A Curious Case

If you want to innovate for the world, you need the room to do it. And what else? Sanjiv Das and a tale of machine learning, mortgages, and mistaken identity.

DAS

WHEN SANJIV DAS was a young child in Bombay, India, and wanted to learn about a subject—Bengal tigers, chess, the history of India’s partition—his mother’s father, Abbas Basrai, would send him to a hallway cabinet.

It was filled with row upon row of files: news and research clippings on hundreds of topics. Das’ grandfather’s ravenous curiosity and love for writing—including book-length biographies of his two wives and three daughters—made quite the impression on the boy.

“He was definitely one of the most curious people I knew,” says Das, whose family lived with his grandfather until Basrai died when Das was six.

Now the William and Janice Terry Professor of Finance and Business Analytics at SCU, Das has spent his life chasing intellectual pursuits along a distinctly winding and varied path.

His desire to get into the innards of problems that interest him led to an ill-advised stint in accounting school in Bombay … to a fascination with using algorithmic models to set prices and decrease risk in complex financial instruments … to a bank job selling models, and then a Ph.D. so he could create such models … to a tenure-track position at Harvard Business School … to a post-Ph.D. master’s in computer science. All that has led to years-long explorations of natural language, data analytics, neural networks, machine learning, artificial intelligence, and financial technology.

He has also been pursued for tenured positions at UC Berkeley, New York University, and others. But he values Santa Clara—to which he was recruited in part by SCU behavioral finance pioneer Meir Statman—for the intellectual freedom it affords scholars.

There’s a value in sharing one’s thought processes: “Here’s a problem, and here’s how I’m thinking about it.”

That, in turn, means SCU attracts smart, collaborative colleagues, Das says, and encourages the kind of interdisciplinary exploration he loves. “My colleagues are pretty incredible,” he says. “Because we are a small University, I’ve been really fortunate to meet and collaborate with all these other people across campus, which doesn’t happen in many schools. I don’t think I’d do better work elsewhere.”

PERIPATETIC MIND

Students also reap the benefits of his peripatetic mind. How so? Das has created new classes (or entire majors) to share his newfound knowledge with students, many of whom revere him for the way he shares his curiosity, thinking, and problem-solving techniques. “He’s very good at explaining the how and the why of what he teaches,” says Jeff Glupker M.S. ’18, who studied machine learning and time series modeling and forecasting with Das—and worked on a practicum at Credit Suisse with him. “Where I get the most value from him is just hearing him and his thought process: ‘Here’s a problem and here’s how I’m thinking about it.’” That, Glupker says, is “training your brain to think.”

The fields in which Das operates are often complex: interest rate derivatives, randomized algorithms, machine learning. But Das conveys a calm and a “lean in” approach that greets virtually every suggestion a student might offer with a pleased smile and a nod and “Yes, and what else?”

Das credits his grandfather with fueling not only kinetic curiosity but a groundedness. The time and home in which he was raised taught him the value of principles, too—and their cost. He grew up in Bombay (now Mumbai) during the late 1960s and 1970s. In the wake of India gaining independence from Britain in 1947, ideas about change were in the air. Economic socialism attracted political attention in large swaths of the country; Indira Gandhi, daughter of independence leader Jawaharlal Nehru, was elected prime minister, but imposed a violent two-year state of emergency when challengers sought to oust her. India entered the age of space exploration and nuclear power.

Das grew up in the home of his grandfather with his brother, Romith, and their parents, who had defied their respective faith traditions by intermarrying. His mother, Zubeida, was Shia Muslim; his father, Sujit, was Bengali Hindu. Das’ mother was excommunicated from the local Muslim community over the union; her family—reform-minded like the Hindu Das side of the family—stuck by her.

His grandfather Basrai occasionally ran afoul of certain hardline Muslim leaders as well—for trying to get what he saw as fairer treatment for poor worshipers. That sort of religious acrimony turned Das into an atheist by his teens, so much so that he tried to avoid enrolling in the Jesuit high school he would later attend—St. Mary’s School in Bombay—by refusing to meet one-on-one with the principal as required for every incoming student. The principal admitted him anyway, based on the strength of his test scores.

The Jesuit ethos he learned at St. Mary’s helped shape him into a person who cares about social justice. (He’s now a Buddhist.) So did his parents’ insistence on ethical living—which gave him a different side of the role of religion.

“My mother was always pointing out verses in the Quran that spoke to ethics, and saying, ‘Here are some interesting principles, you should keep them in mind.’” His mother earned a master’s in sociology. Her first work was in the home. But she also educated herself to successfully invest her inheritance in the stock market as a way to supplement household income. Though his father never attended college, he was the breadwinner—a corporate salesman for the likes of Dunlop Tires and British Paints—and valued the intellectual and cultural leaders helping shape the country’s identity.

The Das family home was a hub for reform-minded people of Muslim and Hindu faiths, who would drop by for informal salons. They might delve into the work of Rabindranath Tagore, the Nobel Prize–winning Bengali writer who was friends with Mahatma Gandhi. Das and his brother were taught to listen and not talk during these sessions—but they soaked them up.

‘OH WOW, SOMETHING USEFUL.’

After high school Das thought an accounting “commerce college” degree would solve a persistent boredom he had felt in elementary and high school. “Oh wow, something useful,” he remembers thinking. But by the time he realized he found accounting tedious, it was too late to leave under India’s rigid educational system. So he soldiered on—and played a lot of cricket, earning a spot on the school’s competitive cricket team. Still, he decided he had a better shot at success in academics than sport.

Armed with an accounting degree, he promptly enrolled in the Indian Institute of Management business school in Ahmedabad, modeled closely on Harvard. (That included not only case studies but room layout, he would later learn.)

His analysis of the home mortgage crisis, “The Principal Principle,” helped find new solutions for banks and borrowers alike.

He didn’t find business school itself especially exciting. But he did find a group of faculty pursuing what fired their imaginations—sparked by work at Stanford and MIT, where they were schooled—in operations research, mathematics, computer science, and programming “compilers”—decidedly un-business-school fare.

Together they formed a half-dozen strong crew of faculty and students, including Das and another young renegade named Raghu Sundaram, who has remained one of Das’ closest lifelong friends, and who now is dean of NYU’s Stern School of Business. The group found ways to create classes and conduct research—into operations research, multi-criteria decision-making, graphical programming, and simulation methods—while the students finished their MBAs.

Das’ first job was helping Citibank trading rooms set up and implement interest rate derivatives pricing models in India, Australia, Japan, Hong Kong, and Singapore. He didn’t like not being able to create the models, which required a knowledge of advanced physics-based mathematics and finance. He was supposed to explain the end result to customers. He wanted to understand how they worked, and maybe find ways to make them work better. So he got a Ph.D. at NYU.

A few years later, after six years on a tenure track at Harvard Business School, he took a sabbatical in Berkeley and got a master’s from Cal in computer science. That field has continued to open up more discoveries—such as his observation that everyone from Google advertising customers to large lending banks were using machine learning to unearth optimal prices for their products. So he decided to teach such topics, often at the same time he himself was learning them.

His colleagues at SCU’s Leavey School of Business credit him with co-launching the school’s graduate business analytics program—which currently enrolls 70 students. He put the curriculum together so students would have marketable skills, and he put the board together. “There were a lot of administrative hurdles,” says Seoyoung Kim, a finance professor who frequently collaborates with Das. “Without him, it wouldn’t have happened.”

THE IMF, THE CRASH, AND THE OTHER SANJIV

Once Sanjiv Das gains mastery in a subject, he’ll also lend his expertise to corporations, regulators, and others—connections he often makes at conferences where he has been asked to speak. In return, he’ll gain access to real-world laboratories in which he can test out and advance his work, bring in students, and see his conclusions take flight.

He has lent his expertise to the International Monetary Fund, teaching fintech, quantitative models for banking, and machine learning for macroeconomics. At the San Francisco District Attorney’s office, he helped flag misrepresentations by banking officials testifying about failed loans. At the Federal Deposit Insurance Corp., where he serves as an academic fellow, he helped analyze loan-restructuring models. Recently he teamed up with SCU Professor of Mathematics Dan Ostrov to conduct research with global investment firm Franklin Templeton, helping find complex new formulas to increase the likelihood that wealthy clients’ existing portfolios will actually achieve their financial goals. Their first of three published papers won the prestigious Harry Markowitz Award for best paper published by the Journal of Investment Management.

His curiosity can be sparked by random encounters. Take his focus, from 2010 to 2012, on how to restructure mortgages after the 2008 financial crash. This was a hugely problematic area, but not one that Das—then focused on options-pricing models and measuring market liquidity—would normally have undertaken. But in 2009, he began getting dozens of angry emails and calls from Citibank clients complaining about the financial institution’s failure to provide mortgage relief after the housing crisis. Das was puzzled. Why him?

It turns out that Sanjiv Das is also the name of the then-CEO of CitiMortgage. The barrage of emails were meant for that Das. SCU’s Das found the situation humorous. He also became fascinated with the problem the bank was trying—and apparently failing—to solve: Which loan terms should banks relax—and by how much—to avoid foreclosures while maintaining profitability in the loans? He reached out to his Citibank namesake, who invited him to his New York office and gave him access to loan data and to his own office to work. Das set to work using the kind of modeling he previously had used on options and other financial securities, this time for mortgages.

“I realized that the choice to walk out and not pay your mortgage is an option I can value,” Das explains. “I could mathematically see what features of that option I could tweak to prevent somebody from exercising it to default on their mortgage.”

The result: a 2012 paper, “The Principal Principle,” that argued the most economically feasible tactic for banks was to forgive outright a portion of underwater home loans—rather than shortening or lengthening the loan term or reducing interest rates. The idea caught on. Das presented it at eight conferences globally. When the U.S. Department of the Treasury updated in 2012 the Home Affordable Mortgage Program, Das’ idea was incorporated as a “principal reduction alternative.” Numerous hedge funds that bought distressed mortgages have also downloaded Das’ paper to help them restructure loans.

Sanjav Das

‘WHAT IF WE TRIED THIS?’

In 2018, in a speech to Santa Clara colleagues after they had named him Faculty Senate Professor of the Year, Sanjiv Das remarked that they were all “fortunate to be in a position of making a living from being curious.” He talked about adoring the freedoms the University affords: freedom to be bored, because it unleashes curiosity; freedom to take risks and to set one’s own standards for success; and freedom to work across disciplines.

His colleagues say they enjoy his collaborative and adaptable “What if we tried this?” approach to co-teaching and to the papers they write together. He has a habit of inviting colleagues to join him in person for marathon paper-writing sessions—rather than each scholar retreating to their own office to tackle discrete sections.

“Sanjiv doesn’t stay still for too long,” says his finance colleague Kim. Far from being averse to changing paths, “As the world changes, he’s happy to learn new things and adapt.”

His mathematics colleague Ostrov notes that Das has twice completely revamped a math-finance class they’ve co-taught; it was originally created to be taught in the Octave computer programming language. Das reengineered the course to teach it in the programming language R; when he discovered Python, he revamped it again—to make it a better experience for students.

Where will Das’ curiosity take him next? Currently it’s fintech—the application of cutting-edge technology to make financial products more efficient, less costly, or more valuable to banks or consumers. He has worked on modeling to use natural language queries to help consumers set individualized financial portfolios—cutting out the advisor middle man. He’s doing that in part with a company called Betterment. com. He’s also helping a Bay Area company called PayActiv with a banking service for low-income workers to access their paychecks before payday for a flat $5 fee. The company currently has processed more than $1 billion for users, with machines in Walmart and other locations.

Once a week, you can find him at the year-old “moonshot” lab at Credit Suisse. There, he and four data analytics graduate students work with Credit Suisse researchers on advanced projects, such as using virtual reality to illustrate the interrelatedness of banks or venture capital relationships. The goal? Provide a new way to help people understand the importance and scope of such relationships.

He also has been working with a group of staff and faculty including Colleen Chien in law, Irina Raicu J.D. ’09 in the Markkula Center for Applied Ethics, and Shannon Vallor in philosophy to tackle the complex ethical issues involved with machine learning and AI. Das is already working with students to craft algorithmic models to help quantify the impact of bias in a number of data-heavy applications, such as hiring or bank lending, and to find ways to “de-bias” the data.

“Here’s a person who is incredibly curious, and he’s at a university that’s allowed him to pursue that curiosity without bounds,” says Ostrov. “The payoff for that has been tremendous—for him, his students, and the University.”

DEBORAH LOHSE is associate director of media and internal communications at SCU. She was previously a staff journalist at the Mercury News, Wall Street Journal, and Money Magazine. Help fuel curiosity in teaching and learning: campaign.scu.edu.

Make AI the Best of Us

What we get out of artificial intelligence depends on the humanity we put into it.

The Co-Op

Santa Clara University has long been a bastion of interdisciplinary learning. A new fund is taking cross-collaboration to new heights.

Human at Heart

How Santa Clara University is distinguishing itself as a leader in one of the fastest-growing industries in the nation.

A Campus on the Rise

New buildings on campus—count ’em, six in total—aren’t the only changes brought by a successful $1 billion fundraising campaign. Come explore what’s new.