www.responsiblelending.org

Joshua M. Frank, Senior Researcher

May 2012

Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Endnotes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

The Center for Responsible Lending (CRL) examined marketing and pricing practices prevalent in the credit card industry before implementation of the Credit Card Accountability, Responsibility and Disclosure Act (CARD Act) of 2009. CRL then examined the connection between these practices and company performance during the recent economic downturn.

For the study, CRL examined 23 marketing and pricing practices that were common among the top 100 credit card issuers during the summer of 2009, when the CARD Act outlawing or curbing many of these practices was passed but not yet implemented. Common practices included
*     imposing high-cost penalty interest rates even when a consumer paid his card on time,
*     using indexes to calculate interest rates to the disadvantage of card holders, and
*     assessing late fees that the Federal Reserve recently ruled were unreasonably high.

We then tracked credit losses for these companies from the beginning of 2006 through 2010.

Complaints to the Better Business Bureau correlated with specific practices, regardless of an institution’s type or size.

In fact, credit card practices were the best predictor of how quickly losses would mount compared to competitors—better than a lender’s location, type or size.

The research shows that unfair, deceptive or abusive lending practices were clustered within firms as part of an over-all strategy, suggesting that credit card issuers either believed in pricing that was stated clearly up front and with an eye on long-term customer relationships, or in pricing that was purposely complicated, hard to understand and intended to maximize short-term revenue.

Banks embracing the latter strategy also tended to be the most aggressive marketers, using high-volume mailings to existing customers to offer high-cost cash advances from a credit card or check. And, the larger the financial institution was that engaged in deceptive or misleading practices, the worse the practices tended to be.  In general, regional or smaller banks or credit unions tended to have clearer, fairer pricing.

The research also shows that a strategy of maximizing short-term revenue by using unfair, deceptive or abusive lending practices led to higher risk and lower profits during the downturn, undermining

a bank’s safety and soundness. Downturns occur fairly frequently.

Since 2000 there have been two recessions, and since 1980 there

have been five. A case in point is mortgage lending: Predatory

products seemed profitable in the short term and seemed to help

fuel economic growth but led to a disproportionate escalation in

losses when housing markets slowed and the economy soured.

Our new research shows this has also been true in the credit

card arena.

And the report shows credit card issuers' past claim is wrong that consumer protections have the

unintended consequence of undermining the safety and soundness of financial institutions. Common

sense curbs on abusive lending practices increase market transparency and bolster firms’ financial

strength. This benefits customers, investors, shareholders and, ultimately, taxpayers, who have had to

pay hundreds of billions of dollars to bail out banks after regulators allowed unreasonable mortgage

practices to flourish.

Lobbyists repeatedly have used the flawed argument that consumer safeguards undermine financial

health. They used it to persuade Congress and regulators to refrain from adopting the common-sense

mortgage lending rules that would have prevented the recent recession and continuing downturn.

And they are using it now to try to roll back reforms adopted in response to the mortgage meltdown.

This study shows that measures to stop deceptive and unfair lending practices promote market

transparency and enhance the health of lenders—and the economy—long term.

Finally, the report identifies practices the CARD Act did not eliminate but are potentially unfair,

deceptive and overly aggressive. Bank supervisors should further scrutinize these practices.

This new CRL research looked at credit card industry practices and financial statements in the

limited the size of late fees and how quickly they are triggered, and curbed many other practices

customers have complained about for years. Specifically, this research looked at credit card industry

practices that were common before the CARD Act and their relationship to customer complaints,

to type of financial institution and to changes in loss rates in the downturn.

issuers imposed penalty interest rates and a variety of fees. The financial data supplied information

on loss rates and the size of the card issuer. The analysis considered data from ten large issuers and

Bureau for the same period, which was just before full implementation of the CARD Act made many

of the practices we examined illegal. Most of the provisions of the Act were fully implemented in

February 2010.

The results provide important information about the relationship between a bank’s consumer policies

and its safety and soundness and for assessing the CARD Act’s success. Lawmakers intended the

CARD Act to create a more transparent market by getting rid of deceptive fees and pricing. Many

of the practices discussed in the report have been outlawed or limited by the CARD Act, though

some remain legal, including high cash-advance fees, excessive international transaction fees, minimum

finance charges that can exceed balances, aggressive marketing, and wide interest-rate spreads.

Earlier research shows that the CARD Act has improved market transparency without raising the

clustering of the bad practices the CARD Act addressed and whether they were connected to

complaints and market outcomes.

We tracked credit losses from 2006 through 2010. Losses for most credit card issuers did not start

to rise substantially until 2008. The next two years, 2009 and 2010, brought elevated losses for the

entire credit card industry. We calculated the increase in losses during the recession by comparing

average losses in pre-recessionary 2006 and 2007 with the average during 2009 and 2010.

We tracked a variety of credit card practices that existed prior to implementation of the CARD Act

and often indicated overly aggressive marketing or unfair or deceptive lending strategies. (See Table

1). Many of the pricing practices were deceptive because they were not closely tied to an issuer’s

costs or risks, and consumers had a poor understanding of how they worked. Some practices led to

wide gaps between the rate quoted in solicitations and the rate consumers actually paid. Clearly,

many issuers applied charges and fees based on research showing which costs consumers pay

Examples of practices analyzed include:

• Penalty interest rates imposed by a hair-trigger—a single,

minor infraction such as being late on a payment by a single

day or, in what is known as “universal default,” paying another

• Manipulation of variable rates using a “pick-a-rate” practice, where an issuer tied interest rates

to an index but picked a date on which changes in the index caused the biggest increase. Prior

CRL research found this to be one of several hidden pricing strategies that consumers understood

the same conclusion and in January 2010 banned “pick-a-rate,” deeming it a deceptive manipulation

of a variable-rate index.

• Charging minimum finance charges of as much as two dollars even if a consumer owed only a

penny of interest that month.

• Charging a foreign transaction fee even when the foreign company billed in dollars, causing

the issuer to incur no extra expense.

• Aggressively pushing debt protection insurance products, which are an extremely poor value

Practices on this list often indicated an issuer relied on price gimmicks and tricks and traps in

what consumers expected. (See Table 1 next page.)

We created a measure that scored unfair and deceptive acts and

was produced when we used factor analysis to gauge how practices

differed among issuers. The UDAP factor score does not

rely on a researcher’s judgment but instead tallies a score based

on the similarity of variables and how they tend to be grouped.

Description (with source for further research)

Cash/Purchase APR spread Large spreads between prices are often used to exploit price complexity and behavioral

biases (see “What Does the Credit Card Market Have In Common with a Peacock?”) No

Maximum Late Fee High late fees can be a way to raise price components that consumers tend to discount

(see “A Just Fee or Just a Fee?”) Yes

Penalty/Regular APR spread Large spreads between prices are often used to exploit price complexity and behavioral

biases (see “What Does the Credit Card Market Have In Common with a Peacock?” ) Partial

Minimum Finance Charge Amount Another type of hidden pricing used by many issuers where a $2 charge may be imposed

if there is only a penny in finance charges (see “Dodging Reform: As Some Credit Card

Abuses Are Outlawed, New Ones Proliferate”) No

Cash advance fee ceiling Consumer-friendly issuers tend to more often limit the maximum cash advance fee

(see “Dodging Reform: As Some Credit Card Abuses Are Outlawed, New Ones

Proliferate”) No

Hairline penalty trigger to end teaser rate Penalty rates are a form of hidden, back-end pricing, especially when they are easily

triggered even in the teaser period (see “Priceless or Just Expensive? The Use of Penalty

Rates in the Credit Card Industry”) Yes

Tiered Late Fee Structure Tiered late fee structures have been used to create a deceptive fee structure with

hidden fee increases (see “Dodging Reform: As Some Credit Card Abuses Are Outlawed,

New Ones Proliferate”) Yes

Teaser/Regular APR spread Large spreads between prices are often used to exploit price complexity and behavioral

biases (see “What Does the Credit Card Market Have In Common with a Peacock?”) Partial

Whether issuer uses a penalty APR Penalty rates have been used as a form of hidden, back-end pricing (see “Priceless or

Just Expensive? The Use of Penalty Rates in the Credit Card Industry”) Partial

Cash advance fee floor amount Cash advance fee floors have been raised by some issuers as a form of hidden fee

increase (see “Dodging Reform: As Some Credit Card Abuses Are Outlawed, New

Ones Proliferate”) No

Cash advance fee amount Cash advance fees add to price complexity and may be discounted by consumers when

choosing to take a cash advance (see “Dodging Reform: As Some Credit Card Abuses

Are Outlawed, New Ones Proliferate”) No

Hairline trigger for penalty APR Penalty rates are a form of hidden, back-end pricing, especially when they are easily

triggered (see “Priceless or Just Expensive? The Use of Penalty Rates in the Credit

Card Industry”) Yes

International transaction fee amount International fees are a hidden fee charged for transactions in foreign currencies, or

sometimes for any transaction with an international merchant (see “Dodging Reform:

As Some Credit Card Abuses Are Outlawed, New Ones Proliferate”) No

Whether penalty rate cure is "hard" Penalty rates are a form of hidden, back-end pricing, especially when they are difficult

to cure (see “Priceless or Just Expensive? The Use of Penalty Rates in the Credit

Card Industry”) Yes

Issuer utilizes "pick-a-rate" This deceptive and obscure tactic is used to manipulate the prime rate used for a

variable APR (see “Dodging Reform: As Some Credit Card Abuses Are Outlawed, New

Ones Proliferate”) Yes

Whether penalty rate cure is unknown Penalty rates are a form of hidden, back-end pricing. Typically, an unknown

(undisclosed) cure means that an account rarely or never cures (see “Priceless or Just

Expensive? The Use of Penalty Rates in the Credit Card Industry”) Yes

Cash advance check volume/total loans Encouraging the taking of cash advances with aggressive mail volumes is a way for

issuers to exploit inattention to price and fee differences and to maximize consumer

debt levels No

Solicitation volume/total loans Overly aggressive use of mail solicitations tends to push debt on consumers rather

than respond to true need and is rarely done by the more consumer-friendly issuers. No

Issuer allows extra time before late fee Consumer-friendly issuers often give consumers a break by allowing extra time before

charging high late fees. No

Length of grace period Consumer-friendly issuers offer more time between issuing a statement and the due date. Yes

Whether issuer charges intl. fee if in $ Charging an international fee on transactions denominated in dollars is a deceptive

fee-maximization tactic (see “Dodging Reform: As Some Credit Card Abuses Are Outlawed,

New Ones Proliferate”) No

Whether teaser is short The use of teasers—particularly short teasers—to attract consumers is an attempt to

exploit known consumer biases (see “What Does the Credit Card Market Have In Common

with a Peacock?”) Yes

Sells fee-based add-ons in card offer Add-on debt protection products have been found to be an exceptionally poor value

for the money by the GAO (see Consumer Costs for Debt Protection Products Can Be

Substantial Relative to Benefits but Are Not a Focus of Regulatory Oversight) No

Note: Full references to all of the related research cited here are available in the report's footnotes. Practices are ranked based on how closely tied they are to a general measure of bad practices.

The practices higher on the list are the ones that are the most powerful predictors of a tendency to have bad practices in general. The construction of this scale is discussed in detail later in

the report.

Addressed by

CARD Act?

The methodology used to construct this score, called factor analysis, attempts to identify underlying

drivers that explain a pattern of correlations within a set of observed variables. Factor analysis

identifies a small number of common factors that can explain much of an observed variance. This

methodology has the advantage of not relying on a researcher’s assumptions about how variables

interrelate. Instead it lets interrelationships among the variables speak for themselves. It is used

in psychology to determine what factors define intelligence, in business to understand consumer

behavior, in economics to study what drives differences in wealth between nations and in the physical

sciences to relate the spatial distribution of different chemicals to possible sources. Factor analysis

is useful when a researcher wants to learn what forces or patterns drive a large number of variables.

For example, say we wanted to compare cars. We would identify a number of car features—power

seats, leather upholstery, built-in navigation systems, anti-lock brakes, a high crash-test score—and

catalog which cars have which. We then would apply factor analysis to determine if features tend to

appear together in certain cars and not in others. We might find cars with anti-lock brakes tend to

also have a high crash-test score and good mileage but not leather upholstery, or that cars with leather

upholstery tend not to get good mileage but are likely to have power seats. The way these features

group together can be described by more general characteristics: safety features, performance features,

In our credit card study, the features we analyzed were practices such as imposing high late fees or

having a hair trigger for penalty rates. We analyzed these features to see if they occurred in groups

and what dimensions—such as a business strategy—explained the clusters.

Typically, a factor analysis highlights multiple factors or dimensions that can explain observed

behavior. A researcher’s job is to interpret the meaning of those dimensions. However, in this case,

assumed that practices might cluster in several strategic niches. Not only does a single dimension

dominate, but that dimension is very easily interpreted. Practices that were consumer friendly all

were associated with low scores on the scale, while practices that could be considered unfair, deceptive,

that the relationships among the 23 variables can be primarily explained by one dimension and that

this one dimension can be unambiguously interpreted as a UDAP factor score. A low score indicates

consumer friendly practices and a high score indicates the opposite. This implies that issuers that

tended to have one consumer-friendly practice tended to have other consumer friendly practices,

while issuers that use one unfair, deceptive or abusive tactic tend to use other similar tactics.

Practices indicating aggressive marketing tactics, such as heavily promoting the use of cash advance

checks or mailing high volumes of offers, lined up in a single dimension along with practices indicative

of cost-obscuring product features. This suggests aggressive marketing and deceptive pricing are

closely linked strategies. This makes sense. Both focus on pushing products that maximize income

regardless of whether they provide a benefit to the customer. Aggressively pushing cash advance

checks in particular seeks to tempt consumers into increasing their level of high-cost debt.

Table 2 splits the credit card issuers into five groups based on their UDAP Factor Score. While the

groups are roughly equal in size based on the number of issuers in each group, it is noteworthy that

this is not true of the number of accounts involved. About 90% of credit card balances were held

by the top 10 issuers when the data was collected. Six of these top 10 issuers were in the group with

the most unclear pricing and aggressive marketing practices. Nine of the top 10 were in the worst

two groups. The average or most common practice for each group is shown in the table. Issuers with

the best practices tended to have more transparent, straightforward pricing and less dependence on

complicated fee policies or pricing gimmicks such as big disparities between teaser interest rates and

final rates.

Good

Practices

Middle

Practices

Bad

Practices

Worst

Practices

(defined as

factor score

2nd quintile)

(defined as

factor score

4th quintile)

(defined as

factor score

3rd quintile)

(defined as

factor score

2nd quintile)

(defined as

factor score

1st quintile)

Regional Bank

or Credit Union

Credit Union Credit Union Regional Bank

or Credit Union

Specialist

Large Bank

or Credit Union

Specialist

Practices Issuer Attributes

Number of Top Ten Issuers Group 0 0 1 3 6

BBB Complaints per billion dollars 51.6 51.3 119.7 90.0 266.7

Average Factor Score -1.3 -0.8 -0.1 0.7 1.4

Maximum Late Fee $15 $25 $31 $39 $40

Used Tiered Late Fees No No Evenly Split Yes Yes

Difference Between Promotional and

Regular Rate if any (percentage points) 2.1% 2.9% 7.6% 12.6% 11.3%

Difference Between Cash and Regular

Rate if any (percentage points) 0.0% 0.9% 3.5% 8.2% 9.8%

Used a Penalty APR No Evenly Split Yes Yes Yes

Penalty APR had a hair trigger NA No No Evenly Split Yes

Once at penalty rate it was very hard

to return to normal rate No No Evenly Split Yes Yes

Average International Transaction Fee 0.3% 0.8% 1.2% 1.7% 3.1%

Charged International Transaction Fee

for Transactions Based in Dollars NA No No Evenly Split Yes

Manipulated the Variable Rate Index

Using the "Pick-a-Rate" trick NA No No No Yes

Cash Advance Checks mailed

per million $ 24 44 80 362 995

Average Cash Advance Fee 0.8% 1.7% 2.4% 3.2% 3.4%

Used floor on Cash Advance

Fee Amount No Evenly Split Yes Yes Yes

Best

Practices

Issuer Most Often is…

Table 3 shows regression results indicate the UDAP factor score was closely linked to issuer type.

The negative coefficients in the second and third rows of the table mean that when the issuer was a

regional bank or credit union, their UDAP factor score tended to be lower. In other words, regional

banks and credit unions tended to have better practices. As the last row shows, being a larger credit

card issuer correlated with a higher UDAP factor score. That means larger issuers tended to have the

worst practices, even after taking into account the type of institution. Even being a credit card

bank—defined as an institution focused primarily on credit cards—proved an insignificant factor

credit card loss rate before the recession, in 2006 and 2007, was also unrelated to practices. This

suggests that while some issuers with bad practices focused more on sub-prime consumers, others

Dependent Variable is UDAP Factor Score

(r-square=.765)

Coefficient p-value

Pre-recession Loss Rate -0.005 0.746

Credit Union -1.919 0.000

Regional Bank -0.433 0.019

Credit Card Bank -0.205 0.286

Managed Loans ($ billions) 0.004 0.025

We tracked complaints to the Better Business Bureau about each issuer.

To measure the frequency of complaints after accounting for the size of

the credit card issuer, we divided the number of complaints by the issuer’s

managed loans. The UDAP factor score explained the level of complaints

much better than the variables for issuer type or size (see Table 4 for

with complaint rates when practices are in the model. This can be

seen by comparing the p-value column in the second column to the fourth column of Table 4, with

a low p-value (often using a cut-off of 0.05 or 0.10 in research) indicating a statistical relationship

strong enough for us to be confident that it is not random error. While complaints show quite a bit

of variation that remains unexplained, bad practices are a good predictor of complaint levels.

Interestingly, loss rates did not significantly relate to frequency of complaints. This suggests that

complaints are not driven by whether a customer has a prime or subprime credit rating but by how

the customer is treated.

Dependent Variable is Complaints to BBB divided by managed loans

(r-square=.153)

Coefficient p-value

Pre-recession Loss Rate -0.009 0.774

Credit Union 0.023 0.964

Regional Bank -0.224 0.506

Credit Card Bank -0.232 0.503

Managed Loans ($ billions) -0.006 0.114

UDAP Factor Score 0.459 0.031

We calculated the increase in loss rate between the average for the pre-recession period of

Using regression analysis, we estimated how much of the change in losses is attributable to practices

measured by the UDAP factor score after controlling for other variables.

The results indicate that practices are a significant and powerful predictor of increases in losses

during a downturn. To give a sense of the magnitude of the relationship between issuer practices

and the change in loss rates, consider two issuers, one with very consumer-friendly practices

(10th percentile of UDAP factor score) and one that scores poorly (90th percentile of UDAP

factors score). If both issuers start with a loss rate of 3.2% before the recession, the consumer friendly

issuer’s loss rate would increase to 5.0% during the downturn, while the other issuer’s loss rate would

increase to 11.3%.

12%

3.2% 3.2%

11.3%

5.0%

10%

8%

6%

4%

2%

0%

Issuer with Safer Practices Issuer with Deceptive Practices

Pre-Recession Losses Recession Losses

In addition to the UDAP factor score, the regression includes variables related to the type and size of

the issuer as well as two additional variables. The first additional variable takes into account regional

economic distress. Specifically, it tracks Arizona, California, Florida and Nevada—states hardest hit

by the foreclosure crisis. There are many ways a region’s rise in foreclosures and drop in home prices

flags issuers that in 2006 and 2007, before the downturn started, had a high concentration of new

card holders rather than older, seasoned account holders. Young accounts tend to have low loss rates,

and during it would likely overstate the impact the recession had on loss rates for issuers that entered

it with high numbers of accounts less than a year old. The loss rate would probably have increased

anyway, even if the economy remained stable, as new card-holder accounts aged.

Regression results are shown in Table 5. The key result is that issuers’ practices were a statistically

significant and powerful predictor of the change in credit losses, as shown by the low p-value in the

last row of the table. The worse an issuer’s practices were, the greater its spike in losses during the

recession. In fact, issuer practices were the best single predictor of how much losses would increase

Expanding on the example given above, consider comparing an issuer who is in the worst 10 percent

in terms of unfair, deceptive and abusive practices to an issuer in the best 10 percent. These two

issuers would have UDAP factor scores of +1.4 and -1.3 respectively. This implies that the increase

in loss rate during the recession would be 6.3 percentage points higher for the issuer with the bad

loss rate of 3.2%., The consumer friendly issuer would see its loss rate increase to 5.0% while the

A loss rate difference of this magnitude has implications not just for the issuer, but also for consumers

and the whole economy. Even in strong years, the return on assets (ROA) for a credit card issuer

of 6.3 percentage points is enough to threaten the safety and soundness of the issuer.

Dependent Variable is the increase in the annual loss

rate during the recession (in percentage points)

(r-square=.468)

Coefficient p-value

Pre-recession Loss Rate 0.100 0.243

Credit Union -0.724 0.620

Regional Bank -1.982 0.039

Credit Card Bank -0.355 0.717

Managed Loans ($ billions) -0.009 0.389

Issuer is in a bubble state 4.265 0.000

Issuer new in 2006-2007 2.343 0.162

UDAP Factor Score 2.049 0.001

The second most important predictor of the jump in losses was whether the issuer focused on

lending in states hardest hit by the foreclosure crisis. This is not surprising. Financial distress

that starts with a family falling behind on a mortgage easily can spread to other debts. And the

greater a region’s housing price crash, the higher its unemployment.

Regional banks had a significantly lower increase in loss rate. This was beyond the impact of

specific practices or size and could be caused by differences in marketing, underwriting, or other

dimensions. All of the regional banks that are in the dataset of top 100 credit card issuers also

happen not to have significant business in states hardest hit by foreclosures. This might be

the cause of the significance of the regional bank variable. As the regression results indicate,

geographic focus is closely linked to the loss rate increase. A bank that reaches across the entire

nation in their consumer base includes some concentration in hardest hit states as well as in

states that have fared better. However, a regional bank that is not in a hardest-hit state is

sheltered from the stronger decline that occurred in these areas and therefore should experience

Neither an institution’s size, whether it was a bank or credit union, or its loss rate going into

the recession had a significant impact on its change in credit losses during the downturn.

Issuers' explanation for why there is a link between predatory practices and higher

loss rates in the recession is that some of the practices were needed to mitigate risk. The idea

is that these practices didn’t cause the risk, but in fact had to be used by issuers to manage the

risk posed by some borrowers. On that reasoning, the banking industry claims that the CARD

Act eliminated some useful pricing tools used to dampen losses. In particular, issuers point to

the CARD Act’s having sharply curtailed their ability to re-price existing balances with a high

CARD Act interfered with the ability of issuers to reprice for borrower risk is not consistent

with the data.

Prior CRL research shows that penalty repricing was not in fact used

much revenue as possible from consumers who had very minor infractions

but who did not have a substantially higher risk profile. Used this

way, interest rates no longer reflected risk, interest rates had become

the risk. Issuers used disproportionate price increases on consumers who

were, for example, late a single day, typically more than doubling their

interest rate. They also sometimes manipulated due dates and payment

processing to increase the likelihood that consumers would pay late and

incur a penalty rate. The result was that penalty rates in practice did

not result in an issuer’s repricing in a manner well-correlated with risk. The documented behavior

of major credit card issuers belies their claim they correlated price with risk: They imposed

penalty rates for extremely minor violations, even those they caused through manipulation.

They used penalty rates that were greatly disproportionate to the risk flagged by the violation.

And they made it exceedingly hard for a customer to get back down to a non-penalty rate.

There are a number of other reasons that this alternative hypothesis is not consistent with the

evidence presented here. The pre-recession loss rate was not significantly correlated with the jump

in losses after the recession. This suggests that the link between practices and the change in loss

rates was not caused by the general riskiness of borrowers, which would be clearly evident in the loss

rate during periods of normal economic growth. The results show that the baseline risk of loss alone

does little to explain what causes some issuers to have a much larger jump in loss rates when the

economy deteriorates. Instead, the change in losses is closely tied to a company’s practices regardless

of the risk profile of its customers during normal economic periods.

It is also important to note that the practices all clustered along a single dimension. There were

a number of practices examined in this study that it is impossible or imprudent to use as a way to

manage risk. For example, some argue the difference between a cash advance APR and a purchase

APR—with the typical interest rate on a cash advance at a large bank 10 percentage points higher

than a purchase rate—are somehow linked to risk-based pricing. But other practices such as manipulating

the prime rate using the “pick-a-rate” tactic have nothing to do with adjusting price to reflect

risk. If some issuers used practices now banned to manage risk, while others issuers used them merely

to maximize short-term income, factor analysis would separate these two distinct corporate strategies

there is no relationship between how closely a practice is linked to the UDAP factor score and

whether it is related to risk management. Practices clearly not used for risk management were

just as likely to be closely linked to the factor score as practices banks have argued were riskmitigation

To further eliminate the possibility that risk mitigation is driving issuer choices, we do a second

factor analysis to create what we will call a UDAP Risk-Unrelated Factor Score. This second factor

analysis only includes practices that logically would not be used to mitigate risk. The smaller group

of variables in the risk-unrelated factor score include 12 out of the 23 variables from the broader

measure while eliminating 11 and correspond again to the practices in Table 7 of the Appendix. This

creates a factor score similar to the one previously created, only it should not be at all correlated

with risk management techniques used by the issuer.

The new factor score turns out to be almost identical to the one for the broader list that includes

risk management factors, with a correlation coefficient between the two of 0.974 (1.0 indicates two

variables correlated perfectly). In addition, the UDAP Risk-Unrelated Factor Score shows the same

link between practices and the jump in loss rate that the full UDAP Factor Score shows.

Dependent Variable is the increase in the annual loss rate

during the recession (in percentage points)

(r-square=.452)

Coefficient p-value

Pre-recession Loss Rate 0.109 0.203

Credit Union -1.083 0.462

Regional Bank -1.890 0.054

Credit Card Bank -0.159 0.873

Managed Loans ($ billions) -0.007 0.504

Issuer is in a bubble state 4.208 0.000

Issuer new in 2006-2007 2.772 0.095

UDAP Risk Unrelated Factor Score 1.676 0.004

The results for both the full UDAP Factor Scores and the UDAP Risk-Unrelated Factor Score turn

out to be virtually identical when assessing the relationship of practices to risk. Table 2 would look

almost the same if the only practices discussed were those clearly unrelated to risk. This eliminates

the possibility that cause and effect are reversed when we study the risk-practices relationship; that

is, it eliminates the possibility that issuers adopted these pricing practices to mitigate risk. To the

encourage them to get in over their heads makes those consumers more likely to default.

The research shows the CARD Act has succeeded in stamping out many bad practices that

undermined the soundness of credit card institutions. Before the CARD Act, bad practices clustered

within firms as part of an over-all strategy. And the largest issuers in what is a very concentrated

market tended to have the worst practices. These practices—many of which were targeted by the

CARD Act—were linked to complaints and to negative economic outcomes. Customers hated them,

and they were bad for business. The credit card market functions much better for consumers and for

the long-term health of banks’ bottom line.

Our study shows practices prohibited by the CARD Act were linked

to an unstable financial market where bank losses were more volatile

and likely to jump rapidly in down-cycles. In addition, as prior CRL

research has shown, the CARD Act has made rates more transparent

with poorly structured bonuses and pay incentives might benefit

from a strategy that focuses on the short-term, that strategy leads to

higher risk, more defaults and lower profits long term. Such shortsightedness

harms consumers, the financial system and the economy.

The results confirm that regulators and lawmakers should think of

consumer protections and of safety and soundness as flip sides of the

same coin. As we’ve shown in this paper with credit cards, net income

obtained through questionable lending tactics does not imply a greater

level of safety and soundness. These tactics are unsustainable and

eventually lead to higher loss rates. And the higher loss rates from

these tactics tend to come at the bottom of the economic cycle, just

when the financial system is most fragile.

In the debate over whether to scale back the power of the new

Consumer Financial Protection Bureau (CFPB), Wall Street lobbyists

claim that consumer safeguards have unintended consequences and

an unfair practice is bad because it cuts into bank revenue. CFPB’s

supporters include many state and federal regulators and industry leaders such as Citigroup CEO

Regulators responsible for the safety and soundness of the financial system should encourage rules

that favor strong consumer protections, such as transparent pricing and products without deceptive

practices. Seeing these consumer protections as conflicting with bank safety and soundness is an

illusion. In short, the CFPB’s mission is not at odds with the safety and soundness mission other

regulators have. Ending practices that undermine market transparency is essential for strong banks

and a vibrant economy.

The continuing financial crisis is rife with evidence that lenders and regulators forgot that

transparency and sound underwriting are cornerstones of good lending. Mortgages failed largely

because federal regulators allowed products to proliferate that consumer advocates warned were

harmful to homebuyers and that reduced—often on purpose—consumers’ understanding of their

mortgages. For example, prepayment penalties, coupled with high adjustable rates, forced buyers to

refinance over and over into bigger and bigger loans. These features are directly linked to significantly

Law professor Ronald Mann has argued that predatory credit card issuers seek to put consumers in a

escalating interest rates and, before the Card Act, were the most profitable segment of the market for

issuers. This implies that issuers want consumers to fail to pay their bills to a certain extent so banks

can charge them higher rates and more fees than they could consumers who pay off their bill each

month and are never late. But keeping consumers in the sweat box is a delicate balance because they

often are on the verge of default and in many cases because of the penalty fees and interest rates the

banks impose. Banks don’t want them to actually default. The balancing act Mann describes is consistent

with our finding that issuers with an aggressive strategy of maximizing short-term net income

at the expense of consumers’ financial well being—that is, those who purposely make pricing unclear

and then aggressively market cards with cost-obscuring features—experience a disproportionate rise

in losses when the economy falters.

Historical evidence shows predatory credit card lending not only backfires during downturns but can

extend the slump. In the late 1990s, Providian Financial’s return on assets made it seem the most

profitable large issuer. Some analysts declared Providian one of the safest credit card companies to

invest in, even as consumer advocates decried many of its practices. When the economy deteriorated

after 2000, Providian’s losses were far greater proportionally than its peers, ending in a company

restructuring. Advanta also seemed to be a profitable issuer that used practices strongly criticized by

consumer advocates. Its losses in the recent downturn started earlier and have been far more severe

An alternative explanation of Providian and Advanta’s woes is that they simply underestimated

and therefore mispriced for the risk of their portfolio. But our new research shows their rise in losses

doesn’t correlate well with the baseline loss level in strong economic times. That baseline risk of loss

doesn’t explain what causes some issuers to have a much larger change in loss rates when the economy

sours. But change in losses is closely linked to a firm’s practices, regardless of the risk profile of its

customers in normal economic periods.

The report also identifies practices the CARD Act didn’t eliminate but that appear linked to

strategies based on potentially unfair, deceptive and overly aggressive practices. Examples include

a reliance on charges consumers often overlook. These include balance transfer fees, high-cost cash

advances, fees for international transactions, and reliance on large spreads between APRs, with

interest rates consumers pay least attention to being systematically higher. These practices deserve

further regulatory scrutiny.

Table 7 shows which variables were included in the reduced risk-associated-free practice scale

described in the text. The reduced scale only contains items that should not be related to risk.

The table also shows the coefficient from the component matrix for each practice in relationship

to the dominant factor in the full UDAP factor score (i.e. the one that is equivalent to a consumerfriendliness

scale).

Included in risk-unrelated

factor analysis

Principal Component

Analysis Matrix Coefficient

for Dominant Factor

Cash/Purchase APR spread 0.842

Maximum Late Fee 0.825

Penalty/Regular APR spread 0.804

Minimum Finance Charge Amount x 0.786

Cash advance ceiling x -0.783

Hairline penalty trigger for teaser rate 0.782

Tiered Late Fee Structure 0.758

Teaser/Regular APR spread x 0.746

Whether issuer uses a penalty APR 0.737

Cash advance floor amount x 0.715

Cash advance fee amount x 0.712

Hairline trigger for penalty APR 0.699

International transaction fee amount x 0.675

Whether penalty rate cure is "hard" x 0.628

Issuer utilizes "pick-a-rate" x 0.538

Whether penalty rate cure is unknown x 0.525

Cash advance check volume/outstndgs 0.495

Solicitation volume/outstandings 0.441

Issuer allows extra time before late fee -0.343

Length of grace period x -0.332

Whether issuer charges intl. fee if in $ 0.330

Whether teaser is short x 0.313

Sells fee-based add-ons in offer x 0.265

(r-square=.462) (r-square=.468) (r-square=.452)

Coefficient p-value Coefficient p-value Coefficient p-value

Pre-recession Loss Rate 0.107 0.216 0.100 0.243 0.109 0.203

Credit Union -1.137 0.417 -0.724 0.620 -1.083 0.462

Regional Bank -2.033 0.035 -1.982 0.039 -1.890 0.054

Credit Card Bank -0.363 0.712 -0.355 0.717 -0.159 0.873

Managed Loans ($ billions) -0.007 0.463 -0.009 0.389 -0.007 0.504

Issuer is in a bubble state 4.296 0.000 4.265 0.000 4.208 0.000

Issuer new in 2006-2007 2.572 0.124 2.343 0.162 2.772 0.095

UDAP Scale or Score 0.129 0.002 2.049 0.001 1.676 0.004

As discussed in footnote 15, a UDAP scale was constructed before viewing the results of the factor

analysis. The UDAP scale was based on judgment regarding the consumer impact of the practices,

rather than coming as an output of a statistical technique like factor analysis. To construct the

UDAP Scale, all of the factors from Table 7 were rescaled so that they each had a mean of zero and

the same standard deviation. This allows each factor to have equal weight when summed together.

In addition, some variables were inverted by multiplying by minus 1 so that higher numbers for all

variables was indicative of an unfair/deceptive/abusive practice while lower numbers were indicative

of more consumer-friendly practices. The variables inverted are those with a negative sign in Table 7.

However, the decision to invert these variables were made based on judgment regarding the nature

of the practices rather than these factor scores and was done before the factor analysis was ever

conducted. As shown in Table 8, all of the measures of UDAP scores (the Scale, the Factor Score,

and the Risk-Unrelated Factor Score) turn out to be close to identical with a very high correlation

coefficient. The correlations were also all highly statistically significant.

UDAP Scale UDAP Factor Score (full)

UDAP Factor Score (full) 0.997

UDAP Factor Score (risk-unrelated) 0.972 0.974

Regression with

UDAP Factor Score

(risk unrelated)

Regression with

UDAP Factor Score

(Full)

Regression with

UDAP Scale

Table 9 below shows results for the same regression shown in Table 5 with the other measures of

practices (the UDAP Scale and the Risk-unrelated UDAP Factor score) added. The results are

generally similar across regressions.

Dependent Variable is the increase in the annual loss rate during the recession (in percentage points)

Table 10 is similar to the Table 9 regressions except it uses standardized beta coefficients.

Standardized coefficients take into account the fact that some independent variables show a lot

of variation while some show very little in the data. It puts the coefficients for the independent

variables on a comparable scale by converting the standard deviation of each variable to be the

same. This allows comparisons to be made across variables in terms of their importance. A higher

absolute value for a standardized coefficient implies that more of the variation in the dependent

variable can be explained by the variation observed in an independent variable. In all of these

regressions, the variable with the greatest explanatory power based on the standardized coefficient

is the UDAP scale or score.

Regression with

UDAP Factor Score

(risk unrelated)

Regression with

UDAP Factor Score

(Full)

Regression with

UDAP Scale

Dependent Variable is the increase in the annual loss rate during the recession (in percentage points)

(r-square=.462) (r-square=.468) (r-square=.452)

Standardized Standardized Standardized

Coefficient p-value Coefficient p-value Coefficient p-value

Pre-recession Loss Rate 0.120 0.216 0.113 0.243 0.123 0.203

Credit Union -0.173 0.417 -0.110 0.620 -0.166 0.462

Regional Bank -0.275 0.035 -0.268 0.039 -0.260 0.054

Credit Card Bank -0.045 0.712 -0.044 0.717 -0.020 0.873

Managed Loans ($ billions) -0.079 0.463 -0.094 0.389 -0.072 0.504

Issuer is in a bubble state 0.455 0.000 0.452 0.000 0.446 0.000

Issuer new in 2006-2007 0.151 0.124 0.137 0.162 0.162 0.095

UDAP Scale or Score 0.563 0.002 0.614 0.001 0.534 0.004

1 Credit Card Accountability, Responsibility and Disclosure Act (CARD Act) of 2009, Pub. L. No. 111-24, 123 Stat. 1734

2 Although the Top 100 issuers were used in the analysis, due to missing data in some fields, about 80 data points are used in each regression or correlation coefficient. All statistics cited had a sample size of over 70, with about 40% of the sample coming from credit unions. The top 10 issuers hold about 90% of all balances, so that the top 100 issuers include many medium-sized credit card portfolios. Solicitation and mail volume data both came from Mintel Comperemedia, a proprietary database and from examination of individual offers posted online by the issuers.

3 This same data, aside from more recent credit loss information, was used in a prior CRL report: Joshua M. Frank, “A Just Fee or Just a Fee?” Center for Responsible Lending, June 8, 2010, available at
http://www.responsiblelending.org/credit-cards/research-analysis/a-just-fee-or-just-a-fee.html  . The report found that higher late fee levels were closely linked to other questionable credit card practices and in fact that when these practices are accounted for, late fees were not linked to credit loss levels as was previously claimed.

4 We end up with less than 100 issuers in the dataset because of incomplete data for some issuers. Most results are based on

about 80 issuers. So if we define any issuer outside of the top ten as a medium issuer, then our final analysis included 10 large

issuers and approximately 70 medium issuers. The smallest issuers in the dataset analyzed each had about $90 million in

credit card receivables.

5 Joshua M. Frank, “Credit Card Clarity: CARD Act Reform Works,” Center for Responsible Lending, February 16 2011,

available at http://www.responsiblelending.org/credit-cards/research-analysis/FinalCRL-CARD-Clarity-Report2-16-11.pdf .

6 This interpretation of pricing spreads is consistent with Ryan Bubb & Alex Kaufman, “Consumer Biases and Firm

Ownership,” Harvard University working paper, May 2009.

7 “Universal default” is defined differently by different experts, but generally refers to raising rates based on factors

completely separate from the performance of the consumer’s credit card account with the issuer. “Universal default” can

refer to rates being raised based on being late on a card with a different issuer or simply based on an decrease in the consumer’s

credit score due to other factors. Whether “universal default” was outlawed by the CARD Act depends on the definition

of universal default used. If “universal default” refers specifically to using these triggers to raise the interest rate on a consumer’s

new balances, even if it has nothing to do with the performance of the consumer’s credit card account with that issuer.

8 Joshua M. Frank, “Priceless or Just Expensive? The Use of Penalty Rates in the Credit Card Industry,” Center for

Responsible Lending, December 2009, available at: http://www.responsiblelending.org/credit-cards/research-analysis/

priceless-or-just-expensive-the-use-of-penalty-rates-in-the-credit-card-industry.html

9 Lenders calculate variable rates by adding a fixed amount to a previously identified index rate, typically a market-based

rate such as the prime rate that changes over time. Traditionally, issuers have specified the index rate on a certain date as

the rate used for calculating that billing cycle’s interest rate. But in the period leading up to the CARD Act, a growing

number of credit card issuers had begun to charge customers a higher interest rate by making a very minor and obscure

change in the fine print of the agreement. More specifically, instead of stating that the index rate “will be the maximum

prime rate reported on the last day of the billing cycle,” many issuers changed the language to state that the index rate

“will be the maximum prime rate reported in the 90 days preceding the last day of the billing cycle.” This practices was

found to raise interest rates by 0.3 percentage points on average (Joshua M. Frank, “Dodging Reform: As Some Credit Card

Abuses Are Outlawed, New Ones Proliferate,” Center for Responsible Lending, December 2009, available at: http://www.

responsiblelending.org/credit-cards/research-analysis/Dodging-Reform-As-Some-Credit-Card-Abuses-Are-Outlawed-New-

Ones-Proliferate.html).

10 United States Government Accountability Office, “Consumer Costs for Debt Protection Products Can Be Substantial

Relative to Benefits but Are Not a Focus of Regulatory Oversight,” March 2011, GAO-11-311.

11 This pricing philosophy in the credit card market is discussed in detail in: Joshua M. Frank, “Why Free Markets

Can Sometimes Turn into ‘Peacock Markets’: The Evolution of Credit Cards,” Journal of Economic Issues, June 2010; and

Joshua M. Frank, “What Does the Credit Card Market Have In Common with a Peacock?” The Lydian Payments Journal,

24-40, 2009.

12 “UDAP refers to Unfair or Deceptive Acts or Practices” Both federal law and laws in all 50 states have some variant of

laws banning such practices. The existing federal UDAP law is the core of the Federal Trade Commission’s enforcement

authority against non-banks, (see 15 U.S.C. §45). Prior to the Dodd-Frank financial reform law, the Federal Reserve Board

had the authority to issue UDAP rules applicable to banks, which compose the majority of credit card issuers, under the

FTC ACT. The new Consumer Financial Protection Bureau (CFPB), when fully operational, will have authority to issue

rules to prevent unfair, deceptive and abusive practices that will apply to all credit card issuers. 12 U.S.C. 5531.

13 The descriptions of the dimensions do not appear spontaneously from the data and are interpretations made by the

researcher. In some cases, by seeing what original variables (in our example, the car product features) rank high in certain

dimensions, it becomes clear and unambiguous what the dimension represents. For example if airbags, anti-lock brakes,

traction control, and high scores on crash tests are all linked to one dimension of auto features while other features are not,

the dimension seems to somehow represent safety. Sometimes the results of factor analysis can result in dimensions that are

hard to translate into a concrete meaning. However, in the case of credit card practices, the dimension that resulted from

the analysis was very easy interpreted

14 The first factor explained 38.3% of the entire variance between the 23 practice-based variables examined. No other

factor explained more than 7.1% of the variance. Table 7 in the Appendix shows more statistical details of how each

variable related to this first, dominant factor. The single factor interpretation is based in part on the Cattell Scree Test, a

very commonly used method for determining how many factors to include in factor analysis results. Eigenvalues for the first

5 components were: 9.586, 1.787, 1.660, 1.385, and 1.785, showing a large drop-off after the first factor, therefore creating a

clear “elbow” graphically.

15 Typically a factor score dimension is interpreted after it is constructed. However, in this case the factor score dimension

turned out to correspond perfectly with a scale that was separately constructed based on “good” or “bad” practices (from

the consumer’s perspective). The scale was not used in the final report because, unlike factor analysis, it relies on the

subjective judgment of the researcher for its construction. But the fact that the factor score so closely mirrored a scale that

was constructed separately, with every variable showing the same sign as the scale is an extremely rare event, and suggests

that the interpretation of the factor score dimension is quite clear. The UDAP factor score and the scale ended up having

the same signs for each practice studied and being highly correlated with each other.

16 Which issuers were “credit card banks” was defined using the FDIC’s labeling of bank peer groups. If a bank holding

company was or had within it a bank designated by the FDIC as a “credit card specialty bank,” then it was given that

designation in the dataset. The FDIC defines a credit card specialty banks as an “Insured commercial banks whose lending

activities are focused on credit cards are in one of three peer groups within this special line of business peer group.” Most of

the top 10 issuers are designated as credit card specialty banks, including banks with a broad range of product lines such as

Bank of America and Chase. However, other banks with many lines of lending but that are less prominent in credit cards

such as Fifth Third Bancorp and SunTrust Banks, Inc. have no credit card specialty bank.

and soundness risk, even if losses in strong economic times appear to be under control.

18 When the frequency of complaints is analyzed as a function of issuer type and size without including practices, it

appears that issuer type matters somewhat, with credit unions getting significantly fewer complaints and regional banks

showing a tendency to possibly get fewer complaints. The credit union variable was statistically significant at the 5% level

while the regional bank variable was only marginally statistically significant—at the 10% level. This suggests that there is

between complaints and issuer type or size is because practices are correlated with issuer type and size and practices are

the underlying driver of complaint levels.

19 The loss rate calculation was done in terms of percentage points—i.e. if losses in the 2006-2007 period averaged

4.5% and losses in the 2009-2010 period averaged 8.7%, the increase in losses would be 4.2 percentage points. Since the

percentage point increase in losses could in part be a function of how large losses were to start with, the pre-recession

loss rate was included as an independent variable in all regressions. In none of the regressions was the starting loss level

statistically significant.

Urban Institute, available at http://www.metrotrends.org/commentary/jobs-housing.cfm.

ch12.html#5sub3.

23 Table 10 of the Appendix gives detailed standardized coefficients for the regression which is the basis for this statement.

The results were very similar if we use the scale discussed in footnote 15 rather than the UDAP factor score. Table 9 of the

Appendix gives regression results using several different measures for practices.

24 This is based on taking the difference in factor scores (1.4 + 1.3) and multiplying that sum by the regression coefficient

for the impact of factor score on the change in loss rates.

25 The starting loss rate and the average increase used in the example are not arbitrary. They are the average of all of the

top 100 issuers.

available at: http://www.federalreserve.gov/boarddocs/rptcongress/creditcard/2009/default.htm

27 If the geographic concentration of the bank in terms of bubble/non-bubble states is important, then one way to

conceptualize the situation is that there would be three levels of loss increase associated with geography. Banks concentrated

in a bubble state would have the largest loss increase. Banks that have a national presence would have a mid-level loss

increase (because they are in both the bubble and non-bubble states). Banks that concentrate their lending in a non-bubble

regions would have the lowest loss increase. All of the “bubble state” issuers in the dataset happened to be credit unions.

28 The argument generally made by issuers is not that the CARD Act raised losses, but that it took away some of the tools

they used to create revenue to offset those losses.

29 Joshua M. Frank, “Priceless or Just Expensive? The Use of Penalty Rates in the Credit Card Industry,” Center for

Responsible Lending, December 2009, available at: http://www.responsiblelending.org/credit-cards/research-analysis/pricelessor-

just-expensive-the-use-of-penalty-rates-in-the-credit-card-industry.html

30 There would be some overlap in these dimensions, but if a group of issuers just used the risk-related practices as a risk

management technique, and rarely used the other questionable practices, then a second primary dimension would be needed

to explain this pattern of behavior. If all the issuers that tended to have bad practices were just using it to manage risk, the

one would see the practices that are not related to risk unexplained or poorly explained by the primary dimension (and likely

a second dimension emerging unless these other practices occurred completely randomly).

31 It could be argued that a few other practices such as cash advance fees and a lack of cash advance fee ceilings, and use of

hard penalty rate cure are also used to manage risk, but this is not the case. First, just like with penalty rates, the evidence of

how issuers used these practices suggests that they were trying to maximize income from hidden pricing rather than trying to

align risk with price. In the case of cash advance fees, the same issuers who tend to charge a high cash advance fee and not

include a fee ceiling already tend to charge cash advance APR's that are disproportionately higher than the purchase APR

relative to the difference in risk between the two groups. Furthermore, the APR is much more rational risk management tool

to use to manage the ongoing risk of carrying a balance than a high one-time fee for a cash advance that may be paid off

by the next statement . Regarding penalty APR cures, given the minor vDescription (with source for further research)

Addressed by CARD Act? Cash/Purchase APR spread Large spreads between prices are often used to exploit price complexity

and behavioral biases (see “What Does the Credit Card Market Have In Common with a Peacock?”) No Maximum Late

Fee High late fees can be a way to raise price components that consumers tend to discount (see “A Just Fee or Just a Fee?”)

Yes Penalty/Regular APR spread Large spreads between prices are often used to exploit price complexity and behavioral

biases (see “What Does the Credit Card Market Have In Common with a Peacock?” ) Partial Minimum Finance Charge

Amount Another type of hidden pricing used by many issuers where a $2 charge may be imposed if there is only a penny

in finance charges (see “Dodging Reform: As Some Credit Card Abuses Are Outlawed, New Ones Proliferate”) No Cash

advance fee ceiling Consumer-friendly issuers tend to more often limit the maximum cash advance fee (see “Dodging

Reform: As Some Credit Card Abuses Are Outlawed, New Ones Proliferate”) No Hairline penalty trigger to end teaser rate

Penalty rates are a form of hidden, back-end pricing, especially when they are easily triggered even in the teaser period

(see “Priceless or Just Expensive? The Use of Penalty Rates in the Credit Card Industry”) Yes Tiered Late Fee Structure

Tiered late fee structures have been used to create a deceptive fee structure with hidden fee increases (see “Dodging Reform:

As Some Credit Card Abuses Are Outlawed, New Ones Proliferate”) Yes Teaser/Regular APR spread Large spreads between

prices are often used to exploit price complexity and behavioral biases (see “What Does the Credit Card Market Have

In Common with a Peacock?”) Partial Whether issuer uses a penalty APR Penalty rates have been used as a form of hidden,

back-end pricing (see “Priceless or Just Expensive? The Use of Penalty Rates in the Credit Card Industry”) Partial Cash

advance fee floor amount Cash advance fee floors have been raised by some issuers as a form of hidden fee increase

(see “Dodging Reform: As Some Credit Card Abuses Are Outlawed, New Ones Proliferate”) No Cash advance fee amount

Cash advance fees add to price complexity and may be discounted by consumers when choosing to take a cash advance

(see “Dodging Reform: As Some Credit Card Abuses Are Outlawed, New Ones Proliferate”) No Hairline trigger for penalty

APR Penalty rates are a form of hidden, back-end pricing, especially when they are easily triggered (see “Priceless or Just

Expensive? The Use of Penalty Rates in the Credit Card Industry”) Yes International transaction fee amount International

fees are a hidden fee charged for transactions in foreign currencies, or sometimes for any transaction with an international

merchant (see “Dodging Reform: As Some Credit Card Abuses Are Outlawed, New Ones Proliferate”) No Whether penalty

rate cure is "hard" Penalty rates are a form of hidden, back-end pricing, especially when they are difficult to cure (see

“Priceless or Just Expensive? The Use of Penalty Rates in the Credit Card Industry”) Yes Issuer utilizes "pick-a-rate" This

deceptive and obscure tactic is used to manipulate the prime rate used for a variable APR (see “Dodging Reform: As Some

Credit Card Abuses Are Outlawed, New Ones Proliferate”) Yes Whether penalty rate cure is unknown Penalty rates are a

form of hidden, back-end pricing. Typically, an unknown (undisclosed) cure means that an account rarely or never cures

(see “Priceless or Just Expensive? The Use of Penalty Rates in the Credit Card Industry”) Yes Cash advance check volume/

total loans Encouraging the taking of cash advances with aggressive mail volumes is a way for issuers to exploit inattention

to price and fee differences and to maximize consumer debt levels No Solicitation volume/total loans Overly aggressive

use of mail solicitations tends to push debt on consumers rather than respond to true need and is rarely done by the more

consumer-friendly issuers. No Issuer allows extra time before late fee Consumer-friendly issuers often give consumers a break

by allowing extra time before charging high late fees. No Length of grace period Consumer-friendly issuers offer more time

between issuing a statement and the due date. Yes Whether issuer charges intl. fee if in $ Charging an international fee on

transactions denominated in dollars is a deceptive fee-maximization tactic (see “Dodging Reform: As Some Credit Card

Abuses Are Outlawed, New Ones Proliferate”) No Whether teaser is short The use of teasers—particularly short teasers—

to attract consumers is an attempt to exploit known consumer biases (see “What Does the Credit Card Market Have In

Common with a Peacock?”) Yes Sells fee-based add-ons in card offer Add-on debt protection products have been found to

be an exceptionally poor value for the money by the GAO (see Consumer Costs for Debt Protection Products Can Be

Substantial Relative to Benefits but Are Not a Focus of Regulatory Oversight) No Note: Full references to all of the related

research cited here are available in the report's footnotes . Practices are ranked based on how closely tied they are to a

general measure of bad practices. The practices higher on the list are the ones that are the most powerful predictors of a

tendency to have bad practices in general. The construction of this scale is discussed in detail later in the report. Violations

involved leading to most penalty APR's and the APR increase that typically was disproportionate to risk, a responsive

mechanism to return consumers to a regular rate when they show a lower risk level would be consistent with trying to

align risk with price. Keeping a consumer at a high penalty rate long-term despite evidence that they have reduced their

risk is consistent with using the practice as a trap rather than with trying to align price with risk.

32 This was confirmed using a t-test comparing the factor score coefficients for the two groups of practices with no

significant relationship found. Similar results where found when using a scale, with both types of practices best explained

by a single scale based on a correlation matrix of all practices.

33 Joshua M. Frank, “Dodging Reform: As Some Credit Card Abuses Are Outlawed, New Ones Proliferate,” Center for

Responsible Lending, December 2009, available at: http://www.responsiblelending.org/credit-cards/research-analysis/Dodging-

Reform-As-Some-Credit-Card-Abuses-Are-Outlawed-New-Ones-Proliferate.html

34 This dimension of the debate is discussed in Derek Thompson, “Safety and Soundness vs. Consumer Protection”,

The Atlantic, March 2, 2010, available at: http://www.theatlantic.com/business/archive/2010/03/safety-and-soundness-vsconsumer-

protection/36941/

35 Pandit recently stated that “behavioral arbitrage” where consumers select products that “look cheap but turn out

expensive” creates systematic risk as was seen in the credit crisis and housing bubble. Pandit called for stronger consumer

protection and product regulation as the solution. Donna Borak, “Citi's Pandit Calls For Greater Consumer Regulation,”

PaymentsSource, September 26, 2011.

36 Oct. 23, 2009 Bernanke speech at a Boston Fed economic conference, "...The recent crisis clearly illustrated the links

between consumer protection and the safety and soundness of financial institutions. We have seen that flawed financial

instruments can both harm families and impair financial stability. Strong consumer protection helps to preserve household

savings and to provide families access to credit on terms that are fair and well matched with their financial needs and

resources. At the same time, effective consumer protection promotes healthy competition in the marketplace, supports

sound lending practices, and increases confidence in the financial system as a whole." http://www.federalreserve.gov/

newsevents/speech/bernanke20091023a.htm

37 Lei Ding, Roberto G. Quercia, Wei Li, and Janneke Ratcliffe. “Risky Borrowers or Risky Mortgages? Disaggregating Effects

Using Propensity Score Models.” Chapel Hill, NC: University of North Carolina at Chapel Hill, Center for Community

Capital, 2008. U.S. Department of Housing and Urban Development, Office of Policy Development and Research (2010).

Report to Congress on the Root Causes of the Foreclosure Crisis, available at http://www.huduser.org/Publications/PDF/

Foreclosure_09.pdf. Roberto G. Quercia, Michael A. Stegman, Walter R. Davis, “The Impact of Predatory Loan Terms on

Subprime Foreclosures: The Special Case of Prepayment Penalties and Balloon Payments,” Housing Policy Debate 18(2):

311-346. Anthony Pennington-Cross and Giang Ho, The Termination of Subprime Hybrid and Fixed-Rate Mortgages, Real

Estate Economics, Volume 38, Issue 3, pages 399–426, Fall 2010

38 Ronald J. Mann, “Bankruptcy reform and the “sweat box” of credit card debt,” University of Illinois Law Review, 2007.

39 Because Advanta went into bankruptcy too soon to get a true reading of what its 2009 and 2010 losses would have been,

it is not included in this report’s statistical analysis.

The Center for Responsible Lending is a nonprofit, nonpartisan research and policy

organization dedicated to protecting homeownership and family wealth by working to

eliminate abusive financial practices. CRL is affiliated with Self-Help, one of the nation’s

largest community development financial institutions.

302 West Main Street

Durham, NC 27701

Ph (919) 313-8500

Fax (919) 313-8595

1330 Broadway

Suite 604

Oakland, CA 94612

Ph (510) 379-5500

Fax (510) 893-9300

910 17th Street NW

Suite 500

Washington, DC 20006

Ph (202) 349-1850

Fax (202) 289-9009

Copyright © 2012 by Center for Responsible Lending