U.S. Credit Card Charge-Offs Will Likely Face An Uphill Climb

Transcription

1 U.S. Credit Card Charge-Offs Will Likely Face An Uphill Climb Primary Credit Analysts: Ildiko Szilank, New York (1) ; James M Manzi, CFA, Charlottesville (1) ; james.manzi@spglobal.com Jun Wang, New York (1) ; jun.wang@spglobal.com Darrell Wheeler, New York (1) ; darrell.wheeler@spglobal.com Kirsten R Mccabe, New York (1) ; kirsten.mccabe@spglobal.com Table Of Contents Scenarios Summary For Our Regression Model Discussing The Variables Used To Project Charge-Offs A Short Description Of The Regression Details How Does The Credit Card Bond Structure Allow For Higher Charge-Offs? Charge-Offs Will Slowly Rise As Revolving Credit Expands Related Research SEPTEMBER 6,

2 U.S. Credit Card Charge-Offs Will Likely Face An Uphill Climb Higher revolving credit card debt and greater holdings on bank balance sheets have led S&P Global Ratings to revisit its projections for credit card charge-offs. Using our multiple regression and our economists' unemployment forecasts, we update our base-case U.S. bank credit card charge-off projections to 2.5% by year-end 2016 for our more-seasoned/higher quality securitized card index and drop the late 2017 charge-off forecast to near 3%. The current results are lower than a year ago, when we projected 3.5% and 4.2% for 2016 and 2017, respectively. For the industry as a whole, we predict charge-offs will increase to 3.1% and 3.4%, respectively, for year-end 2016 and late These levels are 0.4% to 1% higher than the current figures of 2.0% and 3.0%, but are still well below the historical averages of 5.2% and 4.9% for the trusts and the overall industry, respectively. Overview We have updated and augmented our credit card charge-off analysis, which provides results for a wide range of economic scenarios. In addition to using our own series of securitized charge-offs, we also now model the Federal Reserve credit card charge-off series to create regression results for a larger data segment, as the percentage of securitized receivables has been decreasing in relevance over the past six or seven years. Our base-case projection, which uses 24-month averages for the chosen regression variables and our economists' unemployment rate forecasts, predicts credit card charge-offs from our securitized bankcard series at 2.5% by year-end 2016 (versus 3.1% for the industry result). The securitized rate approaches 3% in late 2017 (versus 3.4% for the industry result). The forecasts are up from the current level of 2.0% for trusts and 3.0% for the industry but still well below the historical averages of 5.2% and 4.9%, respectively. We project a higher rate and faster acceleration of charge-offs for the industry as newer accounts season on the managed books. Similarly, if new and less-creditworthy accounts are added to trusts or higher credit lines are granted to weaker obligors already in trusts, we could also see higher charge-offs in the securitization pools that we track. Our first two downside forecasts show that charge-offs could reach 8%-9% for both the securitized bankcard and the Federal Reserve Board series, which could potentially climb to over 12% in our high stress scenario that applies the worst conditions of the previous recessions and uses a 15% unemployment rate. We first published credit card charge-off projections in July 2015 (see "Understanding The Economic Factors Affecting U.S. Credit Card Charge-Offs," published July 13, 2015). At that time, we developed a regression to project future U.S. securitized credit card charge-offs under a variety of economic scenarios. This report updates that analysis by utilizing 24 (rather than 23) years of historical data tracked in S&P Global Ratings' Credit Card Quality Index (CCQI) and, in addition, now examines the Federal Reserve Board's (FRB's) credit card charge-off series, which covers a wider sample of loan data. We believe the addition of the Fed's data provides a good comparison between the smaller universe of securitized receivables that we track and the managed books of commercial bank credit card lenders, and note that the regression using the FRB data had a moderately less "goodness of fit" versus the regression using our CCQI data. Per the Securities Industry and Financial Markets Association data, of the $961 billion in revolving credit outstanding SEPTEMBER 6,

3 in the first half of 2016 (according to the FRB estimate), only 13% ($126 billion) was securitized as of the end of second-quarter 2016, down from roughly 33% in second-quarter The decline is largely attributable to the loss of capital relief from securitizing receivables for generally accepted accounting principles purposes after regulation changes in Our loss rate projections based on Fed data are higher than the securitized universe, and we think loss rates will also increase faster for the industry than the trusts as these newer less-seasoned accounts age in the bank portfolios. In contrast, the securitized universe has seen limited new and unseasoned account additions to its master trusts in the past few years, creating pools with a higher percentage of seasoned and high-credit-quality accounts, which have also contributed to lower delinquencies and charge-off rates compared with the new loans on the managed books. The Fed data set also include charge-offs from new accounts, which may also exhibit lower FICO scores as underwriting loosens, while the majority of the bankcard trusts have at least 80% of their securitized receivables seasoned at least 60 months. In addition, average FICO scores for securitized receivables over the past few years have shown a progressive percentage increase in FICO score buckets of 720 and above, as well as a gradual decrease in FICO score buckets 660 and below. As of first-quarter 2016, the portion of securitized receivables in pools from prime accounts (FICO above 720) was at the highest level since first-quarter 2006 at 61%, while the portion from subprime accounts (FICO below 660) was at the lowest level at 12%. Low interest rates, favorable consumer debt levels, job gains, and moderate wage growth should keep consumer bankruptcies low over the next year, though some nascent signs of loosening lending standards exist. Our first two downside forecasts, which mirror previous recessions, pushes charge-offs to reach as high as 8%-9% for both the securitized bankcard and FRB series. They could also climb to over 12% in an extremely high stress scenario in which the worst conditions of the previous recessions apply and the unemployment rate hits 15%. Meanwhile, our upside projection, in which the unemployment rate reaches and stays at 3.5%, offers only moderate separation (generally around 30 basis points [bps]) from our base-case forecast, which we attribute to current positive consumer credit conditions. However, we believe that charge-offs are not likely to stay in the 2% range for a meaningful length of time as these positive conditions typically lead to greater availability of credit. In turn, this ultimately creates rising credit card charge-offs. Finally, we note that our regression results reflect averages and broad macro conditions, so individual credit card pool performance will vary. Scenarios Summary For Our Regression Model Table 1 summarizes our five projection scenarios by listing the annual movements in the regression variables. The wide range of variable outcomes presented in the scenarios is intended to demonstrate a full sensitivity range of possible charge-off outcomes for both the CCQI securitized and FRB portfolio series. SEPTEMBER 6,

4 Table 1 Summary of Regression Results And Economic Scenarios (Year-End December Figures) Upside scenario Unemployment rate (%) Revolving credit (bil. $) ,029 1,060 1,091 1,124 Household debt service ratio (%) Personal bankruptcies (000s) Predicted CCQI charge-off rate (%) Predicted FRB charge-off rate (%) Base-case scenario Unemployment rate (%) Revolving credit (bil. $) ,025 1,056 1,087 1,120 Household debt service ratio (%) Personal bankruptcies (000s) Predicted CCQI charge-off rate (%) Predicted FRB charge-off rate (%) Stress scenario 1 Unemployment rate (%) Revolving credit (bil. $) 1,063 1,204 1,310 1,408 1,503 1,640 Household debt service ratio (%) Personal bankruptcies (000s) Predicted CCQI charge-off rate (%) Predicted FRB charge-off rate (%) Stress scenario 2 Unemployment rate (%) Revolving credit (bil. $) 987 1, Household debt service ratio (%) Personal bankruptcies (000s) Predicted CCQI charge-off rate (%) Predicted FRB charge-off rate (%) Super stress scenario 3 Unemployment rate (%) Revolving credit (bil. $) 1,063 1,204 1,310 1,408 1,503 1,640 Household debt service ratio (%) Personal bankruptcies (000s) Predicted CCQI charge-off rate (%) Predicted FRB charge-off rate (%) CCQI--Credit Card Quality Index. FRB--Federal Reserve Board. Base-case scenario: current economic variables In the base-case economic scenario, we use the coefficients from the regression equation and apply the average condition from the past 24 months for all the variables except unemployment. For unemployment, we use S&P Global SEPTEMBER 6,

5 Ratings' latest base-case forecasts of 4.7% for 2016 and 4.5% for Household debt service remains at 10%, while revolving debt continues to grow at about 3% annually. The resulting projections for credit card charge-offs increases modestly through 2021 but remains well inside of 5%. Stress scenario 1: duplicating the loose credit environment of 1995 and the 2001 recession This stress scenario looks at the period, when credit expanded before a moderate recession. For the first three years, we mimic the macroeconomic environment, with the unemployment rate dropping below 5%, household debt service climbing above 12%, and revolving debt increasing significantly. In the early years, credit expansion outweighs the fairly good employment conditions but eventually pushes charge-offs to around 5% in (just above 4% for the FRB series). Later in the series, we create a recession similar to 2001, which pushes charge-off rates to nearly 7% for the CCQI series and 6% for the FRB series by the end of Stress scenario 2: mimicking the recession The second recessionary scenario starts with an economic environment similar to that in Then conditions rapidly deteriorate like they did between 2008 and 2011, with unemployment up more than five percentage points and bankruptcies doubling. Charge-offs rise to 9% for both series by the end of the third year (2018). Super-stress scenario 3: combining the worst factors from the previous two recessions with a 15% unemployment rate The two historical recessions provide a good stress baseline, but we felt that an additional vector would help assess a "perfect storm" scenario. We took the worst levels from our first two recessionary stresses and then pushed the unemployment rate up to 15% (we refer to this scenario as a super stress because the U.S. unemployment rate has not reached 15% since 1940.) In fact, while we consider the most recent recession a moderate, or 'BBB', economic scenario in which the unemployment rate reaches 10%, a 15% rate would signify a substantial, or 'A', economic stress (see "Understanding Standard & Poor's Rating Definitions," published June 3, 2009). Under this stress scenario, the charge-off rate for both series increases to just over 12%. Upside scenario: staying the course In this scenario, we continue to move variables as they have progressed over the past 24 months, with the unemployment rate reaching a low of 3.5%. This scenario demonstrates how positive recent credit conditions have been for consumers, as charge-offs in this scenario only improve 10 bps to 60 bps versus the base case. The limited upside demonstrated in this scenario suggests that charge-offs have little room to tighten. Overall, we found that since our last report in July 2015, charge-offs have generally run about 30 bps to 40 bps inside our base-case scenario, as the recent post-crisis collateral still, in our view, has higher credit quality than the full historical range used to derive our regressions. SEPTEMBER 6,

6 Chart 1 SEPTEMBER 6,

7 Chart 2 Discussing The Variables Used To Project Charge-Offs To project credit card losses, we researched the relationship between credit card charge-offs in our CCQI and key economic and credit indicators. Many of these variables can also be correlated with each other (see "Understanding The Economic Factors Affecting U.S. Credit Card Charge-Offs," published July 13, 2015). For the regression with the FRB series, we utilized the same variables. While the unemployment rate relationship with charge-offs was strong for both regressions (nearly one-to-one in the FRB series given a sharp increase in the unemployment rate), growth in revolving credit and the household debt service ratio were not statistically significant within the regression using the FRB series. One possible explanation is that the unemployment factor overwhelmed those factors in that analysis. We note that the overall goodness of fit was moderately lower for the FRB series data versus the CCQI regression, but still quite robust (adjusted R-squared of 0.79 versus 0.9, respectively). Table 2 lists the variables used in the two regressions, along with their coefficients and t-statistics. SEPTEMBER 6,

8 Table 2 Regression Details CCQI series FRB series Applicable value Estimate T-stat(i) P-value(ii) Estimate T-stat(i) P-value(ii) Intercept (1.72) (2.89) Annual change in unemployment rate Annual change in unemployment rate Annual change in revolving credit (12-month lag) Annual change in revolving credit (12-month lag) Household debt service ratio (six-month lag) Credit card charge-off rate (12-month lag) Credit card charge-off rate (24-month lag) Annual growth in personal bankruptcy (three-month lag) If >= 2 percentage points If < 2 percentage points If < 0 (decline) If >= 0 (growth) Not statistically significant Not statistically significant (0.09) (3.5) (0.1) (2.84) (0.23) (8.73) 0 (0.15) (4.74) Adjusted R-squared(iii) Root mean squared error(iv) (i)the ratio of coefficient estimate over its standard error. (ii)indicates statistical significance of the coefficient estimate. Typically, the coefficient estimate is thought to be different from 0 and, hence, significantly associated with the dependent variable if its p-value is no more than (iii)a measure of the goodness of fit of the regression model, with a typical range from 0 to 1. Usually, the higher the adjusted R-square, the better the model fits the data. (iv)root mean squared error. Usually, the lower the RMSE, the better the model predicts the dependent variable. CCQI--Credit Card Quality Index. FRB Federal Reserve Series. The unemployment rate U.S. employment is a key variable in projecting consumer credit health (chart 3). Among the independent variables, the unemployment rate has the highest correlation to credit card charge-offs in both regressions. When the overall year-over-year change in unemployment is less than 2%, our study suggests that the impact on charge-offs is smaller, which translates to a lower coefficient (see table 2). The relationship between U.S. credit card charge-offs and the unemployment rate is weaker post-recession. We believe this is caused by several factors, such as defaulted obligors leaving the pools and tighter underwriting standards, while regulatory constraints have limited banks from expanding credit to nonprime accounts. When we compared our projections from our 2015 report with actual charge-off performance over the last year, we found that actual credit card charge-offs were closer to our upside scenario forecast while actual unemployment rate was almost identical to our base-case forecast. Looking forward, we expect charge-offs to normalize somewhat and the relationship between the two factors to strengthen as underwriting loosens and charge-offs increase as new accounts season. As that happens, actual performance should approach our base-case charge-off forecasts and exceed our upside scenario forecast. In addition, we maintain our view that the correlation should hold stronger in a downturn scenario in which an increasing number of newly unemployed obligors and bankruptcy filings would likely contribute to higher charge-offs. SEPTEMBER 6,

9 Chart 3 Household debt service to disposable income level (six-month lag) Consumers' ability to service their debt also helped predict credit card charge-offs in the regression that utilized our CCQI series but was not statistically significant for the regression using the FRB series. This credit measure improved from 2009 to 2012 (see chart 4) as consumers deleveraged or defaulted and their debts were written off. Today, consumers remain historically underleveraged, and consumer default metrics are at 12-year lows (according to the S&P/Experian Consumer Credit Default Indices). In the near term, we expect the recent low household debt service levels to continue, but they will eventually start to increase with the credit cycle and (potentially) higher interest rates. SEPTEMBER 6,

10 Chart 4 The revolving credit growth rate (12-month lag) The growth rate in revolving credit could signal a change in the credit quality related to collateral underwriting standards. In chart 5, we see that the 12-month-lag level is positively correlated with charge-offs, suggesting that it provides a gauge of credit momentum and is a leading indicator of credit card charge-off behavior. SEPTEMBER 6,

11 Chart 5 While both a decline and growth in overall revolving credit were statistically significant in our CCQI regression, only a decline in overall credit was found to be significant for the regression using the FRB data series. With recent employment and economic growth, consumers have started to use more revolving credit. Overall revolving credit was up nearly $50 billion year over year in June 2016, which followed smaller full-year gains of $12 billion, $34 billion, and $46 billion in 2012/2013, 2013/2014, and 2014/2015. However, a growing economy and an improving labor market may help delay potential higher charge-offs until Increased revolving credit is also related to underwriting quality, which usually loosens as any post-recession period progresses. This should eventually result in higher charge-offs. The most recent FRB Senior Loan Officer Opinion Survey (in July 2016) has already indicated that some banks reduced minimum required credit scores for credit card loans and widened spreads of interest rates charged on outstanding credit card balances. The survey also indicates higher lines granted to credit card obligors. However, given the high cost associated with lower-quality receivables from a regulatory perspective, a powerful constraint to loosening underwriting likely exists. In other words, new loans may not experience the same level of charge-offs as in the past, at least from highly regulated lenders. SEPTEMBER 6,

12 Personal bankruptcy Consumer bankruptcy is correlated with charge-offs in our regression equation (0.01 coefficient). This variable is volatile because of its seasonal nature of the collateral and the 2005 changes to bankruptcy legislation. As a result, we excluded the distorted period (from the fourth quarter of 2005 through the entire year of 2006) from our analysis. In recent years, bankruptcies have trended down, corresponding to shrinking debt obligations and positive conditions for consumer credit, namely low interest rates and moderate job and wage growth. Charge-off rate levels (12- and 24-month lag) Not too surprisingly, the direction of credit card charge-off rates themselves also indicates their momentum. To account for this, our regression analysis incorporates the charge-off rate lagged 12 months. Any wide variability is then generally eased with the incorporation of 24-month lagged figures. We include both because while the charge-off rates can be persistent over a relatively short period of time (12 months), over a longer term (24 months), they exhibit mean reversion. A Short Description Of The Regression Details For our regression analysis, we used S&P Global Ratings' CCQI for bankcards from January 1992 to June 2016 and the FRB charge-off series from first-quarter 1985 to first-quarter 2016 (interpolated into monthly figures). To mitigate seasonal effects and the nonstationarity exhibited in the charge-off rate series, the regression analysis uses the year-over-year changes in charge-off rates as the dependent variable. For the independent variables, the study examined 24 macroeconomic variables in various combinations and with different time lags. For further discussion of the model, the regression equations, pairwise correlation, and more, please see "Understanding The Economic Factors Affecting U.S. Credit Card Charge-Offs," published July 13, How Does The Credit Card Bond Structure Allow For Higher Charge-Offs? Understanding the bond structures of revolving credit card asset-backed securities (ABS) transactions helps illustrate the likely impact charge-offs could have on credit card ABS deals. These structures typically include performance-based amortization triggers, sequential payment priorities for senior notes, and credit protection to allow for an environment that may simultaneously stress multiple variables. For example, when monitoring and rating credit card ABS transactions, S&P Global Ratings assumes that a pool's excess interest is negative and that the payment rate and yield are at their stressed levels in the beginning of the amortization scenarios in higher rating categories. In addition, we established a 'AAA' peak charge-off rate of 33% for our U.S. Bankcard CCQI, which we use as an industry benchmark. While we periodically adjust our base-case assumptions based on economic and pool-specific factors, we generally hold the stressed peak assumptions constant during normal economic cycles, which effectively increases the loss coverage multiples for higher rating categories and creates higher rating stability. Chart 6 illustrates our assumption that, on average, annualized bank credit card charge-offs increase to their stressed peak levels of 34% for 'AAA', 20% for 'A', and 12.75% for 'BBB' rating categories over months and stay at that level until the securities mature. To clarify, the 12.75% rate is simulating a 'BBB' rating scenario, and the peak stays at that level until the bonds mature, while the greater than 12% rate in our combined super stress scenario is simulating an 'A' economic stress, SEPTEMBER 6,

13 and our projected rate only reaches that peak level for less than a year. When rating and monitoring bank credit card ABS transactions in our 'A' rating category, our peak loss rate assumption averages 20%, which is well above the 12% rate in our combined stress scenario. Chart 6 We caution that while our projected charge-off numbers represent both bankcard averages based on our CCQI and the more broad FRB series, individual credit card programs could see significant variance from these averages. To highlight this potential outlier risk, we chose two separate credit card loss events that have transpired in the past: the failure of NextCard and First Consumer National Bank in , which exceeded 35% charge-offs when the industry average was only about 7%, and the Advanta card program default, which creates a similar a 30%-plus charge-off rate in 2010, when average charge-offs were about 9.5% (see chart 7). SEPTEMBER 6,

14 Chart 7 In all three cases, each program's charge-offs sprang from their average rates (well above the industry average) before amortization and then peaked at 35%-40% after amortization. (The 61.52% charge-off in Advanta's credit card program in one month was the result of charging off closed accounts earlier than 180 days.) The senior notes in the related transactions paid out in full within three years, mainly because of the sequential payment structure, in which the senior notes received payment collections before the subordinated notes. Nevertheless, these case studies present a good example of how individual credit card pools may differ from the industry average and our benchmark. These three credit card programs were managed by small unrated, monoline banks that relied mostly on securitization for funding purposes. We believe that benchmarking and using specific economic scenarios to calibrate our ratings criteria enhance ratings comparability. However, our use of these tools does not diminish the importance of issuer- and portfolio-specific analyses. Credit card ABS master trust pools are actively managed, and some card issuers employ account origination and management strategies that may be more successful at mitigating risk in challenging economic environments than others. SEPTEMBER 6,

15 Charge-Offs Will Slowly Rise As Revolving Credit Expands While we don't know which economic scenario may eventually occur, these wide range of outcomes provide market participants a better understanding of the potential impact that each of our selected variables may pose as a credit card charge-off risk. We have seen robust growth in revolving credit over the past year or so, and some emerging signs of looser underwriting standards. That said, we believe relatively tight underwriting practices, low levels of consumer leverage, and moderate job and wage growth are positive for credit card collateral quality in the near term. We do expect some of these prevailing conditions to deteriorate, however, as consumer spending continues to increase. This could cause new credit card receivables to grow as consumers demand more debt. Credit card lenders may respond by increasing credit limits, expanding their lending to newly employed borrowers, and widening their nonprime obligor base to some extent. Our regression projects a higher rate and faster acceleration of charge-offs for the industry as new accounts season on the managed books. However, the potential addition of new and less-creditworthy accounts to trusts or higher credit lines granted to weaker accounts already in trusts could ultimately also result in higher charge-offs for the securitization pools that we track. Related Research Understanding The Economic Factors Affecting U.S. Credit Card Charge-Offs, July 13, 2015 Understanding Standard & Poor's Rating Definitions, June 3, 2009 Only a rating committee may determine a rating action and this report does not constitute a rating action. SEPTEMBER 6,

16 Copyright 2017 by Standard & Poor s Financial Services LLC. All rights reserved. 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. 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 acknowledgment 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 non-public 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 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 STANDARD & POOR S, S&P and RATINGSDIRECT are registered trademarks of Standard & Poor s Financial Services LLC. SEPTEMBER 6,