Recently, SIFMA hosted its 46th annual Operations Conference & Exhibition over several days of presentations and events and nearly 1,000 attendees. Inside a note, SIFMA recapped some of the highlights, including the future of fintech.
FINRA EVP and chief information officer Steven Randich shared with us how FINRA as an SRO is adopting fintech into its everyday operations. FINRA’s goal is to rebuild technology to improve market participants’ digital experience and make it easier for firms to navigate through FINRA’s many divisions and systems, i.e. decrease costs and burden for firms.
As a large data processor for surveillance – monitoring 100% of the equities markets and 70% of the options markets – FINRA processes 135 billion events per day, with trillions of nodes recreating market events. Some of the fintech capabilities they are now incorporating include:
- Data Storage – With the migration from data centers to cloud starting in 2013, the vast majority of data has moved as FINRA continues to increase the amount of automation for clients. Cloud increases capacity 10x and enables continuous tech refreshment, which should decrease costs to users and increase availability of data.
- Registration System – FINRA replaced its web-based CRD with a cloud-based platform, which should decrease duplicative tasks, increase automation and, therefore, reduce industry compliance burden and costs. (5,500+ individual, 50% of firms registered)
- Digital Experience – FINRA continues its digital experience transformation. This should end fragmentation across FINRA divisions and increase the amount of self service, which should be faster than waiting on the phone to speak with a human. The estimated benefit is >$100 million in savings per annum for the industry.
- Innovation – FINRA Createathon, which begin three years ago. is an annual two and one-half day event where cross-disciplinary teams generate ideas and present solutions (NLP, ML, AI, etc.). In 2018, 57 teams and 500 participants across 6 categories led to 12+ ideas advanced to the R&D stage. Idea stats: 57 feed ideas, 29 initiated, 12 developed, 9 presented for rating, 6 selected for funding and 1 completed and to be presented for business case funding.
The industry continues to analyze how fintech solutions can increase efficiencies in the back office, serve clients, manage risk or meet regulatory reporting requirements. While much of a firm’s technology spend is on ways to increase operational efficiencies and decrease costs (for the firms but also their customers), many firms also sponsor investments for fintech innovations.
As such, SIFMA highlighted key fintech topics heard at the conference:
- The Pecking Order – Capital markets exist to serve the needs of corporations, governments, investors and the economy. It is a complex ecosystem which needs multiple components to work in tandem. And while changes surrounding this ecosystem continue to be faster, faster, faster, the bedrock of financial services remains trust. As financial institutions are responsible to regulators and must maintain the trust/confidence of their clients, their approach differs from some fintechs which do not have these obligations and take a fail fast and often approach to innovation. Financial institutions can work to develop their own fintech capabilities or partner with fintech firms, i.e. build, partner or buy strategies. The challenge is how to transform legacy systems to modern ones, keeping the economics in place and not disrupting client services. When looking at fintech capabilities, the question becomes how do we use this to run the business and gain efficiencies? Modernization of operations can be end-to-end updating of systems/processes or developing individual tools (not end-to-end, ex: RPA for automation of specific tasks). In the financial services fintech maturity ladder, not all technologies are on the same rung: most firms are digitizing (RPA, robotics); some are using cloud technology; AI to scale is further out; and DLT is even further out in time to large-scale implementation.
- Robotics, Robotic Process Automation (RPA) – Robotics and RPA have a shorter time frame to implementation. Automation via these fintech capabilities benefits employees and employers, by decreasing time it takes to perform tasks, and clients, by improving straight through processing and other procedures, while increasing transparency and reducing operational risk. While RPA continues to gain maturity and is implemented in multiple areas, a presentation from PwC and UIPath indicated the current automation scope is only 3-7%, out of a 40%+ potential to automate. Opportunities reside in robotics as well –one panelist indicated there are >1 thousand bots in production, with 4 million bots expected by 2021.
- Cloud – Cloud architecture can come in several forms: internal (private), public or hybrids. Initial use cases were to gain raw computing power. Now, firms use clouds to access and utilize data in real time in a flexible and efficient manner. Firms can use multiple cloud sources together while controlling how they interact, with many financial institutions using 3-4 private and public clouds. Yet, a panelist indicated only 20% of workloads in financial services have moved to cloud technology. What will the modern cloud architecture look like – secure, fast, real-time access – when firms ramp up for a world requiring more and more data be pulled in? How does cloud technology play into firms’ commitment to resiliency and risk management?
- AI – AI has been in use at financial institutions for many years, for example: using it for credit scoring in retail banking; using it in CCAR/stress testing to forecast and view changes in results under different macro variables; using it in call centers, where evidence has shown query resolution increases “massively”; and other, more “simple” processes. To continue to grow usage, firms are exploring how to embed AI into the workflow/decision process, for example: KYC/AML monitoring to automate the closing out of cases; or trader conduct surveillance to predict behavior before it happens. Since AI is the technology in the raw, firms need to consider where the data came from and how the model operates, understand how to supervise it and ensure AI is working with professional decision making, not replacing it.
- Cyber – Financial institutions use and store extremely sensitive client data, and there are people (criminals, bad actors, nation states) constantly trying to exfiltrate or compromise this data. As crucial players in capital markets, financial institutions cannot risk operational disruptions due to a cyber attack. Firms continuously test their systems’ resiliency, both on an individual-firm and an industry-wide basis. They continue to spend billions on information security and data protection, with teams solely focused on cybersecurity. It becomes a 24-hours-a-day, 7-days-a-week, 365-days-a-year war firms wage to protect client data and their systems (see the SIFMA Insights Spotlight, Building Resilience with a Culture of Cyber Awareness).
- Data – Data, or the new oil, is key but often needs cleaned up to be useful. Dirty data can drive bad outcomes, and if there is no data to “feed the beast”, then you will not find solutions. More data is becoming available than ever before and can solve problems it could never solve before. Firms can now standardize data, with opportunities to join disparate data sets across business units. Often ~75-80% of the problem is solved by moving data around (data wrangling), yet other data sets need cleansing. Speakers indicated there is no silver bullet – onboard data, make use of it, entitle people internally to use data, etc. What is important is the ability to turn unstructured data (data that is usually not as easily searchable) into structured data (clearly defined data types whose pattern makes them easily searchable) to leverage the power of data. Panelists and speakers further note that data governance is essentially the “table stakes” in financial services today, i.e. data protection is key.