Mind the (Data) Gaps

Keynote Address at the 8^th Annual Conference on Financial Market Regulation (CFMR)

Good afternoon. It’s great to be here at the annual Conference on Financial Market Regulation. I’ve enjoyed the discussions so far, and I am looking forward to hearing more. And I want to welcome Jessica Wachter, our new Chief Economist, to the SEC. I am very pleased that you are joining us, and I am looking forward to working with you.

Before I begin my remarks, I need to mention that the views that I express today are my own and do not necessarily reflect the views of the Commission or its staff.

To start, I want to note that I am thankful for the work that economists do inside and outside the SEC to help us understand the markets we regulate. It’s vital in terms of providing insight and analysis to help shape our regulatory approach. As those of you who have spoken to me may have noticed, I am not an economist. But I do have an economist’s love of good data and the data-driven rulemaking that can result.

Data are central to what we do at the SEC. Economists like you are on the front lines in terms of analyzing, interpreting, and using the data that we have, but without data, no one at the SEC would be able to do their jobs well.

Of course, when we collect data, especially data that contains personal or proprietary information, we need to protect and manage it carefully. Serious harm can come from the mismanagement of data. And we need to be cognizant of the impact on market participants of reporting and disclosure requirements.

However, serious harm can also come from regulating in the absence of relevant data. Without information about the markets that the SEC regulates, we may fail to address problems in our markets, or even make them worse. There is a cost if we fail to obtain the data we need to analyze and to understand the markets – and while the cost may be difficult to quantify, it is very real.

I’m speaking to you remotely today, as I have for all of my speaking engagements since I became a Commissioner at the SEC. That is, of course, because of the COVID-19 pandemic, which has transformed our lives in innumerable ways and resulted in terrible illness and loss of life.

The reason I raise this at the beginning of my remarks, however, is that I want you to think back to the beginning of the COVID-19 crisis, when we still knew little about how the disease spread from person to person. At that point, lacking information about where the true risks lay, I took up the practice of washing anything that entered my house with hot soapy water, and letting it air-dry before permitting anyone to touch it. This rule applied to all my groceries, deliveries, and other new items. Anything that couldn’t be washed was placed in quarantine for three days in a designated corner of my house.

Looking back, of course, that seems silly – we now know that COVID-19 spreads mainly from person to person through the air, not from surfaces like a cereal box.[1] But at the time, with incomplete information, it seemed very reasonable, and I know I wasn’t the only one taking those types of precautions.

Now – where am I going with this? Well, imagine if we had never obtained the data that allowed scientists to determine that COVID spreads mainly through close contact from person to person. In that case, it would have seemed very reasonable – maybe even necessary – to have rules requiring everyone to wash their groceries after bringing them home from the store. However, such a rule would have resulted in a huge expenditure of time and energy, for very little benefit. Furthermore, we might not have put in place the rules that we now know do protect us – like requirements for social distancing, mask-wearing, and of course, working remotely when possible, as most of us are doing today.

So hopefully you see now where my story was headed. Bad information, or a lack of information, can lead to bad regulation, which can result in both unnecessary burdens and missed opportunities. It can also lead to a failure to regulate where regulation would be appropriate, which can result in a failure to address real problems and mitigate harms.

With that preamble, I want to delve into a few areas where I think the SEC can and should obtain more information to facilitate the kind of data-driven regulation I think we could all agree yields better results.

Private Markets

First, over the last several years, the SEC proposed and adopted a number of new rules and other actions relating to the private markets.[2] For the most part, these new rules loosened the restrictions on issuers, investors, and intermediaries, with the goal of increasing investor access.[3]

For example, the SEC expanded the definition of Accredited Investor, allowing a larger group of investors access to private market investments.[4] The SEC also amended its rules to permit larger and more frequent private offerings.[5]

However, what we know about the private markets seems to indicate that when compared with public markets, they are less liquid, more opaque, impose higher transaction costs, are more susceptible to valuation errors, and are more prone to fraud.[6] So do the actions the Commission took to expand access to private markets further our agency’s mission? It is a question that is harder than it should be to answer, because for the most part, we lack good data on private issuers and offerings.[7]

What we do know is that the private markets have increased in size over the years, both in absolute terms and relative to the public markets. The amount of capital raised via exempt offerings now far outpaces the amount raised on the public markets.[8] Offerings under Regulation D alone accounted for $1.5 trillion of proceeds in 2019, compared with $1.2 trillion in the public markets.[9]

And yet, while these markets have been expanding, the information we are collecting about them has not. For the most part,[10] we do not know who invested in these private market offerings or how their investments performed.[11]

In 2013, the SEC proposed to amend Regulation D to strengthen the filing and disclosure requirements.[12] I believe it should be a priority to finalize those proposed changes, as well as to consider what other information we need.

With better information, we can make better-informed regulatory decisions, because we will know what the true benefits and risks are of investing in the private markets. We can better assess how well our existing rules – including the recent changes – work to facilitate capital formation, as well as whether investors are protected sufficiently.

And I hope that economists like yourselves will use that data to provide the SEC with even greater insights into how the private markets operate.

Investor Testing of Form CRS and Regulation Best Interest Disclosures

Another area in which I think we need more information is with respect to the effectiveness of some of the disclosures we require from investment professionals. In June 2019, the SEC finalized a package of rules and interpretations regarding the standards of conduct applicable to broker-dealers and investment advisers.[13] The overall approach relies heavily, though not exclusively, on disclosure.

One piece of that package, Form CRS, is a disclosure that is intended to provide investors with information about the nature of their relationships with their investment professionals.[14] I will say that I very much support the idea behind Form CRS. Investors should be provided with information about the relationship before they engage with a financial professional, so they can make informed decisions about who to work with and what services to choose.

When we rely on disclosure to do the work of protecting investors, though, we need to ensure that investors can read and understand the information they are given.[15] One good way to do that is to engage in “investor testing” – an assessment of investor comprehension including interviews of real investors.[16] The goal should be to ensure that the disclosures are accessible and understandable in a way that enables informed decision-making by all.

To this end, the SEC commissioned some investor testing on a sample Form CRS before finalizing the rule.[17] However, the SEC did not follow up on indications that key aspects of the disclosures were not well-understood by investors: specifically, the distinction between a relationship with a broker-dealer and one with an investment adviser.[18] Further, the SEC provided significant flexibility to financial professionals in designing their forms.[19] There is a lot of variation and the forms can look quite different from the sample that was tested. But the SEC hasn’t done any testing of real forms to see whether they are working as intended. And reports from industry experts and others seem to indicate that they may not be.[20]

Another piece of the rulemaking package, Regulation Best Interest, requires broker-dealers to make certain disclosures to their retail customers. For example, they must make disclosures about the nature of the relationship between the broker-dealer and the customer; the fees and costs the customer will incur; the type and scope of the services to be provided; and conflicts of interest.[21] However, the SEC has never performed investor testing on the effectiveness of these disclosures. The materials provided to investors are often lengthy and technical, using industry terms that may be unfamiliar to retail investors. Whether this truly enables informed decision-making is a crucial question in terms of the effectiveness of our regulatory approach.

I believe we need to engage in investor testing of the actual forms and disclosures that investors receive in order to determine whether or not they are effective. This will allow us to improve the disclosures and make them more useful for, again, all investors. We can also use the testing to assess the limits of disclosure as our primary means of investor protection in this area. Disclosure should not be used as an end run around the obligation to act in a client’s best interest. If there are areas in which even the best disclosures do not appear to provide a full understanding of key information, we should consider other ways to protect investors. This could potentially include eliminating certain practices that create conflicts of interest – particularly if conflict mitigation measures are not sufficiently protecting investors.

Our reliance on disclosure in Regulation Best Interest, Form CRS, and other documents like Form ADV, is based on the idea that it is an effective investor protection tool. But in the absence of investor testing, we lack information on how effective it really is.

Consolidated Audit Trail

The final area I want to talk about today is relevant across a large number of rulemakings and other policy initiatives. There have been many calls for the SEC to respond to the GameStop-related market activity that took place in January of this year.[22] However, before we can respond to any market event, we need to understand what happened. Without good information about what took place, we may respond in ways that fail to address the real issues. And the key to understanding a market event like GameStop is a complete, accurate, and accessible source of market data.

Almost nine years ago, the SEC adopted a rule requiring the exchanges and FINRA to begin the process of building a Consolidated Audit Trail – also known as the CAT.[23] The CAT was conceived in response to the fragmentation of the equity markets – the proliferation of equity exchanges, along with alternative trading systems such as dark pools and over-the-counter broker-dealers acting as market makers.

The May 6, 2010 Flash Crash[24] exposed how this fragmentation had impacted the SEC staff’s ability to understand activity in the equity markets in a timely fashion.[25] While the SEC staff could obtain data about activity on the various different venues, it was time-consuming to collect and labor-intensive to try to piece it all together in order to get a full picture of the market. In order to do so, the SEC staff needed to combine trade and order data from a variety of different sources – and variations in the scope, required data elements, and format sometimes made the data difficult or impossible to merge.[26]

I am happy to say that today, much of that information is now available in one place, in a consistent format – as of this past December, the CAT is combining data from all of those disparate sources to provide a comprehensive picture of trade and order activity to the SEC staff. However, the CAT still lacks the customer and account information needed to get a full understanding of what is happening in the markets.

I am confident the CAT was helpful in our efforts to understand the GameStop-related market activity. However, I know that we would be in an even better position to understand if the CAT was complete. We would also be in a better position to understand and address myriad other issues, including issues related to market manipulation, market structure, and order routing. I therefore believe it should be a priority to finish the CAT, so that it finally reaches its full potential as a tool for understanding and analyzing the markets we regulate.

Other Areas

Of course, private markets, the effectiveness of our disclosures, and trade and order activity are not the only areas in which the SEC needs to address gaps in our data. The GameStop-related events I just mentioned also raised the question of whether we, and the public, have sufficient information regarding short selling activity and the securities lending market more generally.

The more recent events surrounding the failure of the Archegos family office raised similar questions about derivative positions, and highlighted the importance of implementing our long-overdue security-based swaps reporting framework. We should also carefully assess whether the data reported under this framework are sufficient to allow us to detect the buildup and concentration of risk exposures.

As part of our proposed amendments to Rule 144, which governs the sale of restricted and control securities, I think we should consider adding a requirement for corporate insiders to disclose electronically when they sell restricted shares under Rule 144 pursuant to a 10b5-1 plan, and to disclose the date the plan was adopted. This would allow the SEC to more easily detect abuse of 10b5-1 plans.[27] Finally, I would like to see us take action to improve the information available to investors in the municipal bond and corporate bond spaces. And of course I would be remiss not to mention the need for comparable, reliable, and decision-useful disclosures by corporate issuers on Environmental, Social, and Governance measures, which Chair Gensler discussed in his opening remarks yesterday.

As I conclude, I want to note the importance of academic work that analyzes the data we collect, like some of the papers presented yesterday and today. That type of work often informs our regulatory activity. For example, along with a number of papers by outside academics, a DERA white paper on investor outcomes in the OTC markets was cited extensively in the release adopting our recent changes to Rule 15c2-11.[28] Our recent proposal to amend Regulation ATS relies on academic work examining the Treasury markets.[29] The SEC’s earlier rulemaking on NMS Stock ATSs is another good example;[30] in that case, when the Commission proposed the rule, it relied on academic research and studies to show that the market for NMS stock execution services had resulted in the fragmentation of trading volume, which impacted price discovery.[31] There are countless other examples of academic work informing our regulatory activity, and not just in the rulemaking space.

While we have relied on data to the degree we have data, we need to ensure that we have sufficient and accurate information to regulate effectively across the board. And in the areas I have identified, I think we need more data to ensure that the actions we take to advance the SEC’s mission – protecting investors, maintaining fair, orderly, and efficient markets, and facilitating capital formation – accomplish those goals. In other words, I want to make sure we are not making rules requiring people to wash their groceries. My hope is that we can close the data gaps I’ve described, and that the SEC – and academics like yourselves, in some cases – can use the data to better analyze and understand our markets. With that improved understanding, I hope we can engage in more effective, data-driven regulation.

Thank you, and I look forward to your questions.