Subprime lending is back, but this time new characters are entering the fray.
One such company is online lender ZestFinance, which received $150 million last week. It will use that money to lend to subprime borrowers, a group considered untouchable by banks because of their higher risk of default.
The Hollywood financial tech firm said it has developed a machine-learning software program that assesses the risk of lending to individuals based on a number of seemingly quirky traits. Previously, Zest had sold that software to financial institutions, but now it aims to use the program to manage the risk around originating loans to near-prime borrowers, the top slice of the subprime category.
Zest’s software uses nontraditional, often nonfinancial factors to judge credit worthiness.
“One of the things we look at is the way you fill out an application,” said Zest Chief Executive Douglas Merrill. “If you are typing your name in on an application in all upper cases you are slightly higher risk than someone who types in all lower case or upper and lower case.”
Merrill said they don’t know exactly why that is true, but the company’s software has established a correlation between the two. In order to determine credit ratings, the company sifts through more than 70,000 variables, including how long applicants have had their current phone number, social network connections and Web-browsing habits. Traditional factors such as credit scores and credit history are also considered, including data from before the 2008 market crash.
The company launched its subprime Basix loan product this summer in 15 states. Early performance convinced New York investment management firm Fortress Investment Group to boost the effort with the extra funding last week.
Zest loans are often used for consumer needs such as credit card consolidation or large-ticket medical items, said Merrill. Its Basix program offers an unsecured installment loan for a three-year term at fixed annual percentage rates between 26 percent and 36 percent. Monthly payment amounts and APR are fixed over a loan’s lifetime. Loans range from $3,000 to $5,000.
Tech lenders
Zest is part of an emerging financial sector that believes it can nimbly avoid the risks of subprime lending through savvy data crunching.
Last month, two other data-driven lenders raised significant funding: Social Finance Inc. of San Francisco raked in $1 billion, while Chicago’s Avant, which has a research and development office in Playa Vista, took in $325 million.
Armed with heaps of data and the new cash infusion, Merrill said his company can take a piece of a highly profitable sector deemed too risky and left for dead by big banks.
“Ever since the crisis in 2008, banks have really tightened lending. The amount of credit for near prime has really dwindled,” said Merrill. “It’s no longer served well by the financial services.”
Burned by subprime consumer lending after the financial crisis, many banks have recoiled at lending to consumers with poor credit. Others have found that new financial regulations make the practice too difficult.
“It’s become hard to do these subprime products,” said David Choi, director of the Fred Kiesner Center for Entrepreneurship at Loyola Marymount University in Westchester. Before the recession, it used to be possible to find a bank that would give you a loan.
“These days,” he said, “They all work with the same rules.”
Because many of these same consumers also don’t qualify for credit cards and don’t want the onerous terms of a payday loan, Zest is trying to capitalize on a potentially lucrative gap in the market.
“The most profitable customers are often at the lower scores,” said Choi. “By far the most profitable customers are the ones that are struggling.”
In addition to sharp data analysis, Zest tries to avoid diving too deep into the subprime category.
Zest says its Basix software can also adapt to consider different factors as needed.
“We rebuild our models monthly and restructure quarterly, so we can respond very quickly to a bad cycle,” said Merrill. “Through the data, we’ll be able to see a downturn coming and will adjust quickly.”
Still, there are risks.
Banks, credit card companies and other financial institutions have been data crunching for years, said Choi, adding that these same firms also thought they could predict the 2008 downturn and change course.
“There are few economists that were able to predict 2008,” he said. “Things could happen in a few days, in a few months where things go really bad. (In a downturn), all the fiscal analysis that works 95 percent of the time is useless.”
Ultimately, lending in the subprime category is about walking a fine line, said Choi, because it’s impossible to pinpoint a borrower’s risk with 100 percent accuracy.
“The behavior of someone who is a great customer to the subprime industry and the behavior of someone who will default on a loan are very similar,” he said.
