A bank can know a great deal about its own mortgages and still have a poor view of the market around it. That gap became Dan Feshbach’s business. LoanPerformance brought mortgage records together so institutions could compare performance and examine risk. Later, at MeasureOne, he tried to apply the same approach to student lending. The second venture supplied an awkward lesson: experience can tell you what information would be useful, without persuading its owners to hand it over.
Feshbach’s career makes an interesting case for the less theatrical kind of entrepreneurship. His products dealt with records, comparisons, repayment and verification. These are subjects that rarely inspire a standing ovation at dinner. They do, however, determine whether someone understands a loan portfolio or merely feels rather confident about it. Confidence is considerably easier to manufacture.
The story turns on a practical question. How do you get information out of the place where it is stored and into the hands of someone who can use it? His companies approached that question in different ways. LoanPerformance assembled cooperative mortgage databases. MeasureOne built a student-loan data business, introduced an academic-data platform, and developed into a broader service for access and verification. Along the way, the person authorizing the exchange became part of the product itself.
An education outside the usual founder script
Feshbach grew up around Palo Alto after an early childhood that included upstate New York and Iowa. His father worked through several businesses, from selling dresses and cars to stockbroking and dealmaking. The family supplied an early introduction to selling, persistence and personal connections. Later, mentors Mac McQuown and Dave Kaplan helped Feshbach develop his business thinking. He credits McQuown with substantial help on his first business plan.
His own academic route went through the University of California, Berkeley, where he studied city and regional planning. Planning and lending might seem like distant occupations. Both require a person to think beyond an individual transaction. A mortgage belongs to a household, a lender, a neighborhood and a larger market. A record becomes more informative when you can compare it with other records, rather than admire it in isolation.
That comparison is an editorial way of reading his career, rather than a claim about what he intended at Berkeley. What is clear is that he went on to build financial information businesses. His public self-description places him at the intersection of finance, education and consumer lending. It is an unusually useful description: each of those fields contains decisions made about people, and each keeps some of the information needed for those decisions behind an institutional door.
A business built between competitors
LoanPerformance’s cooperative databases offered mortgage institutions something their own books could not: a basis for comparison beyond the organization. Contributors supplied information; the combined data supported benchmarking and analysis. The arrangement depended on a bargain. An institution would disclose part of what it knew in return for learning something about the wider field. Software could process the records. People still had to agree to the exchange.
The company’s importance can be seen in how its data was used. Federal Reserve researchers acquired LoanPerformance data to study loans sold into the non-agency secondary market. The information included interest rates, borrower credit scores and characteristics of loans and properties. It allowed questions about mortgage performance to be examined at the level of the underlying loans, where differences between borrowers and contracts could actually be studied.
A later federal housing-finance analysis used LoanPerformance records to examine mortgages originated from 2001 through 2008. It also adjusted the data to make comparisons with another mortgage dataset. That detail matters. Large datasets do not release anyone from the obligation to ask what they cover. The collection needs a boundary, and the comparison needs a method. Otherwise, a very impressive quantity of information can give a very precise answer to the wrong question.
In April 2005, First American announced its acquisition of LoanPerformance. Feshbach remained president of the business in 2006. Its databases subsequently became part of CoreLogic. The acquisition gives this chapter a firm date, but the more revealing achievement is the infrastructure underneath it: a company had made pooled mortgage information useful enough for financial institutions and researchers to work with.
Student loans were a different negotiation
MeasureOne took the cooperative-data idea into student lending. By 2015, the company described a standardized database covering more than 500 securitizations and a cooperative drawing contributed loan data from major private lenders and holders. Its reports gave participants and observers a view of repayment, delinquency and lending practices. The product was partly the database and partly the shared language in which the market could discuss itself.
The distinction between private and federal loans was essential. MeasureOne’s June 2017 report examined private student-loan data, not the federal market. At the time, private loans represented roughly 7.7 percent of total outstanding student loans. That left 92.3 percent in federal loans. A finding about the smaller group could be valuable without describing the experience of every student borrower in the country.
This is where a profile of a data entrepreneur benefits from a little arithmetic. The impressive headline number for the whole market cannot stand in for the population inside a particular report. Feshbach was working with a specific slice of lending. Its underwriting, cosigners and school-certification practices were part of the explanation for its performance. Those features belonged beside the repayment numbers, rather than in small print far away from them.
The published results also show the rhythm of this business. In December 2016, MeasureOne reported an early delinquency rate of 2.7 percent and a later delinquency rate of 1.9 percent, each with its own definition. A data company earns its place by returning to those measurements consistently. One dramatic chart attracts attention; comparable reporting over time gives someone a chance to make a decision.
The difficult part happened before the spreadsheet
In September 2016, MeasureOne announced a $2.3 million seed financing led by Socratic Ventures, with Colchis Capital and University Ventures participating. The mix brought education and consumer-finance experience into the company. Feshbach’s stated aim was to give lenders, schools and families better information for decisions. The financing was a discrete milestone, rather than a substitute for demonstrating that customers would contribute or use the data.
His work also reached discussions about the wider student-loan information problem. At a Federal Reserve Bank of Philadelphia conference, he emphasized data cooperatives as a response to limited information. The conference account described practical obstacles: trust takes time, and lenders and other holders of records can be reluctant to share them. Public and private institutions bring different incentives to the same table.
It is easy to picture a database as an engineering project. This part of Feshbach’s career suggests a second picture: a lengthy negotiation about who benefits from participation. Someone must explain why contributing information is worth the trouble. Someone must make the definitions usable across organizations. Someone must stay with the project after an encouraging meeting produces no actual records. The spreadsheet is the visible result of agreements made elsewhere.
His later recollection of MeasureOne’s evolution included resistance from student-loan data holders and a customer’s interest in academic information. The company began exploring data contributed by consumers themselves. That changed the route into the records. The founder’s experience with one cooperative market had given him a starting point; the next market required him to reconsider the mechanism.
“I think I’m a good connector for people.”Dan Feshbach
The transcript becomes a usable record
By September 2019, MeasureOne was introducing an academic-data API. The platform extracted information from inconsistent transcripts and standardized it for applications. It included analytical products intended to connect academic records with credit performance and persistence. The company described possible uses across lending, academic institutions, employment, insurance and residential real estate. A document familiar to a registrar was being translated into something another system could process.
The attraction was understandable. Young people can have academic achievements before they have much of a conventional financial history. A platform that makes those records usable opens another set of questions for businesses serving them. The usefulness of any particular prediction still depends on the application. Standardization makes analysis possible; it does not make every decision based on that analysis wise.
The same announcement named Elan Amir CEO and Marita Scarfi CFO. Feshbach, the former CEO, would serve as chairman and advisor. That is an important turn in a founder’s story. The company was changing its product and its leadership at the same time. Feshbach’s contribution could continue through advice and governance while another executive took responsibility for operating the business.
MeasureOne’s present platform extends across income, employment, education, insurance and other consumer records. Its model puts consumer permission at the center of access. For this story, the continuity lies in turning information kept in separate places into something a business can use. The change lies in whose authorization helps make that movement possible. A cooperative database and a consumer-authorized connection solve related problems through different relationships.
A connector with more than one company
Feshbach’s work has also included co-founding the educational-software company TeachTown and founding Multiple, which connects startups, entrepreneurs and investors. Multiple’s current team page places him on its board and names Keevin O’Rourke as CEO. His career has therefore continued beyond MeasureOne’s executive office, with founder experience carried into support for other ventures.
There is a tempting version of this biography in which every company becomes proof that the founder always knew what would happen next. The actual sequence is more interesting. A mortgage-data business found a market. A student-lending venture built reports and a cooperative, then expanded toward academic records and broader access. Leadership changed. New organizations followed. The record contains adaptation, which makes it useful to someone facing a decision of their own.
His description of himself as a connector fits a career that keeps bringing separate parties into the same transaction: lenders and information, consumers and businesses, founders and investors. A connection only becomes valuable when both sides have a reason to use it. Much of his work has involved building that reason into a company.
Knowing where the data sits is a beginning. Getting it shared, making it comparable and finding someone willing to pay for its usefulness takes longer. Dan Feshbach’s career has unfolded in that interval. It is a place with fewer ribbon-cuttings than a founder might hope for, and rather more conversations. His businesses offer a reminder that information acquires value through the patient work of making it available to a decision.