To Switch Or Not To Switch: An Examination of Consumer
Behavior in the Credit Card Industry
David M. Frank, Iona College1
The credit card and credit lending industry is one of the most competitive financial industries in the United States. In this industry we e a paradox between supply and competition. One could expect that competition would be abundant becau many firms compete in this market. Ironically, even though competition is inten, the industry fails to offer consumers the traditional benefits arising from competition. The benefits are: low price, market incentives to switch and “sweeteners” such as lower rates, and other perks associated with credit cards.
Credit card companies compete in two different markets, a primary market and a condary market. The primary market is the first level of competition within the industry; it is where consumers first come into the market eking credit. It is at this level that firms vie for first-time customers. Since most people need to establish credit and since there is an abundance of banks and credit companies, the supply is elastic. Credit firms as well as commercial banks are more than willing to offer consumers lines of credit. For instance, the firms frequently visit various college and university campus to solicit
under-employed full-time students. The firms do this fully cognizant that the students are at a high risk of default.
Financial institutions also nd out mailings to almost every houhold in the country with offers of credit. Becau credit is so easy to obtain, this paper will not focus on the elasticity of supply. Instead, this paper investigates why people cho to switch credit cards within the condary market. Conquently, the condary market level of competition is defined here to be the level where credit card issuers and commercial banks compete for each other’s existing customers. The target consumers at this level of competition are established
customers with balances on their credit cards. Consumers at this level have a certain degree of brand loyalty. Since competition at this level is basically centered on established customers, the market ba contains fewer people.
This paper also explores the reasons why people switch credit cards. Specifically interest rates, fees and card balances, and perks such as frequent flier miles and shopping discounts. The market structure that makes up this vast credit card and credit lending industry will also be discusd. In addition, a correlation will be established showing a link between consumer opinion towards the cred
it card industry and the fierce competition between rival firms and commercial banks. Finally, a logit regression model will be ud to predict consumer behavior and why the participants would switch to other credit cards. It is expected that balance and interest rates have the greatest influence on why consumers switch credit cards. The results should also show some relationship as to why people switch and whether or not they carry a balance from month to month.
I. Literature Review
The credit card industry is comprid of 4,000 firms that ll similar rvices to over 200 million customers nationwide. The market is not highly concentrated. That is, the top ten firms control two–fifths of the market, while the next ten, share one–tenth of the asts. Figure 1 below illustrates the holdings of credit card debt by the type of financial institution. The disparity between the commercial banks and the finance companies is especially noteworthy.
Table 1: Major Holders of Credit Card Debt (Billions of Dollars)
1998
1997
Total1264.11331.7
Commercial Banks512.6508.9batter
Finance Companies160.0168.5
Credit Unions152.4155.4beij
Savings Institutions47.251.6
Non–financial Business78.9 74.9
Source FRB2
Further, the industry has virtually no barriers to entry. In 1982, Baumol, Bailey, Panzer and Willig introduced the Contestable Markets Theory. This theory suggests that firms are able to enter the industry freely and unhindered by any barriers to entry.3
1. Entry is free and without limit.
2. Entry is absolute.
3. Entry is perfectly reversible. 4
The credit card and credit lending industry fits into this model as new firms and banks enter and leave this market freely. They have no barriers to entry and costs are minimal. Once in, they compete fiercely and show little if any reluctance to compete with already established companies. They even compete directly with companies that have the largest market share.5 This fierce competition has influenced how consumers cho credit cards, and will be discusd below.
II. Survey and Regression Model
bugtrap
In order to develop a fuller understanding as to why consumers choo to switch credit cards, or move balances from one card to another, a survey was distributed to faculty and staff at Iona College.6 The survey was designed to ascertain the extent of why people would switch credit cards, and asked the following questions.
1. How many credit cards do you have?
2. What is the balance on your credit card?
3. Do you know your interest rate?
口译官网
4. Would you switch cards if the interest rate were to increa?
5. Are your credit cards “maxed-out”?
6. Would you switch to another credit card if your credit were at its limit?
星期一
7. Do you consider your credit cards as another source of available cash
(extension of liquidity)?
The survey was designed to ascertain primarily yes and no respons. A positive or (yes) respon would be assigned a value of one, while a negative (no) respon would be assigned a value of zero.7 The survey also included two subjective questions to obtain statistical answers (i.e., How many cards do you have and what are the balance amounts). The latter question regarding balance was divided into 4 categories of:
1. Balance between $0-1500
2. Balance between $1501-3000
3. Balance between $3001-4500
4. Balance between above $4501
Each category was assigned a number from one to four in order to define average balance. For example, a mean balance of 1.70 can be equal to a balance of $1525-1550.8
The survey results provided some interesting results. Ninety eight percent of the people surveyed had credit cards while only 52% knew their interest rates.
Table 2: Descriptive Statistics Variable Mean Std. Deviation Balance 1.69 1.13 Maxed out
.09 .29 Switch if maxed out
.46 .49 Pay balance in full every month .56 .49 Extension of liquidity
.38 .48
When looking at the statistical results, it is important to remember the sample and how it originated. Typically most colleges, faculty and staff are not highly compensated. This explains why the mean balance, as shown in Figure 2 is between $1525-1550. Also of interest are consumers who credit lines are at their limits. Only nine percent of tho surveyed are at their limits, but 46 percent would switch if their cards were maxed out.
To estimate the probability that consumers would switch credit cards, a logistic
regression model was ud and the following logit model was estimated:
(1) S =
β1B + β2IR + β3M + β4P+ β5L + E
coqui
where:
•
留学英语S is the probability of switching •
B is Balance •
IR is interest rate •
M is credit limit maxed out •
P is do you pay your balance in full every month? •
L is extension of your liquidity.
Table 3: Regression Model
Error t-Ratio Prob. |t| ≥ X Variable Coefficient9 Std.
.00000
5.746
.2486E-01
B .1428
olivia.00004
4.089
商务英语考试.5899E-01
IR .2412
M.1815 .1165 1.558 .1193
羊的英文
2.014
.0440
.5627E-01
P .1133
.3849
2.833
.6929E-01
L .1963
At the 5 percent level, the following coefficients have significance: Balance, interest rate, pays balance in full, and extension of liquidity.
A logistic regression model was ud to determine the probability of switching for veral reasons. When using dummy variables, the dependent variable is not linear or continuous but dichotomous.10 That is the dependent variable has only two values (1,0) or is either a responder or non-responder to the question, "Would you switch." As a result, logistic regression analysis is the most appropriate type of examination.11 The logit model is bad on cumulative logistic probability of switching explained by balance, interest rate, maxed-out, pay balance in full every month, and extension of liquidity.
The model’s coefficients can be explained as follows12:
1. For every unit increa in balance, the probability of switching increas
by .034.
2. As IR increas by one unit, the probability of switching increas by
.057
3. If a person pays his/her balance in full every month, the probability of
switching increas by .268.
The regression results fit my theory fairly well. As expected balance and interest rates have the greatest influence on why people switch credit cards. However, paying balance in full
