It has been said that data are to this century what oil was to the last one—a driver of growth, change and disruption. The financial services industry is at the forefront of this trend, and new technologies, new applications and even new forms of intelligence are emerging to help financial companies make the most of this new resource.
The growing interest in harnessing the power of data was clearly visible in the most recent Innovators Pavilion, FIA's annual showcase for fintech startups. More than half of the companies that participated in the 2017 Pavilion are connected to the new data economy. Some specialize in extracting new data from non-traditional sources such as satellite images and social media. Some provide innovative tools for making sense of the data that firms already have. Others provide platforms for organizing data more efficiently so that more information can be analyzed more quickly and with greater precision.
Data is not the only area where innovation is emerging in fintech, however. The 2017 cohort included two early-stage companies in the bitcoin and blockchain space and three companies in regtech, a branch of fintech that specializes in regulatory and compliance solutions. Closer to home for the derivatives industry, several companies came to the Pavilion with innovative solutions designed for exchange connectivity, quantitative trading of futures, and post-trade reporting.
This article profiles each one of the 19 companies that participated in the 2017 Innovators Pavilion, describing the solutions they offer and the potential value for the derivatives industry. Some of these companies are at a very early stage of development and are just beginning to roll out their products. Others are more advanced, with several years of sales, funding from venture capital investors, and partnerships with well-known institutions in financial services. What they have in common is an innovative approach to solving problems and the potential to drive growth, change and disruption in derivatives markets.
This year's Pavilion included three companies that have found innovative ways to extract data from nontraditional sources and use that data to provide insights for market professionals.
Prattle, a Boston-based company founded in 2014, has developed a patent-pending combination of natural language processing, sentiment analysis and machine learning. Its first application was to "Fed speak," the arcane art of deciphering statements by Federal Reserve governors and other central bank officials. Rather than looking for key words and generating a simple positive/negative signal, Prattle’s system evaluates all the words within every communication to assess the tone of each speaker and distill the meaning of their statements. The system then generates a quantitative sentiment indicator that can be used to predict a statement's impact on fixed income, foreign exchange and other financial markets. The company's system currently covers more than a dozen central banks and has been recognized by authorities such as the Federal Reserve Bank of San Francisco for its accuracy.
Applying natural language processing, sentiment analysis and machine learning to Fedspeak and earnings calls.
In September 2017 Prattle extended its business model to the equity markets by adding real-time analyses of earnings calls from nearly 3,000 publicly traded companies. The company has created a unique lexicon for each company and uses that information to quantify the market impact of an earnings call based on the historical relationship between the company’s language and corresponding market activity. “Analysts and market participants are moving from traditional, opinion-based decision making to quantitative or quantamental analysis," said Evan Schnidman, the founder and chief executive officer of Prattle. "Our technology empowers users to efficiently and cost-effectively run quantitative models or conduct fundamental research that results in optimal investment decisions.”
Predata takes an innovative approach to extracting insights on political risk from social media. The New York-based company was founded in 2015 by James Shinn, a former senior official at the CIA and the Defense Department. He set out to build a system to collect digital conversations from all over the world and then generate predictions about the probability of certain types of events, such as civil protests, labor strikes, election results and terrorist attacks. Rather than trying to parse the meaning of each message, Predata focuses on the metadata—the number of messages in a conversation, for example—and uses that information to generate predictive signals. The company scrapes metadata from tens of thousands of digital conversations on social and collaborative media platforms, including Twitter, YouTube, Reddit and Wikipedia, and measures the volume and intensity of those conversations against a database of historical events. This allows Predata to ascribe a numerical value to the near-term probability of events. Despite being little more than two years old, the company has already signed up a number of hedge funds and government agencies as clients, and teamed up with Bloomberg to broadcast a set of political volatility indices on its terminals. The company also received $3.25 million in funding from four venture capital investors in July 2016. Chris Sugden, managing partner at Edison Partners, one of the firms that participated in the funding round, commented that Predata "is revolutionizing the predictive analytics landscape" with its unique use of social conversation data and machine learning.
TellusLabs aims to derive meaningful insight from satellite data. The Boston-based company combines decades of satellite imagery with a machine learning platform to answer critical and time-sensitive economic questions. The company focused initially on the agriculture industry, providing crop yield estimates for U.S. corn and soy harvests that have been as accurate as the official USDA estimates but produced weeks ahead of the government. The company is now rapidly expanding the scope of its platform with the goal of building a truly global view of grain production, and has begun working on similar analytics for other crops such as sugar, cotton and coffee.
Revolutionizing the predictive analytics landscape with a unique use of social conversation data and machine learning.
David Potere, the former U.S. navy officer who co-founded the company, explains that what sets TellusLabs apart from other remote sensing companies is that it uses images from government satellites rather than commercial satellites. Those images are better for analyzing plant growth not only because they go back farther in time but also because they are captured more frequently and include light from beyond the visible spectrum. A human eye may not be able to see infrared and ultraviolet light, but machines can, and TellusLabs uses data from the full color spectrum to generate its predictions of agricultural production. Mark Friedl, the other co-founder, is a professor at Boston University's earth and environment department and one of the world's leading experts on the use of remote sensing to study biogeophysical patterns and processes on the surface of the Earth. The team also brought in agronomists to identify the key "vital signs" for plant growth and build those into the company's machine learning model. That allows the model to answer questions such as what colors provide the best information about soil moisture.
"The agricultural commodities space has yet to have a Moneyball moment," commented Fernando Rodriguez-Villa, the company's head of business development. "There are literally quadrillions of data points of satellite data from the past 15 years that can be used to paint a nuanced picture of the health and progress of the global agricultural balance sheet, but no one has built the right easel. At TellusLabs, we are building that easel and proving that it works by painting a couple of paintings ourselves."
Gathering new data is only the first step of the process. A bigger challenge for many financial firm is making sense of the massive amount of data already available. Three of the companies in the 2107 Pavilion are aiming to address this problem by building platforms that harness the latest advances in artificial intelligence, natural language processing and highperformance software engineering.
Boston-based Amenity Analytics specializes in helping its customers pull information out of any type of text. The company's technology combines the most recent advances in machine learning with proprietary natural language processing technology developed by one of its co-founders, Ronen Feldman, who is recognized as one of the world's foremost experts in text mining. Load a mass of documents into the system, and Amenity's "visual information extraction platform" will pull meaningful insights from phrases and patterns in those texts. For example, financial users can use the system to sift through conference call transcripts and SEC filings and identify key players, events, relationships, and sentiment.
Democratizing world-class natural language processing with a self-correcting artificial intelligence platform.
Amenity was founded in July 2015 and raised $7.6 million in August 2017 from early stage investors, including the venture capital arm of Intel. “Amenity Analytics is rewriting the rulebook on analytics and democratizing world-class natural language processing with its self-correcting AI platform,” an Intel executive commented in October when the funding round was announced.
Iguazio has developed a "unified data platform” that is well-suited for companies that need to process huge amounts of data in real-time, particularly if they are investing in time-sensitive applications of artificial intelligence and machine learning. The company's name comes from an enormous waterfall on the Iguazu River in South America, which the company's founders see as a metaphor for the volume, velocity and variety of data available for analysis in today's environment. The Israeli company was founded in 2014 by several veteran tech entrepreneurs and has received funding from several large strategic investors, including CME Group, Dell and Verizon. Its technology combines an innovative high-performance architecture, a cloud-based platform-as-a-service model, and an interface that enables users to manage and consume data without the need for IT involvement. That is a big differentiator from other platforms, the company's executives say. Users of the iguazio platform do not need an in-depth expertise in artificial intelligence; they just load their models into the platform and capture the data.
Market surveillance is one example of a use case in finance. Rather than using a simple rules engine, an exchange can use iguazio's platform to support machine learning models that can process large amounts of streaming and historical data. This improves the ability to detect attempted manipulation in real-time with many fewer false positives, the company says. In addition, the platform allows for the rapid updates and the deployment of new models without requiring long engineering cycles.
The third company, ForwardLane, has a more focused mission: using artificial intelligence to help wealth managers and financial advisors provide better advice for more clients. The New York-based company is targeting a common problem for front office staff—coping with a blizzard of information relevant to their clients' portfolios, including investment research, real-time news, market data, product briefs, ratings changes, and compliance updates. ForwardLane addresses that problem by using AI to sift through the information in real time and generate talking points and investment insights tailored to each client. "In essence, we help financial service professionals overcome cognitive overload, helping to get the right information to the right people at the right time," says Nathan Stevenson, the company's founder and chief executive officer. ForwardLane's system is based on an advanced natural language processing and reasoning platform that aggregates, organizes and synthesizes information from automated feeds and then generates quantitative portfolio insights and investment recommendations. Although the system is not aimed specifically at the derivatives industry, Stevenson says it can help bring more investors into the derivatives markets by making it easier for advisors to understand what drives the performance of these products and how to explain them to their clients.
Another way to make sense of data is to create new ways to visualize that data—creating images that reveal patterns, trends and anomalies. Two companies in the 2017 Pavilion are building innovative visualization tools designed specifically for financial applications.
London-based Financial Network Analytics uses network algorithms to create dynamic maps of the connections within complex financial systems. The company's founder, a Finnish economist named Kimmo Soramaki, is an expert at applying network theory to finance. He spent the first part of his career working at central banks, building models to show connections within markets and payments systems. His work proved its value during the financial crisis, when he helped policymakers get a better understanding of the chains of relationships that can create systemic risk in the financial markets. In 2014 he founded FNA to leverage this work. The company has built a data analytics and visualization platform that automatically translates complex financial data into visual maps and dashboards, which substantially reduces the time, cost and resources required for analysis. Its platform is designed to capture any type of structured financial data, including trades, payments, asset prices, and equity volatilities. Soramaki says the platform has been used by regulators, market infrastructure operators and market participants to answer questions in such areas as stress testing, market surveillance, interbank contagion analysis, and liquidity risk analysis.
Virtual Cove takes the emerging technology of virtual reality and applies it to the problem of understanding market data. The company, which is based in the town of Natick just outside Boston, gives its customers the ability to see data in three dimensions instead of two. Rather than looking at a chart on a screen, a customer can hover over data displays in three dimensional space and explore the data from all angles, using the same 3D googles that consumers use for video games. Bob Levy, the company's co-founder, compares using the technology to driving a car. Just as a driver can process many factors at the same time, such as the curve of the road, the movement of other cars, the speedometer and the rear view mirror, Virtual Cove's technology allows market professionals to apply their spatial awareness to data exploration. Its immersive data visualization solutions portray high data volumes and many relationships in a single multidimensional view. Levy sees the technology as being especially valuable for quants, analysts, portfolio managers and traders by helping them quickly detect anomalies and outliers.
Three of this year's cohor t of startups are focusing specifically on improving the trading process. California-based Quantiacs aims to break open the doors to quantitative trading by making it easier for anyone with programming skills to develop their own algorithmic trading strategies. The company's platform allows users to write the code for an algo and test it against up to 25 years of market data for more than 80 futures contracts. The company provides access to the platform for free; the profits come from identifying successful trading strategies and then offering the developers access to capital from institutional investors. Once a strategy is backed by capital, Quantiacs takes 20% of the returns, and then splits that 50/50 with the developers. The company launched in 2014 with a competition at Stanford University and has been growing by leaps and bounds ever since. Martin Froehler, the company's founder, says that more than 8,000 quants have contributed more than 2,500 algorithms to the platform over the last three years.
Most futures contracts trade on only one exchange, but that's the exception to the general rule in financial markets. In most markets, traders and investors need to have access to many trading venues. That is where TransFICC steps in. The London-based company provides a single point of access to more than 200 venues in the fixed income world. Its technology is designed to translate different API standards to a common format, and uses simple binary encoding and high-performance messaging to ensure fast processing of market data and order messages.
Providing a single point of access to more than 200 trading venues across multiple types of markets and trading protocols.
For banks and asset managers in these markets, connecting to the TransFICC API costs much less than connecting to all those venues separately and keeping up with the constant stream of updates to the interface. The range of venues includes bond platforms such as Brokertec, MarketAxess and MTS, futures exchanges such as Eris, and swap execution facilities run by BGC, Tradeweb, Tradition and trueEx.
Getting the trade done is only half the story, however. Then there's the post-trade process. Theorem Technologies, a Chicago-based company launched in October 2017, offers a suite of cloud-based tools to streamline and automate several key middle-office and back-office processes. The software was originally built by Thales Trading Solutions, an introducing broker set up by several former Newedge executives, for internal use with its clients. The software turned out to be so popular that Thales spun off the business unit into a separate company. Theorem's software covers reporting, analytics, trade flow, automation, reconciliation, aggregation, data transformation, and straight-through processing. It can consolidate futures, foreign exchange, and other trading data across multiple counterparties, helping even small end-users manage a multi-broker environment without significant incremental cost per broker. That fits in well with an industry-wide trend toward using more brokers to execute trades and an increased regulatory focus on accurate reconciliation of fees and trade records.
Swap Market Innovations
Two companies in this year's cohort of startups are focused on the over-the-counter derivatives markets. Synswap, a London-based company started in 2016 by Sophia Grami, a former derivatives trader at BNP Paribas, is seeking to leverage distributed ledger technology to build a better infrastructure for credit default swaps and other OTC derivatives. Its platform is built on Hyperledger Fabric, a blockchain framework hosted by The Linux Foundation. It is designed to support a peer-to-peer network run by its members, where each trading counterpart contributes nodes and deploys smart contracts to process its derivatives trades. Once a credit default swap is booked, smart contracts model the payoff and several steps of the post-trade workflow. The company also is building a clearing solution, where the risk mitigation techniques applied today by clearinghouses are replicated and automated via smart contracts. Synswap is still in an early stage of development, but its prototype has been received favorable reviews from market regulators and accelerator programs including the Innovation Hub run by the U.K. Financial Conduct Authority and the Barclays accelerator program in New York.
Exploring applications of distributed ledger technology to swap trading workflow.
Quantile Technologies, also based in London, solves a more pressing problem. The company's swap compression technology helps market participants reduce the number and value of their swap positions by matching offsetting trades and then netting them down. That is a key issue for banks looking to mitigate the impact of the Basel III capital standards on their derivatives businesses. The company's founders, Andrew Williams and Stephen O'Connor, previously worked in senior roles at Morgan Stanley and can draw on a deep understanding of the business workflow and an industry-wide network of relationships. Although they are competing with several other providers of compression services. Quantile says it has an edge because its modelling approach, which combines linear and non-linear programming and machine learning techniques, produces more efficient compressions. In October, Quantile completed its first compression run for swaps cleared at LCH with five large banks. The company also has two industry partners: First Derivatives, which provides the technology underpinning its compression service, and AcadiaSoft, which uses Quantile's service for its margin management products.
RegTech and Compliance
As the industry faces an unprecedented burden of new rules and regulations, a host of companies have emerged with novel solutions for compliance. This year's Innovators Pavilion featured three startups that are tackling the compliance challenge in innovative ways.
LogicGate has created a platform to organize and automate the process of managing compliance. No specialized knowledge or coding expertise is required, the company says. Users can simply draw a flowchart of their business processes and within minutes deploy a fully auditable, controlled process application. Rather than using a mish-mash of excel spreadsheets and email messages, companies can use LogicGate's system to organize their compliance workflows and ensure that nothing falls through the cracks. The company was founded in 2015 by several risk, compliance and legal technology consultants, and recently raised $1.9 million in funding from eight venture capital firms.
Prosparency addresses a more narrow problem in the market data area—the need for a more efficient way to track the use of market data so that subscriber fees can be assessed accurately. Exchanges are now requiring vendors to qualify and verify their non-professional subscribers accurately or pay audit fees associated with non-compliance. Prosparency has developed a solution for collecting, updating and monitoring subscriber data in an efficient and sustainable manner. The New York-based company launched its product in 2016 and has landed at least one major client. Interactive Brokers, which provides access to more than 120 trading venues around the world, began using Prosparency's "know your subscriber" product in May 2017 to reduce manual processing of subscriber classifications, leading to greater accuracy, reduced administrative burdens, and lower risk of audit liabilities.
Communications on the trading desk is another area where compliance needs have ramped up. The challenge here is that traders use a wide range of communications tools, which means that compliance staff need a way to track conversations taking place on every type of platform and scan them to make sure that the traders are not breaking the rules. Anish Parikh and Evan Caron, the two founders of Whistler Technologies, know the problem first-hand. Both come from the energy industry and they have tailored their system to commodity traders. Whistler's patentpending system captures email, instant messages and voice communications and brings all of those conversations into one dashboard so that compliance staff can get a holistic view of what traders are saying. Customers say the system stands out for the breadth of its coverage—it can track communications even on encrypted platforms such as WeChat to WhatsApp. In addition, the Denver-based company has developed an innovative prevention mechanism that uses machine learning to identify problematic messages as they are being typed and blocks those conversations from continuing. That helps strengthen internal defenses against spoofing, market manipulation, wash trades and other types of violations.
Last but not least, the 2017 Pavilion included three startups that are focused on product innovation. One of these companies, BCause, is focused on bitcoin. That is a crowded space, but BCause is hoping to carve out a niche based on its domain expertise in derivatives, multi-part business plan, and support from a strategic partner. The Virginia-based company has several executives with experience in futures markets, including Fred Grede, the chief executive officer, who formerly was an executive vice president of the Chicago Board of Trade and chief executive officer of the Hong Kong Futures Exchange. The company's strategic partner is SBI Group, a Japanese financial services company that bought 40% of BCause in October. BCause is currently seeking regulatory approval from the Commodity Futures Trading Commission for a cash-settled futures contract based on an index of bitcoin prices and is working with several industry tech vendors to build its trading platform.
Creating an intelligent multi-issuer platform to disrupt the structured note market.
The company also is developing a spot market for bitcoin trading as well as the infrastructure to "mine" bitcoin, a computationally intensive process used to create new bitcoins and add transaction records to the bitcoin ledger. BCause's mining plans got a big boost from the partnership with SBI; when the Japanese firm bought its stake in the company in October, it also agreed to provide "significant amounts of crypto-currency mining gear" to be co-located at BCause’s mining operations in Virginia Beach, Virginia.
Halo Investing is seeking to transform the $400 billion structured note market by creating a more efficient mechanism for investors to find the right product. The Chicago-based company was founded in 2015 by Jason Barsema, a former partner at Credit Suisse's private banking business, and Biju Kulathakal, a serial entrepreneur who co-founded the company that ultimately became Redbox, one of the largest movie rental companies in the U.S. Together they have created an independent multi-issuer technology platform for structured notes that dramatically reduces the friction in the investment process. Structured notes are debt obligations typically issued by major financial institutions that track the performance of an underlying asset and often use derivatives to provide downside protection. The traditional creation and issuance process is complex and time-consuming, and that makes them too expensive for ordinary investors. Halo's platform uses technology to streamline the process and disrupt the economics of distribution. Investors can choose from a menu of over 6,000 ideas that are updated every day, or customize their own strategy for low notional minimums and near-instant execution. Barsema and Kulathakal say that more than 50% of their customers have never purchased a structured note before, a strong indication that the platform is paving the way for more participation.
Peak Soil Indexes is focused on the financialization of farmland. The Georgia-based company has developed a set of farmland price indexes derived from actual farmland sales that it believes can serve as the basis for a whole new asset class. The company was founded in 2014 by Paul Kanitra, a former fixed income trader. PSI currently offers eleven indexes that cover farmland prices in six U.S. states and the Canadian province of Saskatchewan. Unlike existing survey-based indexes, all of PSI's indexes are constructed from publically available records, county assessor data and other sources. Kanitra says his indexes can serve as the benchmarks for derivatives contracts, exchange-traded funds and other financial products. Although that has not happened yet, Kanitra anticipates demand from institutional investors looking for alternative investments, and points to the increasing number of funds specializing in farmland investments as a harbinger of the trend.