My career path from HFT cop to investment banker to hedge fund PM to data startup
John Collins had a baptism by fire when he started out in finance. As an investigator for NYSE Euronext tracking tick data for suspicious trading, he was working at the exchange in 2007 when the infamous 'quant meltdown' kicked off, which was swiftly followed by the Bear Stearns' and Lehman Brothers' collapse of 2008 and then the financial crisis that followed. After three years, he decided that it was high time he took time out to study an MBA.
“Those events, among others, created a wildly dynamic regulatory landscape that has only recently returned to something approaching a steady state,” Collins, who is a co-founder of artificial intelligence and data startup, Thasos Group, says.
But those experiences also stoked his interest in finance and investment management. Collins decided to enter MIT Sloan's two-year MBA program, where he worked with the Center for Collective Intelligence to assess the impact of social sentiment on the stock market, specifically publicly listed renewable energy companies. He coded a fully automated trading system to test the sentiment signals with live capital.
“My trading system placed real-money trades for me while I attended classes,” Collins says. “After some early successes, I wrote a whitepaper describing my strategy and results and circulated it among my finance and hedge fund contacts, which eventually ended up in the hands of my soon-to-be Thasos co-founders, Greg Skibiski and Wei Pan.
Thasos is an artificial intelligence platform that turns location data from mobile phones into insights that can inform trading decisions. It is one of the plethora of new firms looking to tap into hedge funds' appetite for third-party data to inform trading decisions during the current era of low volatility that is leaving so many hedge funds reeling.
Before landing there in 2015, however, Collins decided to join Credit Suisse's financial sponsors group in New York as an associate. But, despite closing seven transactions during his 12 months there, the idea of returning to coding and trading systems was too much to resist.
“After my investment experience at MIT and on the sell-side, it was clear to me where my interests were strongest," he said. "I had found nothing more intellectually rewarding than transforming large-scale data exhaust from disparate businesses – social media, location and web-scraped data – into actionable alpha signals that increased transparency into the fundamentals of publicly traded companies."
This was in 2013. In the early days, Collins wrote approximately half of the code base for the Thasos technology platform and geofenced stores and other locations of interest. He also managed a $1m portfolio for the company to build a track record on trading its location-based alpha signals.
In 2015, Thasos executed a two-part deal with its first major hedge fund client, a New York-based firm with $10bn-plus under management for project-based research and forecasting work, as well as the creation of a $50m managed account over which Thasos had full discretion. All investment decisions were based on its location data alpha signals.
After closing the deal, Thasos moved from Boston to New York, where Collins became portfolio manager for the $50m managed account with an event-driven equities long-short strategy, while also doing data analysis for the hedge fund. After seven months, the managed account was performing well enough that the hedge fund doubled the AUM to $100m.
There was a fierce debate among the co-founders about whether to commit to being a fund manager and try to grow AUM or make its technology platform and location data insights available to a broad range of clients. They opted for the latter.
Collins transitioned from a location data analyst and hedge fund PM using location data alpha signals to chief product officer catering to the data needs of hedge fund clients.
“Working onsite at the fund for a full year provided me with invaluable daily feedback from other PMs regarding what types of location data signals were most valuable for their investment processes and how the location data should be structured for efficient use,” Collins says.
Photo courtesy of Thasos