For years now, the SEC has quietly advanced its data analytics so that the staff is better equipped to examine registrants and identifier outliers. For example, the Office of Risk Assessment analyzes data in documents such as Forms ADV and PF. Information from this analysis is used to direct risk-based examinations of registrants, including investment advisers. The Office of Compliance Inspections and Examinations also uses data-rich tools such as the National Exam Analytics Tool (NEAT) to improve its review of areas such as trading.
For policy-making efforts, the various SEC divisions rely on the Office of Data Science and the Division of Economic Research and Analysis, which collectively maintain the Quantitative Research Analytical Data Support (QRADS) program. The rule-making divisions can use the data to direct rulemaking or enhance analysis to make rules more effective (and litigation proof).
Enforcement also uses data analytics to monitor huge amounts of trading data in search of insider trading or market manipulation; it has also used artificial intelligence to identify insider trading networks. It all sounds very Orwellian, but it’s nothing compared to what Amazon, Facebook, and Google do to consumers every day.
The Division of Investment Management uses an internal tool called “Monitoring and Analytics GUI for Investment Companies” or MAGIC to pull together data sets from registrants and other sources. The staff can use the output of these tools to test portfolio compliance against disclosure so that it can identify outliers, and provide that information to OCIE or Enforcement. New forms of information, such as Form N-PORT and N-CEN, will no doubt give the staff the ability to monitor other aspects of fund portfolios, such as liquidity. And if this is what the U.S. Government can do with data, imagine what today’s cutting-edge companies are doing with it.
The impact of technology on compliance
Technology allows compliance officers to quickly access the data they need to test the various policies and procedures in place. For firms that don’t have robust databases, testing is highly manual and, therefore, slower and more expensive. Additionally, poor data management results in less organization transparency, which can harm internal reporting and oversight.
In today’s compliance environment, transparency and data are directly related and essential to managing risks. Better data means better transparency and clear visibility into risks, trends, and other insights that allow compliance officers to manage risks. Easily accessible data results in improved monitoring, which in turn allows for quicker identification of issues. By detecting issues earlier, compliance officers can react before the issue creates a reportable event. Additionally, better data allows compliance officers to cut through the noise and evaluate helpful information.
Technology not only helps with performance, but it also acts as a cost reducer. As a firm grows, they generate more activity and data. If your data entry and oversight processes are manual, then growth requires more people to manage these processes, and costs increase. With scalable technology, the adviser’s current systems can manage the increases in activity and data and still provide the overseers with valuable reporting; in other words, the scalable technology makes the overseers more scalable. Together, the people and their machines can do more without adding substantial costs to the firm’s operating structure.
Technology can also reduce turnover rates among staff. Generally speaking, no one enjoys data entry and manual testing. People are more interested in analyzing trends and other developments so they can use their powerful cognitive skills to create solutions. They are also freed to do other work that can both enhance their satisfaction and add more value to the business. For example, they can participate in client relations, product development, and marketing in a more effective way.
How compliance officers are deploying machines
Robotic process automation (RPA) is well established in today’s technology. Whether it’s chatbots, smart inboxes, or other software that create virtual fill-in-the-blank, RPA is already impacting today’s asset managers. Marketing and CRM tools increasingly use these processes, as do social media, which is increasingly important to growth.
Machine learning is the current frontier of compliance. Many Fintech firms, Joot included, are looking for ways to leverage the large data sets that their clients produce so that rules-based algorithms can be further leveraged to improve compliance oversight and risk management.
Without getting into too much technical detail, techniques such as machine learning or deep learning can train a computer to identify trends that may be too complex for the human mind to track. Additionally, it can eliminate certain low-level tasks, such as image identification and text recognition. Instead of a compliance officer who needs to review every piece of advertising, an algorithm could be used to identify potential issues and notify a compliance officer that further investigation is needed. Similarly, predictive analytics can be used to forecast activity in ways that are difficult for people to do with their current tools. Compliance could apply these techniques to processes such as risk assessments and annual reviews.
Both machine learning and predictive analytics require huge data sets to provide helpful results, and that’s where the biggest issues lie for many asset managers. Most of the industry is comprised of small to middle-sized managers that do not produce enough data to develop a better process. Firms like Joot seek to eliminate this barrier by aggregating data from many small firms.
There are a few questions that you should consider when evaluating technology for your firm.
Who on your team is needed to evaluate and onboard the technology?
If it’s a cross-functional team, is everyone committed to the project?
What are some of the issues that could derail the integration?
Who owns the technology, you, or the vendor?
Who at your firm is responsible for the technology, even if its ownership resides with a vendor?
Who will lead the vendor due diligence at the outset and on an ongoing basis?
How easy is the integration? Will it take days, weeks, or months?
How much training is needed to get your staff used to the tool and its functions?
What data formats, systems, and sources does the technology rely upon?
How are those various points connected, and what securities protocols are in place at each node?
If I could leave you with one thought, it is this: The asset management industry is being permanently changed by data analytics, and small and middle-sized firms need to figure out how they are going to keep up with the accelerating use of these tools. These firms need to realize that building internal systems is not adequate because the data sets are too weak and the systems too expensive to build on a small scale. Big data is just that, big! Further, it’s just the first phase of the data revolution occurring in our industry.
The next phase will increasingly leverage artificial intelligence that layers onto machine learning and predictive analytics. This advanced layer will bring the machines closer to human intelligence but with the huge advantages of longevity and scale. More importantly, these tools will be freed from well-structured data sets. Instead, artificial intelligence will, as people, be able to take unstructured information and glean insights. These insights will improve the previously developed tools like robotic process automation and machine learning. Phase one is like a student reading a library of books to learn about a subject. Phase two is like the student finding a master instructor to accelerate their learning and understanding of all the information on that subject. Welcome to the future.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.