HOME Banking and Finance  PROGRAMS SUBJECT AREAS Skills Case Study

CAPITAL ONE FINANCIAL CORPORATION
CASE STUDY

By Bruce Egsieker

Richard Fairbank (Stanford Business School 1981) and Nigel Morris (London Business School 1985), the founders of Capital One Financial Corporation, had taken Capital One-a major credit card subsidiary of a small, regional bank-and turned it into the nation's seventh largest credit card issuer. From its initial public offering in 1994 to 2000, Capital One's stock price increased more than 1000%, while the S&P (Standard and Poor's) 500 index had increased just under 300 %.

The company's average annual growth rate of 46% was the largest in the industry.  They also have an excellent "information-based strategy". They are a technology company first and a credit card company second.

Credit cards provided three services: means of payment, consumer loans, and product marketing.

Credit card interchange fees run 1.7%.

As long as credit card holders did not exceed their credit limits, they could borrow and pay off balances at will, thus providing them with revolving lines of credit. A cardholder who regularly paid the entire balance was known as a "transactor"

In 1987, nearly 4,000 banks offered credit cards.  The top four issuers of association cards were Citibank ($15.3 billion in outstanding balances, Chase (now J.P. Morgan Chase) with $5.4 billion, Bank of America with $5.2 billion and First Chicago with $4.6 billion.

Their mass mailings might generate a 5-10% response rate at 19.8% and a $20 annual fee.

In 1987 association members offered their cardholders the chance to earn frequent flyer miles based on their purchases.  This was very innovative…

In the 1980's, credit card spending rose from 4 to 8% of all consumer spending…by 1989…credit cards accounted for 70% of revolving debt…

The founders of the company, Fairbank and Morris, concluded that credit cards are not banking but information. They moved to customize the products to each customer and turned their business into a scientific laboratory.  Using thousands of statistical experiments illustrate the power of data mining and provide the company with a profile of cardholder behavior.

Fairbank and Morris intended "to build an organization from the ground up in which an information-based strategy was the strategic fulcrum."

"We believe technology is going to be the central nervous system of our information-based strategy."

Signet Bank in Richmond Virginia executed an agreement and merged with Capital One in October 1988. The merger allowed Capital One to access to Signet's customer base. Signet had the lowest charge-off rate in the industry (lowest number of defaults). Dan Oelrich, of Signet used statistical models to distinguish different kinds of credit worthy cardholders.  Capital One incorporated his model.  The model uses various tests to assign creditworthiness to the cardholder.

The test involved transferring a cardholder's other credit card balances to Capital One, for which the customer would receive a low introductory interest rate known as a "teaser" or 1.9% (1992).  No one in the industry had ever done this before.  It was a great success and Signet Bank spun off Capital One in October1994 with a $1.1 billion dollar IPO.

Capital One was hiring 100 people per week just to keep up with the transfer demand.  A. T.& T. responded by offering to drop annual fees for new Universal Card holders.

Creating the New Canon…A Religion of Science

The philosophy of the new organization was to exploit information by constructing scientific models that could be used to both assess the creditworthiness of potential cardholders, and to customize product offerings for existing ones.

They were building creditworthiness scores for individuals from credit bureau information, if they could forecast the creditworthiness of potential cardholders it would keep them from danger, i.e. defaulted card balances…the lemon problem.

So they are identifying risk profiles from publicly observable data which represents your behavioral characteristics to access you individual risk and making a judgment to issue you a card or to not issue you a card.

From your profile they can and do customize your APR (annual percentage rate on the card), your annual fee, the amount of your credit line, the color of the card, the initial "teaser" rate, co-branding with "affinity" partners like Delta Sky Miles Program, co-signer requirements, deposit requirements, etc. all of this based on a statistical actuarial calculations.

One reason for the "short life" of credit cards is that customers have learned how to game this system to get the best card out there.

Also all cohorts are tracked by their marginal contributions to the companies profit over time. Sudden drops in profitability of a product will lead to a withdrawal of a product by the company.

This entire test design has worked for Capital One and now the entire industry has followed their lead but it takes a scalable, robust system.  Capital One spent $100 million dollars in 1997 on a new (world's largest at the time) Oracle database with twenty-three terabytes of data or the equivalent of 40 single typed spaced pages of text for every American.  They handle 30 million cardholder inquiries per year.

Hiring on for Capital One is tough, three roundtrip interviews and a test of problem-solving skills.  They look to performance candidates.

For your information Capital One considered Auto Insurance at one time but rejected it.  They do not like any business where regulation inhibits freedom to test and mass customize.

Their business magic lies in the algorithms of customization.  Today Capital One executives consider cell phones a credit card with an antenna.

Credit cards are used on the Internet B2C (business to customer).  Forty percent of the transactions are through Visa and 29% use Mastercard.   It will be interesting to watch Capital One in the future with digital currencies, free C2C dot.coms, PayPal, Virtual Escrow providers and the like.

Capital One is very much involved in venture capital for new startup companies, particularly for firms with information-based strategies and firms competing in the financial industry.  The Internet isn't about commerce, its about information.