By QUENTIN HARDY – NYT
MOUNTAIN VIEW, Calif. — Some of the engineers who used to help the Central Intelligence Agency solve problems have moved on to another challenge: determining the value of every conceivable investment in the world.
Five years ago, they started a company called Addepar, with the aim of providing clear and reliable information about the increasingly complex assets inside pensions, investment funds and family fortunes. In much the way spies diagram a communications network, Addepar filters and weighs the relationships among billions of dollars of holdings to figure out whether a portfolio is about to crash.
Professional wealth managers are going to be seeing a lot more of big data. Last spring, Addepar raised a substantial sum to take this mainstream, and although it is not the only one bringing big data to a portfolio statement, its cast of characters sets it apart.
“One of the most foundational questions in finance is ‘What do I own, and what is all of this worth?’ ” said Eric Poirier, the chief executive of Addepar. “ ‘What is my risk?’ turns out to be an almost intractable problem.”
Although the list of wealth managers who use Addepar is confidential, Mr. Poirier says it has already grown from people like Joe Lonsdale, its tech-billionaire founder, and Iconiq Capital, which manages some of the Facebook co-founder Mark Zuckerberg’s money, to include family offices, banks and investment managers at pension funds.
“In this state, some people are just getting wealthier,” said Joseph J. Piazza, chairman and chief executive of Robertson Stephens L.L.C., a San Francisco investment adviser that manages about $500 million using software from Addepar. Ten years ago, he said, “it might be a young entrepreneur with $50 million. Now it could be 10 times that, and they are thoughtful, bigger risk-takers.”
Investing used to be a relatively simple world of stocks, bonds and cash, with perhaps some real estate. But deregulation, globalization and computers have meant more choices. For a wealthy person, this could mean derivatives, private equity, venture capital, overseas markets and a host of other choices, like collectibles and Bitcoin.
And for all the computers on Wall Street’s trading floors, a lot of money management is surprisingly old-fashioned. Venture capitalists may invest in cutting-edge technology, but they sometimes still send out quarterly reports on paper. Financial custodians, which hold securities for people, often have custom-built computer systems. That makes it hard to compare a trade at one with a trade at another.
“The market is much more complicated than it used to be,” said David G. Tittsworth, president and chief executive of the Investment Adviser Association, a trade group of 550 registered firms. “The rich have bigger appetites for futures, commodities, alternative investments. There’s a lot of demand for helping them keep track of what their holdings actually are.”
Mr. Poirier, 32, a New Hampshire native who started a coding business at 14 before heading to Columbia University, worked on analyzing fixed-income products at Lehman Brothers from 2003 to 2006, before that Wall Street firm collapsed from mismanagement of its own risk. “Trying to figure out a yield, I’d work with a dozen different computer systems, with different interactions that people didn’t understand well,” he said.
He then took a job with Palantir Technologies, a company founded to enable military and intelligence agencies to make sense of disparate and incomplete data. He went on to build out Palantir’s commercial business, managing risk for things like JPMorgan Chase’s portfolio of subprime mortgages.
There were plenty of parallels between the two worlds, but instead of agencies, spies and eavesdropping satellites, finance has markets, investment advisers and portfolios. Both worlds are full of custom software, making each analysis of a data set unique. It is hard to get a single picture of anything like the truth.
Even a simple question like “How many shares of Apple do I own?” can be complicated, if some shares are held outright, some are inside a venture fund where the wealthy person is an investor and some are locked up in a company that Apple acquired.
Finance “was the same curve I encountered in the intelligence community,” Mr. Poirier said. “How do you make sense of diverse information from diverse sources, when the answer depends on who is asking the question?”
The parallel was also evident to Mr. Lonsdale, a Palantir co-founder. From an earlier stint at PayPal, he had millions in cash and on paper is a billionaire from his Palantir holdings. He also knew lots of other young people in tech who could not make sense of what was happening to their money. “Wealth management is designed for the 1950s, not this century,” he said.
Mr. Lonsdale left Palantir in 2009, starting Addepar with Jason Mirra, another Palantir employee, in 2009. “It didn’t make sense for Palantir to hire 20 or 30 people to work in an area like this,” Mr. Lonsdale said. Mr. Mirra is Addepar’s chief technical officer. Mr. Poirier joined in early 2013 and became chief executive later that year.
Besides Mr. Lonsdale, early investors in Addepar included Peter Thiel, a founder of both PayPal and Palantir. More money came from Palantir’s connections to hedge fund investors. Addepar’s $50 million funding round last May was led by David O. Sacks — another PayPal veteran, who sold a company called Yammer to Microsoft for $1.2 billion in 2012 — and Valor Equity Partners, a Chicago firm that has also invested in PayPal, SpaceX and Tesla Motors, among other companies.
Despite the pedigree, Mr. Lonsdale says Addepar, which has 109 employees, is not meant just as a tool for rich tech executives or family money. They are, he said, “just the early adopters.”
Karen White, Addepar’s president and chief operating officer, says a typical customer has investments at five to 15 banks, stockbrokers or other investment custodians.
Addepar charges based on how much data it is reviewing. Ms. White said Addepar’s service typically started at $50,000, but can go well over $1 million, depending on the money and investment variables involved.
And in much the way Palantir seeks to find common espionage themes, like social connections and bomb-making techniques, among its data sources, Mr. Lonsdale has sought to reduce financial information to a dozen discrete parts, like price changes and what percentage of something a person holds.
As a computer system learns the behavior of a certain asset, it begins to build a database of probable relationships, like what a bond market crisis might mean for European equities. “A lot of computer science, machine learning, can be applied to that,” Mr. Lonsdale said. “There are lessons from Palantir about how to do this.”
A number of other firms are also trying to map what everything in a diverse portfolio is worth. One of the largest, Advent Software, in 2011 paid $73 million for Black Diamond, a company that, like Addepar, uses cloud technology to increase its computing power and more easily draw from several databases at once.
“We’ve been chipping at the problem for 30 years,” said Peter Hess, Advent’s president and chief executive. “There is a lot more complexity now, and the modernization of expectations about how things should work is led by the new tech money. But because of Apple and Google, even my parents have expectations about how easy tech ought to be.”