Count Gilad Woltsovitch among those entrepreneurs whose personal experience played a critical role in the formulation of his company.
Mr. Woltsovitch is co-CEO and head of product at Backed, a platform providing young borrowers with accessible and affordable loan products which help them build both their financial independence and their credit score.
Mr. Woltsovitch moved to New York City in 2008. He had a job, savings, and a history of financial responsibility.
But he had no credit score. That made it challenging to get approved for a credit card or to rent an apartment.
Looking at Mr. Woltsovitch’s background one could be forgiven for thinking he posed little credit risk. He is a co-founder of iAlbums, a semantic media player curation engine and a former entrepreneur in residence at Cyhawk Ventures, where he created consumer products for the ad-tech industry. During this period he also established the Israel Ethereum meetup group, which researches and develops the Ethereum Decentralized Apps ecosystem.
With detailed knowledge of machine learning, computer programming and algorithmic synthesis, it was only a matter of time before Mr. Woltsovitch found the right alt-fi opportunity.
He initially considered cryptocurrency but deemed it too early in its development with too much risk. He looked instead to fin-tech.
“In 2014 there was a lot of hype around Lending Club and Prosper,” Mr. Woltsovitch explained.
Not only was there plenty of hype, there was also plenty of publicly available data, thanks to Lending Club.
That data showed decline rates ranging between 86 and 90 percent, Mr. Woltsovitch said. At those rates he theorized there had to be a nice layer of borrowers who were still good credit risks. So he began reverse engineering profiles to find them.
And he did. Millions of them. During the process he learned that many people deemed to be poor credit risks were actually people with minimal experience.
Climbing on the credit carousel can be a delicate process. To receive credit it helps to have a credit history, which, by definition means you need to be approved at some point. As traditional lenders move upstream to bigger fish they have less interest in dealing with new entrants into lending channels.
That has produced 26 million American adults who the Consumer Financial Protection Bureau deems “credit invisible,” those with no credit history whatsoever. That is roughly one out of every nine adults.
A further 19 million adults could not be effectively scored by current models, either because they have an insufficient history or because they displayed little recent activity. To generate a history, you need activity. To be active you need a history.
Combined, those totals represent nearly one in five American adults.
“Back then there was no difference between people with poor credit and those with a thin file,” Mr. Woltsovitch explained.
One group with a limited credit history which provides a unique opportunity are college graduates and they are Backed’s focus, Mr. Woltsovitch said.
He learned 60 percent of millennial college graduates receive some sort of financial aid from their parents, including allowing them to live at home. That is common because some are not aware of other options while those who are seldom qualify for any reasonable one.
Handouts, while understandable, do not foster independence. In solving a short-term solution they create a longer term problem.
Millennials gravitate to mom and pop because they are dependable, unlike those early jobs, which may be short-term, low-paying, or otherwise less than ideal. Yes they may have money now but is less certain.
Backed allows for co-signers like other lending options but in the event the borrower misses a payment, the co-signer gets 15 days notice to assist before default.
“Normally, the co-signer is notified after default when it is too late and two credit scores are damaged,” Mr. Woltsovitch explained. “Then you are responsible for the full debt plus penalties and interest all at once.”
Think of a Backed loan as a middle ground between financial dependence and dire financial consequences. Break glass only in an emergency, with no fees or penalties for doing so.
Mr. Woltsovitch said Backed had an impressively short time to market, thanks in part to regulators understanding the seriousness of the problem Backed addresses.
“Regulators are trying to find solutions. College graduates are soon going to be the largest base in the country and regulators are well aware of the problem their delayed entry into the economy will cause.”
Backed does provide loans without a “backer.” Borrowers with a minimum FICO score of 660 can qualify on their own.
Borrowers can also get a progressive loan which can reduce their rate by as much as 12 percent by providing the names of “vouchers,” people who will attest to their trustworthiness, and the backers, who agree to cover loan repayment.
Like many platforms, Backed relies heavily on the collection and interpretation if big data, Mr. Woltsovitch said. Unlike many, they partnered with a data-mining pioneer.
Dr. Jacob Zahavi built AMOS, one of the first commercial data mining systems. Active in business as Chief Analytics Officer for DMWay, a firm which builds state-of-the-art and fully automated platforms for developing predictive models more effective than those built by humans alone, Dr. Zahavi has been a professor at Tel Aviv University for 40 years.
Backed took the trove of publicly available data from Lending Club and reversed engineered it, creating a machine learning algorithm. Data from credit bureaus and other sources was added and they trained the model to operate as if hundreds of thousands of real loans had performed over three years.
Backed also adds to its database by encouraging borrowers to complete short quizzes in exchange for a reduced origination fee.
“The more data we have the better the model is,” Mr. Woltsovitch explained.
As the economy heads toward the down part of the cycle, the industry must prepare itself for the different environment, Mr. Woltsovitch advised. Many platforms may discover they were operating under a positive selection bias during healthy times. Left undiscovered, platforms simply add to the biases the longer times are good.
The problem is only discovered when times are less good. Unemployment rises and defaults might too.
“Until we experience it in real time no one will know how to model it,” Mr. Woltsovitch said.