CME Group, the world’s largest operator of derivatives exchanges, ended an era with the closure last summer of its remaining open-outcry futures trading pits. The move, an acknowledgment of the near-total changeover in capital markets to electronic trading, was one that many in the financial community knew was inevitable but nonetheless was jarring on the warm July day when the few lingering traders who still gathered in the downtown Chicago pits decamped for the last time.
CME’s massive riverfront headquarters is like a museum of past trading eras, some of which overlap in the present. On the floor where CME’s raucous trading pits once operated is the Global Command Center, a cavernous, calmly humming space where a smattering of technologists stare at screens, monitoring the company’s global technology infrastructure. A full half of the Global Command Center is occupied by Globex, CME’s electronic trading platform, which is in large part responsible for displacing the live traders.
Just a few floors away is the Innovation & Acceleration Lab, created to push CME into yet another era. Opened in 2013, the lab is a fitting approximation of the start-up culture CME has recently decided it wants to channel. The company’s technologists, security experts and other employees gather there to meet with clients or outside experts to discuss challenges and their possible solutions. The lab’s Easter-egg color palette, shining whiteboard walls covered with cut-out clouds and artfully arranged Post-it notes, and well-stocked kitchen help establish its kinship to the world depicted in HBO’s Silicon Valley.
CME Group CEO Phupinder Gill sees that his industry is evolving at an ever-increasing rate. The central element of the innovation effort he has unfurled represents an attempt to leapfrog his company an era or two ahead of the pack. In 2014, CME launched a wholly owned fund called CME Ventures. Unlike the Innovation Lab, CME Ventures isn’t designed to answer today’s — or even tomorrow’s — problems but, as Gill puts it, those that could appear “the day after tomorrow.”
“We see the pace of change becoming faster and faster,” says the 55-year-old CEO, who goes by “Gill” because of the difficulty most people have pronouncing his first name. “We saw an acute need to look beyond just the next year or two to understand how the world is evolving around us.”
Gill is positioning CME to take an informed view on the direction in which the future is hurtling. As technological progress both within and outside the trading industry continues at an ever-quickening clip, the new in-house venture fund functions partly as insurance policy, partly as crystal ball: CME Ventures’ eight portfolio companies represent a range of technologies and applications that some in the trading industry find surprising and confounding. Whereas some portfolio companies have clear applications to CME’s business, others edge closer to the realm of science fiction — and that is absolutely by design.
One of CME Ventures’ portfolio companies, San Diego–based Nervana Systems, is applying deep learning, a form of artificial intelligence, to complex analytical problems. Another, Vancouver, Canada–based 1QBit Information Technologies, develops software for quantum computing, a technology still in its infancy that could eventually operate at speeds that would render current machines obsolete.
Three dedicated employees run CME Ventures, including executive director Rumi Morales. The trio reports to Robert Zagotta, CME’s head of strategy and execution. CME Ventures also relies on a five-member investment committee of employees pulled from divisions across the company to approve investment decisions. The CME Ventures team researches roughly ten to 20 companies per week; Morales estimates that for each of the eight companies in which the fund has invested as much as $5 million apiece, her group has looked into about 50 strong possibilities.
Rumblings of the desire to stay on top of transformational innovations have recently emanated from other major exchange operators. Chicago Board Options Exchange expanded the purview of its chief strategy officer in 2013 and has recently taken stakes in trading-related companies. German exchange operator Eurex Group is hoping to invest soon in a venture capital fund that will provide an arm’s-length connection to stakes in fintech start-ups. Jeffrey Sprecher, CEO of Intercontinental Exchange (owner of the New York Stock Exchange), says he serves as a board member of the Georgia Institute of Technology in Atlanta in part because it offers him a close view of the university’s technology commercialization and entrepreneur center.
Though these companies share the urge to innovate, CME is the only one among them that has created an in-house corporate venture arm. That seemingly slight variation in its innovation-capturing strategy is significant.
“CME Ventures is not doing what you’d expect from a strategic investment group in this space; they’re not going out and just investing in a bunch of trading and derivatives-related start-ups,” says Mark Longo, a Chicago-based former options trader who now runs the Options Insider, a website that offers options users free news and analysis. “They’re taking a longer view. They’re doing a lot of esoteric stuff — things that may or may not accrete to their bottom line in the next five years or even a decade.”
Longo notes that CME may have the luxury of taking a long view in part because of its dominant position in futures trading. Many of the company’s peers, especially those in the crowded options-trading market, are under more pressure from shareholders to swiftly demonstrate the results of investments and acquisitions.
Of course, CME also faces pressure to innovate. The 2011 implosion of futures brokerage MF Global Holdings led to a decrease in trading volume and revealed CME’s dependence on futures trading. Although CME’s stock price is up about 80 percent since Gill became CEO in May 2012, trading at $93 a share in mid-December, it is still well below the high of $141 it reached in late 2007. The firm also has had to deal with the sudden departure of the executive who helped spark the CME Ventures idea and ran the operation before leaving in April 2015 to become chief executive of one of its portfolio companies.
CME Group has had a long and sometimes fraught relationship with innovation. Over its centurylong life the Chicago Mercantile Exchange (CME’s predecessor) prided itself on developing new futures products, in an effort to unsettle the dominant position of its larger and older crosstown rival, the Chicago Board of Trade. But over most of the 20th century, nothing altered the way both exchanges’ derivatives products were traded. A time-traveling trader jumping from the CBOT’s 1880 open-outcry pits to a 1980 version might notice some confounding fashions and newfangled electronic boards, but otherwise the scene would be largely the same. The trader members of CME and CBOT, who owned their respective firms, appreciated that constancy.
In 1988, CME entered into an agreement with Reuters Holdings that would eventually — dramatically — alter the trading scene. The two announced they would jointly develop a new, 24-hour electronic trading network called Globex that would allow CME traders and their clients to buy and sell financial futures and options contracts on Reuters terminals when the Chicago-based trading pits were closed.
CME traders, along with those at CBOT, hated the idea. Michael Gorham, who joined CME in 1979 to run the commodities research group and left in 1997 as vice president of international market development, says a line he heard again and again from CME traders discussing Globex was “If it ain’t broke, don’t fix it!”
“They felt floor trading was superior,” adds Gorham, who now runs the Illinois Institute of Technology’s Center for Financial Innovation. But certain higher-ups at CME were eager for the future they were inviting. The Globex project was spearheaded by former CME executive chairman Leo Melamed, who said at the time that the partnership would usher in “a new era of futures and futures options.” He would be proved correct, but it was a slow ushering and one that many in the industry resisted, especially in the U.S. By 1997 only 1 to 2 percent of CME’s trading was done via Globex. Fully electronic exchanges were accepted more quickly in Europe. Electronic German exchange Eurex appeared on the scene in 1998 with the merger of German derivatives exchange Deutsche Terminbörse and Swiss Options and Financial Futures Exchange (SOFFEX). By 1999, Eurex had supplanted CBOT as the world’s largest derivatives exchange.
In 2001, CME surpassed its Chicago rival to become the U.S.’s No. 1 futures and options exchange, largely because its members had begun to fully embrace electronic trading. By 2007, CME had amassed the size and clout to attempt to purchase its onetime chief competitor. Gill played an integral role in driving what proved to be a transformational merger. Born in Malaysia and raised in Singapore, Gill had joined CME in 1988 as a trading clerk in the Canadian-dollar pit and from there moved to the firm’s clearinghouse, where he rose through the ranks for 16 years. By 2007 he was president of CME Group. It was in this role that he made the phone call to CBOT’s largest shareholder, which had voted against CME’s merger offer, to keep the deal from dying. That year CME paid $11.9 billion for CBOT.
CME Group continued its push toward single-handedly consolidating its industry with the purchases of the New York Mercantile Exchange for $9.4 billion in 2008 and the Kansas City Board of Trade for $126 million in 2012. Gill took over as CEO in May of that year. It was a heady time for the firm: Volumes had reached 2.9 billion contracts traded in 2012, compared with 1.3 billion in 2006, and annual revenue had almost tripled, to nearly $3 billion, in the same period.
But Gill had inherited formidable challenges from his predecessor Craig Donohue. The collapse of futures brokerage MF Global in October 2011 had shaken client confidence — $1.6 billion in client money was missing from the imploded firm’s accounts — and raised the specter of new regulations.
Upon stepping into the CEO role, Gill shuttled around the world to meet with investors in the U.S., London, Tokyo and Singapore. He noticed that his clients’ concerns could be summarized in two central questions: Where would the group generate future growth, and what would happen to CME and its customers as a result of the ongoing regulatory changes?
Gill believes CME’s growth must come from outside the U.S., with the best opportunities in Asia, especially China and India. Today, CME has clients in 189 countries, but 70 percent of its revenue still comes from the U.S.
To deepen CME’s global client base, Gill says, the company must do a better job of understanding the rapidly evolving needs and expectations of derivatives traders, and how they differ among markets. Specifically, he wonders how CME can help investors slash the two days it takes to settle a foreign transaction and how it can help customers more seamlessly stay abreast of regulatory compliance, risk management and capital requirements. In pursuing technological solutions, Gill stresses, CME must do more than just create new products. If another Globex-level disruption is in the offing to meet these needs, he wants to be the one behind it.
“In terms of reaching a client base, technology is a great enabler,” he says. “In enhancing any kind of relationship with any kind of client on almost any basis, whether it’s front office, middle office or back office, we are beginning to realize that technology must be the one element that binds all three of these things together.”
Gill says in 2000 he and CME’s other directors held up cell-phone maker Nokia Corp. as a company whose capacity for technological innovation their firm should emulate. But a few years later, Nokia was displaced by Motorola, then Motorola by Research in Motion, maker of BlackBerry devices. But, Gill says, “they never kept up with that innovation.” He started to instead conceive of the trajectories of these companies as cautionary tales. “You have to start thinking in an urgent way, ‘What might be next?’” he says.
By early 2013, Mark Fields, a CME managing director working in mergers and acquisitions, corporate finance and development, found himself entertaining similar questions, and he took them to Gill. Fields had been reading technological futurists like Ray Kurzweil, as well as Nicholas Carr’s The Big Switch, about the revolution portended by cloud computing. He wondered what the technological future could hold for his industry and whether calling up folks in Silicon Valley — as he’d done in a previous role as a corporate and investment banking group relationship manager at TD Securities — to see what was brewing within the start-up world could provide a clue.
“I made the case that if you want to know what’s possible, you can go to the universities,” Fields says. “But if you really want to know what’s commercial — and what’s keeping people up at night — go look at what the VC community is investing in.’”
Fields suggested CME establish a venture capital fund to “seek out brave new worlds,” as he puts it, and make strategic investments in the companies developing them. Gill liked the idea. The new CEO told Fields he’d have adequate resources to ensure the project was effective, but he imposed a stipulation: “Make sure you bring these learnings back into the firm to make us smarter.”
CME’s board moved swiftly. In November 2013, CME Group approved the creation of a fully owned subsidiary called Liquidity Ventures I. It was established in March 2014. The board authorized initial seeding in the tens of millions of dollars, with the understanding that the company would continue to seed the fund every year as part of its budget process.
From the start the operation — later renamed CME Ventures to align itself with high-profile corporate venture arms like Google Ventures and Citi Ventures — has behaved more like a generalist high-tech venture capital fund than like an exchange operator’s corporate fund focused on financial technology and other immediately applicable trading innovations. The venture capital structure institutionalizes a focus within CME on a part of the tech landscape that Gill considers particularly vexing: the transformational innovations still too nascent to be commercialized or acquired. CME Ventures’ sweet spot is technology that could disrupt the financial services industry in three to five years.
This is where CME’s approach deviates from the strategic investments and innovative initiatives of other exchanges. Thomas Ascher, chief strategy officer at Eurex-owned International Securities Exchange — the first fully electronic U.S. options exchange — has overseen strategic venture investing for his firm since 2005. But unlike CME Ventures’ portfolio, ISE’s three strategic investment stakes are in technologies with direct applications to trading.
John Deters has served as Chicago Board Options Exchange’s chief strategy officer since late 2013. He believes that as innovation has become more broadly distributed, firms like his need to be more outward-looking to capture potentially disruptive developments. Under his purview CBOE has taken strategic stakes in established trading technology. But Deters is not looking for the day-after-tomorrow developments that CME Ventures seeks.
ICE’s Sprecher has yet another philosophy on innovation. The Atlanta-based exchange operator buys start-ups and sets them up to develop their innovations in-house. “We have a lot of little groups in here that we give a lot of leeway to in order to try new things that are disruptive,” says Sprecher, who was personally involved in the NYSE’s minority investment in Bitcoin wallet company Coinbase in January 2015.
The idea at CME Group was that CME Ventures would be the net to capture the best of the innovations being developed out in the world. With Fields at the helm, CME Ventures began taking stakes of between $500,000 and $5 million in portfolio companies, racking up four investments during his tenure.
The line between entrepreneurs and venture capitalists can prove permeable, and in April 2015 Fields crossed over. He left CME to become chief executive of San Francisco–based Wickr, developer of a secure communications app built on military-grade encryption technology and CME Venture’s first investment.
The departure somewhat rankled Gill, but the CEO makes light of it now. “We still see [Fields] more than we want because he’s running one of the portfolio companies,” he half-jokes. Fields says Gill warmly wished him the best when he announced his decision to leave the company.
Morales, executive director of CME Ventures since its founding, took over the fund’s operations after Fields left. At the beginning of her career, in the mid-1990s, she spent two years at private venture firm Petra Group and three years as a principal at Beechtree Capital, where she negotiated new venture capital investments and managed $220 million in debt and equity transactions.
Morales is quick to point out that nothing changed within the fund, strategically or otherwise, after Fields left. Most recently, CME Ventures has been focusing on start-ups that offer promising solutions in real-time payment technologies, precision big data, security, deep learning and next-generation computing, though its areas of interest are constantly evolving. With each investment the fund is making a bet that a given company’s technology will not only prove successful but that it will address the client sticking points that Gill identified after taking the job.
For nearly as long as he can remember, Naveen Rao has been obsessed with how the brain solves problems. By age nine he was tearing through science fiction about artificial intelligence and starting to learn how information is stored in circuits, both neural and artificial. After earning his BS in electrical engineering and computer science at Duke University, he spent ten years designing processors and specialized chips for a series of Silicon Valley start-ups and tech companies, but he still wanted to know more about how the brain solves problems and processes data. Rao figured that if he better understood the brain’s method of parallel distributed computation — that is, how it operates through many tiny, interconnected computational units called neurons — he could vastly improve the synthetic systems that serve as computers’ brainpower. In 2007 he left Silicon Valley for Brown University to pursue a Ph.D. in neuroscience.
In 2011, doctorate in hand, he joined brokerage and fintech firm Investment Technology Group to work in algorithmic trading. He was eager to bring new computational approaches to ITG’s processes and products, but despite “pushing that agenda quite a lot there,” he says the company wasn’t ready to hear it.
Rao, however, was more than ready for the changes he saw converging in his area of expertise. In February 2014 he co-founded Nervana to build an infrastructure for synthetic neural networks that can make sense of data — the futuristic computational niche dubbed deep learning.
CME Ventures added Nervana to its portfolio in May 2015, soon after Morales took the helm. The power of deep learning may eventually prove transformational for exchanges, helping them locate fraud and abnormalities, not to mention develop new data-driven products to sell to clients.
This capability is desperately — and increasingly — needed within capital markets. As trading volumes have exploded over the past decade, the opportunities to commit fraud and make mistakes have dramatically increased. The May 2010 “flash crash” demonstrated that one trader with an intimate understanding of trading algorithms has the ability to bring a mighty exchange like CME screeching to a stop. In that case a trader placed an enormous order for E-mini S&P 500 stock index futures contracts that he planned to cancel, creating a dramatic market imbalance, halting CME trading on the E-mini and causing a brief stock market crash. The CME is eager to avoid another such disruption.
Nervana meets weekly with a data science team at CME, sifting through the exchange’s hundreds of terabytes of daily data and discussing which types of neural networks might find usable patterns within it. “When [CME] first approached us, we didn’t think they’d be quite as forward-thinking as they are,” Rao says.
CME Ventures’ investment in 1QBit underscores the venture fund’s determined forward view. The Canadian company develops software for quantum computing hardware that’s still in its infancy. CEO and co-founder Andrew Fursman studied financial engineering at Stanford University and the future of technology at Singularity University, a Silicon Valley–based accelerator and educational organization. Through his small, British Columbia–based venture capital group, Minor Capital, Fursman took a stake in D-Wave Systems, a quantum computing device manufacturer. In 2012, Fursman and a partner from Minor Capital approached D-Wave to propose a new start-up focused exclusively on applications for quantum computers. They got the green light, and 1QBit was born.
D-Wave’s machines — currently, the sole commercially available quantum computing devices — can run only when the temperature of their environment is kept close to absolute zero, or –459.67 degrees Fahrenheit. Nonetheless, their potential uses have already captivated visionaries in many industries. Quantum computers solve problems using a process called quantum annealing, which does not depend on classical mathematical functions. Instead, it takes advantage of the fact that each quantum bit, or qubit, holds two pieces of information. Once quantum computer chips expand their size and power, this process opens up the possibility of solving complex optimization problems for which classical computers can only find approximations.
Recently, 1QBit collaborated on a white paper to describe one application of quantum computing for finance, providing a potential solution for what is called the optimal portfolio trajectory problem: Is it possible to consider not just today’s information for determining the best asset allocation but to take into account best guesses for the future, thereby avoiding the transaction costs involved in rebalancing each day? Quantum computing software could allow this.
CME took a stake in 1QBit in the summer of 2014. Bluford (Blu) Putnam, CME’s chief economist and head of its data science operation, sits on 1QBit’s board. Putnam says that over the course of his 38-year career he has spent considerable time working with dynamic Bayesian systems (which apply a specific interpretation of probability) to manage portfolios and assess financial risks, but that quantum computing could quickly make those methods obsolete.
“By and large, I rejected many interesting paths to portfolio optimization and risk management just because they were incredibly hard to solve, if not impossible with the tools at hand,” Putnam says. “It is a mind-set change to now seek out the nearly impossible to solve problems [with quantum computing] rather than to avoid them as a waste of research time.”
For CME portfolio company Dwolla, innovation is centered on providing an immediate solution to a problem that the exchange operator, its clients and the rest of the world have long lived with: the delay in transferring money between financial institutions. Dwolla resulted from the frustration of its co-founder and CEO, Ben Milne, over web credit card fees when he was running a company that manufactured speakers. Milne sold that business and launched Des Moines, Iowa–based Dwolla in late 2010.
Dwolla’s digital payment network transfers money between banks instantaneously and securely. For the moment, its technology moves only dollars, though Milne says he plans to add other currencies and types of assets. Bitcoin operates on a similar infrastructure, called blockchain, but the blockchain ledger is by design a totally transparent and permissionless means of storing information. Dwolla’s system allows financial institutions and customers to decide who sees their payment and account information.
Transferring assets in real time could mean opportunities to increase liquidity and reduce risks for clients of exchange operators like CME. In October, just over a year after investing in Dwolla, CME announced that it had established a commercial partnership. The first application of Dwolla’s technology within CME will be in its broker and clearing firm payment systems. By integrating Dwolla’s technology, CME will allow its clients to process payments faster.
Portfolio company Ripple, like Dwolla, is using a proprietary distributed ledger technology to move funds in real time. Morales says she views Ripple as complementary to Dwolla, though the San Francisco–based company has positioned itself more as settlement infrastructure than as the payment system Dwolla considers itself to be. Ripple is helping to solve the problem of delays in international settlements by tackling the issue caused by the existence of many correspondent banking networks, all built on different types of systems.
Though real-time settlement may be the most immediate problem within CME Group, security is likely to be the most persistent concern of CME Ventures’ investment committee. “CME is a company that has to be 99.99999 percent reliable because you’re protecting client information and data in transactions that have an impact on the world economy,” says Fields, CEO of Wickr, in which CME first invested as part of a $30 million Series-C funding round in June 2014. Fields acknowledges that he’s no cryptography specialist, but he understands security needs from an enterprise perspective: “To me Wickr is the next generation of encryption technology that is an essential part of the defense architecture” at any company.
CME Ventures added Powerlytics to its portfolio in October 2014 for a foothold in what Morales calls precision big data. Powerlytics’ CEO Kevin Sheetz is a former partner at KPMG, where he developed an inside understanding of the data that’s available to financial services firms — and of where the gaps are. Sheetz co-founded the Doylestown, Pennsylvania–based company in 2011 to offer much more granular, reliable data on all sectors of the U.S. economy. Powerlytics integrates 20 to 30 different governmental data feeds that can offer insights into historical performance in the markets over the past 15 years.
In late October, CME Ventures announced a new investment, in New York–based Digital Currency Group, whose mission is to build a better financial system using blockchain and digital currencies. DCG manages a portfolio of 56 digital currency companies in 17 countries. A few days after the DCG announcement, the corporate venture capital fund struck again, adding to its roster Fortscale, a San Mateo, California–based, behavior-analytics-focused security start-up that combines expertise from the Israel Defense Forces’ elite security unit, big-data analytics and machine learning.
As CME Ventures’ portfolio continues to grow, it offers insight into the future direction of the exchange operator. “As we look forward to a truly globalized and interconnected world, we’re trying to see what clients want,” Gill says. “We’re looking at the value chain in which our clients operate and understanding how it might change, and we’re investing in those changes now instead of three to five years from now, when it might be a little late.”
