The credit card industry is highly concentrated. Ninety percent of the lending market is controlled by the top 10 issuers, yet there still appears to be fierce competition for balances and customers among these top issuers. However, there is a widespread belief that pricing is unfair despite this competition, resulting in reforms implemented by the Federal Reserve and passage of the May 2009 Credit CARD Act. The question explored here is how a market like the credit card market can fail to function properly despite heavy competition. This article offers a new conceptualization of a market structure, defined as a “peacock market,” characterized by a disproportionate mix of form over substance. It then shows how practices within the credit card industry can be better understood through the new structure.
I. Defining a Peacock Market
Charles Darwin first wondered how traits that clearly negatively impact survival rates can continue to exist in an evolutionary process. One classic example he used is the elaborate feather displays of male peacocks. Generating and maintaining these feathers takes considerable energy away from other purposes and adds the cost of slowing the males down if they attempt to escape from a predator. To explain this phenomenon, Darwin created the theory of sexual selection. While male peacock displays are inefficient in terms of personal survival, they have great signal value, which attracts mates and increases reproductive success. Early peacocks initially showed a modest display of relatively low cost that became self-perpetuating, and eventually grew to absurd proportions (from a survival standpoint). The peacock remains a classic example in evolutionary biology of how signal value can displace efficiency and be evolutionarily self-perpetuating despite intense competition.
Similarly, defined here a peacock market is a marketplace that has evolved to become dominated by elaborate signals (such as price or quality signals to consumers) no longer connected in a meaningful way to price, quality, or efficiency. As in biological evolution, peacock markets can become self-perpetuating because the signal retains value in creating demand despite being dysfunctional. The signals are designed to take maximum advantage of information shortages, limited attention, hard-wired behavioral biases, and socially-reinforced beliefs.
Just as male peacocks are rewarded in the natural selection process for their signals, peacock markets will reward firms that structure products and services around powerful but distorted signals that tend to suggest prices are low or quality is high. Firms will then ensure they are compensated for the prominently signaled pricing through complex, deferred, and other pricing mechanisms generally underweighted by consumers. These pricing mechanisms may be buried deep into contractual terms and conditions. They may be hidden in practice due to the complexity of the pricing structure and the scarce attention of consumers. They also may take advantage of well-known consumer biases. The known tendency for people to discount the distant future too heavily can result in prices that are backloaded, with the initial price being artificially low. Biases towards excessive optimism and overconfidence can result in prices that exploit contingencies that consumers underweight in likelihood (for example, the probability of being charged a late fee or receiving a penalty interest rate). Generally, firms in a peacock market focus their marketing and product design on signaling strongly in those dimensions that appeal to these biases, while utilizing mechanisms to maximize profitability in underweighted areas, rendering consumers overconfident and misinformed.
Why Peacock Markets Form
Certain consumer purchases are less inherently transparent both in quality and in pricing. These are the markets more likely to evolve into peacock markets. When the underlying desirable qualities of a product are not transparent and when pricing is inherently complex, cues are more easily converted to a signal that has little relation to true quality or price.
Complex ongoing services, including many within the financial marketplace, may be more likely to morph into peacock markets. Important characteristics of quality are often difficult to measure, making selection among alternatives difficult. It is easy to add multidimensionality to pricing due to the complicated nature of the service. In addition, multiple contingencies often exist that require additional service or can be used as an excuse to change the cost of a service. The book Gotcha Capitalism warns consumers about a variety of markets where hidden pricing is common, and tends to be dominated by multidimensional services received over an extended period of time such as cellular phone service, cable television, credit cards, and checking accounts.
There may also be an arbitrary element in the determination of which markets become peacock markets, with minor historical events shaping the industry’s future in important ways. It is now well-established that both the use of technology in the economy and institutions related to the economy are subject to path-dependence, where the path of future change and innovation depends heavily on past events. Often, dimensions along which progress is defined can get “locked-in” to a particular path based on prior events and choices. In the case of peacock markets, the dependence on inefficient signals can become self-perpetuating for a number of reasons. First, the technology for creating future signals can be cheaper and more efficiently produced when signals with a shared technological basis have been used in the past. For example, one “innovation” in the credit card industry has been the ability to price different account balances separately. This technology could then be expanded from two separate balances to three and higher. It also spawned a variety of complex pricing strategies that depended and built on this technology. Peacock markets can also be self-perpetuating because the ability to shroud new terms increases based on the number of terms already in existence. Behavioral economics suggests that people have limits on how many dimensions of price or quality they can consider at once when evaluating goods or services. For example, if a credit card has a single annual percentage rate (APR), adding a single fee that must be clearly disclosed is likely to receive more consumer attention when evaluating the product than if there are a half dozen potential fees and multiple APRs on the account already.
Peacock markets are also self-perpetuating from a cultural/institutional perspective of the firm. While individual tactics may vary, over time the strategic philosophy of a firm and even an industry can become deeply entrenched. If industry leaders have seen their historical success driven by price reduction through increased efficiency or core product improvements through innovation, then future strategies will tend to gravitate towards these themes. For example, historically innovation in the computer industry focused on increasing core product improvements (i.e., processing power, memory, lighter laptops, flat screens, CD and portable memories instead of floppy disks, wireless capacity, etc.). In contrast, if the leading firms have grown in profit and size by producing deceptive signals that suggest low prices while actually charging higher prices, this strategic philosophy will become just as deeply ingrained and can become self-perpetuating. For example, in the mortgage industry preceding the financial crisis, deceptive pricing practices grew in prevalence and were reflected in product features such as “exploding” adjustable rate mortgages with large built-in price increases.
For all of these reasons, markets that have peacock-market-tendencies will likely tend to become more and more differentiated into predominantly peacock markets over time.
A Field Guide to Spotting Peacocks
Since some level of distortion may occur in many markets, it is important to have applied indicators that differentiate a peacock market from other markets. Signs of a peacock market include:
- Signals are no longer strongly correlated to their underlying intended purpose. Often in traditional signaling theory, an intended signal is accompanied by random distortion.  However, in a peacock market the distortion is intentional. In fact, a price signal may yield little information about the overall price and may even be inversely correlated with it.
- Increasing disparity over time between the signaled information and the underlying factor it is intended to represent. A peacock market is an evolving market. Therefore, over time, the signals may get further and further removed from the information they are intended to signal. For example, there may be growth in the disparity between the average signal price and non-signal price over an extended period.
- The signal comes to dominate and drive product design. Signals in peacock markets are no longer merely marketing to enhance a core product, but become the primary driver of demand and profitability. Therefore, products evolve to be designed around the signal. This is a clear sign of market inefficiency since innovation is no longer focused on making fundamentally better products.
- The evolution of “fragmentation.” Fragmentation occurs when firms intentionally break price or quality into smaller pieces for the purpose of creating signals. The traditional economic literature on signaling theory indicates the signal is generally used to improve information in an uncertain environment. Fragmentation works in the opposite direction, where a segment of price (or quality) is intentionally broken off from the whole to create an artificial signal. Therefore, fragmentation is actually an information reduction strategy.
- “Shrouded” costs (or reductions in benefits) are prevalent. These shrouded attributes are used to compensate for losses due to extreme signals and to generally enhance profits. Shrouded costs can be naturally underweighted due to behavioral biases, or hidden through complexity of contracts or complexity of the mechanism itself. A shrouded reduction in benefits could, for example, be an obscure but important exception to an insurance policy that reduces firm costs and consumer value.
Inefficiency is often difficult to conclusively prove in economics, but intuition suggests that certain pricing and product structures are likely to be inefficient. If a snack food manufacturer decides, instead of raising prices per unit, to reduce the contents by weight of each box sold while maintaining the same box size, this seems to likely be inefficient packaging designed to distort consumer perceptions.
II. Applying the “Peacock Market” Concept to Credit Cards
A Federal Reserve paper described the evolution of the industry as follows: “[T]he relatively straightforward pricing model of a single APR, an annual fee, and modest penalty fees has been replaced by a model with a complex set of APRs, new and increased fee structures, and sophisticated finance charge computation techniques.” Much of price complexity in credit cards is related to an attempt by firms to create the illusion of low prices as well as an attempt to raise demand through differentiation of a product that is a fairly uniform commodity in reality. In a June 2009 speech, President Obama described problematic credit card issuers as companies that “compete not by offering better products, but more complicated ones, with more fine print and more hidden terms.”
Some of the practices discussed here may be limited in scope or prohibited by the Credit CARD Act passed in May 2009 as well by Federal Reserve Unfair and Deceptive Acts or Practices (UDAP) Rules announced in December 2008. However the majority of these changes have not yet taken effect. In addition, the behavior of companies prior to implementation of these rules lends insight into issuer behavior and market structure in general. As previously noted, even if specific practices change because Congress and the Federal Reserve have determined that many issuer pricing policies are unfair and deceptive, the strategic philosophy and corporate culture at firms in this market climate may be less easily altered.
Signals No Longer Are Strongly Correlated to Their Underlying Intended Purpose
The most obvious credit card price signal is the low introductory “teaser” rate. Often this price will be displayed very prominently in large font on the first page of the card offer, as well as several other places, and sometimes even on the envelope. Sometimes “no annual fee” will also be used as a prominent price signal.
Both of these prove to be poor signals for overall product price. The presence of a low teaser rate is not a useful indicator of an overall low cost of credit. In fact, in 2009 there was no significant correlation between the teaser APR and the regular purchase APR in credit card solicitations. Likewise, “no annual fee” says little to nothing about the size of other fees or total price paid.
As one should expect in a peacock market, these signals are also distorted towards known biases. There is an emphasis on short-term teasers that takes advantage of excessive discounting of future costs and an overly optimistic view of how much the consumer might borrow in the future. The annual fee, which consumers cannot avoid, is typically avoided in recent product offerings, while back-end contingent fees and APR changes that consumers underestimate in likelihood of occurrence are prevalent. Credit card terms also take advantage of cognitive limits and limited attention by adding many dimensions to price that are mathematically complicated (such as payment allocation methods) or simply go unnoticed.
Increasing Disparity Over Time between the Signaled Information and the Underlying Factor It Is Intended to Represent
There is strong evidence of a growing disparity between the signal price and other dimensions of credit card industry costs. The number and magnitude of penalty and miscellaneous service fees have increased over time. When the use of penalty fees in the credit card industry is studied, their use was shown to be increasing over time both in its appearance in the product contract (Figure 1), and in terms of how many people were at a penalty rate (Figure 2).
Figure 1: Share of Credit Card Solicitations with a Penalty Rate
Figure 2: Share of Credit Card Balances at Penalty Rates
There was also a growing disparity between the penalty rate and the regular purchase APR, something we referred to as “penalty shock” (see Figure 3). In addition, it turns out that the penalty rate is an underlying price dimension that remains hidden, with the Survey of Consumer Finances suggesting most people (54%) were unaware they were being charged this dramatically higher rate.
Figure 3: Penalty Shock
The Signal Comes to Dominate and Drive Product Design
While there are certainly signal distortions in the purchase of a new car, basic car design still is centered on useful attributes such as space, power, fuel efficiency, reliability, etc. The majority of innovations likewise focus on improving these product features. For credit cards, however, much of what is called product innovation comes in the form of new pricing structures that create perceived value without creating underlying benefits (or conversely that maintain perceived cost while actually increasing the underlying cost). An example of a product designed around creating perceived value through a distorted signal is a “0% for life” offer on certain credit card account balances. The offer is only profitable due to repricing mechanisms such as penalty APRs and payment allocation rules that create higher prices for the majority of consumers than what the offer would seem to imply.
The pervasive use of teasers also drives the design of credit card products, other dimensions of pricing, and the technology used to support those products. Other examples of products designed around the price structure include “fee harvester” cards that are targeted to the subprime market and often feature a moderate APR, but have a low line of credit and numerous front-end fees. These fees, including application fees, monthly maintenance fees, and similar recurring charges often consume the majority of the consumer’s credit line. Even card products targeted to the prime market may let fee income considerations influence credit lines. Some issuers might offer low lines while issuing multiple accounts to the same household, a strategy that seems unlikely to be driven by risk (given the aggregate line extended to the household) but instead may arguably be an attempt to collect higher fee income.
One of the most effective shrouded pricing mechanisms issuers have used has been structuring the allocation of payments to maximize revenue without changing the stated APR for any balance category. This mechanism has greatly influenced product design, resulting in the proliferation of multiple balance categories on a single account as well as teaser rates and other price gimmicks. When analyzed, payment allocation practices are found to be powerful, poorly understood by consumers, and inefficient. Figure 4 shows just how powerful the issuer practice of allocating payments to the lowest rate balance first can be using a particular scenario, where a consumer starts with a teaser rate balance and makes regular payments, purchases, and cash advances. In this example, the promotional rate balance disappears in six months, rendering it useless at that point even if the offer gave the promotion for a full year. After month six, the purchase balance rapidly declines leaving almost all of the customer’s balance at the cash advance rate which is typically much higher.
Figure 4: Affect of Lowest Rate First Method of Allocating Payments
The Evolution of “Fragmentation”
With fragmentation, a uniform price or quality is broken into smaller pieces for the purpose of creating signals. Unlike the traditional literature on signaling where information is naturally fragmented and the signal is used to improve information, in a peacock market fragmentation is actually an information reduction strategy. A firm manipulates a product design to shroud underlying pricing or quality while presenting prominently the factors optimized to draw in consumers.
Credit card pricing has become dominated by prominent (but misleading) price signals combined with shrouded revenue enhancement mechanisms to compensate for the cost of these signals. While a uniform price — or at least not highly disparate price — over time is more informative and more consistent with product cost, the price of credit cards over time is intentionally fragmented into two general components. The first is a highly prominent price signal that consumers are known to disproportionately attend to, while the second is a less obvious but larger component of price that is adjusted to make up for losses in the signal price.
Perhaps the archetypical example of fragmentation for any industry could be the credit card teaser rate. In a perfect market with rational consumers using a fixed, reasonable discount rate, there would be no reason to price products at 0% for the first 6 months or year, and then a much higher interest rate thereafter. It makes little sense based on the cost of lending or any other economic rationale. APRs on credit cards were at one time uniform. The primary purpose of disaggregating the near term interest rate and charging a separate price appears to be to create a powerful price signal. But this is not an information-enhancing signal as the traditional economic literature suggests. Instead, unnecessary complexity is introduced to create a signal that is lower than the long-term cost of credit. Due to cognitive constraints and limited attention, the information level digested by the consumer is probably lower due to the creation of the signal. This is what the signal is intended to do in peacock markets. It creates a compelling impression of low prices despite the fact that the total price paid by the typical consumer is much higher. In some cases these signals will be constrained by their cost or by reduced effectiveness beyond a certain point. However, in the case of teaser rates, the signaled price evolved over time to become lower and lower until a natural limit is reached; currently the most common teaser rate is 0%.
The penalty shock from penalty rates (shown in Figure 3 above) is evidence of growing fragmentation. Penalty rates create a false dichotomy in pricing by typically charging much higher rates to customers with only very minor violations (such as being a single day late). The triggering mechanisms of these penalty rates have grown broader over time to include less risky consumers while at the same time the disparity between regular and penalty rates has increased. This suggests that penalty rates have probably risen due to the fact that they are underweighted in consumer decisions, creating a higher level of fragmentation. There is also a growing disparity between the regular purchase and cash advance rates, with the difference between these two rates more than doubling between 2001 and 2009 (see Figure 5). While there is a risk difference between cash advance and purchase balances, it is highly unlikely that the probable loss rate on cash advances is 10 percentage points higher than the loss rate on purchase balances. It is also highly unlikely that this risk differential has more than doubled in the span of a few years. The more likely explanation is that this is an example of fragmentation that takes advantage of which price signals are overweighted and underweighted. While the most highly-weighted dimension of APR is probably the teaser rate, the regular purchase rate likely receives the second most weight. Cash advance APRs and other conditional APRs probably receive little attention given the number of fees and other terms consumers must simultaneously digest. Previous research backs this up, finding that half of people, even when they are cash advance users, do not know that they are charged a different rate for these balances.
Figure 5: Cash versus Purchase: APR Difference
Further evidence of fragmentation can be observed in a study by Ryan Bubb and Alex Kaufman. The authors compare the credit card practices of banks to credit unions and find substantial differences. As shown in Figure 6, credit unions have much more uniform pricing, while bank pricing is highly fragmented. This is not surprising since the mission of credit unions is to act in the interest of their members, who are also their borrowers while the mission of for-profit banks is to maximize profits. Banks have an incentive to fragment prices in a way that sends powerful signals while raising prices in underweighted areas. Credit unions may not have as strong an incentive to do so because such fragmentation would in effect be deceiving their own members.
Source: Bubb & Kaufman, 2009. Reprinted with permission.
“Shrouded” Costs (or Reductions in Benefits) are Prevalent
Credit cards clearly include numerous shrouded costs. Two that have already been mentioned include penalty APRs and payment allocation mechanisms. The payment allocation policy utilized by issuers was found be a shrouded attribute of the product. Consumer understanding was extremely low. When surveyed, only 3.4% of the public got three basic questions right that were necessary for understanding payment allocation’s impact. These were multiple questions with only 2-4 answers. Randomly guessing would have resulted in answering all the questions correctly 6.25% of the time. Even respondents with a high self-reported level of personal finance knowledge scored worse than if they had guessed at random.
Figure 7: Percent of Respondents Who "Passed" Payment Allocation Test
There are also numerous fees that raise the total cost for credit card services. Some of these late fees and over-limit fees are “penalties” for violating an agreement. However, these have evolved from being used as a tool for stopping undesirable behavior to becoming an important revenue stream. In effect, card issuers now want, and in some cases have been found to actually encourage, behavior that causes a cardholder to pay late or go over their credit limit. Other non-penalty fees are incurred for a wide range of other reasons such as engaging in foreign transactions, taking a cash advance, or account inactivity.
Shrouded reductions in benefits also frequently occur in the credit card industry. These frequently occur in rebate programs. One dimension of rebate benefit reductions may come in the form of restrictions on cashing in rebates. Even more shrouded may be the ability to cancel rebate programs arbitrarily (as some issuers have done recently) or rebate points due to a consumer’s late payment.
If a cardholder benefit is liquidity in case of an emergency, a shrouded benefit reduction also occurs when issuers change terms arbitrarily. Many issuers have recently reduced lines and closed accounts due to current economic conditions. This leaves consumers without emergency liquidity that they relied on when they are most vulnerable.
Figure 8: The Impact of the Identity of the Plaintiff on Arbitration Results
The use of arbitration clauses is an important shrouded benefit reduction. Many borrowers are unaware of the arbitration clause hidden in almost all credit card terms and conditions, or its potential impact on their ability to settle disputes. Looking at a database of arbitration disputes compiled by Public Citizen and consisting primarily of credit card cases, we found evidence of bias against consumers. All types of firm plaintiffs got better results than consumer plaintiffs, but firms that appeared frequently got the best results (see Figure 8). In addition, arbitrators had a clear incentive to be biased, with arbitrators that sided with firms more receiving a larger number of cases in later periods (see Figure 9). The bias in arbitration combined with lack of access to the court system can result in a shrouded benefit reduction in a number of ways. For example, credit card issuers often tout the limited fraud liability consumers have when they use a credit card. However, that right depends on a credit card issuer not ignoring claims of fraud. And in some cases when issuers persisted in pursuing a debt despite claims that is was incurred through fraud, arbitrators have sided with the issuers and ignored the fraud claims.
Product and Market Efficiency is Reduced as a Result of the Above
According to economic theory, sending the wrong price signals generally will cause market inefficiency. Teaser rates, for example, cause inefficiency by creating the wrong incentives for borrowing funds and encourage excessive short-term borrowing. Prices that are too high can also send the wrong signals to the extent that consumers are aware of them.
Previous research has also found other evidence of inefficiency. Figure 10 shows the impact of various methods of allocating payments made by consumers on risk-based pricing. Although the practice will soon be outlawed by the enactment of the Credit CARD Act, issuers have almost universally selected the lowest rate first (LRF) method of allocating payment (i.e. the teaser balance is paid first, before the purchase balance, which is paid before the cash advance balance). We found that issuers choose this method despite the fact that it is the least efficient payment allocation mechanism. In fact, this method of allocating payments resulted in inverse risk-based pricing, where the least risky consumers are charged the highest rate.
Figure 10: How APR Changes with Risk under Different Payment Scenarios
Considering penalty repricing, it was found that issuers typically did not notify consumers of the price change. (As mentioned previously, a related finding is that most borrowers did not know when they were at a penalty rate.) While legally permissible right now, this policy counters arguments that penalty repricing reduces moral hazard. Minimizing consumer knowledge certainly creates inefficiency, especially if it perpetuates moral hazard.
Credit cards have evolved from a product with one price to a product with several interest rates for different types of activity that can change for a number of reasons related to timing, index rates, triggering events, or simply the issuer’s choice. Credit cards also have a wide range of fees and add-on products. Much of price complexity in credit cards as a lending product is related to an attempt by firms to create misleading signals that cater to biases. This mix of prominent price signals in overweighted domains combined with revenue enhancement mechanisms in underweighted domains describes an archetypical peacock market.
Since this market structure is inherently inefficient, market intervention has been and continues to be appropriate. Attempts to deceive consumers and exploit biases should not be costless. Issuers need to fear some consequences for these actions other than that they will eventually be painlessly phased out. For example, the Credit CARD Act had a nine-month delay for implementation of most provisions and the Federal Reserve rules allowed a full year and a half for issuer transition. The goal (particularly with the Federal Reserve rules) appears to have been to make the transition as easy and costless for issuers as possible. But this leaves the incentives in place for issuers who have a peacock-market strategic philosophy to try new tricks without fear of adverse consequences.
Furthermore, while very valuable to consumers, the improvements from the Credit CARD Act are static and are not a long-term solution to a rapidly evolving peacock market. Ongoing monitoring and intervention (as necessary), such as with the proposed Consumer Financial Protection Agency (CFPA), could offer a long-term solution.
Joshua M. Frank is Senior Researcher at the Center for Responsible Lending.
 Senior Researcher, Center for Responsible Lending.
 “Shrouded attributes, consumer myopia, and information suppression in competitive markets,” describes how biases and information shortages may be exploited and how it may not be in any firm’s interest to reveal additional information. Xavier Gabaix and David Laibson, working paper 05-18, MIT Department of Economics, April 11, 2005.
 Bob Sullivan, Gotcha Capitalism (New York: Ballantine Books, 2007).
 To some readers, this point may seem obvious. However, much of economic theory is based on assumptions of path independence, where markets always evolve towards the most efficient configuration. See W. Brian Arthur, “Competing Technologies, Increasing Returns, and Lock-in by Historical Events,” Economic Journal (1989): 99, 116-131 and Paul A. David, "Clio and the Economics of QWERTY," American Economic Review (Papers and Proceedings) (1985): 75, 332-37 for the case for technology. See Douglass North, Institutions, Institutional Change and Economic Performance (Cambridge: Cambridge University (1990) for the case regarding institutions.
 See, e.g., A. Michael Spence, “Signaling in Retrospect and the Informational Structure of Markets,” (Nobel Prize Lecture, December 8, 2001).
 Mark Furletti, “Credit Card Pricing Developments and Their Disclosure,” discussion paper, Philadelphia Federal Reserve Payment Card Center, January 2003.
 Barack Obama, “Remarks by the President on 21st Century Financial Regulation Reform,” June 17, 2009, available at http://www.whitehouse.gov/the_press_office/Remarks-of-the-President-on-Regulatory-Reform/
 Results are based on analysis of Mintel Comperemedia credit card solicitations through July 2009. Mintel Comperemedia is a searchable competitive database tracking direct mail and print advertising in the United States and Canada, as well as e-mail in the U.S. Mintel Comperemedia tracks information to analyze eight vertical markets: Banking, Credit Card, Investments, Insurance, Mortgage and Loan, Telecom, Travel and Leisure, and Automotive. The regression coefficient between the two variables was slightly negative, suggesting if anything an inverse relationship. However, the relationship was not statistically significant. It should be noted that the purchase rate is merely another signal of cost rather than the true cost of credit, which may vary considerably from the purchase rate due to default APRs, fees, and different APRs for other balances.
 Joshua M. Frank, “Priceless or Just Expensive? The Use of Penalty Rates in the Credit Card Industry,”
Center for Responsible Lending, December 16, 2008, 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
 Id. at 6.
 For further discussion of fee harvester cards, see Rick Jurgens and Chi Chi Wu, “Fee-Harvesters: Low-Credit High-Cost Cards Bleed Consumers,” National Consumer Law Center, Nov. 2007, available at www.consumerlaw.org/issues/credit_cards/.../FEE-HarvesterFinal.pdf .
 Joshua M. Frank, “What's Draining Your Wallet? The Real Cost of Credit Card Cash Advances,”
Center for Responsible Lending, December 16, 2008, available at: http://www.responsiblelending.org/credit-cards/research-analysis/what-s-draining-your-wallet-the-real-cost-of-credit-card-cash-advances.html
 In this scenario, the consumer starts with a $1,000 promotional balance and makes purchases and cash advances of $100 a month each, while making payments in an amount equal to this monthly activity ($200 a month).
 See, e.g., Spence, “Signaling in Retrospect.”
 Conclusion based on analysis of data from Mintel Comperemedia.
 In addition, the price difference between these two balances is even higher due both to cash advance fees and to payment allocation methods that cause the cash advance rate to persist longer than it normally would.
 Frank, “What's Draining Your Wallet?”
 Ryan Bubb and Alex Kaufman, “Consumer Biases and Firm Ownership,” working paper, Harvard University, September 2009, available at http://www.people.fas.harvard.edu/~ryanbubb/papers/BubbKaufman_ConsumerBiasesandFirmOwnership.pdf
 While credit unions have much less fragmented prices, this practice is not universal. Some credit unions have offered credit cards in partnership with some of the issuers with the most aggressive pricing (possibly without being aware of the full implications of their practices). Likewise, in other areas (such as overdraft fees), a number of credit unions use the same policies as the large banks.
 Frank, “What's Draining Your Wallet?”
 Credit card issuers have even sometimes intentionally not credited payments in a timely manner in order to cause consumers to be late. Linda Punch, “Getting Tough?” Credit Card Management (Feb. 1, 2005).
 John O’Donnell, “The Arbitration Trap: How Credit Card Companies Ensnare Consumers,” Public Citizen, 2007, available at http://www.citizen.org/documents/ArbitrationTrap.pdf.
 Frank, “What's Draining Your Wallet?”
 Frank, “Priceless or Just Expensive?”
 If a borrower knew they were at a penalty rate and further knew that certain behavior (such as paying on time for six months) would cause their account to return to the regular rate, this would encourage responsible behavior.