For years now, the concept of “Big Data” has been wildly popular in the business world. It seemed that everyone who was anyone was using big data to improve performance, boost revenue, and break into new markets. For true technology innovators, big data was the first step to something bigger: artificial intelligence.
Machines need enormous amounts of data to develop artificial intelligence. AI developers use data to improve the algorithms that form the underlying intelligence of machines. In other words, the big data movement was like primary school for artificially intelligent systems. Today, such systems have moved on from primary school and are entering their adolescent years in secondary school.
This post is the first of a three-part series that discusses how AI systems are or will be used by industry participants. Part 1 discusses the SEC’s development of these systems and the factors driving that development. Part 2 discusses how AI systems are going to disrupt service providers to asset managers. Part 3 discusses the use and benefits of AI systems for asset managers.
Artificial Intelligence in Investment Management Part 1: the SEC.
The SEC is still focused on big data. It’s 2009 XBRL rule, and the more recent Form ADV, mutual fund modernization, and liquidity rules, are each an example of the SEC’s effort to obtain data on the asset management industry. All three rules developed requirements for industry participants to send structured data to the SEC and investors. In his recent speech at the RegTech Data Summit, Commissioner Michael S. Piwowar discussed the usefulness of the structured data to industry participants, investors, academics, and the SEC. During his speech, Commissioner Piwowar pithily used a quote from Chaucer, “Out of the ould fields must spring and grow the new Corne.” In the old fields of financial innovation, big data is old corn but the SEC is still trying to harvest it.
The newest crop of innovation, which is growing and budding as we speak, is artificial intelligence. The SEC recognizes that the data it collects allows machines to quickly and easily analyze such data, which can enhance automation and business information processes. These analyses can then be used by other machines or people, such as investors or asset managers, to make better decisions.
On April 11, 2018, at a KPMG sponsored an Investment Management Regulatory, Matt Giordano, former Chief Accountant at the Division of Investment Management, discussed how the SEC is using big data to monitor performance outliers, asset flows, and systemic risk. He noted that the 2014 Financial Stability Oversight Council (FSOC) report on the asset management industry highlighted the SEC’s needs to collect more data on registrants so that the SEC could effectively meet its mission objectives of investor protection and capital formation. It was this report that prompted the SEC to develop its rules that collect registrant data on Forms ADV, N-CEN, N-LIQUID, and N-PORT.
- Evaluate systemic risk based on an entity’s products and activities, not its size.
- Coordinate oversight activities across regulators such as the SEC and CFTC.
- Revise its liquidity risk management rule.
The central to the initial and updated FSOC reports is the SEC’s need to obtain certain data on the asset management industry so that the Commission can develop new regulations or modify existing regulations. As the SEC continues to obtain this information from registrants, it will need to develop systems to categorize and evaluate this data. Therefore, it’s only a matter of time before the old fields of the SEC result in new uses of AI systems for regulatory monitoring. But first, it must graduate from primary school.