Loss Reserves are Falling, But They Could Soon Be on the Rise

Transcription

1

2 FEBRUARY 2013 WHITEPAPER Authors Christian Henkel Enterprise Risk Solutions Director Jan Larsen, CFA Enterprise Risk Solutions Associate Director Contact Us AMERICAS EUROPE ASIA (Excluding Japan) clientservices.asia@moodys.com JAPAN clientservices.japan@moodys.com Visit: Info@moodys.com Loss Reserves are Falling, But They Could Soon Be on the Rise In December 2012, the Financial Accounting Standards Board ( FASB ) and the International Accounting Standards Board ( IASB ) released their proposed update to the accounting standards for the Allowance for Credit Losses ( ACL ). The standards are intended to address weaknesses in the prior guidance that were revealed by the financial crisis. Specifically, the guidance seeks to facilitate more timely recognition of likely losses, and also seeks to simplify the prior guidance by establishing a single standard for impairment. In this whitepaper, we argue that the banks who best adapt to the new standards will be at a competitive advantage, and not only because of the edge inherent in setting more accurate provisions. The same systems and processes that can help you comply with the new standards can be leveraged to improve a wide array of core business activities.

3 The ACL is one of the most significant estimates on a bank s financial statements and regulatory reports. Because it has a direct impact on earnings, the quantification of ACL is a frontburner concern for banks and their stakeholders. Since 1987, loan loss provisions have consumed 13 percent of banks net operating revenue (net interest income plus total noninterest income). By comparison, that is about half of what banks have spent on total salaries and benefits over the same period. The banking industry saw provision expenses climb to as much as 38 percent of net operating revenue during the recent financial crisis, compared to as little as five percent during expansionary periods. 1 PROVISION FOR LOAN LOSSES TO NET OPERATING REVENUE ( ) Source: FDIC 1 The proposed standards update, called the Current Expected Credit Loss Model ( CECL Model ), is meant to replace the existing incurred loss estimation approach. According to FASB, the CECL estimate of ACL is neither a worst case scenario nor a best case scenario, but rather reflects management s best estimate of the contractual cash flows that the bank does not expect to collect over the life of the loan. It takes into account the historical loss experience for like assets under similar economic conditions, and incorporates reasonable and supportable economic forecasts. 2 Because economic conditions are to be explicitly accounted for, it will no longer be acceptable to base the ACL on the borrowers creditworthiness alone. The new CECL concept of ACL is therefore closely aligned with the familiar concept of expected loss, which is the credit loss the bank expects over a given time horizon. In this case, the time horizon is the full life of the loan. Below, we discuss three best practices that we believe put banks in the best position not only to achieve compliance with the proposed regulatory standards, but also to leverage the investment in improved provisioning systems to enhance core business activities. By adopting the three best practices advocated in this paper, your bank will be positioned to achieve superior business results through a number of mechanisms including: Increased efficiency; Improved data accuracy and consistency; Better-informed underwriting; Earlier detection of credit deterioration; Empirically-based pricing; Enhanced portfolio management capabilities; and Competitive advantages arising from more accurate provisioning. 1 Quarterly Banking Profile, FDIC, September 30, Financial Instruments Credit Losses (Subtopic ) Exposure Draft FASB, December FEBRUARY 2013 LOSS RESERVES ARE FALLING, BUT THEY COULD SOON BE ON THE RISE

4 In our opinion, the most successful banks in the new regulatory environment will be those that foster a symbiotic relationship between risk and the front office. Anchoring risk management processes to core business activities encourages a risk-conscious culture, which, in turn, tends to lead to stronger institutional support for improved risk management. At the same time, the information supplied by improved risk management systems can drive more profitable business decisions. In this way, risk management processes and business outcomes are simultaneously improved. THE THREE BEST PRACTICES Standardize Your Data Collection and Consolidate Your Data Align Your Risk Measurement System With Your Business Leverage Improved ACL Assessment to Achieve Better Business results Best Practice #1: Standardize Your Data Collection and Consolidate Your Data We ve seen that the CECL concept of ACL can be equated with expected credit loss over the life of the loan. In the next section we will see that several methods exist for estimating expected loss, but a common characteristic of these methods is that they require data. Ideally, these data should (a) span a sufficiently long time period (e.g., at least one full economic cycle), (b) include information on both defaulted and non-defaulted loans, (c) include obligor-level data and obligation-level data for each loan, and (d) include predictive macroeconomic variables. Typically, such a robust dataset must be compiled from a broad spectrum of public and private sources. To facilitate this effort, banks of all sizes have taken great strides over the last decade to standardize their data collection processes and to consolidate their data into central repositories. We firmly believe that one of the best internal investments a financial institution can make is in its data infrastructure. When welldesigned and properly implemented, financial templates, rating platforms, and scorecards each capture key pieces of information during the loan origination phase, throughout the life of the loan and beyond. A robust data infrastructure leads to decreased data error rates, enables improved reporting capabilities, and provides the foundation for a best-in-class risk measurement framework. The successful implementation of a sound data infrastructure will enable your organization to (1) ensure that it is comparing apples to apples by utilizing consistent data inputs across loans and obligors, (2) improve the reliability of any calculations or decisions your organization makes based on internal data by standardizing and vetting it beforehand, and (3) improve the efficiency and thoroughness of your bank s work by centralizing all internal data in one secure location. Moreover, your bank s ACL estimates will improve because of the increased quality of the inputs. Under the proposed CECL standards, banks will likely record larger upfront provisions to capture losses over life of the loan. Depending on how the proposed standards are ultimately implemented and phased in, the biggest hit may come as the new standards first come into effect, because increased provisions may need to be recognized for loans already on the books. Consequently, there is a pressing need to improve loss forecasting abilities in advance of a potential one-time increase. Going forward, ongoing maintenance of the expected loss models will be critical to ensure accurate provisioning as new loans are booked and as the risk of existing loans changes. Integrated data platforms are the foundation for all of these activities. The vast majority of banks face unique challenges in this regard, for two reasons. First, they have less internal data, which limits their ability to develop internal models. Second, the upfront investment in building out data infrastructure can be daunting. Data limitations can be overcome by judiciously augmenting internal data with external data sources, both public and commercial. The upfront investment 3 FEBRUARY 2013 LOSS RESERVES ARE FALLING, BUT THEY COULD SOON BE ON THE RISE

5 in building data infrastructure can be mitigated by turning to a pre-packaged, third-party software solution. But in any event, banks of all sizes should be working towards capturing complete and accurate information about their loans in a central location. Best Practice #2: Align Your Risk Measurement System With Your Business The identification of an appropriate ACL methodology does not occur in a theoretical vacuum. The appropriate level of complexity is institution- and portfolio-specific; one size does not fit all. For one bank, simple loss rates applied to a pool of loans of a given risk rating may be appropriate. For another, a bottom-up, comprehensive internal risk measurement system may be in order. 3 These determinations can only be made in light of careful consideration of the nature and scale of the bank s operations in each of its portfolio segments, and must be able to withstand regulatory scrutiny. Banks with adequate internal data frequently adopt a dual ratings process, where the probability of default ( PD ) is estimated separately from loss given default ( LGD ). The expected loss for a given loan is then calculated as its PD times its LGD (times its exposure at default, for revolving loans). This method is among several valid options for estimating expected credit loss explicitly contemplated in FASB s proposed standards update. For banks with sufficient internal data, PD, LGD and EAD models are typically based in large part on the bank s own historical default and loss experience. However, most banks lack sufficient data for creating such models. Banks that find themselves in this situation have several options. First, it is often possible to build robust custom models based on external data, which can be sampled in such a way as to be representative of the bank s own portfolio. Another option for banks facing data constraints is the use of vended models and scorecards to estimate PD and LGD, and in turn, expected loss. 4 Banks often prefer to rely on vended models regardless of internal data limitations, because such models are market-tested, vendor-maintained and embody best practices. Finally, for community banks with relatively straightforward portfolios, it may be appropriate to estimate expected loss directly, based on their historical loss experience for obligors of like risk and the their assessment of the current economic environment. Regardless of the ACL estimation methodology employed, to be in the best position to prove compliance with the proposed CECL standards, banks should conduct a rigorous assessment of the quality and relevance of the underlying data and properly document their findings. Banks should also have welldocumented evidence that their ACL methodologies are relevant and accurate. The first step towards addressing these issues is to segment the portfolio into homogenous pools of risk so that the accuracy and stability of the models can be tested for each segment. This typically means dividing the portfolio along lines of common risk characteristics such as industry and size. Where exactly those lines get drawn can be driven by qualitative and/or statistical considerations, but regardless of the approach taken, the bank must be prepared to disclose and support its segmentation procedures to regulatory authorities. The performance of the risk models must then be validated for each portfolio segment, which typically relies on statistical tests. Another important issue to consider is whether the expected loss estimate is point-in-time ( PIT ) or through-the-cycle ( TTC ). 5 PIT estimates attempt to measure the level of risk as of the estimation date, whereas TTC estimates attempt to measure the level of risk assuming mid-cyclical economic conditions. Where feasible, we recommend that banks rely on PIT risk measures for ACL purposes and TTC risk measures for economic capital purposes. The use of PIT measures for ACL is consistent with the CECL requirement that current and forecasted economic conditions be captured in the ACL estimate. Unlike the rare, extreme losses represented by economic capital, the losses represented by the ACL are expected 3 That said, regardless of the scale and complexity of your bank s lending activities, the credit loss estimation methodology you employ must also take into account the global aspects of the revised standards. For instance, according to the proposed update, it would be inappropriate to estimate the expected credit losses by simply multiplying an annual loss rate by the remaining years of the loan term, because loss experience is often not linear. See Financial Instruments Credit Losses (Subtopic ) Exposure Draft, FASB, December Even banks with ample internal data frequently utilize vended models for benchmarking purposes. 5 In reality, most risk measures fall somewhere between being purely PIT and purely TTC. 4 FEBRUARY 2013 LOSS RESERVES ARE FALLING, BUT THEY COULD SOON BE ON THE RISE

6 to occur in the ordinary course of business. It therefore makes sense for the ACL to be calculated in a manner that reflects current market conditions, as opposed to long-term average conditions. That said, the particular circumstances of a given bank may alter these recommendations. For example, community banks with less developed infrastructure may find it impractical to implement both PIT and TTC models. In this case, a careful assessment must be made as to all of the intended uses of the models aside from ACL, in order to arrive at the best compromise. One alternative is the adoption of a hybrid model, incorporating certain aspects of PIT and TTC models. The challenges facing a bank in estimating ACL do not stop with model implementation. The continuous need to prove that models remain relevant over time requires a rigorous program of ongoing model maintenance. This includes repeating model validation exercises, which focus on statistical comparisons of predicted and actual experience, on at least an annual basis. In addition to validation exercises focusing on the rank-ordering of risk, the bank must also keep a close eye on the level of risk predicted by the models. Default and loss rates fluctuate meaningfully over time, necessitating periodic recalibration. Lending practices, peer behavior, and credit conditions change as well. Furthermore, banks are required to weave their risk management methodologies into the fabric of their routine business practices. Banks should have well-documented and consistently applied policies, procedures and tools in place to effectively estimate expected credit losses each reporting period. 6 The provision for loan losses on the bank s income statement must capture risk migration during the period, and must be accurately reflected in the ACL on the balance sheet. Moreover, the required disclosures are not limited to numerical estimates of expected credit losses; they also include qualitative information about the credit risk inherent in the portfolio and how management monitors that risk. Finally, model predictions are only as good as their inputs, so it is vital to ensure that model users are well-trained, and that spreading and grading are done in a consistent manner across the organization. To this end, we have found that it is critical to engage model users throughout the development and implementation process. This ensures that they have a sense of ownership of the models and understand how they work. Best Practice #3: Leverage Improved ACL Assessment to Achieve Better Business Results In the current environment of regulatory uncertainty and rapid change, one thing we can be fairly sure of is that the emphasis on rigorous risk management will not decline in the foreseeable future. The ability to adapt to this new environment will likely separate winners from losers. We believe that the best option for banks is to lever improved risk management as a competitive advantage. The new ACL guidance provides an ideal opportunity to lay the foundation for this approach. Before enumerating the numerous ways that ACL systems can be leveraged, let us first note that a superior ACL methodology provides a competitive advantage in and of itself. Banks are engaged in constant competition for the most attractively priced risks. By overestimating the provisions for certain prospective loans, banks risk losing out on profitable business and/or understating earnings. On the other hand, by underestimating the provisions for prospective loans, banks risk taking on unprofitable business and/or overstating earnings. The banks with the most accurate upfront assessments of expected loss are in the best position to come out on top. Moreover, the same machinery that underlies an appropriate ACL methodology can be leveraged to support core business activities relating to every stage of the loan lifecycle, starting with origination. In our experience, many banks have unnecessarily cumbersome underwriting processes that take into account factors that bear little relationship to credit risk. A sophisticated ACL methodology can help isolate the factors that are truly predictive of credit risk, thus enabling the bank to streamline the underwriting 6 See the interagency guidance of 2006, the standards codification of 2009 (including proposed amendments), SFAS 5 (ASC ) and SFAS 114 (ASC ). 5 FEBRUARY 2013 LOSS RESERVES ARE FALLING, BUT THEY COULD SOON BE ON THE RISE

7 process. In addition to reducing costs by improving efficiency, this can help the bank gain a competitive edge by enabling it to act more quickly on attractive opportunities. Risk-based pricing is another important innovation that can be implemented with robust risk systems in place. Risk-based pricing seeks to determine the theoretically correct price of a loan given the riskiness of the obligor and current market conditions. The most sophisticated risk-based pricing models typically take PD, LGD and EAD as inputs, which are precisely the components of expected loss. Therefore, by investing in a state-of-the-art ACL system, banks obtain the foundation for a risk-based pricing platform. The CECL standards require ongoing assessment of credit migration, so that ACL can be adjusted appropriately and in a timely manner. Investing in the systems to comply with this requirement can pay off in the form of earlier detection of credit deterioration. The ability to bring an obligor back to the table in the early stages of trouble is critically important for minimizing losses and protecting profitability. Early detection also enables early impact analysis, so that where losses are unavoidable, the bank is able to plan appropriately. Banks that base their ACL estimates on the output of a dual-rating system can use these same outputs to drive economic capital calculations. Far from being strictly a risk concept, a robust economic capital system can enable the bank to make better portfolio allocation decisions, and can also serve as the foundation for improved performance measurement. Improved risk measurement capabilities can also be used to ensure that the bank s master rating scale meaningfully differentiates risk. This is true whether the bank has separate scales for PD and LGD or a single scale based on expected loss. Ensuring the effectiveness of the rating scale is of great importance, because loan officers tend to think in terms ratings rather than PD, LGD and expected loss. The bank must therefore ensure that it is providing its loan officers with an effective tool for discriminating risk. 7,8 Finally, as alluded to in connection with Best Practice #1, standardizing data collection processes and storing data in a secure, central repository has two business benefits. First, being able to access all relevant data in a central location improves the bank s efficiency. Second, standardizing data collection processes, including establishing clear financial statement spreading guidelines, improves the reliability of the bank s data and enables more accurate hence more profitable estimates. By seizing on these opportunities to leverage risk management systems to improve business decisions, banks can increase the prominence of risk management within their firms, promote a more risk-conscious culture and thereby build stronger institutional support for enhanced risk management systems. 7 We also note that the proposed guidance requires that banks provide a qualitative description of how their internal ratings related to the likelihood of loss if they use internal risk ratings in their ACL analyses. 8 For more on the importance and process of creating an effective master rating scale, see C. Henkel, Mastering Your Rating Scale, The RMA Journal, October FEBRUARY 2013 LOSS RESERVES ARE FALLING, BUT THEY COULD SOON BE ON THE RISE

8 Appendix A Sample Services Related to the Quantification of Expected Credit Loss and ACL Service Name Project Planning & Scope Refinement Portfolio Segmentation Review Existing Policies & Procedures Scorecard Workshops Data Collection & Research Dual Risk Rating Scale Design Overview Critical review of client risk-rating framework requirements. Scope of work can be highly variable, ranging from a set of scorecards to a broader internal risk-rating system. The goal is to define the desired end state and develop a detailed plan to get there. Focus areas typically include gap analysis, conceptual and functional design, data management, integration with credit processes, and sufficiency of testing. Assessment of portfolio segments. The loan portfolio will be divided into homogeneous pools of credit risk to enable effective risk modeling and reporting. Segmentation often takes into account statistical considerations in addition to expert judgment. Gathering internal documentation for study of existing practices. Policies and procedures should serve as living documents that set forth the activities to be performed and rules to be followed in the day-to-day course of business. Accordingly, Moody s will work corroboratively with your team to ensure they are reflective of leading practices while preserving your institutional culture. Focus will be centered around providing specific and actionable guidance for improved risk measurement in credit, lending, loan review, and ALLL activities. Collaborative sessions with business line representatives and senior managers. Risk factor selection is not guided by statistical results alone. Input from the business lines is critical to ensure that the models are accepted and understood by those that will use them. Incorporating insights from the business lines and senior management often leads to better models than those that can be derived using statistics alone. Development of developmental, validation and calibration data samples. For banks that possess sufficient internal data for custom model development, great care must be taken in assembling, cleaning and documenting the datasets used for model development, validation and calibration. In the absence of robust internal data, appropriate external datasets must be identified, cleaned and documented, and every effort should be made to utilize any internal data that may exist for purposes of performance testing. The Moody s Analytics Credit Research Database ( CRD ) and Ultimate Recovery Database ( URD ) are leading examples of external data sources that can be used to augment a bank s internal data. A Master Rating Scale (MRS) should facilitate the goal of uniformly discriminating risk for activities such as loan approval, risk rating, loan loss provisioning, pricing, portfolio and capital management. Keeping with best practices which separate the components of EL, this effort will typically focus on creation of a PD scale and a LGD scale, and in turn, an EL scale to be utilized across the loan portfolio. The goals are improved granularity, low concentration of risk grades, and improved objectivity. 7 FEBRUARY 2013 LOSS RESERVES ARE FALLING, BUT THEY COULD SOON BE ON THE RISE

9 Calibration Prototype Development Backtesting/Validation Benchmarking Spreading and Grading Consistency Maintenance Manual/ Model Documentation Custom Risk Model/ Scorecard Training Calibration is the process of setting the appropriate level of risk predicted by a given risk model and mapping model scores to the master ratings scale. Model calibration can be evaluated according to how closely the model s predicted probabilities match realized default and loss rates. Proper calibration also provides a distribution of borrowers across ratings without excessive concentrations, a regulatory requirement. Standalone mockup of the new risk models for testing and benchmark purposes. A fully functional mockup of the newlydesigned model is typically provided in Microsoft Excel format. The model prototype can be used for testing prior to implementation. It can also serve as a useful benchmark to check that the model is properly implemented in the bank s internal systems. Testing the new model on the bank s historical data. The specific validation exercises performed depend critically on the availability of internal data. Classic examples include estimation of accuracy ratios, r-squared values, and correlation tests. Comparison of the performance of an internal model against an external model. A portfolio sample and a relevant external reference model are identified. Typically, the reference model is purely quantitative, and does not include judgmental factors. The portfolio sample is processed through both models the internal model and the reference model and the performance of the two models is compared based on the bank s historical data. The goal is to ensure that the new model performs at least as well as the benchmark. Benchmarking can be viewed as a type of validation exercise. Standardized spreading and risk grading techniques supporting model methodologies. Development of uniform spreading and grading procedures helps ensure accuracy and consistency. These procedures compliment the construction of financial/quantitative model factors to provide users with a clear understanding of how scorecard inputs are to be derived and how other considerations (e.g., guarantor treatment) will be managed. Spreading techniques may be designed with a view towards automating the scoring process to the greatest extent possible. Reference guide and how to manual for conducting internal validation of borrower risk models. Content is typically customized based on client requirements. Topics include benchmarking, data collection, data sampling, position analysis, univariate testing, multivariate testing, and ongoing management of validation and calibration activities. Many clients also use the maintenance manual as a policy document for internal model maintenance. Practitioner-focused training in a Train-the Trainer format. The training seminar is custom-designed to help the bank learn how to use risk models and scorecards (e.g., RiskCalc, CMM, Moody sdesigned custom models, etc.). Seminars typically focus on the use of scorecards in the normal course of business and include theory and practical application through case studies. At the conclusion of the seminar, Moody s provides the institution with the presentation materials for re-use in future internal training sessions. 8 FEBRUARY 2013 LOSS RESERVES ARE FALLING, BUT THEY COULD SOON BE ON THE RISE

10 Appendix B Products Related to the Quantification of Expected Credit Loss and the ACL Moody s Analytics helps capital markets and risk management professionals worldwide respond to an evolving marketplace with confidence. The company offers unique tools and best practices for measuring and managing risk through expertise and experience in credit analysis, economic research and financial risk management. Moody s Analytics provides a number of products and data sets that could be very useful for financial institution s who wish to optimize their ACL allowances while also complying with the related supervisory guidance. RiskCalc Plus provides probability of default measures for private firms. RiskCalc, with 28 country and industry-specific models, leverages a proprietary default database to enable greater accuracy, consistency and efficiency than other commercially available models and internal bank models when estimating private firm credit risk. CMM (Commercial Mortgage Metrics) is the leading analytical model for assessing probability of defaults and loss given defaults for CRE loans. CreditEdge Plus provides probability of default measures for public firms. CreditEdge Plus offers daily updates on changes in the probability of defaults and a robust pricing framework for interpreting signals between equity and debt markets for over 30,000 publicly traded firms around the world. LossCalc is a loss given default model that incorporates both static and forward-looking dynamic drivers of recovery, ranging from firm specific variables to broader geography and industry factors. RiskFoundation includes a common datamart optimized for managing risk and delivers business and regulatory capabilities, an administrative console, customization tool kits, grid computing and scenario analysis software that can all be adapted to fit your bank s unique situation. RiskOrigins is a workflow-driven origination solution that leverages RiskFoundation and allows commercial lenders to streamline and standardize their complete underwriting process for commercial loan classes, create a single system of record, and a platform for probability of default and loss given default analyses. RiskFrontier produces a comprehensive measure of risk, expressed as Credit VaR or Economic Capital, which comprises the basis for deep insight into portfolio dynamics for active risk and performance management. Moody s Analytics Credit Research Database (CRD) is the world s largest and cleanest database of private firm financial statements and defaults, built in partnership with over 45 leading financial institutions around the world. Moody s Analytics Ultimate Recovery Database (URD) features three different approaches to recovery calculation and nearly 5,000 defaulted bond and loan instruments since The URD empowers clients to use the same data as Moody s analysts to develop predictive loss given default (LGD) models. 9 FEBRUARY 2013 LOSS RESERVES ARE FALLING, BUT THEY COULD SOON BE ON THE RISE

11 Speakers Bios Christian Henkel Christian Henkel is a Director in the Enterprise Risk Solutions group within Moody s Analytics, based in New York City. In his current role, Christian leads consulting engagements with clients throughout North America and is an experienced credit practitioner. Having spent most of his career in commercial banking, Christian has a unique blend of business and academic experience across the financial services industry - including expertise in commercial credit and financial analysis, portfolio management, asset quality, loan loss reserve methodologies, credit administration, process redesign, safety & soundness examinations, and credit risk modeling. Christian has also served as a credit risk instructor for the graduate banking school at Southern Methodist University (SMU). His formal education includes an M.B.A. from the University of Texas at Dallas; a B.A. from UTD; and graduated Valedictorian from the Southwestern Graduate School of Banking at SMU. Jan Larsen Jan Larsen is an Associate Director in the Enterprise Risk Solutions group at Moody s Analytics, based in New York City. In his current role, Jan manages engagements with clients throughout North America. He has been applying quantitative techniques and economic reasoning to help financial firms navigate regulatory, legal and business challenges for almost nine years. His recent engagements have focused on topics including PD modeling, complex securities valuation and economic capital calculations. Jan holds a BA and MS in Physics from Reed College and New York University, respectively. He is also a CFA charterholder, and earned the PRM designation from PRMIA Moody s Analytics, Inc. and/or its licensors and affiliates (collectively, MOODY S ). All rights reserved. ALL INFORMATION CONTAINED HEREIN IS PROTECTED BY LAW, INCLUDING BUT NOT LIMITED TO, COPYRIGHT LAW, AND NONE OF SUCH INFORMATION MAY BE COPIED OR OTHERWISE REPRODUCED, REPACKAGED, FURTHER TRANSMITTED, TRANSFERRED, DISSEMINATED, REDISTRIBUTED OR RESOLD, OR STORED FOR SUBSE- QUENT USE FOR ANY SUCH PURPOSE, IN WHOLE OR IN PART, IN ANY FORM OR MANNER OR BY ANY MEANS WHATSOEVER, BY ANY PERSON WITHOUT MOODY S PRIOR WRITTEN CONSENT. All information contained herein is obtained by MOODY S from sources believed by it to be accurate and reliable. Because of the possibility of human or mechanical error as well as other factors, however, all information contained herein is provided AS IS without warranty of any kind. Under no circumstances shall MOODY S have any liability to any person or entity for (a) any loss or damage in whole or in part caused by, resulting from, or relating to, any error (negligent or otherwise) or other circumstance or contingency within or outside the control of MOODY S or any of its directors, officers, employees or agents in connection with the procurement, collection, compilation, analysis, interpretation, communication, publication or delivery of any such information, or (b) any direct, indirect, special, consequential, compensatory or incidental damages whatsoever (including without limitation, lost profits), even if MOODY S is advised in advance of the possibility of such damages, resulting from the use of or inability to use, any such information. The ratings, financial reporting analysis, projections, and other observations, if any, constituting part of the information contained herein are, and must be construed solely as, statements of opinion and not statements of fact or recommendations to purchase, sell or hold any securities. NO WARRANTY, EXPRESS OR IMPLIED, AS TO THE ACCURACY, TIMELINESS, COMPLETENESS, MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE OF ANY SUCH RATING OR OTHER OPINION OR INFORMATION IS GIVEN OR MADE BY MOODY S IN ANY FORM OR MANNER WHATSOEVER. Each rating or other opinion must be weighed solely as one factor in any investment decision made by or on behalf of any user of the information contained herein, and each such user must accordingly make its own study and evaluation of each security and of each issuer and guarantor of, and each provider of credit support for, each security that it may consider purchasing, holding, or selling. Any publication into Australia of this document is pursuant to the Australian Financial Services License of Moody s Analytics Australia Pty Ltd ABN AFSL This document is intended to be provided only to wholesale clients within the meaning of section 761G of the Corporations Act By continuing to access this document from within Australia, you represent to MOODY S that you are, or are accessing the document as a representative of, a wholesale client and that neither you nor the entity you represent will directly or indirectly disseminate this document or its contents to retail clients within the meaning of section 761G of the Corporations Act JANUARY 2013 SP21905/101215/IND-104