Standing on the Shoulders of Giants: the Effect of Passive Investors on Activism Authors: Ian R. Appel, Todd A. Gormley, and Donald B. Keim Discussant: Tianyang Zheng 1/12
Motivation Activism is inhibited by free-rider problem: Activists bear the entire cost. Yet all other shareholders share the benefit of improved firm value. Ownership by passive (index) funds has quadrupled over the past 15 years. The voting power that comes with their ownership makes them an important type of investor in corporate activism. There is debate about whether increased passive ownership will help or hamper activism. Positive: their inability to sell their positions may make them more motivated to help activist campaigns by reducing the cost of activism. Negative: conflict of interests and different goals due to different investment horizons may cause index funds reluctant to help activists. 2/12
Research Question and Results How passive investors affect activism by other investors? During 2008 to 2014, the overall effect on the likelihood of activism campaign is insignificantly negative. What type of activist campaigns get affected by more passive ownership? Campaign: Likelihood of seeking board representation and proxy fight and number of board seat sought. All other types of campaigns ~. Campaign outcome: Likelihood of proxy fight settlement and M&A related actions which improve firm value. Likelihood of all other firm policy change actions and the activists winning proxy fight ~. 3/12
Identification Strategy Endogeneity concern: omitted variable For example, passive ownership and the likelihood of activism can both be correlated with past performance. Poor past performance removed from certain index less passive ownership higher likelihood of activism. Solution: fuzzy regression discontinuity/instrument setting Russell 1000/2000 indices: the two indices include the largest 1000 firms and the next 2000 firms respectively. They are reconstructed at the end of June each year. The firms around the index cutoff have exogenous variation in passive ownership conditioning on their market cap, past index assignment and whether their market cap falls within a certain range of the 1000 th largest firm. 4/12
Banding Policy for Index Assignment Banding policy for index assignment since 2007: a stock with an end-of-may market cap below (above) the market cap of the 1,000th (1,001st) largest market cap will be included in the Russell 2000 (Russell 1000) index unless: that stock was included in the Russell 1000 (Russell 2000) last year. its market cap is not below (above) the market cap of the 1000th (1001st) largest market cap by more than 2.5% of the cumulative market cap of the Russell 3000E Index. 5/12
First Stage Regression F-stats are above 10 Conditional regression: conditional on all campaign events (13D forms). This measure has the problem of inflating the number of campaigns. 6/12
Second Stage Regression Passive ownership does not have significant effect on the likelihood of overall activism campaign. The presence of passive investor might facilitate activism: activism. Managers might internalize this possibility and act to preempt activist campaigns: activism. What are these observations? 1000 firms*7 years (2008-2014) 6797 obs? 7/12
Likelihood of Different Types of Activist Campaigns Passive investors tend to focus on board related issues so that activists might tailor their campaign to attract the support from passive investors. 8/12
Proxy Fight Likelihood and Board Seats Sought Without restricting sample to campaigns that seek board seats, results from column 4 to 6 are less informing (it is mechanically similar to the result in previous table). Proxy fights are costly: average cost over $10MM. Support from passive investors makes proxy fight less costly: actual fight. Managers might internalize this possibility and settle with activists: actual fight. Activists are more ambitious about the number of seats sought. 9/12
Proxy Fight Outcomes Managers might internalize the risk of losing the fight and choose to settle with activists: settlement. For actual fight, passive ownership has no effect on the chance of activists winning the fight. Shouldn t column 4 be significant? Why results in column 4 and 5 are both positive? I m less convinced by the authors claim that results in column 4 and 5 support their argument. Maybe apply a test for results in column 4 is larger than column 5? Similarly, samples are confusing. Should restrict sample to campaigns which involve intended proxy solicitation (this could lead to settlement or actual fight). 10/12
Non-Proxy Fight Campaign Outcomes Negative coefficient for column 3: passive investors have large ownership in both parties in the merger deal (maybe). Insignificant results from column 5 to 7: passive investors don t want to meddle with these types of campaigns. Threat of fight is not credible because potential gain is not greater than the cost of proxy fight (maybe). 11/12
Concluding Remarks This paper has done a thorough job documenting the effect of increased passive ownership on different types of activism campaigns. However, there are always two competing forces on the effect. The channel that managers try to avoids actual fight (settlement) can be further explored. It is still ambiguous that why passive funds are willing to help activists. It seems that index funds have neither cost nor benefit from supporting activism. Maybe index funds don t internalize the benefit for themselves, but they leverage their information about potential voting results to help the non-passive funds under the same family? 12/12