The concept of alternative data and credit card scoring needs to center on “why.” Credit is saturated in the United States (and other mature countries like Australia, Canada, and the United Kingdom). Write-off and delinquency rates are predictable, and operational expenses require interest rates to be in the range of 20%.
Each market has fairness requirements and disclosure standards that ensure lending is available to everyone who meets a relatively low standard of ability, integrity, and stability.
The Wall Street Journal reports on the potential for creditors to shift from traditional lending standards to softer measures to increase credit availability. It is counter-intuitive. Consider today’s credit card market in the U.S., where revolving debt is at an all-time high, at more than $1 trillion outstanding. With 450 million cards outstanding for 130 million households, we have more than enough credit in play.
So, why add weaker credits?
- For decades, banks and other financiers have relied primarily on consumers’ borrowing history to make lending decisions. Now revenue-hungry companies are considering metrics both mundane and peculiar, like whether applicants shop at discount stores, subscribe to magazines or pay their phone bills on time.
- Those experimenting with new metrics range from big-name banks like Goldman Sachs Group Inc., Ally Financial Inc. and Discover Financial Services to upstart financial-technology firms.
Yes, we do have an unbanked population in the U.S. These fall into three groups: 1) people who have poor credit history; 2) people who have new credit history; and 3) people who do not want the documentation and prefer to transact in cash.
- The field of potential new borrowers is huge: About 53 million U.S. adults don’t have credit scores, according to Fair Isaac Corp., creator of the widely used FICO scores. Another roughly 56 million have subprime scores. Some have a checkered borrowing history or high debt loads. But others, banks point out, don’t have traditional borrowing backgrounds, often because they are new to the U.S. or pay for most expenses with cash.
Perhaps it does make a little sense to open the credit standards a marginal amount, but there will be credit loss implications. Today, credit card issuers can be precise in their pricing by knowing a specific credit score will require an explicit interest rate because of anticipated credit losses.
Lower credit standards and you will undoubtedly drive credit losses. Then, do we shift the expense to the new credit candidates, and price credit in the high double digits as is the case in Latin America, or raise everyone’s interest pricing?
Alternative data is interesting. Sure, you might enhance data by adding in someone’s reading habits or store preference, but you also degrade a proving model that links credit risk and opportunity. And for developing markets, consider how the scoring of social practices is out of control in China.
Perhaps alternative data makes sense in some markets where credit infrastructure is lacking, but for credit cards in mature markets, aggressive lending is already in force.
Overview by Brian Riley, Director, Credit Advisory Service at Mercator Advisory Group