On October 12, 2015, an activist hedge fund we’ll call John Doe Management filed a Form 13D, disclosing 5.5 percent ownership and an intent to pursue an activist campaign in a target firm we’ll call Industrial Corp (IC). The next day, IC’s stock rose 7 percent. John Doe Management’s last purchase of IC prior to filing the 13D occurred on October 3. The following day, an IP address owned by a different hedge fund, referred to here as AAA Group, suddenly became interested in IC, logging onto the Securities and Exchange Commission’s EDGAR database and downloading IC’s financial statements. This IP address had not accessed any information pertaining to IC in the preceding months.

Perhaps this was merely a coincidence. However, just two months earlier, on August 15, 2015, someone from the same IP address belonging to AAA Group went to EDGAR and displayed a similarly sudden interest in the financial statements of what we’ll call Aviation Management (AM). The very next day, John Doe Management again filed a 13D disclosing an activist stake in AM, and AM’s stock rose by 6 percent. One might suggest that AAA Group is highly skilled at predicting activist campaigns, but the only campaigns that it seems to predict are those of John Doe Management.

Although we have changed the names and the dates in the above scenario, it describes actual events. They are, in fact, typical of many activist campaigns, where a particular outside investor’s IP address consistently predicts the targets of a given activist.

Why might an activist be willing to leak their pending campaign information? Over the course of a campaign, activists need to build a coalition of shareholders to win any voting disagreements with the incumbent management team. By leaking this information, and allowing the informed shareholders to build positions in the targeted stock ahead of the disclosure-related price jump, activists have added a friendly base of voting support to their campaign, increasing their likelihood of winning any fight with managers (Dimson, Karakas, and Li, 2015; Crane, Koch, and Michenaud, 2019; Brav, Jiang, Li, and Pennington, 2019; Li and Hi, 2020).

In a new study, we identify institutional investors with non-public information about pending activist 13D filings. Because these filings typically cause a target’s stock price to rise, prior knowledge of these events is immensely valuable. The records of SEC EDGAR downloads present a unique opportunity to investigate this type of leaking and potential informed trading. Rarely in finance research can we document whether traders actually have valuable information, or what that information might be, but the EDGAR log files allow us to identify the link between an activist and the investment firm that persistently has non-public information pertaining to that activist.

We investigate search activity on the SEC’s EDGAR website during the 10 days prior to an activist campaign disclosure (once passing the 5 percent ownership threshold, activists have 10 days to disclose the ownership via the 13D). The search activity is reported by IP address, which we then match to specific institutional investors. We look for investors that download information pertaining to the activist’s target during this 10-day period, but that have not accessed information on the same firm during the preceding 50 days. Our attention is therefore focused on investors that display a sudden interest in a specific firm immediately prior to a 13D filing.

We find that certain institutional investors are especially adept at predicting the targets of a specific activist, downloading information pertaining to multiple campaign targets from that activist during the 10-day window prior to public disclosure. This pattern of informed EDGAR access suggests that these investors have information pertaining to the activist’s targets. We note that these informed investors are unaffiliated with the activist and are not named on the 13D filing. After establishing the presence of these investors, we next analyze the effects that their presence has on the market and on the activist’s strategy.

We first define our measure of interest. We consider an IP address to be a Suspect IP if 1) it downloads information pertaining to a campaign target during the 10-day window for at least two campaigns belonging to a single activist, 2) it has not accessed information on that same target firm for the 50 days preceding the 10-day window, and 3) it belongs to an investment firm, which includes hedge funds, mutual funds, and investment banks. Our variable of interest in our main tests, run at the activist campaign level, has  a value of one if a Suspect IP is present at the campaign.

Our analysis begins by looking at the market trading activity in the 10-day pre-13D window. If the activist has leaked the plans for the activist campaign, we would expect to see substantially greater trading activity in the stock of the targeted firm. We indeed find this to be the case: The presence of a Suspect IP is associated with significantly increased trading volume in the underlying stock; average turnover increases by approximately 0.5 percent of shares outstanding.

We next examine the post-13D returns associated with the Suspect IP campaigns. If the leaked campaign information has any value for the receiving party, we would expect to see evidence of that. We find that the Suspect IP is associated with significantly higher returns in the 10 days after the 13D filing. Returns are 2.5-2.9 percent higher when a Suspect IP is present. To be clear, we do not argue that this result is causal. Instead, it merely documents the financial benefits to the informed investor. An investor who consistently receives these profits from the activist is more likely to support the activist in a proxy fight.

If the activists are facing difficulty earning shareholder support over the course of the campaign, they may be forced to purchase additional shares in the target firm to augment their voting power. However, their leaked plans may bring on investors friendly to their cause, reducing the likelihood of additional purchases. We find this trend to be true; the probability that activists increase their stake beyond the level reported in the initial 13D filing decreases by roughly 10 percent when a Suspect IP is present.

We might also see evidence of leaked plans in increased institutional ownership after the 13D disclosure, as the informed investors purchase stakes in the target firm. Using 13F filings, we find that institutional investors do increase their ownership more for Suspect IP campaigns between the quarter before and the quarter after the 13D filing than for non-leaked campaigns. The increase in institutional ownership is approximately 3.3-3.4 percent higher for campaigns with a Suspect IP.

Finally, we examine the effects of leaked plans on the activist’s strategy. If an activist has additional voting support from investors who bought the stock prior to the 13D disclosure, they should have more confidence in spending the formidable resources required for a proxy fight. We indeed find that this is the case; campaigns with a Suspect IP are more likely to enter into a proxy fight in the year following the campaign. Further, we also find evidence that, conditional on entering a campaign, the activist is more likely to win a board seat or otherwise accomplish its stated goals. These results speak to the activist’s incentives for leaking the plans; they allow other firms to share in the profits in exchange for additional voting support.

Is this type of shared information and subsequent trading illegal? The laws on this are murky. The SEC does not regulate this type of information; there is therefore no obligation to keep the information private, as there would be with, for example, a pending earnings report. However, the SEC does require disclosure of alliances between investors to be included on the 13D filing. If the activist and the Suspect IP have not disclosed any arrangements or agreements, they could face action from the SEC. The SEC recently opened investigations into several hedge funds for allegedly not disclosing these arrangements.

This post comes to us from professors Ryan Flugum at the University of Northern Iowa, Choonsik Lee at the University of Rhode Island, and Matthew E. Souther at the University of South Carolina’s Darla Moore School of Business. It is based on their recent article, “Shining a Light in a Dark Corner: EDGAR Search Activity Reveals the Strategically Leaked Plans of Activist Investors,” available here.