Wharton's Daniel Taylor discusses his research on the effects of undisclosed SEC investigations.

In November 2019, The Wall Street Journal reported that the Securities and Exchange Commission was investigating Under Armour, a Baltimore-based maker of footwear, sports and casual apparel. Though the investigation had begun more than two years earlier, the company had not disclosed the fact that regulators were looking into its accounting practices. On the day the story broke, Under Armour’s shares fell by 19%. Clearly, the market’s reaction showed that investors believed this was relevant information and material to their perception of the shares’ — and the company’s — value.

Should companies like Under Armour go public sooner about the fact that the SEC is investigating them? If so, when is the right time to disclose such potentially damaging information? And before the company makes the disclosure, do insiders — who may or may not be direct targets of the probe — gain an unfair advantage in being able to sell shares before the information hits the market?

Daniel Taylor, a professor of accounting at Wharton, and his co-authors — Terrence Blackburne of Oregon State University; John D. Kepler of Stanford University; and Phillip J. Quinn of the University of Washington — investigated these questions and more in their research paper titled, “Undisclosed SEC Investigations.” The researchers reviewed data obtained from the SEC on every formal investigation between 2000 and 2017, which was previously not public, to come up with their findings.

Taylor spoke with Knowledge at Wharton about this research and the implications for SEC investigations, share prices and insider trading. An edited transcript of the conversation appears below. (Listen to the podcast at the top of this page.)

Knowledge at Wharton: What prompted your study on undisclosed investigations by the SEC?

Daniel Taylor: Two events prompted the research. Back in January 2016, the Southern District of New York had a case in front of it, Lionsgate Entertainment, which went all the way through the court system. The ruling was that corporations are not under any obligation to disclose SEC investigations or receipt of a Wells Notice. (A Wells Notice tells the corporation that the SEC is likely to enforce against them.) The court held that “the defendants did not have a duty to disclose the SEC investigation and Wells Notices because the security laws do not impose an obligation on a company to predict the outcome of investigations. There is no duty to disclose litigation that is not ‘substantially likely to occur.'”

That was surprising to me. I work at the intersection of accounting and finance and economics. We teach students that materiality is about whether a reasonable person would alter their valuation of the firm or their investing decision based on the information. In accounting we regularly recognize and account for uncertain negative events. For example, banks typically make an allowance for loan losses or for doubtful accounts, even though the loans haven’t actually defaulted; they provide an accounting contingency. We have loss contingencies related to litigation and environmental issues. That is why I was surprised that in this case, the Southern District of New York seemed to set the bar to disclosure only when there is enforcement. That is at odds with how we typically think in economics about materiality. That got me thinking, “This is an interesting case where the bar for disclosure is different than what we see in other accounting and finance settings.”

Less than 50% of firms disclose SEC investigations eventually.

Then, last November, The Wall Street Journal broke the news that the SEC was investigating Under Armour’s accounting practices. Under Armour had, in fact, been cooperating for two and a half years but had not disclosed that investigation or that cooperation to shareholders. On the day that story broke, Under Armour’s [stock] price fell 19%. Clearly, the market thought the investigation was sufficiently material to alter valuation by a substantial amount.

This got me thinking – along with my coauthors — about the frequency with which firms are not disclosing material SEC investigations. If The Wall Street Journal had not broken that story, would Under Armour have eventually disclosed it, or would shareholders never have learned about the investigation unless it actually led to an enforcement action?

The next question is, if the company has an ongoing SEC investigation and chooses not to disclose it – which they may not be obligated to — do they also not trade on that information? If the firm chooses not to disclose the investigation, and the investigation is material, that potentially means that insiders have an information advantage over normal shareholders. We typically think that fiduciary duty compels managers to either disclose the information or abstain from trading. So, they’re not required to disclose, [but] do they abstain from trading?

When we started our research, we didn’t quite appreciate how topical, relevant and far-reaching it was. Let me illustrate with a simple example; on February 13, Tesla disclosed it received an SEC subpoena on December 4 and it was cooperating with the SEC in their investigation. That puts about a two-month gap between when Tesla was hit with a subpoena and when it disclosed the subpoena. The natural question is whether the investigation is material. We don’t necessarily know yet. Insiders at Tesla might, but outsiders don’t. Given the disclosure occurred on February 13 and the subpoena was back in December, did any insiders in Tesla trade in the intervening period between when they learned of the subpoena and when they disclosed it? This is very, very topical and touches a lot of different companies out there right now.

Knowledge at Wharton: Considering that the SEC is highly secretive when a company is being investigated, how challenging was it to gather the information you needed? How did you go about gathering the data?

Taylor: This paper relies on data that the SEC provided to us. SEC investigations are shrouded in secrecy. The SEC doesn’t comment on ongoing investigations. Only the firm being investigated, the firm’s counsel and the SEC and investigators are typically aware of the investigation. And, as we just discussed, the firm doesn’t have any obligation to disclose that investigation. So prior work typically only observed investigations when the firm disclosed the investigation or when enforcement was brought. You couldn’t observe instances of SEC investigations that were undisclosed.

We wanted to figure out what fraction of firms had undisclosed investigations. This presents a challenge, because the firm isn’t disclosing it, and the SEC may not have an enforcement action on it. So how do we uncover these undisclosed investigations?

We found that one year after the investigation is opened, the median investigated firm underperforms the market by about 6%. Two years later, it underperforms the market by about 10%.

Over the course of six months, one of my coauthors went back and forth with the SEC, requesting information under the Freedom of Information Act on closed investigations — not open investigations but those that had closed. As part of the process of that back-and-forth with the SEC, we negotiated with them to release data on all closed, formal SEC investigations between 2000 and 2017. It’s about 17 or 18 years of data on whom the SEC has been formally investigating.

The data includes everyone that is investigated regardless of the outcome. In some sense, we observe the master list of those individuals and entities that were formally investigated by the SEC. The key adjective is “formally.” The SEC has many different scopes of investigation, “matter under inquiry” being one that’s not formal. A formal investigation implies subpoena power by the SEC. We wanted to focus on instances in which things are serious enough to warrant subpoenas. The data entailed 299 pages of records of those individuals and entities, more than 12,000 investigations that the SEC undertook over that period. So now we know whom the SEC was investigating over the period, independent of the outcome of the investigation and independent of whether the firm disclosed it.

Knowledge at Wharton: What are the main takeaways from your research?

Taylor: As I mentioned, the SEC turned over records to us of more than 12,000 investigations. Most of those investigations are not against publicly traded entities. They’re against individuals, broker dealers, they’re investigations for Ponzi schemes for non-publicly traded entities … [there are] all sorts of investigations out there.

One of the fun aspects of working on this paper is that almost every one of the observations is a story; each investigation has a back-story about what is being investigated, who is being investigated and why they are being investigated. We took the 12,000-plus investigations and focused on investigations of entities that are traded on the three major exchanges – the New York Stock Exchange, the NASDAQ and the American Stock Exchange (AMEX). That got us down to fewer than 4,000 investigations from 2000 to 2017.

The first thing we did once we had the set of investigations is look at what fraction of those firms that were under a formal investigation actually disclosed it and the timing of that disclosure. What we found is that about 19% to 20% of firms disclose some sort of SEC investigation within ten days of the investigation opening. By day-plus-ten after the investigation opened, 20% of firms have been forthcoming and have said look, the SEC is looking into something. The disclosures are not necessarily detailed but at least we have some evidence the company was forthcoming. Interestingly, less than 50% of firms disclose the investigations eventually. So, a large chunk of investigations appear to go undisclosed by the firm. We can disagree whether it should be allowed or not, but it seems to be allowed under the law.

If you knew that the firm was under investigation, and you sold your shares immediately upon learning that, you would have avoided significant losses across the entire sample.

The next question is, let’s look at who’s under investigation. The first thing we looked at was industry distribution. It looks like the distribution of industries that the SEC tends to investigate lines up well with the distribution of industries across the three major exchanges. You don’t see the SEC investigating disproportionately into any one industry.

After that, we looked at the distribution of size, the company size. There we found something pretty interesting. Some 20% of the SEC’s investigations tend to target the largest 10% of firms. I found that very surprising, because my belief coming into this was that potentially the largest 10% of firms are going to have the most well-resourced defense teams and going to be the most experienced. Our research suggests that the SEC doesn’t shy away from taking on those firms that can mount, potentially, the best defenses. It also is consistent with the notion that the SEC targets those firms that have the largest scope of malfeasance. If you’re a large firm and you have a potential fraud, there’s likely going to be more harm done, and so potentially that’s something that goes into the SEC’s decision to investigate larger firms. Conversely, very few investigations occur in the smallest firms.

And then we looked at future performance. We have this investigation data. The investigations are undisclosed. Well, are the investigations in fact material? We decided to get at that by looking at what happens to the firm over the course of the investigation. We found that a year after the investigation is opened, the median investigated firm underperforms the market by about 6%. And two years later, it underperforms the market by about 10%. That’s the median. What this means is that more than 50% of investigated firms have declines in their stock price even larger than 10% market-adjusted after two years. We found also that the investigations tend to foreshadow drops in the firm’s earnings and increases in stock price volatility.

The results of the stock price performance and operating performance suggest that the investigations are economically material. If you knew that the firm was under investigation, and you sold your shares immediately upon learning that, you would have avoided significant losses across the entire sample.

Now, this doesn’t mean that every investigation is material. There could certainly be immaterial investigations. But most of the investigations in the sample, the majority of those 4,000 or so investigations, do appear to be clearly economically material events.

It’s important to point out that it’s not necessarily the individual being investigated who is doing the trading.

Knowledge at Wharton: Let’s drill a little deeper. As your paper says, senior managers or insiders of a company that is being investigated learn about the investigation sooner than the outside investors. How does this insider knowledge influence their behavior? Do insiders exploit this advance knowledge and trade based on their access to this confidential information?

Taylor: One can view this research as having four effective steps. Step one is acquire data from the SEC on whom they have investigated regardless of whether the firm discloses it or the SEC discloses it. So, we acquire data on all investigations. Step two is [to determine] what percentage of firms actually disclosed the investigation, so we looked at the rate of disclosure of all the investigations. Step three is then to establish whether the investigations are in fact material. We know for a large number, they’re undisclosed. The natural question is of these undisclosed investigations, are they economically material? All right, so we’ve established the vast majority [of investigations are] undisclosed, we’ve established that the vast majority are economically material.

The fourth step is [to ask], if an investigation is undisclosed and it is material, do insiders abstain from trading? Fiduciary duty compels insiders to either disclose material information or abstain from trading. We’ve established that disclosure isn’t necessarily required in this setting. Among those firms that do not disclose, we then look at whether the insiders in fact trade. And the answer to that is yes, they do appear to be trading during the investigation. Now we know not only that the investigation is undisclosed, that the investigation is material, but that insiders don’t seem to be abstaining from trading during the period of the investigation. That’s really interesting because what it suggests is that there is some gray area as to legality here, and a gray area as to whether we should even be allowing that to occur.

It doesn’t appear that companies are locking down and preventing their insiders from trading during the investigation. Now, it’s important to point out that it’s not necessarily the individual being investigated who is doing the trading. You can imagine certain individuals at the firm learning that the SEC is investigating one of their colleagues for, say, accounting fraud or foreign corrupt practices or some bad behavior. It’s those other individuals — who aren’t actually the subject of the investigation — that appear to be trading once they learn that one of their colleagues or some division of the firm is being investigated.

The SEC should be examining insider trading during the course of investigations, even if the investigation itself is unrelated to insider trading.

Knowledge at Wharton: Were these individuals penalized for insider trading?

Taylor: There are cases, obviously, in which executives who are engaging in fraud start cashing out. The SEC is good at catching those — because if they’re investigating for accounting fraud, they’re looking at incentives to do so, and one of the incentives is to inflate the stock price to cash out the shares. The next question is, what about the lower-level employees, what about some of the other insiders whom the SEC isn’t necessarily investigating for the fraud, but once they learn that the CEO or the CFO is being charged, they start selling their shares? That’s what our research is diving into. It suggests that both firms and the SEC need to do a better job of monitoring incidental trading in firms that are being investigated by the SEC.

Knowledge at Wharton: What are the implications of your research for regulators, for managers and for investors?

Taylor: The first thing is that the SEC should be examining insider trading during the course of investigations, even if the investigation itself is unrelated to insider trading. For example, the SEC might serve a subpoena for bribing a foreign official, a Foreign Corrupt Practices Act violation. Individuals privy to the subpoena at the firm may take that opportunity to liquidate some of their shares, even though those individuals had nothing to do with the original offense, with the original bribe. Our findings suggest that the SEC should be monitoring for insider trading even when the investigation doesn’t necessarily relate to insider trading or doesn’t necessarily relate to accounting fraud.

We have not examined trading in, say, the options market or hedge funds or trading by outside parties. We don’t seem to find any evidence of abnormal trading in the market on the firms’ shares in the NYSE, the NASDAQ or the AMEX. One thing that the SEC should be doing is looking at trading during periods of investigation. This is because that provides an opportunity — if the investigation is material — for insiders, either corporate insiders or insiders at the law firm that’s helping the firm. It provides them with the opportunity to potentially trade profitably.

With respect to managers and the firm itself, the implication here is clear. Boards and general counsels need to lock down the trading of anyone involved with ongoing regulatory investigations and comport with best practices. This isn’t a question of legality. This is a question of corporate governance, of best practices. If executives at the firm that know their division is under investigation by the SEC, and that has the potential to turn bad, then good corporate governance would be to prevent those individuals from trading. As the general counsel or the board, [they need to say] if we’re not going to disclose this information to shareholders then we need to stop these corporate insiders from trading.

…We seem to be seeing notable exceptions to this. When Boeing was under investigation for issues related to its 737 Max, did it clamp down on the trading of its insiders? I have written a case study that would suggest otherwise. There are other examples out there. I have done related research on product recalls. Does the firm clamp down on the trading of insiders while it is investigating for product defects and product recalls? There needs to be a recognition that there are things other than accounting information that can be material information that would warrant the general counsel locking down the trading of corporate insiders.

There are many, many examples of this. One of the popular examples now making its way through the court is CBS. [Former chairman and CEO] Les Moonves was investigated as part of the Me Too movement, and the board seemed to be aware of this before the public knew about it. Litigants have alleged that the board knew about this and traded in the intervening period before the disclosure. That would be another case where the board knows about an investigation, even if not from the SEC — [it could be] from the Department of Justice, from the FBI, from the police, or from other regulators. In such cases, the board and the general counsel need to up their game and clamp down on trading.

Insiders — if the investigation is not disclosed — should be abstaining from trading. That evidence is an open-and-shut case.

Knowledge at Wharton: Based on your research, do you think the rules of disclosure should be modified?

Taylor: Well, one can think about whether it should be mandatory for firms to disclose any investigations. That is one path to go down. Here, I don’t think the evidence is necessarily clear that we should require all investigations to be disclosed. The evidence suggests that many investigations are immaterial. They end relatively quickly. You don’t want to require disclosure in such cases because investors may overreact to the disclosure of what might turn out to be an immaterial investigation. For me, I don’t think there are necessarily implications for whether we should require disclosure of these investigations.

But I absolutely think that insiders — if the investigation is not disclosed — should be abstaining from trading. That evidence is an open-and-shut case for me. This is especially [true] in the investigations we found that are material. I think the most surprising aspect of the paper was that there is a substantial fraction of firms that aren’t disclosing the investigation, where the investigation is material, and where managers are not abstaining from trading in the intervening period between when they receive the subpoena and when the subpoena or the investigation is disclosed. The evidence suggests that in the absence of disclosure, managers aren’t abstaining from trading, and instead seem to be potentially profiting handsomely from their trades.

The general counsels and the boards need to step up their game in terms of clamping down on trading by those who are aware of investigations into their firms.

Imagine a scenario in which your firm has been served a subpoena. The firm doesn’t disclose it for another six months. If it’s a material investigation, that gives you potentially six months’ window to begin liquidating your shares before that information hits the market. You may have incentives to liquidate even if you are not the person who engaged in the fraud or the underlying bad activity in the first place. In the paper, we calculate some statistics for how large the losses are that insiders can avoid. What we find is that by selling at the outset of the investigation when they first learn of it, they’re able to avoid losses in excess of 15% over a six-month period. And that’s only six months after the investigation starts.

That to us suggests that the general counsels and the boards need to step up their game in terms of clamping down on trading by those who are aware of investigations into their firms. The issue of large losses is substantial and kind of eye opening. But it is intuitive when you walk through it. The SEC is investigating you for something very bad. You now have a six-month window to sell your shares before that hits the news. Right? So when it hits the news — in the case of Under Armour [the stock price] went down 19%. This was also the case with CBS and Les Moonves; the stock price went down a substantial amount and generated lawsuits. When one recognizes that the average investigation that is material has a stock price drop of 10%, loss avoidance of 15% seems to be in line with the notion that these investigations are in fact material.

Knowledge at Wharton: What policy recommendations follow from what you have just been saying?

Taylor: Here the recommendation is for the SEC. I don’t really think it’s necessarily a disclosure route. The SEC provided us with the data. They have even richer data: key dates of the investigations; they know when the subpoenas are served; whom they’re served to; various points in time of the investigation of when they make their discoveries, when they alert the board, when they alert certain managers. For each of those dates, I think our research would say the SEC has sufficient grounds to investigate trading of officers and directors at the firm — and even potentially outsiders, such as outside legal counsel, for example — around those dates.

Our recommendation would be for the SEC to drill down during each investigation — even if it’s not an insider trading investigation — to drill down, to look, to investigate insider trading in conjunction with the actual investigation. Look to see whether the outside law firm was involved, if there was trading there. Look to see whether there’s trading by lower-level employees, by officers and directors and potentially come up with additional charges based on what they find. Maybe they will find that one of the employees who was in charge of delivering the subpoena actually was one of the individuals who traded. With every investigation, the research suggests, comes the opportunity for opportunistic trading. So consequently, the SEC has justification for investigating trading with every investigation that they open.

With every investigation, the research suggests, comes the opportunity for opportunistic trading. The SEC has justification for investigating trading with every investigation that they open.

Knowledge at Wharton: What future research do you plan to pursue based on what you’ve learned?

Taylor: Getting unique data from the SEC in terms of the total set of formal investigations over a decade and a half has opened some opportunities for future research, even moving beyond insider trading. One can start to map out what are the SEC’s enforcement preferences. Whom does the SEC tend to go after earlier? Does the SEC seem to respond to media articles, so if there are media articles that allege bad behavior, does that trigger an investigation? What are the triggers of formal investigations that we could potentially map out from observable firm characteristics? What are the characteristics of firms that lead to a higher probability of an SEC investigation?

We’ve also recently received data from the SEC on Wells notices. These notices are an even stronger indicator of something coming down the pipeline than an undisclosed investigation. Wells notices are served to individuals and entities against whom the SEC has a high likelihood of pursuing an enforcement action; they provide the recipient with an opportunity to respond to the allegations before the case is brought.

That speaks to the Lionsgate ruling that motivated our research. We’re finding some evidence that Wells notices served against entities have a very high rate of enforcement. More than 85% of Wells notices served to companies get enforced. These are the egregious cases. And so here again, because of that court ruling, are firms disclosing the Wells notices? Do we see trading in the intervening period between when they received the Wells notice and when they disclosed the Wells notice? Again, we can ask, now that we have the full set of Wells notices, what types of firms receive or tend to receive Wells notices? We can try and get a sense of the types of firms that may be heavily scrutinized by the SEC.

So, there are lots of questions that are going to be coming down the pipeline. Readers and listeners [of Knowledge at Wharton] should stay tuned for more.