set of cases that we believe provide a more meaningful context for interpretation of the effects of its application.

      The table is organized in three panels across base Pre-GOA probability of default rates (PD) of 1, 3, and 6 percent (i.e., examples of loss rates as would be determined by the RBCST credit loss module or from the rural utility guarantee fee). The stressed GOA (GOA

      Pre-CRA) is applied to each case and a pre-concentration risk adjusted loss rate provided in column D (Pre-CRA loss rate). The first table assumes a 25-percent concentration ratio (CR) and provides associated final loss rates in column F after the CRA. Column G reproduces the multiples of change cited by Farmer Mac in its comment.

      A

      B

      C

      D

      E

      F

      G

      GOA Pre-

      Pre-CRA

      Loss rate

      Pre-GOA PD

      CRA

      loss rate

      CR

      post-CRAGOA

      = F/D

      (percent)

      (percent)

      (percent)

      (percent)

      (percent)

      AAA...............................

      1

      1.41

      0.0141

      25

      0.261

      18.48

      AA................................

      1

      3.70

      0.0370

      25

      0.278

      7.51

      A.................................

      1

      5.13

      0.0513

      25

      0.288

      5.62

      BBB...............................

      1

      11.48

      0.1148

      25

      0.336

      2.93

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT