Report to the Congress on Reductions of Consumer Credit Limits Based on Certain Information as to Experience or Transactions of the Consumer
Print (316 KB PDF)
- Executive Summary and Recommendations
- Introduction
- Information Used to Prepare the Section 505 Report
- Findings from Interviews with Credit Card Issuers
- Findings from the Survey of Credit Card Issuers
- Findings from Discussions with Other Supervisory Agencies
- Findings from the Assessment of Credit Record Data
- Recommendations
- Appendix A. Section 505 of the Credit CARD Act of 2009
- Appendix B: The Survey of Credit Card Issuer Practices
- Appendix C: The Credit Record Data Used in This Report
- Footnotes
Executive Summary and Recommendations
Creditors consider a wide variety of information in managing and assessing credit risk and account profitability. Over the past year or so, concerns have been raised about the types of information some creditors may consider when setting and adjusting credit terms on credit cards they issue to consumers. For the most part, the types of information that have been the focus of attention relate to the possible use of transaction-specific details, such as the identity or location of the merchant or store involved in a credit card transaction or the types or prices of the items purchased. In order to learn more about the practices of credit card issuers in this regard, section 505 of the Credit Card Accountability Responsibility and Disclosure Act (Credit CARD Act) of 2009 requires the Federal Reserve Board to prepare a report to the Congress. This report was prepared in response to the congressional request.
To prepare this report, Federal Reserve staff held conversations with credit card issuers, their federal supervisory agencies, and entities that provide account management services to issuers. The report also reflects information gathered from a special survey of card issuers and an assessment of credit record information.
Conversations with the leading credit-card-issuing banking institutions (defined hereafter to include commercial banks, savings institutions, and credit unions) and their federal regulators (the Office of the Comptroller of the Currency, the Federal Deposit Insurance Corporation, the Federal Reserve Board, the Office of Thrift Supervision, and the National Credit Union Administration) indicate that credit card issuers consider a wide range of information when deciding whether to reduce the credit limit or increase the interest rate on a credit card account. Card issuers draw on information from (1) their own records on their cardholders' account usage and payment behavior and (2) credit records maintained by the credit-reporting agencies about cardholders' use of credit more broadly, including their payment experience on other debts, records of collections, and new monetary-related public record items such as bankruptcy or garnishments. Card issuers may also reduce credit limits and change interest rates in response to changes in broad economic conditions that affect their cost of funds, actual or expected losses, and account profitability.
The Use of Transaction-Related Practices to Reduce Credit Lines or Raise Interest Rates
Concerns have been raised about whether credit card issuers reduce credit limits or raise interest rates on credit card accounts based on certain types of information about the purchases cardholders make with their cards or the parties from whom they have received a mortgage. The survey of issuers of general-purpose bank credit cards conducted for this report found that few issuers considered the types of information of concern at any time during the three-year period from November 30, 2006, through November 30, 2009.1 The survey also found that very few used the types of information of concern during November 2009, the month before the survey was mailed to issuers.
Specifically, a survey of the 100 largest issuers of general-purpose bank credit cards plus 75 other institutions selected to be representative of all but the smallest remaining bankcard issuers found that during the three-year period preceding the mailing of the survey, 6 commercial banks or savings institutions used one or more of these practices as a basis for adjusting credit limits on active accounts; of those 6 issuers, 2 also adjusted interest rates on accounts, and 4 also closed accounts. During the month of November 2009, only 2 commercial banks or savings institutions used one or more of these practices as a basis for adjusting credit limits on active accounts; of those, 1 issuer adjusted interest rates on accounts, and 1 closed accounts.2 One other commercial bank continues to review accounts for credit-line changes periodically but did not do so in November 2009. No credit unions reported using any of these practices in November 2009 or at any time in the preceding three years.
Card issuers that make use of transaction-related items in adjusting account terms tend to rely on fairly general metrics--for example, the total amount spent on a purchase rather than the amount spent on a specific item; broad merchant category codes, such as hardware stores, entertainment establishments, or grocery stores, rather than a specific merchant identity; or general geographic location of the merchant rather than the street address, census block group or census-tract designation, or local five- or nine-digit Zip code of the business.
Issuers that have used one or more of the practices to reduce credit limits, raise interest rates, or close accounts have large numbers of cardholders, and consequently the number of their cardholders that could potentially be affected is also large. However, card issuers report they have not implemented these practices in isolation; the practices are considered along with many other metrics of credit risk or card use, and these other metrics are typically the determining factors in credit risk management decisions. In the vast majority of cases, when transaction or mortgage-lender-related information has been considered, it has not been an important factor
For those institutions that have used such information, the proportion of cardholders actually affected has been small. The two issuers that used a section 505 practice during November 2009 had roughly 53 million active credit card accounts. Among these, about 340,000 experienced a line reduction for any reason during that month. (Because most cardholders have only one active account from a given issuer, it is reasonable to interpret numbers of accounts reported here as equivalent to numbers of cardholders when focusing on a given issuer.3) Of these, fewer than 1,900 cardholders, or about one-half of 1 percent of those experiencing a line reduction, were triggered, at least in part, by one of the practices of concern. The low incidence of line reductions triggered, at least in part, by the section 505 practices indicates that they are relatively rare events.
The survey also asked each issuer how many cardholders were subject to a review that could have led to a line reduction or increase in interest rates based on one or more of the section 505 practices. In November 2009, of the 53 million active accounts at these two issuers, about 35 million cardholders were subject to such a review and consequently were "potentially affected." (In addition, one issuer with about 2 million active accounts that conducts reviews involving a section 505 practice did not conduct such reviews in November 2009.) As noted, fewer than 1,900 cardholders out of the 35 million reviewed experienced a line reduction that month due, at least in part, to section 505 practices. At about 320,000, the number of cardholders actually affected is notably larger during the 25-month period before the survey was mailed, but here, too, affected cardholders represent a small proportion of the total potentially affected. They also represent a small proportion of the number of cardholders that experienced a line reduction for any reason during this period.
It is important to note that transaction-specific information, such as the identity of the merchant or the amount of a charge, is an essential element of fraud detection and prevention systems. In this context, credit card issuers routinely use transaction-related information to temporarily block access to accounts or to close accounts until the cardholder can be contacted to verify the authenticity of the transaction. The survey excluded the use of transaction-related information for purposes of fraud detection and prevention.
Assessing the Incidence of Credit-Line Reductions across Populations
It is not possible to conduct a study of the potential "adverse effects" of the practices cited in section 505 on minority or low-income cardholders using information from credit card issuers. Under Regulation B, which implements the Equal Credit Opportunity Act (ECOA), creditors cannot collect information on the race or ethnicity of their credit card customers.4 Also, credit card issuers do not generally maintain verified or up-to-date information on the incomes of their cardholders.5
An enhanced set of credit records can be used to identify whether there is any association between line reductions in general and population characteristics.6 The Federal Reserve staff has access to a nationally representative sample of credit records of individuals drawn from the credit records maintained by TransUnion LLC (TransUnion).7 These individual credit record data have been supplemented with information on the race or ethnicity of the individuals, as well as the incomes and racial or ethnic profiles of the residents of the neighborhoods where the individuals reside.8
However, two factors severely limit the ability to draw firm conclusions about whether any of the practices identified in section 505 are at the root of any association between credit line reductions and demographics that may be detected. First, no information about the practices identified in section 505 is included in credit record data. In fact, issuers do not report to the credit bureaus the specific reasons why a credit-line reduction occurred, and therefore such information does not exist in the credit record data. Only limited information that may have played a role in a line reduction, such as failure to repay the balance owed on the account in question or on other credit accounts, is reflected in credit records.
Second, the special survey of card issuers conducted for this report found that only a relatively small number of cardholders that experienced credit-line reductions have been affected by any of the practices identified in section 505. As such, the credit-line reductions observed in the credit record data largely reflect line reductions for reasons other than those identified in section 505, and therefore it is impossible to know whether any of the section 505 practices lie at the root of any association between credit line reductions and demographics. If the section 505 practices had been found to have been much more widely used and to have affected a large proportion of cardholders, the credit record data might have provided a reasonable way to assess the effects of the section 505 practices on different populations.
Although an incidence analysis focusing on the relatively few consumers affected by the practices listed in section 505 is not feasible, the Federal Reserve staff examined the incidence of credit-line reductions in general across population subgroups using credit record data.9 The staff conducted this analysis in the spirit of responding to the specific congressional request and also to help shed some light on a broader, but nevertheless important and related, subject. The findings from this analysis are presented at the end of this report.
Recommendations
Very few issuers of general-purpose bank credit cards currently consider transactionrelated or mortgage-lender-specific information of the type identified in section 505 to reduce credit limits or set interest rates on accounts. Discussions with card issuers and their federal supervisors found that other metrics used to manage credit risk, such as measures of late payment behavior or measures of cash advance activity, are much more important factors in reducing lines or setting interest rates than the practices identified in section 505. Moreover, significant changes in broad economic conditions can be important factors leading to changes in account terms or account closures.
To the extent that transaction-related information is currently used by issuers to reduce credit limits or raise interest rates, it tends to be general in nature and does not rely on the identity of the individual store or merchant, the types of items purchased in a transaction, or the price of any given item. Moreover, in conversations with issuers, it was reported that unfavorable public attention drawn by the use of some forms of this information to reduce credit lines deterred them from adopting such practices going forward. Also, fair lending laws create compliance risk that may deter issuers from using certain types of information, particularly some forms of geographic categorization.
If the Congress were to consider restrictions on card issuer practices in this regard, it would be important to recognize the central role that detailed transaction-specific information plays in fraud detection and prevention systems. For example, the identity of the merchant, the merchant location, and the size and timing of charges are all central elements of fraud detection and prevention systems. Restricting the use of this type of information for purposes of fraud detection and prevention would likely substantially reduce the effectiveness of these systems and raise creditors' costs and, ultimately, the prices consumers pay for credit cards.
Although the current use of transaction-specific information of the type identified in section 505 for purposes of credit risk management appears quite limited, public policy actions prohibiting the use of these practices could still be costly over time. For example, prohibition would preclude potential innovations that would help card issuers manage credit risk more effectively. Such potential costs should be weighed against the possible benefits from restricting these practices.
Introduction
Creditors consider a wide variety of information in managing and assessing credit risk and account profitability. Over the past year or so, concerns have been raised about the types of information some creditors may consider when setting and adjusting credit terms on credit cards they issue to consumers.10 Some of these concerns have been expressed in press accounts, and reference was made to them in a 2009 hearing held by the U.S. House of Representatives.11 For the most part, the types of information that have been the focus of attention relate to the possible use of transaction-specific details, such as the identity or location of the merchant or store involved in a credit card transaction or the types or prices of the items purchased. Concerns have also been raised about whether the identity of a mortgage lender a consumer may have used or that of the firm servicing the mortgage is a factor in setting the terms of the consumer's credit card account.
In order to learn more about the practices of credit card issuers, section 505 of the Credit Card Accountability Responsibility and Disclosure Act (Credit CARD Act) of 2009 requires the Federal Reserve Board to prepare a report to the Congress (see appendix A, "Section 505 of the Credit CARD Act of 2009"). The report is to consider the practices used by credit card issuers at any time in the three-year period preceding enactment of the statute. The report is to include three core items: (1) estimates of the number of credit card issuers that have used the types of information specified in section 505 when deciding whether to reduce credit limits or raise interest rates applicable to an account; (2) an assessment of the extent to which the practices identified in section 505 have an adverse affect on minority or low-income consumers; and (3) based on that assessment, recommendations from the Federal Reserve on any regulatory or statutory changes that may be needed to restrict or prevent such practices.12 This report was prepared in response to the congressional request.
Consistent with the congressional request, this report focuses on open-end credit card plans issued to consumers. Thus, the report does not consider credit cards issued to businesses. It also does not consider "charge cards"--that is, cards issued to consumers whereby amounts charged during a month are due in full at the end of the billing period. Finally, it may exclude some credit cards issued by retailers or other merchants.13
Information Used to Prepare the Section 505 Report
Information used to prepare this report was drawn primarily from three sources: (1) a nationally representative special survey of credit card issuers focusing on certain practices of raising interest rates and reducing credit limits, (2) discussions with the largest issuers of bank credit cards, and (3) discussions with the federal agencies that supervise banking institutions and credit unions about the use of the practices listed in section 505. In addition, credit record data were used to examine the incidence of credit-line reductions for any reason across populations.
The Mail Survey of Credit Card Issuers
To gather information on the number of creditors that have engaged in one or more of the practices identified in section 505, the Federal Reserve conducted a mail survey that included each of the 100 largest issuers (measured by total dollars of outstanding credit card balances on their books or in credit-backed securitizations) and a credit card asset-weighted sample of an additional 75 randomly selected bankcard issuers with credit card assets exceeding $20 million (the survey is shown in appendix B, "The Survey of Credit Card Issuer Practices").14 The banking institutions included in the survey issue more than 98 percent of all the general-purpose bank credit cards extended to consumers.
Card issuers were identified from the Consolidated Reports of Condition and Income, or Call Report, for commercial banks and savings associations as of September 30, 2009, and for credit unions as of June 30, 2009, and included any banking institution that had credit card outstandings and offered credit cards to the public.15 Participation in the survey was mandatory (the reason for mandatory participation is discussed later).
The survey focused on open-end credit cards issued to consumers. For the survey, openend credit card accounts included general-purpose cards (such as MasterCard, Visa, Discover, or American Express), private label cards, and co-branded cards. The survey did not include lines of credit, such as overdraft lines tied to checking accounts, or charge cards. It also did not include credit cards issued to businesses.
The survey asked each issuer whether it had used any of the practices identified in section 505 at anytime in the three-year period preceding the receipt of the survey, which was defined as the period from November 30, 2006, through November 30, 2009. For those that responded affirmatively, a question also asked about the use of these practices in the month preceding the survey, which was defined as November 2009. For institutions that responded affirmatively to either question, follow-up questions asked about (1) the number of cardholders with active accounts that could potentially have been affected by the practices; and (2) the number of cardholders with active accounts that actually experienced an increase in their interest rate, a reduction in their credit line, or an account closure due to each of the specific practices.16 In both instances, issuer calculations were based on the number of cardholders with active accounts and not on dollars owed on those accounts.
To determine the number of account holders actually experiencing a credit-line reduction or interest rate increase due to the practices cited in section 505, the survey instructed institutions to review the reasons provided to cardholders in adverse action notices both for the month of November 2009 and over the 25-month period from November 1, 2007, through November 30, 2009. Regulation B requires that creditors provide customers with adverse action notices that identify as many as four main reasons for an adverse action, and these records must be retained for 25 months.17 If a cardholder received multiple adverse action notices over the 25-month period, the cardholder was counted as having been affected by the policies if any one of the adverse action notices identified a practice cited in section 505 as a reason for the action.
Participation in the survey was mandatory out of concern about the potential for "nonresponse bias." Nonresponse bias can arise if the institutions that choose not to respond to a voluntary survey differ systematically from those that voluntarily respond. In the current application, nonresponders might be more likely to have engaged in the practices identified in section 505.18 Nonresponse bias, if it occurred to any notable degree, would make it impossible to respond to the congressional request to identify the number of credit card issuers that use one or more of the factors identified in section 505.
Telephone Interviews with the Largest Credit Card Issuers
Prior to initiating the mail survey of issuer practices, the Federal Reserve staff conducted telephone interviews with nine of the largest credit card issuers to gain a better understanding of these creditors' interest rate and credit-limit-setting practices. These background discussions covered the techniques and types of information considered in managing credit card accounts both for the purposes of detecting and controlling fraud and for broader credit risk and account profitability purposes. The discussions included a review of the reasons provided to the cardholder when an adverse action notice was sent to explain why the account experienced an increase in interest rate or a reduction in credit limit. The telephone interviews were also used as a sounding board to determine whether it was feasible for card issuers to answer the types of questions ultimately included in the mail survey.
Interviews with Federal Supervisory Agency Staff
To gain further information, the Federal Reserve staff contacted the principal federal banking institution supervisors and asked them about the general practices of credit card issuers in reducing credit lines or raising interest rates and whether they had observed issuers using any of the practices identified in section 505. The supervisors have considerable knowledge of the underwriting and account management practices of the credit card issuers they supervise, and their experiences provided information used to prepare this report.
The Sample of Credit Records Used in This Report
The Federal Reserve obtained from TransUnion the complete credit records (excluding any identifying personal or creditor information) of a nationally representative random sample of 301,536 individuals as of June 30, 2003.19 The Federal Reserve subsequently received updated information on the credit records of these individuals as of December 31, 2004; June 30, 2007; and December 31, 2008. Updates were not available for all individuals in the original 2003 sample, typically because some individuals had passed away. As each update was received, an additional group of individuals was added to the credit record sample to replace those whose credit records were no longer active. This procedure ensured that the updated credit record samples provided both an ongoing panel of individuals and a nationally representative random cross section sample of credit records.
The credit record samples as of June 30, 2007, and December 31, 2008, were used to prepare this report. In the aggregate, the June 30, 2007, sample included information on 302,441 individuals (table 1). Most individuals had records of a credit account in their files, but some had only records pertaining to (1) collection agency actions; (2) monetary-related public actions, such as bankruptcy or garnishment; or (3) inquiries initiated by creditors.
Further details about the credit record data used in this report, including how individuals were categorized by income or race or ethnicity, are provided in appendix C, "The Credit Record Data Used in This Report."
| Type | Number | Percent |
|---|---|---|
| Total | 302,441 | 100.0 |
| Credit account | 252,400 | 83.5 |
| Public record | 40,301 | 13.3 |
| Collection agency account | 106,732 | 35.3 |
| Creditor inquiry1 | 175,315 | 58.0 |
| None of the above | 6,142 | 2.0 |
| MEMO | ||
| Credit account only | 79,316 | 26.2 |
| Public record only | 2,790 | .9 |
| Collection agency account only | 24,892 | 8.2 |
| Creditor inquiry only1 | 7,759 | 2.6 |
1. Includes only inquiries made within two years of June 30, 2007. Return to table
Source: Here and in subsequent tables except as noted, Federal Reserve sample of credit records supplied by TransUnion.
Findings from Interviews with Credit Card Issuers
To learn more about the account management practices of creditors, the Federal Reserve staff conducted telephone interviews with nine of the largest credit card issuers for purposes of gathering background information. As of the end of 2008, these nine companies accounted for about three-fourths of all general-purpose credit cards issued to consumers nationwide.20
The telephone interviews provided information on the techniques and processes that card issuers use to manage their accounts and the types of information they consider in deciding whether to close or block access to accounts, raise or reduce credit limits, or change the interest rate on a consumer's account. Additional information on account management practices and the types of information considered by issuers was obtained from federal banking regulatory agency staff charged with overseeing the safety and soundness of credit card issuers and from discussions with entities that provide account management services to card issuers.
Account Management Practices
Credit card issuers consider information drawn from their own records about a cardholder's account use, as well as information obtained from third parties, primarily the three national credit reporting agencies. Card issuers (or third-party vendors) use these data to construct statistical models that seek to predict a consumer's risk of future nonpayment or other possible outcomes. Some card issuers have sufficient resources and scope to construct and implement their own in-house proprietary systems, while others rely on products made available by other institutions. In some cases, card issuers supplement their in-house systems with products developed by these outside vendors. In addition to these automated systems, issuers also rely on manual processes to manage credit risk. In this context, the card issuer may contact the cardholder to learn more about the circumstances that led to a payment issue. Account managers may elect to override actions, such as a reduction in credit limits, initiated by the automated systems.
Account management systems are used to help determine account-level actions, including setting or changing credit limits and interest rates on accounts and closing accounts. As already mentioned, these systems rely heavily on information garnered from the issuers' own records on the behavior of a cardholder and from data gathered on the customer from third-party records. The card issuer's own records provide high-frequency information on a range of customer activities, including measures of borrowing and repayment behavior and of account use. Data from credit-reporting agencies provide a more holistic picture of the current credit circumstances of an individual than can be gleaned from the more limited information that a creditor has from its own credit-card-account-level or institution-wide data on the customer. Credit record data include detailed account-level information on a customer's various outstanding open and closedend loans with other creditors, information on the cardholder's recent efforts to secure additional credit, records of collection actions taken against the cardholder, and adverse monetary-related public record information. The credit-reporting agencies may also provide a summary measure of the cardholder's creditworthiness, such as a credit score or change in credit score over a period of time.
Many of the account-level behaviors that card issuers consider from their internal records on a cardholder focus on payment experience, including the timeliness and size of payments and the degree of delinquency when a consumer fails to meet a payment obligation. For example, the degree of delinquency might focus on the length of time since the last payment was due, and the magnitude of the delinquency might consider the sizes of the balances that are in arrears. Payment behaviors that are considered also include information on the extent of the payment, such as the amount of the last payment or the payment as a percentage of the amount due; records of payments returned for insufficient funds in the account used to make the payment; and over-the-limit experience, such as the frequency and degree of over-the-limit occurrences. In addition, card issuers may consider the extent and amount of cash advance activity (including timing, dollar amount and frequency, and changes in such activity).
Issuers may also consider internal records they have about a customer from their other account relationships, such as checking or savings accounts or other loans. For example, if a cardholder becomes delinquent on a vehicle loan or personal line of credit extended by the card issuer or an affiliated entity, the card issuer may reduce the credit limit on the consumer's credit card account. It should be noted that provisions of the Credit CARD Act of 2009 prohibit card issuers from increasing the interest rate on balances owed on an account except in limited circumstances (for example, if a cardholder's payment is 60 or more days late). However, issuers may raise interest rates on new balances. The provisions in the act do not restrict issuers' ability to reduce credit limits.
Beyond the types of basic credit record or on-account behavioral models already described, some models consider information drawn from transaction-authorization data. Because such data have many dimensions and possible permutations, only the largest issuers can successfully develop and implement such models using only internal data on their cardholders. Independent third parties that can aggregate data across many issuers can build default prediction models using transaction data more reliably and then make these models available to issuers of all sizes. Some of these products integrate information drawn from both consumer credit records and transaction-related information.21 Others are developed entirely using transaction-authorization data.22 The models may be used in conjunction with other risk-management tools and can be used to help an issuer decide whether to reduce a cardholder's credit limit, close the account, or increase the rate of interest.
Discussions with the providers of such models indicate that the models differ from each other and do not all include the same factors. Some of the factors that may be considered separately or in combination in one or more of the models include, for example, transaction amounts, merchant category codes, time of a transaction, different metrics of cash advance activity, and a designation of the broad areas within a metro area or state where the merchant is located.23 The discussions indicate the models do not target specific merchants or stores by name, the specific items purchased, or their prices. In fact, price and specific product information are not included in the transaction-authorization data of general-purpose bankcards.24 Nor do these models include information on the identity of the mortgage lender. In these systems, the underlying data on consumer activities are updated daily, providing card issuers with actionable information on a high-frequency basis.
Open-end lending products, such as credit cards, differ in one major respect from closedend products, such as vehicle loans or most types of mortgages. For closed-end products, the key judgments are made at the time an application for credit is reviewed. The decision of whether to extend a loan and the pricing of the loan are established at the outset of the process. Verification of the borrower's identity and of the veracity of the information submitted by the applicant or other parties occurs primarily at the initial stage of the lending process. For openend credit products, for which the borrowing and future-use patterns are heavily determined by the consumer, fraud, credit risk, and account profitability management decisions are made both at the start of an account relationship and on an ongoing basis.
Because open-end accounts can be used at the discretion of the consumer and because the consumer's financial and other circumstances are continually changing, constant review of account activity is essential to managing the credit risk posed by the account holder and to ensuring the needs of the customer are met. Moreover, since open-end credit products, including credit cards, are particularly vulnerable to criminal activity, constant monitoring of account activity is necessary for detecting and controlling fraud or other illegal activity.
Fraud Management and Control Systems
Issuers use systems established specifically to help identify and manage fraudulent (illegal) or suspicious use of accounts. These systems tend to be distinct from credit risk systems and often emphasize transaction-related information. Systems built to detect and control fraudulent or suspicious use of accounts monitor account behavior closely, tracking individual transactions to identify uses, or patterns of uses, that suggest a credit card (or number) may have been stolen or used by someone other than the account holder (for example, a family member or former spouse) in an unauthorized manner. Fraud detection systems may use a holistic view of the account holder that considers behavior across different products extended by the issuer, including both credit and debit cards and account holder payment experience on these accounts.25
Many card issuers rely on automated systems provided by third parties to detect and manage fraud.26 These systems sometimes involve sophisticated applications that have the ability to learn and understand individual cardholder spending patterns. Based on the feedback the systems provide to account managers, the issuer can take many actions, including (1) rejecting or challenging transactions, (2) temporarily blocking access to the card until the account holder can be contacted to verify the authenticity of the transaction, (3) closing an account and reissuing a new card to a customer, or (4) creating a chargeback to the merchant.
Detailed transaction and cash advance information are important components of fraud detection systems. Information indicating from whom an item is purchased, when the transaction took place, the location of a merchant (for example, out of the country or a significant distance from the cardholder's residence), the automated teller machine (or ATM) identification number used for cash advances, and the dollar amount involved are all factors considered in fraud detection systems. Fraud detection systems are generally not used to reduce credit limits or raise interest rates on accounts; rather, these systems are used to aid decisions about whether to close an account or temporarily shut off access to the account until the account holder is contacted or other actions are initiated.
Findings from the Survey of Credit Card Issuers
To learn about the extent to which issuers of open-end credit cards to consumers reduce credit lines or raise interest rates based on information such as the type or location of the merchant the consumer transacts with or the types or prices of items purchased using issuers' cards, the Federal Reserve conducted a national mail survey of 175 credit card issuers (see appendix B).27 Taken together, these issuers account for the vast majority of general-purpose bank credit cards issued to consumers and of outstanding balances on such accounts. As of November 2009, these 175 credit card companies had issued almost 677 million cards to consumers, of these 350 million were used by the cardholder in the previous 12 months.
Results of the survey provide a way to estimate both (1) the number of credit card issuers that use each of the practices identified in section 505 and (2) a measure of the how many cardholders have been both potentially and actually affected by each of these practices. The survey design examines both activities that occurred anytime in the three-year period preceding receipt of the survey and practices that were in effect as of November 2009, the month before the survey was mailed to respondents.
Interpreting Issuer Responses
In the responses to the survey and in several follow-up conversations with card issuers, certain practices came to light that, while not specifically listed in section 505, nonetheless generally relate to cardholder transaction activity or the mortgage lender they use. In each case, the Federal Reserve staff decided how to classify the responses for purposes of analysis. The following discussion describes these practices and whether the staff considered them to be within the purview of the report.
Transaction-Velocity-Related Activity
One reported practice is the use of "transaction velocity" for account risk management. Transaction velocity refers to the number or dollar amount of transactions made by a cardholder in a given period (for example, the previous few months for longstanding accounts or the initial few weeks for new accounts). Transaction velocity is not tied to any specific merchant, location, product, or price. Issuers have indicated to Federal Reserve staff that unusually high velocity or a change in velocity may reflect cardholder behavior that is indicative of elevated credit risk.
A few issuers considered transaction-velocity-related metrics as relevant to section 505 and reported using them in their survey responses. In contrast, other issuers did not report this practice, although they do use some form of transaction velocity for risk management (as revealed in follow-up conversations). For purposes of categorizing responses to the survey, Board staff did not consider the issuers that reported transaction-velocity-related actions to have reduced credit lines for the reasons outlined in section 505.28 This decision was made primarily because this practice does not involve the identity of a specific merchant, its location, the type of product purchased, or its price.
Cash Advance Activity
Many card issuers monitor cash advance activity as a risk indicator. The survey instrument instructed respondents that this practice does not fall under the practices of interest outlined in section 505. However, one card issuer indicated that, for some of its new cardholders, it monitors cash advance activity in conjunction with card purchases at certain types of merchants (for example, jewelry and electronic goods stores). According to this issuer, it has found such behaviors among new cardholders to be an early warning of elevated credit risk. Cardholders meeting these criteria were subject to line reductions, and Board staff considered such line reductions to be relevant for the analysis presented here.29
Gambling Activity
Some issuers report that they consider gambling activity as an indicator of elevated credit risk. Federal Reserve staff initiated follow-up telephone interviews with four respondents that reported gambling activity as a factor in line reduction decisions. These issuers reported that they contact the delinquent cardholder to determine the reason for the delinquency. If the delinquency arose from "excessive" gambling, they would cut the available credit line or close the account.30 However, gambling activity would not generally result in such actions if the account was not already in arrears or did not show elevated levels of debt. Moreover, delinquent accounts are very likely to be subject to line reductions or account closures even in the absence of gambling activity. For purposes of categorizing responses, the issuers that consider gambling only when coupled with a delinquency or an exceeding of a credit limit are not considered to have made a line reduction for one of the reasons cited in section 505.
Merchant-Related Activity
One issuer reported that it had reduced the credit lines of cardholders with lower credit scores if they had not shopped at the retail store affiliated with the card issuer. This issuer reported that it found that cardholders with lower scores who did not shop at the affiliated retailer had elevated delinquency rates. For purposes of categorizing responses to the survey, this issuer was considered to have used one of the transaction-related practices of interest.
Two issuers reported using Merchant Category Codes (for example, grocery stores, furniture stores, and so on) or Standard Industrial Classification codes, along with other information on cardholder performance, as factors in behavioral risk-scoring models developed by third parties.31 In one case, these codes were matched to the geographic areas where the merchants were located. The level of geography used is broad but may be smaller than a large metropolitan area. A third issuer grouped all charges into a small number of very broad categories and considered the performance of its other customers within those categories in deciding on line reductions for a given individual. Each of these practices was considered transaction related for purposes of classifying responses for this report.
One issuer, in the past, combined information on the shopping patterns of its cardholders with data on cardholder performance to help assess the risk of individual accounts. More specifically, in considering whether to reduce the credit line of a given cardholder, this issuer considered the performance and spending patterns of other cardholders with similar creditrelated characteristics who shopped at the merchants where the given cardholder had made purchases. Some of the merchant codes were store specific and, consequently, location specific, while others did not focus on a specific store but rather a chain of outlets for a given merchant. In implementing the policy, the issuer did not look to the name of the firm (such as ABC Drugstore), but rather it used anonymous codes. Federal Reserve staff determined that this practice fit within the scope of practices of concern.
Mortgage-Lender-Related Activity
Two issuers reported considering mortgage-lender-related information in the past. Both considered the performance of mortgages at the lenders where their cardholders had mortgages. For one issuer, if a cardholder had a mortgage loan from a lender with a badly performing mortgage portfolio, the issuer considered this factor when making decisions on possible reductions in credit lines. In implementing this risk-management strategy, the issuer used anonymous lender codes rather than the specific name of the lender.
The second issuer also reported considering the mortgage lenders of cardholders in their line reduction decisions. This issuer had acquired the mortgage loan portfolio of a lender that extended mortgages allowing borrowers wide latitude in the amounts of their mortgage payments (often referred to as "pick a pay" loans). A review of the credit card performance of individuals with such mortgages revealed relatively high delinquency rates on their credit cards. This issuer reviewed the circumstances of these individuals (for example by reviewing their credit records) in order to make decisions about possible line reductions. Federal Reserve staff determined that the practices of both of the card issuers that consider mortgage-related information fit within the scope of section 505.
The Number of Issuers Engaging in the Practices
Tables 2a, 2b, and 2c summarize the responses to the survey regarding the number of respondents that reported engaging in each of the practices of concern identified in section 505, both currently and at any time in the past three years. The three tables cover (1) line reductions,(2) increases in interest rates, and (3) account closures. Because practices may differ by issuer type and size, commercial banks and savings institutions are presented in two categories: those with credit card outstandings that place them among the 25 largest such institutions and all others. Credit unions are shown separately, but to help ensure that their responses remain anonymous, the credit unions are not categorized by credit card outstandings.
Few credit card issuers reported that they engaged in any of the practices of concern as of November 2009 or at any time in the past three years. None of the credit unions and only six of the commercial banks or savings institutions that issue cards considered any of these practices when making decisions about line reductions (table 2a). As of November 2009, only two of these issuers considered any of the practices in making such decisions.
Among the largest 25 card issuers, 2 considered the type of transaction when making decisions regarding line reductions in November 2009. None of these large issuers currently consider any of the other factors in making these decisions. In the past three years, these practices were somewhat more common, as 5 of the larger issuers, including the 2 that currently consider the type of transaction, engaged in one or more of the practices when making line reduction decisions. The one card issuer outside the top 25 that considered any of the practices considered the identity of the merchant in the past three years but not in November 2009.
When making decisions about interest rate increases on credit cards, 2 of the issuers in the top 25 reported taking into account the type of credit transaction in the past three years (table 2b). One of these issuers used this practice in November 2009. None of the credit unions and none of the card issuers outside the top 25 considered any of the factors in decisions regarding interest rate increases.
Regarding account closure decisions, 4 of the top 25 card issuers reported that they considered the type of credit transaction in the past three years, and 1 considered the type of transaction as of November 2009 (table 2c). Of the 4 issuers, 2 considered the identity of the merchant in account closing decisions in the past three years, but none did so in November 2009. None of the credit unions and none of the card issuers outside the top 25 considered any of the factors in decisions regarding account closures.
| Congress-specified practice | Period1 | Total | Commercial banks | Credit unions | |
|---|---|---|---|---|---|
| 25 largest2 | Others | ||||
| Geographic location where a credit transaction took place | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 1 | 1 | 0 | 0 | |
| Identity (name) of the merchant involved in a credit transaction | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 3 | 2 | 1 | 0 | |
| Type of credit transaction | Nov.-09 | 2 | 2 | 0 | 0 |
| Nov.-06 to Nov.-09 | 5 | 5 | 0 | 0 | |
| Types of items purchased in a credit transaction | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Prices of items purchased in a credit transaction | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Any changes in the types of items purchased in the consumer's credit transaction history | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Any changes in the prices of items purchased in the consumer's credit transaction history | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Identity of a mortgage creditor that extended a mortgage loan secured by the primary residence of the consumer | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 3 | 3 | 0 | 0 | |
| Identity of a mortgage creditor that holds a mortgage loan secured by the primary residence of the consumer | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 3 | 3 | 0 | 0 | |
| Any other data pertaining to the use of such credit card account | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Total | Nov.-09 | 2 | 2 | 0 | 0 |
| Nov.-06 to Nov.-09 | 6 | 5 | 1 | 0 | |
| Total (any practice at any time) | 6 | 5 | 1 | 0 | |
Note: One issuer that reduced credit limits because of section 505 practices during the period from November 30, 2006, through November 30, 2009, did not reduce limits based on these practices during November 2009 but continues to use this practice on a periodic basis. This information was revealed to Federal Reserve Board staff during a followup interview with the issuer.
1. In this and subsequent tables, "Nov.-09" refers to November 1, 2009, through November 30, 2009. "Nov.-06 to Nov.-09" refers to November 30, 2006, through November 30, 2009. Return to table
2. In this and subsequent tables, "largest" is based on credit card outstandings from the September 2009 Consolidated Reports of Condition and Income. Return to table
| Congress-specified practice | Period1 | Total | Commercial banks | Credit unions | |
|---|---|---|---|---|---|
| 25 largest2 | Others | ||||
| Geographic location where a credit transaction took place | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Identity (name) of the merchant involved in a credit transaction | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Type of credit transaction | Nov.-09 | 1 | 1 | 0 | 0 |
| Nov.-06 to Nov.-09 | 2 | 2 | 0 | 0 | |
| Types of items purchased in a credit transaction | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Prices of items purchased in a credit transaction | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Any changes in the types of items purchased in the consumer's credit transaction history | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Any changes in the prices of items purchased in the consumer's credit transaction history | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Identity of a mortgage creditor that extended a mortgage loan secured by the primary residence of the consumer | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Identity of a mortgage creditor that holds a mortgage loan secured by the primary residence of the consumer | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Any other data pertaining to the use of such credit card account | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Total | Nov.-09 | 1 | 1 | 0 | 0 |
| Nov.-06 to Nov.-09 | 2 | 2 | 0 | 0 | |
| Total (any practice at any time) | 2 | 2 | 0 | 0 | |
Note: See table 2a, notes 1 and 2.
| Congress-specified practice | Period 1 | Total | Commercial banks | Credit unions | |
|---|---|---|---|---|---|
| 25 largest2 | Others | ||||
| Geographic location where a credit transaction took place | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 1 | 1 | 0 | 0 | |
| Identity (name) of the merchant involved in a credit transaction | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 2 | 2 | 0 | 0 | |
| Type of credit transaction | Nov.-09 | 1 | 1 | 0 | 0 |
| Nov.-06 to Nov.-09 | 4 | 4 | 0 | 0 | |
| Types of items purchased in a credit transaction | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Prices of items purchased in a credit transaction | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Any changes in the types of items purchased in the consumer's credit transaction history | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Any changes in the prices of items purchased in the consumer's credit transaction history | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Identity of a mortgage creditor that extended a mortgage loan secured by the primary residence of the consumer | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Identity of a mortgage creditor that holds a mortgage loan secured by the primary residence of the consumer | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Any other data pertaining to the use of such credit card account | Nov.-09 | 0 | 0 | 0 | 0 |
| Nov.-06 to Nov.-09 | 0 | 0 | 0 | 0 | |
| Total | Nov.-09 | 1 | 1 | 0 | 0 |
| Nov.-06 to Nov.-09 | 4 | 4 | 0 | 0 | |
| Total (any practice at any time) | 4 | 4 | 0 | 0 | |
Note: See table 2a, notes 1 and 2.
The Number of Cardholders Affected by the Practices
Tables 3a, 3b, and 3c provide results of the survey questions asking each credit card issuer that engaged in any of the practices identified in the survey to indicate how many of its customers were potentially and actually affected by the practices both in November 2009 or at any time in the past twenty-five months (November 1, 2007 through November 30, 2009). As before, because practices may differ by issuer type and size, the tables group commercial banks and savings institutions in two categories: those with credit card outstandings that place them into the 25 largest such institutions and all others. Once again, credit unions are shown separately.
Because 5 of the 6 issuers that considered one or more of the practices in line reduction decisions are large and because the automated systems that incorporate these factors are widely applied, many millions of cardholders have been potentially affected by these practices. However, the numbers actually affected are much smaller (table 3a). In total, about 320,000 cardholders experienced a line reduction in the past three years due, at least in part, to one or more of the practices. However, issuers report that many fewer cardholders were actually affected as of November 2009: About 1,900 cardholders experienced a line reduction due, at least in part, to one or more of the practices in the month before the survey--a small proportion of the roughly 35 million cardholders potentially affected that month.
Regarding changes in interest rates on accounts, no cardholders were affected in November 2009, although about 194,000 were affected in the past three years (table 3b). With respect to account closings, 416 cardholders were affected over the past three years, and none were affected in November 2009 (table 3c).
To provide additional perspective on the number of cardholders affected by the section 505 practices, we exploit questions posed to card issuers about line reductions, interest rate increases, or account closures that took place for any reason and about the subset of cardholders who received an adverse action notice for any reason, not just those related to section 505 practices. The questions focused on activity over two time periods: November 2009 and the 25-month period from November 1, 2007, through November 30, 2009.
Regardless of how it is measured, the 1,900 cardholders that experienced a line reduction due, at least in part, to the section 505 practices during November 2009 represent only a very small proportion of the credit card accounts issued by the two credit card companies that used these practices at that time (table 4). (Because most cardholders have only one active account from a given issuer, it is reasonable to interpret numbers of accounts reported here as equivalent to numbers of cardholders when focusing on a given issuer.) These 1,900 cardholders represent only 0.00004 percent of all the active accounts extended by the two issuers, 0.56 percent of the accounts that experienced a line reduction (for any reason) from these issuers, and 0.57 percent of the accounts that experienced a line reduction accompanied by an adverse action notice. Proportions for those experiencing increases in interest rates or account closures are also small as shown in table 4.
| Congress-specified practice | Period1 | Extent to which customers are affected2 | Total | Commercial banks | Credit unions | |
|---|---|---|---|---|---|---|
| 25 largest3 | Others | |||||
| Geographic location where a credit transaction took place | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 5,236,919 | 5,236,919 | 0 | 0 | |
| Actually | 12,380 | 12,380 | 0 | 0 | ||
| Identity (name) of the merchant involved in a credit transaction | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 13,815,749 | 13,806,619 | 9,130 | 0 | |
| Actually | 90,388 | 81,258 | 9,130 | 0 | ||
| Type of credit transaction | Nov.-09 | Potentially | 34,842,795 | 34,842,795 | 0 | 0 |
| Actually | 1,890 | 1,890 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 60,431,827 | 60,431,827 | 0 | 0 | |
| Actually | 124,485 | 124,485 | 0 | 0 | ||
| Types of items purchased in a credit transaction | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Prices of items purchased in a credit transaction | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Any changes in the types of items purchased in the consumer's credit transaction history | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Any changes in the prices of items purchased in the consumer's credit transaction history | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Identity of a mortgage creditor that extended a mortgage loan secured by the primary residence of the consumer | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 13,845,518 | 13,845,518 | 0 | 0 | |
| Actually | 107,777 | 107,777 | 0 | 0 | ||
| Identity of a mortgage creditor that holds a mortgage loan secured by the primary residence of the consumer | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 13,845,518 | 13,845,518 | 0 | 0 | |
| Actually | 107,777 | 107,777 | 0 | 0 | ||
| Any other data pertaining to the use of such credit card account | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Total | Nov.-09 | Actually | 1,890 | 1,890 | 0 | 0 |
| Nov.-07 to Nov.-09 | Actually | 323,714 | 314,584 | 9,130 | 0 | |
Note: Components may not sum to totals because if an individual was affected by more than one practice, the individual was counted only one time.
1. In this and subsequent tables, "Nov.-09" refers to November 1, 2009, through November 30, 2009. "Nov-07 to Nov-09" refers to November 1, 2007, through November 30, 2009. Return to table
2. Cardholders potentially affected are those who might have had their credit limits reduced because of a given practice. Cardholders actually affected are those who had their credit limits reduced because of the practice; a cardholder was considered to be actually affected if the cardholder received an adverse action notice, as required by the Equal Credit Opportunity Act, that included the practice as one of the up to four reasons for the action. Return to table
3. See table 2a, note 2. (In this and subsequent tables, "largest" is based on credit card outstandings from the September 2009 Consolidated Reports of Condition and Income.) Return to table
| Congress-specified practice | Period1 | Extent to which customers are affected2 | Total | Commercial banks | Credit unions | |
|---|---|---|---|---|---|---|
| 25 largest3 | Others | |||||
| Geographic location where a credit transaction took place | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Identity (name) of the merchant involved in a credit transaction | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Type of credit transaction | Nov.-09 | Potentially | 2,533,745 | 2,533,745 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 30,246,698 | 30,246,698 | 0 | 0 | |
| Actually | 194,246 | 194,246 | 0 | 0 | ||
| Types of items purchased in a credit transaction | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Prices of items purchased in a credit transaction | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Any changes in the types of items purchased in the consumer's credit transaction history | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Any changes in the prices of items purchased in the consumer's credit transaction history | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Identity of a mortgage creditor that extended a mortgage loan secured by the primary residence of the consumer | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Identity of a mortgage creditor that holds a mortgage loan secured by the primary residence of the consumer | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Any other data pertaining to the use of such credit card account | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Total | Nov.-09 | Actually | 0 | 0 | 0 | 0 |
| Nov.-07 to Nov.-09 | Actually | 194,246 | 194,246 | 0 | 0 | |
Note: See notes to table 3a
| Congress-specified practice | Period1 | Extent to which customers are affected2 | Total | Commercial banks | Credit unions | |
|---|---|---|---|---|---|---|
| 25 largest3 | Others | |||||
| Geographic location where a credit transaction took place | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 5,236,919 | 5,236,919 | 0 | 0 | |
| Actually | 102 | 102 | 0 | 0 | ||
| Identity (name) of the merchant involved in a credit transaction | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 13,806,619 | 13,806,619 | 0 | 0 | |
| Actually | 155 | 155 | 0 | 0 | ||
| Type of credit transaction | Nov.-09 | Potentially | 32,300,000 | 32,300,000 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 56,243,538 | 56,243,538 | 0 | 0 | |
| Actually | 194 | 194 | 0 | 0 | ||
| Types of items purchased in a credit transaction | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Prices of items purchased in a credit transaction | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Any changes in the types of items purchased in the consumer's credit transaction history | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Any changes in the prices of items purchased in the consumer's credit transaction history | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Identity of a mortgage creditor that extended a mortgage loan secured by the primary residence of the consumer | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Identity of a mortgage creditor that holds a mortgage loan secured by the primary residence of the consumer | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Any other data pertaining to the use of such credit card account | Nov.-09 | Potentially | 0 | 0 | 0 | 0 |
| Actually | 0 | 0 | 0 | 0 | ||
| Nov.-07 to Nov.-09 | Potentially | 0 | 0 | 0 | 0 | |
| Actually | 0 | 0 | 0 | 0 | ||
| Total | Nov.-09 | Actually | 0 | 0 | 0 | 0 |
| Nov.-07 to Nov.-09 | Actually | 416 | 416 | 0 | 0 | |
Note: See notes to table 3a
| Account action | Period1 | Percentage of all currently active accounts2 | Percentage of all cards that experienced an account action | MEMO: Number of issuers with practice | |
|---|---|---|---|---|---|
| Total cards | Cards with adverse action notices | ||||
| Credit limit reduction | Nov.-07 to Nov.-09 | … | 1.17569 | 2.3452 | 6 |
| Nov.-09 | .00004 | .55706 | .57346 | 2 | |
| Account closure | Nov.-07 to Nov.-09 | … | .01321 | .02949 | 4 |
| Nov.-09 | … | … | … | 0 | |
| Increased interest rate | Nov.-07 to Nov.-09 | … | .78642 | 2.28400 | 2 |
| Nov.-09 | … | … | … | 0 | |
Note: We assume that each cardholder has one active bank credit card account with each issuer. A review of the credit record sample used for this report found that fewer than 10% of the cardholders with active bank credit cards had more than one card from the same issuer.
1. See table 3a, note 1. (In this and subsequent tables, "Nov.-09" refers to November 1, 2009, through November 30, 2009. "Nov-07 to Nov-09" refers to November 1, 2007, through November 30, 2009.) Return to table
2. All survey issuers reported they have more than 353 million active credit cards in total as of November 2009. Of these active credit cards, more than 74 million card accounts were reported by the six issuers that have any Congress-specified practices during the past three years. The two issuers who have these practices during November 2009 account for 53 million card accounts. Return to table
… Not applicable. Return to table
Findings from Discussions with Other Supervisory Agencies
To gain further information about the practices of bank credit card issuers, the Federal Reserve staff contacted each of the federal banking institution supervisors, including the National Credit Union Administration.32 The supervisors are knowledgeable about the credit risk management practices of the credit card issuers they supervise.
Each of the agencies reported that the card issuers they supervise reduce credit limits or raise interest rates based on changes in broad economic conditions (such as changes affecting card issuers' cost of funds) and on metrics of credit risk drawn from credit-reporting agency records or the card issuers' own experiences with their cardholders. The supervisory agencies noted that a couple of institutions they supervised had used practices of the type identified in section 505 in the past to alter the terms offered to their cardholders but that they had not encountered the use of such practices recently.
Findings from the Assessment of Credit Record Data
As noted, Regulation B prohibits creditors from collecting information on the race or ethnicity of their credit card customers, and credit card issuers do not generally maintain verified or up-to-date information on the incomes of their cardholders. As a result, it is not possible to conduct a study of the adverse effects of the practices cited in section 505 on low-income or minority cardholders using information from credit card issuers on the racial or ethnic status or income of cardholders. It is possible, however, to conduct a more general analysis of the incidence of credit-line reductions by using the credit record data described earlier that the Federal Reserve staff constructed, which combine individual credit records with information on individual race or ethnicity and on the income status and racial and ethnic composition of the individual's neighborhood.
To conduct this analysis, the credit limit of each active bank credit card account for each individual as of June 30, 2007, was compared with the credit limit on that card as of December 31, 2008. Accounts that experienced reductions in their credit limits during the observation period are the main focus of the analysis. Because the credit records are as of two points in time, changes in credit limits reflect only the limits as last reported by the card issuer at each observation point. In some cases, the limit on a given credit card could have changed more than once during the observation period, but only the limits at the beginning and end of the observation period are available.33
Credit Records and the Reasons for a Reduction in Credit Limits
As noted, credit record data do not provide the reason(s) an account experienced a reduction (or increase) in the size of the credit line. A credit line may be reduced for a wide range of reasons related to how a consumer is using an account (for example, missing scheduled payments) or for reasons related to other measures of credit risk or potential account profitability. Sometimes the credit limit on an account may be reduced if an issuer provides a new account to a consumer with a credit limit that, when combined with the limit on the original (unclosed) account, is larger than the issuer wishes to make available. More generally, an issuer may need to reduce its credit risk exposure in response to adverse changes in the broader economy and may choose to do so by reducing credit lines for large numbers of its cardholders. Most important for this report, credit record data do not provide any direct evidence on whether any individual cardholder experienced a reduction in credit limit due, in whole or in part, to the specific reasons identified in section 505. Also, the survey of card issuers found that relatively few cardholders experienced line reductions related to the section 505 practices. Thus, the credit record data largely reflect line reductions of individuals that are entirely unrelated to section 505 practices. For these reasons, it is impossible to know whether any differences in the incidence of line reductions across populations revealed in an analysis of credit records are related to section 505 practices.
Sample of Cardholders Used in the Analysis
The basic unit of analysis is the cardholder. If a cardholder in the sample of credit records had multiple active bank credit card accounts and experienced a line reduction on any of the accounts over the observation period, the cardholder was classified as having experienced a line reduction. This classification approach is relatively extreme in that a cardholder is classified the same way if a line reduction is experienced on one account or on all accounts.
An alternative specification is also considered. The alternative approach is account, not cardholder, based. Under this approach, each active bank credit card account is considered separately and categorized as experiencing a line reduction or not. Thus, a cardholder with two accounts that experienced line reductions and two accounts that did not would, in effect, be represented as half as likely to experience a line reduction as that same cardholder in the individual-based approach. As it turns out, both approaches yield similar results regarding the association between incidence of line reductions and population demographics.
Incidence of Line Reductions across Populations
To assess the incidence of line reductions across populations, cardholders were sorted by income or racial or ethnic characteristics. Once categorized, the proportion of members in each group who experienced a line reduction on at least one of their active credit card accounts over the observation period is calculated.
Grouping Individuals
To conduct the analysis of the incidence of line reductions across populations, individuals were grouped by income and minority status. For the analysis by income, individuals were grouped into three categories using the census-tract median family income of their place of residence relative to the median family income of their metropolitan statistical area (MSA) or the nonmetro portion of the state for those residing outside of MSAs. The three relative income categories are (1) less than 80 percent, (2) between 80 percent and less than 120 percent, and (3) 120 percent or more. These three relative income categories are referred to later as indicating lower-income, middle-income, or higher-income areas.
Minority status was represented in two ways: by census-tract minority population and by self-reported minority status. For analysis by census-tract minority population characteristics, individuals were grouped into three categories based on the proportion of the census-tract total population that was minority. The three groups are census-tract minority populations of (1) less than 10 percent, (2) between 10 percent and 80 percent, and (3) more than 80 percent. The ranges used in the first and third groups were chosen to be relatively narrow because these two groups of census tracts are intended to represent with some degree of certainty the likely racial or ethnic status of the individual. Wider ranges would yield less certain classifications.
For the analysis by self-reported minority status, individuals were identified as either minority or non-Hispanic white using data from their Social Security Administration (SSA) records (with, as noted in appendix C, a few individuals categorized by their census-tract population characteristics).
Incidence of Line Reductions across Populations by Income or Minority Status
Credit record data indicate that about 8.5 percent of individuals with active generalpurpose bank credit cards experienced a reduction in the credit line on at least one of their active credit card accounts between June 30, 2007, and December 31, 2008 (table 5a).34 Across populations, a larger proportion of cardholders residing in lower-income areas were found to experience a line reduction than those in higher-income areas. The analysis also found that a larger proportion of minority cardholders or cardholders residing in census tracts with larger shares of minority population experienced a line reduction than non-Hispanic white cardholders or those in areas with smaller minority population shares. The same patterns are found when the analysis is conducted at the account level rather than the cardholder level (table 5b).
These associations between line reductions and population characteristics do not take into account differences across populations in account use, credit risk profiles, or payment performance on these or other credit accounts or other factors that may lead to a credit-line reduction. Differences across populations may exist along any of these dimensions. Also, although the initial sample of credit records and cardholders is nationally representative, the analysis sample may not be so because the criteria used to select the analysis sample may have excluded cardholders in a nonrandom way.
Most important for this report, the credit records do not include the reason(s) that a cardholder experienced a change in a credit line. Specifically, whether any of the practices identified in section 505 led to changes in credit lines is not known. Consequently, it is not possible to determine whether any relationship exists between cardholder demographics and line reductions due specifically to section 505 practices. For any relationship between cardholde demographics and section 505 practices to be detectable in the credit record data, one or more of the section 505 practices would have to be a principle driver of line reductions. However, the survey of card issuers conducted for this report found that a relatively small proportion of cardholders actually experience line reductions for reasons related to those identified in section 505, meaning that such practices are not principal drivers of line reductions.
Note: Table includes 89,265 individuals. See text for a definition of active accounts. 1. Information on race or ethnicity is from the Social Security Administration (SSA), or, when unavailable from the SSA, it is from the census-tract race (minorities as a percent of population). |
Note: Table includes 145,383 active bank credit card accounts. See text for a definition of active accounts. 1. See table 5a, note 1. Return to table |
Recommendations
Very few issuers of general-purpose bank credit cards currently consider transactionrelated or mortgage-lender-specific information of the type identified in section 505 to reduce credit limits or set interest rates on accounts. Discussions with card issuers and their federal supervisors found that other metrics used to manage credit risk, such as measures of late payment behavior or measures of cash advance activity, are much more important factors in reducing lines or setting interest rates than the practices identified in section 505. Moreover, significant changes in broad economic conditions can be important factors leading to changes in account terms or account closures.
To the extent that transaction-related information is currently used by issuers to reduce credit limits or raise interest rates, it tends to be general in nature and does not rely on the identity of the individual store or merchant, the types of items purchased in a transaction, or the price of any given item. Moreover, in conversations with issuers, it was reported that unfavorable public attention drawn by the use of some forms of this information to reduce credit lines deterred them from adopting such practices going forward. Also, noncompliance with fair lending laws creates risk that may deter issuers from using certain types of information, particularly some forms of geographic categorization.
If the Congress were to consider restrictions on card issuers' practices in this regard, it would be important to recognize the central role that detailed transaction-specific information plays in fraud detection and prevention systems. For example, the identity of the merchant, the merchant location, and the size and timing of charges are all central elements of fraud detection and prevention systems. Restricting the use of this type of information for purposes of fraud detection and prevention would likely substantially reduce the effectiveness of these systems and raise creditors' costs and, ultimately, the prices consumers pay for credit cards.
Although the current use of transaction-specific information of the type identified in section 505 for purposes of credit risk management appears quite limited, public policy actions prohibiting the use of these practices could still be costly over time. For example, prohibition would preclude potential innovations that would help card issuers manage credit risk more effectively. Such potential costs should be weighed against the possible benefits from restricting these practices.
Appendix A: Section 505 of the Credit Card Act of 2009
SEC. 505. REPORT TO CONGRESS ON REDUCTIONS OF CONSUMER CREDIT CARD LIMITS BASED ON CERTAIN INFORMATION AS TO EXPERIENCE OR TRANSACTIONS OF THE CONSUMER
- (a) REPORT ON CREDITOR PRACTICES REQUIRED--Before the end of the 1-year period beginning on the date of enactment of this Act, the Board, in consultation with the Comptroller of the Currency, the Director of the Office of Thrift Supervision, the Federal Deposit Insurance Corporation, the National Credit Union Administration Board, and the Federal Trade Commission, shall submit a report to the Committee on Financial Services of the House of Representatives and the Committee on Banking, Housing, and Urban Affairs of the Senate on the extent to which, during the 3-year period ending on such date of enactment, creditors have reduced credit limits or raised interest rates applicable to credit card accounts under open end consumer credit plans based on--
- (1) the geographic location where a credit transaction with the consumer took place, or the identity of the merchant involved in the transaction;
- (2) the credit transactions of the consumer, including the type of credit transaction, the type of items purchased in such transaction, the price of items purchased in such transaction, any change in the type or price of items purchased in such transactions, and other data pertaining to the use of such credit card account by the consumer; and
- (3) the identity of the mortgage creditor which extended or holds the mortgage loan
- (b) OTHER INFORMATION--The report required under subsection (a) shall also include--
- (1) the number of creditors that have engaged in the practices described in subsection (a);
- (2) the extent to which the practices described in subsection (a) have an adverse impact on minority or low-income consumers;
- (3) any other relevant information regarding such practices; and
- (4) recommendations to the Congress on any regulatory or statutory changes that may be needed to restrict or prevent such practices. secured by the primary residence of the consumer.
Appendix B: The Survey of Credit Card Issuer Practices
PURPOSE, FOCUS AND CONTACT INFORMATION FOR THE SURVEY
Purpose of the Survey
Section 505 of the Credit Card Accountability Responsibility and Disclosure Act of 2009 directs the Federal Reserve Board to conduct a study of the extent to which issuers of open-end credit cards to consumers reduce credit lines or raise interest rates based on certain information, such as the type or location of the merchant the consumer transacts with, or the type or price of items purchased using their cards. The study calls for a review of practices that may have been in place over the past 3 years.
Part of the study will be based on the responses to the attached survey which is being distributed to the largest 25 credit card issuers and a nationally representative sample of an additional 150 issuers. Completion of the survey is mandatory. A copy of the Section 505 requirements is part of the survey documents. We ask that you review the attached Section 505 language prior to completing the survey to help gain a better understanding of the specific practices that are the focus of the study.
Focus of the Survey
The survey seeks information on certain specific practices that issuers of open-end credit card plans may consider when deciding to (1) raise interest rates, (2) reduce credit limits or (3) close credit card accounts. The survey asks specifically whether an issuer has used each specific practice at any time in the past three years and for those that respond in the affirmative whether the issuer has used the specific practice in the month before the survey. For those that respond in the affirmative, follow-up questions ask about (1) the number of cardholders with active accounts that were potentially affected by the practices; and, (2) the number of cardholders with active accounts that actually experienced an increase in their interest rate, a reduction in their credit line or the closure of an account due to each of the specific practices. In each instance, calculations should be based on the number of cardholders with active accounts and not dollars owed on those accounts.
Information about cardholders potentially affected should be based on a review of consumers with active credit card accounts that were subject to either an automated or judgmental review that included the practices identified in the survey. Information about cardholders actually affected should be based on a review of the reasons provided to cardholders that received an adverse action notice both for those receiving adverse action notices in the previous month and separately for those receiving adverse action notices during the 25 month period preceding receipt of the survey. The 25 month period corresponds to the time adverse action notices must be retained under the provisions of the Equal Credit Opportunity Act. Detailed instructions including definitions of terms appear below.
Coverage
The survey asks only about open-end credit cards issued to consumers. The survey does not include lines of credit. It also does not include credit cards issued to businesses or professionals. Consumer credit card accounts include: "general purpose" cards (such as, MasterCard, Visa, Discover or AMEX), "private label" cards and "co-branded" cards.
Definition of "Active" Accounts
Active accounts are those that the consumer used to make purchases or take cash advances at any time during the twelve months preceding the receipt of the survey. If the only activity over the 12 month period was for a cash advance arising from a draw on the credit card line to cover an overdraft on a linked checking account, treat that account as inactive.
Time Frame for Responses
The Congress has established a short time frame for the study. The report to the Congress is due May 22, 2010. As a result, we must receive your response to the survey by February 1, 2010.
For your convenience, we have enclosed a self-addressed envelope you may use to mail in your responses.
THE SURVEY OF CREDIT CARD ISSUER PRACTICES:
SURVEY LAYOUT AND DETAILED INSTRUCTIONS
This survey is divided into 4 sections. All respondents need to complete sections 1, 2, 3 and 4. Section 1 collects contact information and a few general statistics about your institution's consumer credit card business. Section 2 asks about certain specific practices that may have led to reductions in credit limits on accounts. Section 3 asks about specific practices that may have resulted in the closure of an account. Section 4 asks about certain specific practices that may have led to increases in interest rates on accounts. Section 5 collects some additional broad statistics only for institutions that respond in the affirmative to at least one of the questions posed in sections 2, 3, or 4 of the survey.
A copy of the Section 505 statutory language is included at the end of the survey. Please review the Section 505 language to help ensure that your responses to the survey cover all the relevant practices that your institution may consider when deciding whether to reduce the credit limit or raise the interest rate on an account or to close the account. Please keep in mind that the survey is focused only on transaction-related practices or those that involve the identity of the mortgage lender.
The detailed instructions that follow provide definitions and guidance to help you complete the survey.
If you are unsure whether the practices asked about in the survey include those that your institution has used please contact:
Glenn Canner at the Federal Reserve Board by calling (202) 452-2910.
Detailed Instructions for Completing the Survey
This section provides instructions and definitions to help complete the survey.
Instructions: The survey focuses on practices regarding reductions in credit lines, account closures and increases in interest rates on active open-end credit card accounts. The survey asks for the number of cardholders potentially affected and actually affected by a given practice. Cardholders potentially affected were those that might have had the credit line reduced, account closed or the interest rate increased on their credit card due to a given practice. For example, if your institution uses an automated system to determine credit line changes for some cardholders and the given practice was an element of the automated system, then all cardholders subject to that automated system could have potentially been affected. If your institution uses a judgmental system for other customers and the practice was a factor considered for only half of the customers subject to the judgmental system then only half of the cardholders subject to the judgmental reviews could potentially have been affected.
In answering questions about cardholders actually affected, consider only the subset of cardholders who were potentially affected who actually experienced a reduction in their credit limit account closure or an increase in their interest rate due to each of the specific practices identified in the survey. To determine who among the account holders potentially affected were actually affected, consider only those customers who received an adverse action notice as required by the Equal Credit Opportunity Act (ECOA) and one of the (up to as many as) four reasons sighted in the adverse action notice related to the specific practice as the reason for the creditor's action.
If your institution sent cardholders a general notice informing them of a line reduction, account closure or interest rate increase and the notice indicated they should contact the issuer for the specific reasons, the notices, and the reasons that would have been provided to the cardholder if they had contacted your institution should be considered in determining whether a cardholder was affected by each of the specific practices asked about in the survey. Once again, the only reasons that need to be considered are those that relate to the specific practices asked about in the survey questions.
Note: For this survey, questions about cardholders seek information only about one cardholder per account. Thus, if an account involves the issuance of cards to more than one individual (for example, a husband and wife) treat it as a single cardholder.
Note: Only cardholders that were entitled to receive an adverse action notice under the ECOA should be considered in determining whether a cardholder was actually affected by the practice.
Note: Section 505 asks about the use of each of the specific practices during the past 3 years. In answering the "yes or no" questions with regard to whether your institution used the practice or not, consider the 3 year period prior to the receipt of the survey. This time frame covers the period November 30, 2006-November 30, 2009. If you respond affirmatively regarding the 3 year period than a follow-up question asks about the use of the practice in the current month which for this survey is defined as November 2009.
Note: If you respond affirmatively to a question about the 3 year period follow-up questions ask about cardholders potentially and actually affected over the 25 month period November 1, 2007-November 30, 2009. The shorter 25 month period is used instead of the full 3 years because under the provisions of the ECOA adverse action notices are required to be retained only for 25 months.
Definitions: To ensure institutions respond in a consistent manner the following definitions are provided.
- For questions related to the Geographic Location refers to the location of the store or merchant where a transaction took place. Location includes any of the following: 1. specific street address, 2. zip code, 3. census tract, 4. county, 5. state, or 6. any grouping of smaller geographies, such as census tracts grouped by income or racial or ethnic composition.
- For questions related to Type of Credit Transaction--Type of Credit Transaction refers to the broad categories card issuers sometimes use when grouping a customer's charges, sometimes on an end-of year statement or posted to a web site. For example, charges made to buy gas or automobile parts or repairs might be grouped under the broad category "Oil and Gas;" purchases at restaurants and movie theaters might be grouped as "Entertainment."
Note: Type of Credit Transaction does not correspond to whether a customer took a cash advance versus making a purchase with their card. - For questions related to The Type of Items Purchased--The type of items purchased in the context of the survey refers to the different types of items a cardholder might have purchased, such as engine oil versus automobile tires at a auto supply store, or at a restaurant, beverages versus food, or at retailer shoes versus clothing. The type of items purchased also includes distinctions within a category such as generic automobile oil product versus a premium brand automobile oil.
- For question related to Any Other Data Pertaining to the Use of Credit Card--Any Other Data in the context of this survey corresponds only to transaction specific information that a card issuer might consider that is not covered in the prior questions. For example, consideration of the day of the week of a purchase or the number of purchases over the month would meet this definition.
SECTION 1: GENERAL INSTITUTION INFORMATION AND SECTION 505 STATUTORY LANGUAGE
Instructions
Please provide the following general contact and other information about your institution as requested below. Please type or print clearly.
Institution Name:
Contact Information
Name of person responsible for the survey:
Telephone Number:
Fax Number:
E-mail address:
1. Type of institution: (check one)
- _____ Commercial bank
- _____ Savings institution (savings bank or savings and loan association)
- _____ Credit Union
2. Credit Card Data (exclude business and professional credit cards):
- a. Number of credit card accounts issued to consumers as of November 30, 2009:
- b. Number of active credit card accounts issued to consumers as of November 30, 2009
SECTION 2: QUESTIONS ABOUT REDUCTIONS IN CREDIT LIMITS
At anytime in the previous three years, did your company consider the following information when reducing a consumer's credit limit on an active credit card account:
1) Did you consider the geographic location where a credit transaction took place (see Definitions Section)?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
2) Did you consider the identity (name) of the merchant involved in a credit transaction?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
3) Did you consider the type of credit transaction (see Definitions Section)?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
4) Did you consider the type of items purchased in a credit transaction?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
5) Did you consider the price of items purchased in a credit transaction?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
6) Did you consider any change in the type of items purchased in the consumer's credit transaction history?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
7) Did you consider any change in the price of items purchased in the consumer's credit transaction history?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
8) Did you consider the identity of a mortgage creditor which extended a mortgage loan secured by the primary residence of the consumer?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
9) Did you consider the identity of a mortgage creditor which holds a mortgage loan secured by the primary residence of the consumer?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
10) Did you consider any other data pertaining to the use of such credit card account (see Definitions Section)?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
SECTION 3: QUESTIONS ABOUT ACCOUNT CLOSURES
At anytime in the previous three years, did your company consider the following information when closing a consumer's active credit card account:
11) Did you consider the geographic location where a credit transaction took place (see Definitions Section)?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
12) Did you consider the identity (name) of the merchant involved in a credit transaction?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
13) Did you consider the type of credit transaction (see Definitions Section)?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
14) Did you consider the type of items purchased in a credit transaction?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
15) Did you consider the price of items purchased in a credit transaction?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
16) Did you consider any change in the type of items purchased in the consumer's credit transaction history?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
17) Did you consider any change in the price of items purchased in the consumer's credit transaction history?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
18) Did you consider the identity of a mortgage creditor which extended a mortgage loan secured by the primary residence of the consumer?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
19) Did you consider the identity of a mortgage creditor which holds a mortgage loan secured by the primary residence of the consumer?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
20) Did you consider any other data pertaining to the use of such credit card account (see Definitions Section)?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
SECTION 4: QUESTIONS ABOUT INCREASES IN INTEREST RATES
At anytime in the last month or anytime in the previous three years, did your company consider the following information when increasing a consumer's interest rate on a credit card account:
21) Did you consider the geographic location where a credit transaction took place (see Definitions Section)?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
22) Did you consider the identity (name) of the merchant involved in a credit transaction?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
23) Did you consider the type of credit transaction (see Definitions Section)?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
24) Did you consider the type of items purchased in a credit transaction?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
25) Did you consider the price of items purchased in a credit transaction?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
26) Did you consider any change in the type of items purchased in the consumer's credit transaction history?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
27) Did you consider any change in the price of items purchased in the consumer's credit transaction history?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
28) Did you consider the identity of a mortgage creditor which extended a mortgage loan secured by the primary residence of the consumer?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
29) Did you consider the identity of a mortgage creditor which holds a mortgage loan secured by the primary residence of the consumer?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
30) Did you consider any other data pertaining to the use of such credit card account (see Definitions Section)?
- a) At any time in the Last 3 Years? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last 25 months?
- ______ ii. What is the number of cardholders actually affected in the last 25 months?
- If no, move to the next question. If yes, then:
- b) At any time in the Last Month? Yes No
- If no, move to the next question. If yes, then:
- ______ i. What is the number of cardholders potentially affected in the last month?
- ______ ii. What is the number of cardholders actually affected in the last month?
- If no, move to the next question. If yes, then:
SECTION 5: FOLLOW-UP QUESTIONS
Note: ONLY ANSWER questions in this section if you responded affirmatively to any of the questions in Sections 2, 3 or 4 of this survey regarding specific practices considered in deciding whether to reduce credit limits, close an account or increase interest rates on active credit card accounts.
Note: In Sections 2, 3, and 4 of this survey questions were asked about the number of individuals affected by each specific practice. Since a single cardholder might have had their credit line reduced, their account closed or their interest rate increased as the result of more than one of these practices the first three questions in this section ask about the number of different cardholders that were affected by the application of these practices. As before, for these calculations exclude inactive accounts.
1. Number of different cardholders that experienced a reduction in their credit line because of any of the practices identified in Section 2 of this questionnaire:
- a. During the 25 month period November 1, 2007 through November 30, 2009?______
- b. During the month of November 2009?______
2. Number of different cardholders that experienced an account closure because of any of the practices identified in Section 3 of this questionnaire:
- a. During the 25 month period November 1, 2007 through November 30, 2009?______
- b. During the month of November 2009?______
3. Number of different cardholders that experienced an increase in their interest rate because of any of the practices identified in Section 4 of this questionnaire:
- a. During the 25 month period November 1, 2007 through November 30, 2009?______
- b. During the month of November 2009?______
4. Number of credit card accounts issued to consumers that experienced a reduction in their credit limit for any reason:
- a. During the 25 month period November 1, 2007 through November 30, 2009?______
- b. During the month of November 2009?______
5. Number of credit card accounts issued to consumers that experienced an account closure for any reason:
- a. During the 25 month period November 1, 2007 through November 30, 2009?______
- b. During the month of November 2009?______
6. Number of credit cards issued to consumers that experienced an increase in their interest rate for any reason except: (1) changes in interest rates due to a change in an index for variable-rate accounts, (2) due to a switch of consumers from a variable-rate plan to a fixed-rate plan or from a fixed-rate plan to a variable-rate plan, (3) the end of a promotional rate period or (4) an across the board or portfolio adjustment in interest rates:
- a. During the 25 month period November 1, 2007 through November 30, 2009 (if an account received multiple increases during the period count it only once)?______
- b. During the month of November 2009?______
7. Number of credit card accounts for which an adverse action notice was provided for any reason regarding a reduction in the credit limit on an account:
- a. During the 25 month period November 1, 2007 through November 30, 2009 (if an account received multiple notices during the period count it only once):______
- b. During the month of November 2009:______
8. Number of credit card accounts for which an adverse action notice was provided for any reason regarding the closure of an account:
- a. During the 25 month period November 1, 2007 through November 30, 2009:______
- b. During the month of November 2009:______
9. Number of credit card accounts for which an adverse action notice was provided for any reason regarding an increase in the interest rate on an account:
- a. During the 25 month period November 1, 2007 through November 30, 2009 (if an account received multiple notices during the period count it only once):______
- b. During the month of November 2009:______
Appendix C: The Credit Record Data Used in This Report
This appendix provides information about the credit record sample used in this report, including how the credit card accounts of interest were identified and how cardholder income and race or ethnicity were established.
CONTENTS OF CREDIT RECORDS
Credit records generally include five types of information: (1) identifying personal information, including the name of the individual, current and previous residential addresses, and Social Security number; (2) detailed information reported by creditors and others on each current and past loan, lease, or non-credit-related bill; (3) information derived from monetary-related public records, such as records of bankruptcy, foreclosure, tax liens, garnishments, and other civil judgments; (4) information reported by collection agencies on actions associated with credit accounts and non-credit-related bills; and (5) the identities of individuals or companies that request information from an individual's credit record, the date of the inquiry, and an indication of whether the inquiry was by the consumer, for the review of an existing account, or to help the inquirer make a credit decision.
RECORDS OF CREDIT CARD ACCOUNTS
Overall, nearly 84 percent of the individuals in the credit record sample used for this report had information reported about at least one credit account (table 1 in the main text). Among the various types of credit accounts included in credit records, the category with the largest number of accounts is credit cards. Credit cards accounted for 42 percent of all open revolving or nonrevolving credit accounts as of June 30, 2007, and for nearly 56 percent of the revolving accounts (data derived from table C.1).
IDENTIFYING THE CARDS OF INTEREST
Credit card accounts can be grouped in different ways, including by the type of entity issuing the card. Three broad categories are (1) "bank" credit cards (bankcards), issued by banking institutions, including credit unions; (2) "store" cards, issued by retailers and merchants; and (3) "other" cards, issued primarily by travel and entertainment companies, finance companies, and national oil and gas companies.35 This report focuses on bankcards, which include general-purpose credit cards such as MasterCard, Visa, Discover, and American Express. Bankcards accounted for 91 percent of all the credit card accounts included in credit record files as of June 30, 2007.
| Type | Number | Percent |
|---|---|---|
| All revolving | ||
| Credit cards | 535,507 | 56.1 |
| Other1 | 418,574 | 43.9 |
| Total | 954,081 | 100.0 |
| All nonrevolving | ||
| Installment | 178,559 | 56.3 |
| Mortgage | 100,280 | 31.6 |
| Other2 | 38,073 | 12.0 |
| Total | 316,912 | 100.0 |
| MEMO: | ||
| Type of credit card | ||
| Bank | 487,370 | 91.0 |
| Retailer | 42,080 | 7.9 |
| Other3 | 6,057 | 1.1 |
| Total | 535,507 | 100.0 |
Note: Components may not sum to totals because of rounding. Open accounts include open credit accounts that have been verified in the three months preceding June 2007 and have payment history reported in the past year.
1. Includes, for example, a charge account or personal line of credit. Return to table
2. Includes, for example, utility company account records. Return to table
3. Includes, for example, credit cards issued by a finance company. Return to table
Among bankcards, this report also focuses on "open" and "actively reported" credit card accounts (hereafter referred to as "open accounts") that the credit record data indicate were actively being used by cardholders. In general, these credit accounts include credit cards that were being reported on at the time the survey sample was pulled; were reported on during the period from June 30, 2007, through December 31, 2008 (the "observation period"); and were not reported as "closed" with a zero balance during the observation period.36 Open credit card accounts include accounts that the consumer could use for making purchases or taking cash advances over the entire observation period. Some of these accounts may have been used by the cardholders during the observation period, and some of them may not have.
Among the open credit card accounts, the focus of the analysis is on those that were determined to be active. This restriction is important for two reasons. First, the survey of card issuers undertaken for this report focused on the actions card issuers took on active accounts. Second, discussions with issuers indicated that inactive accounts were subject to line reductions due simply to their inactive status.
It is difficult to determine with certainty from the credit record data whether an account is actively being used by a cardholder. The primary difficulty is that the credit record data available for this report does not show balances owed other than at the end points of the observation period. In order to ensure that the accounts used in the analysis were active, the sample was restricted to only cards that showed a positive balance at the end points of the observation period. A card that was used over part of the observation period but showed zero balances at the end points was categorized as inactive even though the cardholder was using the card. Unfortunately, we cannot distinguish such cards from those that were inactive during the entire observation period. Hence, the definition used here to identify active accounts is imperfect and restrictive.
CREDIT LIMITS ON CREDIT CARD ACCOUNTS
Importantly for this report, the credit record data include the credit limits for open-end credit accounts, including credit card accounts. The credit limit represents the largest amount the consumer may borrow under the terms of the account. However, credit card issuers may allow a cardholder to borrow more than the stated limit, usually assessing an "over-the-limit" fee. Also, in some cases, the credit limit reported to the credit-reporting agencies is not up-to-date; in these cases, the amount owed on the card as it appears in the credit records may exceed the last reported limit.
On occasion, credit limits are not reported to the credit-reporting agencies. Nonreporting of credit limits for credit cards was a significant issue in the past but poses much less of a problem now, as most issuers currently report this information to the credit-reporting agencies. When a credit limit is not reported, industry participants and those building credit-scoring models customarily substitute the highest balance ever owed on the account. However, for the analysis conducted for this report, it is critical to know with a high degree of certainty the actual credit limits. Thus, accounts that did not have limits reported at both the beginning and the end of the observation period are excluded. In the Federal Reserve sample of credit records used here, credit limits were missing either in June 2007 or December 2008 for about 10 percent of the bank credit cards (data not shown in tables).37
INTEREST RATES ON CREDIT CARD ACCOUNTS
Unlike credit limit information, credit records do not include the interest rate(s) applicable to credit cards or other credit accounts. For closed-end fixed-rate loans, credit record data can often be used to estimate the interest rate on an account using the monthly payment and term information included in the credit record. Such a procedure is not possible for open-end accounts. Thus, the credit record data can provide only information pertinent to an assessment of the relationship between reductions in credit lines and the race or ethnicity or the income of cardholders.
LOCATIONAL INFORMATION FROM THE CENSUS 2000 DATA
At the request of the Federal Reserve, TransUnion "geocoded," to the extent possible, the June 30, 2007, home address of each individual in the credit record sample to help identify the year 2000 census-block group or census-tract code of the person's residence. These geographic codes were then used to merge Census 2000 data on the income status and racial and ethnic composition status of census-tract residents with the credit record data.38
RACE OR ETHNICITY OF INDIVIDUALS WITH CREDIT RECORDS
The only personal demographic information included in an individual's credit record is the individual's date of birth, and such information is often not available (for example, date of birth was not present in about one-third of the credit records in the Federal Reserve sample). However, credit records contain additional types of personal information--name, Social Security number, and current and previous addresses--that can be used to obtain further demographic information on the individual from other data sources.
For purposes of preparing its August 2007 report to the Congress on credit scoring, the Federal Reserve asked TransUnion to provide identifying personal information to the Social Security Administration (SSA) for each of the individuals included in the June 30, 2003, Federal Reserve credit record sample so that the SSA could provide the Federal Reserve with demographic data about these individuals.39 The SSA gathers demographic information about individuals when they apply for a Social Security card.40
The procedures followed for preparing the 2007 report to the Congress ensured that the SSA received no information included in the credit records of the individuals other than the identifying personal information needed to match the administrative records maintained by the SSA. The Federal Reserve received from the SSA a data file that included the demographic characteristics of the individuals in the sample but no identifying personal information. TransUnion received no information from the SSA or the Federal Reserve on the demographic characteristics of the individuals in the sample.
With the names and Social Security numbers provided by TransUnion, the SSA extracted and provided to the Federal Reserve the following information for each matched individual to the extent available: citizenship status, the date the individual filed for a Social Security card, place of birth, state or country of birth, race or ethnic description, sex, and date of birth. All of the aforementioned information except the race or ethnicity of the applicant is required on the application form for a Social Security card; race or ethnicity is requested on the form, but the applicant is not required to supply it.
Two aspects of the SSA administrative records bear importantly on the analysis in this study. First, some individuals failed to provide some demographic characteristics when completing their applications for a Social Security card. Also, over the years, some individuals applied more than once for a Social Security card and so had more than one opportunity to report their demographic characteristics. The SSA provided the Federal Reserve the information reported by these individuals on each of their applications; however, in some cases, the information was inconsistent.41 For example, some individuals reported different responses for date of birth, sex, or country of origin on their various applications.
Second, the SSA in 1981 changed the options offered to individuals for reporting their racial or ethnic status. For years preceding 1981, individuals had three choices, from which they were asked to select one--(1) "White," (2) "Black," or (3) "Other." Beginning in 1981, individuals have had five options, from which they choose only one--(1) "Asian, Asian American, or Pacific Islander"; (2) "Hispanic"; (3) "Black (Not Hispanic)"; (4) "North American Indian or Alaskan Native"; or (5) "White (Not Hispanic)." To prepare the earlier report to the Congress, the Federal Reserve staff took a number of steps to resolve inconsistencies in the race or ethnicity data provided by the SSA, and these adjusted data were used to prepare this report. At the end of the process, each individual was assigned a probability of being a non-Hispanic white.42
For this report, race or ethnicity information was available only for the individuals in the June 30, 2003, sample of credit records. Individuals who have been added as replacements over time to the Federal Reserve's ongoing panel of credit record data do not have this information. Individuals who have been added as replacements are not random but rather are more likely to be younger or foreign born.
To prepare this report, the race or ethnicity information from SSA records was used, to the extent available, for the individuals in the ongoing panel that were first included in the June 30, 2003, sample. For individuals added to the ongoing panel since that time, the racial or ethnic composition of the census-tract place of residence was used to impute (estimate) the race or ethnicity. Specifically, if an individual lived in a census tract that had x percent minority residents according to the Census 2000, then the individual was assigned an x percent probability of being a minority and a (1 - x) percent probability of being a non-Hispanic white. For this exercise, a minority was anyone who was not non-Hispanic white.
Footnotes
(1) The request for this report asked about the use of the practices identified in section 505 during the past three years. In keeping with the request, the survey asked each issuer whether it had used the practices over the period from November 30, 2006, through November 30, 2009. However, in determining which cardholders were affected by the practices, respondents were asked to focus on activities over the 25-month period from November 1, 2007, through November 30, 2009. Although the Congress requested information about practices over the past three years, creditors are required to maintain adverse action notices for only 25 months. The reasons provided in adverse action notices constitute the method used in this report to determine why a credit line was reduced or the interest rate increased on an account.
(2) As discussed later in the article, participation in the survey was mandatory, and responses were received from all 175 institutions that received the questionnaire. The survey excluded banking institutions with credit card outstandings of less than $20 million. The reason for this exclusion and more details about the survey are provided later.
(3) A review of the credit records used in this report indicate that less than 10 percent of the cardholders have more than one active credit card account from the same issuer.
(4) In general, Regulation B prohibits the collection of such information for any type of credit except home mortgage loans. See Board of Governors of the Federal Reserve System, "B (12 CFR 202): Equal Credit Opportunity," information on the Equal Credit Opportunity Act (Regulation B), webpage, www.federalreserve.gov/bankinforeg/reglisting.htm#B.
(5) Historically, individuals often have been asked to submit income information when applying for a credit card, but issuers generally have not validated such information, nor have they asked cardholders to provide updated information except sometimes when a request has been made by the consumer to increase the credit line. As a result of the Credit CARD Act of 2009, a card issuer cannot open an account or increase the credit limit on an existing account before considering the ability of a consumer to make the required minimum periodic payments based on the consumer's income or assets and current obligations. These new rules may lead issuers to gather and retain more systematic and verifiable information about income. See Board of Governors of the Federal Reserve System (2010), "Subpart G--Special Rules Applicable to Credit Card Accounts and Open-End Credit Offered to College Students," section 226.51, in "Truth in Lending," final rule (Docket No. R-1370), Federal Register, vol. 75 (February 22), pp. 7818-19, available at http://edocket.access.gpo.gov/2010/pdf/2010-624.pdf.
(6) Because credit records do not include information on the interest rates applicable to credit card accounts, such records do not afford an opportunity to address the effects of the section 505 practices on increases in interest rates.
(7) TransUnion is one of the three national credit-reporting agencies. The credit-reporting agencies gather detailed information on consumer use of credit of all types for most adults in the United States. The three national creditreporting agencies are Equifax, www.equifax.com; Experian, www.experian.com; and TransUnion LLC, www.transunion.com. Identifying personal information, such as the names and addresses of individuals and their Social Security numbers, was not made available to the Federal Reserve. Also, the names of creditors or other entities that report information were not included in the credit records received by the Federal Reserve.
(8) To prepare a previous report for the Congress, the Federal Reserve obtained from the Social Security Administration information from its administrative records on the race or ethnicity of individuals in the sample of credit records received by the Federal Reserve (details about this process and the steps taken to ensure the privacy of individuals are provided later). See Board of Governors of the Federal Reserve System (2007), Report to the Congress on Credit Scoring and Its Effects on the Availability and Affordability of Credit (Washington: Board of Governors, August), www.federalreserve.gov/boarddocs/rptcongress/creditscore/default.htm.
(9) Using credit records to assess changes in credit limits is a complex process. During the observation period considered here, not all issuers reported credit limits, and those that reported them at the end of the observation period may not have done so at the beginning of the period. We considered only accounts with reported limits at both ends of the observation period. Issuers that did not report limits may not be similar to other issuers, and their customers may not be similar to the cardholders of issuers that report limits.
(10) See, for example, Charles Duhigg (2009), "What Does Your Credit-Card Company Know About You?" New York Times, May 17; and Ron Lieber (2009), "American Express Kept a (Very) Watchful Eye on Charges," New York Times, January 31. Also see http://frwebgate.access.gpo.gov/cgi-bin/getdoc.cgi?dbname=111_house_hearings&docid=f:48675.pdf.
(11) See Congresswoman Maxine Waters's remarks (2009), in TARP Accountability: Use of Federal Assistance by the First TARP Recipients, hearing before the House Committee on Financial Services, U.S. House of Representatives, February 11, House Hearing 111-4, 111 Cong. (Washington: Government Printing Office), p. 29, available at www.house.gov/apps/list/hearing/financialsvcs_dem/hr021109.shtml. In addition, the Federal Trade Commission (FTC) took action against CompuCredit Corporation in 2008 for a variety of practices deemed to be deceptive, including those related to the company's marketing of credit cards whereby disclosures made to consumers failed to provide adequate information about certain practices concerning cardholders' purchases that could lead to a reduction in their credit limits. Specifically, the FTC found that CompuCredit failed to inform new cardholders that for the first 90 days after card issuance the company would monitor purchase activity and might reduce their credit limits based on a behavioral scoring model. More details are in Federal Trade Commission (2008), "FTC Sues Subprime Credit Card Marketing Company and Debt Collector for Deceptive Credit Card Marketing," news release, June 10, www.ftc.gov/opa/2008/06/compucredit.shtm.
(12) For purposes of this report, a minority individual is any individual who is not non-Hispanic white. Available data do not allow us to reliably categorize cardholders by their actual income. For this report, cardholders are grouped by income using information about the income characteristics of the census tract where they reside (discussed in more detail later).
(13) The report covers co-branded credit cards, which are general-purpose credit cards issued by a banking institution that are designed with a retailer's or merchant's company brand prominently displayed on the card. According to one estimate, co-branded credit cards account for 55 percent of all card spending. See MasterCard Worldwide, "Co-Branded Cards," webpage, www.mastercard.com/us/merchant/solutions/co-branded_cards.html (accessed February 17, 2010). Credit cards that are excluded from the report include those that are issued directly by a retailer or merchant and that can be used only by the cardholder at the specific retailer or merchant. In some cases, store or merchant cards are issued through a finance company rather than a banking institution. Cards issued by finance companies are also excluded from the report.
(14) In recognition of the potential burden on smaller institutions, the survey excluded banking institutions and credit unions with credit card assets of less than $20 million. None of these smaller institutions are significant issuers of credit cards. Among the issuers with less than $100 million in credit card assets, the survey included 37 of 196 with assets between $20 million and less than $50 million, 23 of 62 with assets between $50 million and less than $75 million, and 15 of 30 with assets between $75 million and less than $100 million.
(15) The websites of each of the institutions initially selected for participation in the survey were reviewed, or telephone contact was made, to help ensure the institutions chosen to participate in the survey were currently offering credit cards to the public rather than only having credit card outstandings on their books. If a banking institution had credit card outstandings on its books but did not issue credit cards to consumers, another issuer in the same size category was randomly chosen as a substitute. The September 2009 Call Report for credit unions was not available at the time the survey sample was drawn.
(16) Credit card accounts are potentially affected by the policies if a factor specified in section 505 could have played a role in a line reduction or an increase in interest rate on the account. Thus, if a policy was implemented as a fact or in an automated model applied to cardholders generally, the potential number or share of cardholders affected could be large; if the policy was limited to a subset of customers by some other screening criteria, then the potential number or share of customers could be small.
For this exercise, institutions were asked to consider only "active" accounts. Active accounts are accounts that the consumer used to make purchases or take a cash advance at any time during the 12 months preceding the receipt of the survey. If the only activity was a cash advance arising from a draw on the credit card line to cover an overdraft on a linked checking account, the institutions were told to treat that account as inactive for purposes of the survey.
(17) Regulation B does not require that adverse action notices be sent in all circumstances. For actions taken in connection with inactivity, default, or delinquency as to an account, no adverse action notice is required.
(18) Other factors could cause a card issuer to not respond to a survey. For example, the burden of completing the survey may be greater for some institutions than for others.
(19) Credit-reporting agency files include identifying personal information that allows the credit-reporting agencies to distinguish among individuals and construct a full record of each individual's credit-related activities. Files include the individual's name, current and previous addresses, and Social Security number. Other information sometimes found in credit files includes date of birth, telephone numbers, name of spouse, and number of dependents. Except for date of birth, such information was removed from the sample for this study.
(20) Estimate derived from tables shown in The Nilson Report; for more information, see HSN Consultants Inc. (2009), The Nilson Report, "50 Largest Visa & MasterCard Credit Card Issuers," issue 918 (January), p. 10; and "U.S. General Purpose Credit Cards," issue 924 (April), p. 10; both issues are available at www.nilsonreport.com/recentissues.htm.
(21) For example, Experian and Integrated Solutions Concepts, Inc. (a wholly owned subsidiary of Visa U.S.A. Inc.), offer BankruptcyPredict, an account management-tool that combines consumer credit characteristics drawn from credit records and Visa transaction spending characteristics to predict future cardholder bankruptcy. See Experian Information Solutions, Inc. (2010), "BankruptcyPredict," webpage, www.experian.com/products/bankruptcypredict.html.
(22) For example, FICO offers FICO Transaction Scores, which are developed using transaction-authorization data. See Fair Isaac Corporation (2010), "FICO Transaction Scores," webpage, www.fico.com/en/Products/DMApps/Pages/FICO-Transaction-Scores.aspx.
(23) Every transaction processed by Visa or MasterCard is assigned a merchant category code. The codes cover relatively broad groups of merchants such as jewelry stores, landscaping services, home supply warehouse stores, shoe stores, and restaurants. Merchant category codes are used to classify a business by the types of products or services offered. There are about 600 four-digit codes that group various types of businesses. See Visa U.S.A. Inc. (2004), "Visa Commercial Solutions: Merchant Category Codes for IRS Form 1099-MISC Reporting," http://web.archive.org/web/20070710202209/http://usa.visa.com/download/corporate/resources/mcc_booklet.pdf.
(24) As opposed to bank credit card issuers, retailers that issue cards may have access to detailed transaction-specific information, such as the type of merchandise purchased.
(25) For example, an account holder may have a history of sending in checks to the creditor that is not backed by sufficient funds in their account.
(26) There are many automated fraud detection and management systems available in the marketplace. Some issuers may also develop their own internal systems. For an example of a third party system see information about the Falcon Fraud Manager system distributed by Fair Isaac, inc; www.fico.com/en/Products/DMApps/Pages/FICO-Falcon-Fraud-Manager.aspx.
(27) After receiving the responses to the survey, the Federal Reserve staff learned that five of the commercial banks that reported credit card outstandings on their Call Reports did not issue credit cards to consumers directly but rather through an arrangement with another commercial bank. In four of these cases, the credit card activities related to these institutions were reported by the other commercial banks in the survey. In one case, involving an issuer with a small number of customers, the credit card activities were not reported because the associated company was not included in the survey.
(28) Each of the issuers that reported considering transaction velocity in their survey responses reported that a very small proportion of their account holders experienced a line reduction for this reason.
(29) It is important to note that cardholders who made charges at these types of merchants without the cash advance activity were not subject to line reductions.
(30) The issuers that reported taking such steps all indicated that a very small number of cardholders were affected by this credit risk management practice.
(31) For more information about Merchant Category Codes, see Internal Revenue Service (2004), "Merchant Category Codes to Determine Reportable Payment Card Transactions," Internal Revenue Bulletin 2004-31 (Washington: IRS, August 2), available at www.irs.gov/irb/2004-31_IRB/ar17.html; for more information about Standard Industrial Classification (SIC) codes, see the Occupational Safety and Health Administration's webpage permitting SICrelated searches, available at www.osha.gov/pls/imis/sicsearch.html.
(32) Telephone interviews with representatives of the other supervisory agencies took place in March 2010.
(33) For calculating changes in credit limits, it is assumed that the last reported limit is the limit that was available to the cardholder on the account as of December 31, 2008. If there was no update to the reported limit between June 30, 2007, and December 31, 2008, it is assumed that there was no change in the limit on the account.
(34) The analysis focuses on general-purpose bank credit card accounts that credit record data indicate were actively being used by cardholders during the period from June 30, 2007, through December 31, 2008. The analysis considers only cards that the individual had at both the beginning and end of this 18-month period. A more complete description of the accounts considered in the analysis is provided in appendix C.
(35) Other types of categorizations are possible but are not relevant for this report. In today's marketplace, most store cards are underwritten and issued by banking institutions.
(36) For some accounts, it was not possible to determine with certainty whether the account was open or closed; these accounts were excluded from the analysis conducted for this report.
(37) As of the end of December 2008, credit limits were missing for only about 2 percent of the bank-issued credit cards used in the analysis.
(38) In some cases, complete street address information was not part of the credit record maintained by TransUnion, and so geocoding to the census block or census tract was not feasible without some assumptions. For individuals who had a state, county, and Zip code reported but no census tract, the Federal Reserve staff geocoded the census tract that corresponded to the center point of the Zip code.
(39) More information about the procedures is provided in Board of Governors of the Federal Reserve System, Report to the Congress on Credit Scoring; see note 8.
(40) The application form for a Social Security card is form SS-5 (08-2009), available at Social Security Administration, "Social Security Forms," webpage, www.ssa.gov/online/ss-5.html.
(41) Individuals may have applied multiple times for a Social Security card for several reasons, including loss of the original card or a change in legal name. Individuals are allowed to obtain up to 3 cards in a year and up to 10 over a lifetime except for applications in response to a change in legal name, which are unlimited. Individuals always receive the same Social Security number when they make additional applications.
(42) Further details are in Board of Governors of the Federal Reserve System (2007), Report to the Congress on Credit Scoring and Its Effects on the Availability and Affordability of Credit, "The Data and Model" (Washington: Board of Governors, August), www.federalreserve.gov/boarddocs/rptcongress/creditscore/datamodel.htm#toc8.1.