Let the games begin. In early September, a group of data-centric traders and aspiring hedge fund managers from different backgrounds gathered to compete in the seventh BattleFin competition. The prize? The chance to win seed money to start their own fund or, at the very least, receive publicity to lure curious venture capitalists to join in with their cash. Some in the media have called BattleFin the Hunger Games for hedge funds, and like the dystopian YA novel, BattleFins multi-tiered competition is fierce. After all, as traditional hedge funds see steady if ho-hum returns of three to five percent, the winners of BattleFin are pushed to produce the next giant fund that beats Wall Street.
Thats the aim behind BattleFin, the brainchild of co-founders Tim Harrington and Brian Tomeo along with co-founder, legal counsel and hedge fund specialist Marc LoPresti. BattleFin is not just its own proprietary fund with plans to become a full-fledged hedge fund this fall; Harrington said he and his team are taking a venture capital approach to the hedge fund industry by incubating a number of start-up managers.
This is pretty impressive when you consider that BattleFin held its first tournament in 2012 with six contestants out of 300 entries. Tournament 5 had more than 9,000 models entered from 43 countries. This month, BattleFin launches Tournament 7 with $25 million in potential allocations from two unnamed investors. We held our Big Data Combine [event] last November where we had people from 43 countries compete. I didnt even know there were 43 countries, LoPresti said.
The contestants, mostly men, do not have the same resumes that you would find on a traditional trading floor: M.B.A. degrees from Ivy League schools and other top-shelf institutions. BattleFin contestants hail from across the globe and from unusual backgrounds and disciplines. Traders last month profiled Tim Fligg, a former chef and current options trader who won his category in BattleFin 6.0 earlier this year and has plans to start his own hedge fund. Another contestant, David Bush, a winner in Tournament 4.0, is a concert musician turned hedge fund manager. Yet another BattleFin winner, Mark Angil, is a former U.S. Navy officer who built nuclear reactors before winning the Sharpe Ratio Shootout of the BattleFin 4.0 tournament. In fact, Tomeo, who helps run the competition, competed in the first BattleFin, where he came in sixth. The former JPMorgan vet now oversees the technology arm of the competition, which powers the leaderboard used to tally the traders results.
Along with finding the next top trader and trading strategy, Harrington said that the aim of BattleFin is to push traders beyond the tried-and-true modes of trading. BattleFin believes that the future of hedge fund investment strategies centers on big data and the managers who are willing to look for data beyond their Bloomberg terminals and Thomson Reuters feeds. Harrington cites hedge fund giants Two Sigma and Renaissance for their investment philosophies.
Harrington noticed that many of todays hedge funds were simply gathering assets, and he estimated that 90 percent of the new allocations are going to the large funds that were seeing 3 to 5 percent returns even though the market was up 30 percent.
It just didnt make sense to us because we saw these small and emerging managers in the sweet spot of their funds performance curve and they couldnt get access to capital. We decided to address this issue, put our flag in the ground and incubate the best strategies we could find, Harrington told Traders. There wasnt a way for institutional investors to really get access to the early stage alpha generators due to fears of operational risk and the amount of time needed to vet these early stage managers. With this in mind, he added, BattleFin created a business model that acts as a virtual funnel to rank and quickly evaluate the newer managers.
We use these leaderboards and then a BattleFin score to rank them, Harrington said.
The best strategies wont be found in traditional market data. BattleFin participants look to sources such as Twitter and other social media platforms, as well as scientific and government reports.
Were looking for disciplined, repeatable, differentiated strategies, Harrington said. We want someone who is looking at the world in a different way. We stay away from guys who rely on sellside research or technical analysis. Were looking for the strategies that leverage data streams and managers who can mine large amounts of data and make sense of it. We pay particular attention to strategies that use weather, Twitter feeds and other data sources that have not been picked over and over-analyzed.
One such big data visionary was a trio of Massachusetts of Institute of Technology professors who formed Flyberry Capital, a small Cambridge, Mass., hedge fund that exploits data from government reports, weather forecasts and other sources. They won the first BattleFin tournament and theyre looking to grow into a larger asset management shop. Flyberry Capitals principals also graced the cover of Traders February 2014 edition.
After beating out the other competitors in their heat, Flyberry Capital CEO Michael Chang and his team began quant trading using a big data strategy. The fledgling fund relies on its Flyberry Engine, a big data search tool created by former Google fellow and fellow MIT graduate Sean Chang, whom colleagues call the Michael Jordan of programmers. The Flyberry Engine scours the Internet for news alerts and trends from sources raging from social media, government agency reports and announcements, and findings from weather, earthquake and other geosensors.
This big data push has led to some dynamic and unorthodox ideas for investing strategies, to say the least. We had one group, an astrophysicist from UC Santa Barbara and a professor of astronomy from the University of Washington, that entered the Launch category with a strategy based on how gases form around black holes and related the models to trading stocks, Harrington recalled. He also mentioned a trading team from California called Manifold Partners that utilizes big data strategies and looks at the clustering patterns that emerge.
Tomeo likens BattleFins mission to grooming traders for the emerging manager space, where a trader has less than three years experience and less than $300 million in assets under management. Were trying to redefine the space like the frontier managers [found in Nigeria, Egypt and Vietnam] where we would give capital to early stage guys that are in the idea generation mode, Tomeo said. We would help them build a company so they could take their idea, test it, trade it, and actually raise assets into it. Or, we look to invest in them ourselves, and they run their own infrastructure themselves.
For Tomeo, its about grooming data miners into investors and traders. There are a lot of people who have the ability to mine gold, but they dont have the operational skills to figure out where to look for the gold, he told strategies that we saw being designed by people who had the data mining skills were just not investable because they were either too volatile – they had too much risk in one asset class – or they werent consistent enough in terms of how they were looking at risk management.
We try to determine the rules and objectives for the data scientists. We point them in the right direction, and then they find the gold. What they come up with and produce is actually something that is sellable and meaningful, because the data analysis is only useful if you can use that to generate a return, Tomeo said.
According to BattleFins website, 10 of the best strategies have been chosen to incubate.
The BattleFin Fund
Harrington and Tomeo not only run the competition, they manage the BattleFin fund, too. Based in Rowayton, Conn., in leafy hedge fund country between Greenwich and Westport, the fund started trading its own money this summer. Harrington declined to say how much it had in assets under management but added that two investors were slated to provide seed capital in early September. Tomeo runs the technology side of the business from BattleFins Miami office, where the technology for the BattleFin leaderboard is based. BattleFin has a team of eight employees who help run the fund, the competition and other arms such research and development.
BattleFins fund will be driven by what he called open source investing, which will allow investors to view its operations as transparently as possible on a daily basis. They can log in at any point and see how their investment is doing, what the fund is doing and what the funds doing relative to the market, he said. This is different than whats traditionally been going on in the hedge fund world, where maybe you get a monthly or quarterly statement that really just has a number on it.
The principals of BattleFin started its fund with their own capital and tested their trading front end, risk management and reporting tools for two months before launching the fund. BattleFin has spent a lot of time and money building out the technology to rank and source these emerging managers. In some cases we have been able to build off systems like Realtick which offers multiprime broker trading capabili-ties. This is important to small managers who may have to trade at mul-tiple places where their investors are located. Harrington said.
Inside the BattleFin Games
Heres how the quarterly tournaments work: The seventh BattleFin tournament will be funded by two anonymous investors to the tune of $25 million. Participants must provide their own money or raise the capital to participate. The most recent tournament began on Sept. 2 and will end on Dec. 1, with winners named on or around Dec. 15. Before then, BattleFin and partner TigerDirect will sponsor a live networking event in Miami called Silicon Beach that will take place in the infield of Marlins Park.
During the tournament, BattleFin looks at each teams track records and their pedigrees. Using data analytics, BattleFin backtests to match their track record plus what weve seen on the live trading environment, according to Harrington. Using the data collected, BattleFin is able to quickly evaluate the strategies that enter the tournaments. Since track records can be short we need to use all the data we can. We drill down to position level detail to see how many trades the managers are getting right and what type of risk controls the managers use. The results are monitored by BattleFins risk officer, Kent Baur.
The contestants are divided into three categories based on their trading experience and size of their particular fund. The Launch category is for new investors to pre-launch their fund, sometimes with paper or non-money trading. If they do well, BattleFin and its unnamed sponsors may invest money so that they can move to the second level. The Professional level requires that participants have a six-month track record of trading with their strategy and around $1 million in assets under management. For the top category, Elite, traders must have $5 to 10 million in AUM and a 12-month audited track record, and must survive BattleFins painstakingly thorough vetting process.
We review the presentations and the data streams, Harrington said. Then we have about 100 strategy-related questions where the managers have to tell us about how the portfolio works, the position sizes, what risk limits do they use, and so on.
Is BattleFin upending the way Wall Street in general and the buyside in particular look for promising new talent? LoPresti is modest despite being impressed with the talent that shows up to compete.
I hate to use the word disrupting because its overused, but we are sort of disrupting the traditional process of asset management talent identification and incubation, he told Traders. I dont think we pose a threat to those shops. We have a very nice opportunity set in terms of the number of people that we can help.
