Print
E-MailVyacheslav Fos is Assistant Professor of Finance at Boston College Carroll School of Management. This post is based on a recent paper authored by Professor Fos; Kerry Back, J. Howard Creekmore Professor of Finance at Rice University Jones Graduate School of Business; Pierre Collin-Dufresne, Professor of Finance at the Swiss Finance Institute at École Polytechnique Fédérale de Lausanne; Tao Li, Associate Professor of Finance at City University of Hong Kong; and Alexander Ljungqvist, Ira Rennert Professor of Finance at NYU Stern School of Business. Related research from the Program on Corporate Governance includes The Long-Term Effects of Hedge Fund Activism by Lucian Bebchuk, Alon Brav, and Wei Jiang (discussed on the Forum here); and Pre-Disclosure Accumulations by Activist Investors: Evidence and Policy by Lucian Bebchuk, Alon Brav, Robert J. Jackson Jr., and Wei Jiang.
Activist shareholders play an important role in modern corporate governance. The Economist describes them as “capitalism’s unlikely heroes” and reports that, between 2010 and 2014, half the companies in the S&P 500 index had an activist shareholder and one in seven were the target of an activist campaign (The Economist, February 7th 2015).
Activism comes in many forms. Perhaps the best known involves hedge funds accumulating stakes in firms with the intention to create value by influencing management. Prominent examples include William Ackman, Carl Icahn, Daniel Loeb, and Nelson Peltz. Existing shareholders can turn from being passive to being active when they recognize an opportunity for enhancing the value of their holdings. CALPERS and the Norwegian sovereign wealth fund are well known examples of this form of activism. And activist short-sellers can take actions so as to reduce firm value to benefit their short positions. For an example, see Bloomberg Business on how “Hedge funds found a new way to attack drug companies and short their stock” (March 20th 2015), describing how some activist hedge funds challenge pharmaceutical patents in court to reduce the value of the firms owning these patents, presumably benefitting from previously established short positions. The profitability of activism for investors hinges on their ability to trade before stock prices reflect their intention to become active. Thus, there is a fundamental link between market conditions (liquidity), activism, and firm value.
Liquidity can have both harmful and beneficial effects on activism. Coffee (1991) and Bhide (1993) argue that higher liquidity should be associated with lower economic efficiency, because liquid markets make it easy for large shareholders to “take the Wall Street walk” (i.e., sell down their positions) rather than engage actively in a firm’s corporate governance when intervention might increase firm value. A certain level of illiquidity might then be desirable, to “lock in” large shareholders. Maug (1998) argues the opposite, pointing out that more noise trading enables a potential activist who does not already own a sizeable initial toehold to accumulate more shares and eventually become active. Thus, the size of the initial toehold seems to play a key role in the relation between liquidity and activism.
In our new paper, Activism, Strategic Trading, and Liquidity, we revisit the classic question of the relation between liquidity and economic efficiency by generalizing the activism technology. The existing literature considers only binary forms of activism, in which the activist’s action leads to a fixed increase in firm value. In reality, there are many different types of activism, including those with non-binary effort and those resulting in a non-binary effect on firm value. When an activist seeks to increase payouts (e.g., Carl Icahn and Apple), it arguably requires more effort to induce a larger change in payout policy, which leads to a larger effect on firm value. When an activist risk arbitrageur wants to influence whether an M&A deal is completed, the outcome is likely to be binary but the effort expended by the activist is continuous (Jiang, Li, and Mei, 2016). The probability that the activist is successful is an increasing function of her continuous effort. When an activist seeks to fire a CEO, the outcome is binary though effort may not be. When an activist runs a proxy contest hoping to replace incumbent directors with nominees of her choosing, the number of directors the activist can get elected arguably depends on her effort. Given a wide range of activism technologies, it is an open question whether and how the relation between market liquidity and economic efficiency depends on the activism technology.
In his seminal contribution, Kyle (1985) derives equilibrium price dynamics when a large trader possesses long-lived private information about the value of a stock that will be revealed at some known date and that is independent of the large trader’s actions. We extend the dynamic version of the Kyle (1985) model to a large trader (the potential activist) who can affect the firm’s liquidation value by expending costly effort, but who can also decide to walk away. We work in continuous time, because similar to Back’s (1992) extension of Kyle’s model to non-Gaussian distributions, it affords tractability. We obtain the equilibrium in closed form for a general form of activism technology, which leads to several new insights.
First, the underlying nature of the effort cost function, in particular whether it is binary or continuous, plays a crucial role in the relation between liquidity and activism. In our model, an increase in noise trading volatility increases the variance of the activist’s terminal stake (because the activist on average buys if the price is depressed due to liquidity-motivated selling and on average sells if the price is inflated due to liquidity-motivated buying). As a result, it leads to higher (lower) expected effort if future effort is convex (concave) in the stake accumulated by the activist, irrespective of the activist’s initial stake size. This does not apply to the binary case, in which effort is a step function and hence neither convex nor concave. In the binary case, the relation between liquidity and activism is similar to that obtained by Maug (1998) in a one-period model. Thus, the effect of noise trading on activism can differ considerably depending on whether the activism technology is binary or continuous.
A second new insight is that an increase in noise trading volatility does not necessarily improve market liquidity. This contrasts with the standard Kyle (1985) model in which market depth is an increasing function of noise trading volatility and a decreasing function of private information. The reason for the difference is that private information and noise trading volatility are linked in our model. An increase in noise trading volatility increases the variance of the activist’s terminal stake due to her endogenous trading, as discussed above. Because the size of the activist’s stake is private information, an increase in noise trading volatility increases adverse selection for market makers. This effect on adverse selection can more than offset the natural increase in market depth stemming from increased noise trading. Specifically, an increase in noise trading volatility reduces market depth in the binary model when the mean of the activist’s initial stake is either very small or very large, and the same occurs in the continuous model when the derivative of the firm value as a function of the activist’s stake size is convex in the stake size. Overall, that activism also affects market liquidity is a link that has not been emphasized in the literature, where liquidity is typically taken to be exogenous and one to one with the amount of noise trading.
A third new insight is that, because both market liquidity and activism are jointly determined, the cross-sectional relation between them depends on the source of variation in either. The model parameters that can be varied to change market liquidity and activism are the noise trading volatility, the activist’s productivity, and the mean and variance of the activist’s initial stake. Changes in each of these parameters are also naturally related to various policy measures (such as position disclosure rules, length of the trading window, or the legal environment). The comparative statics of our model are hence informative about the potential implications of changes in these policy measures. To illustrate how the relation between market liquidity and activism depends on the source of variation in each, consider three examples. First, and as explained earlier, variation in noise trading volatility, which could be caused by a new trading tax or by a change in the length of the trading window, can produce either a positive or a negative cross-sectional relation between liquidity and activism, depending on the nature of the activism technology and, possibly, the size of the initial stake. Second, less uncertainty about the activist’s position, which could be caused by increased disclosure requirements, always improves market liquidity since it reduces adverse selection risk, but its effect on activism can be positive or negative depending on the nature of the activism. Finally, variation in the activist’s productivity, which could be caused by a change in the legal environment, induces a negative cross-sectional relation between activism and liquidity, because a greater scope for value-changing activity by the activist increases the importance of private information regarding the activist’s intentions (which depend endogenously on her holdings of the firm’s stock) and hence reduces liquidity. Thus, which parameter is varied (or which policy change is considered) has different implications for the cross-sectional relation between activism and market liquidity.
The full paper is available for download here.
