Exit as Governance

Traditional theories of blockholder governance have focused primarily on blockholder intervention in management decisions. However, recent theories posit that blockholders can govern firms even when they have no intervention power. These theories view blockholders as informed traders who control management through “exit,” i.e., selling a firm’s stock based on private information (Admati and Pfleiderer 2009, Edmans 2009, Edmans and Manso 2011). Blockholder exit in these models exerts downward pressure on the stock price, which hurts management through its equity interest in the firm. Management therefore wants to make sure its actions are such that blockholders are willing to stay with the firm.

When blockholders are informed traders, management undertakes productive effort and investment in order to improve firm value and dissuade blockholders from exiting. The true governance force therefore comes from the threat of blockholder exit, not actual exit. Even if no exit is observed, blockholders could be governing effectively because their exit threat is sufficient to discipline management.

In our paper, Exit as Governance: An Empirical Analysis, forthcoming in the Journal of Finance, we empirically test the governance impact of blockholder exit threats. Since threats cannot be directly observed, this study focuses instead on a key mechanism that facilitates exit threats, namely stock liquidity. The exit threat models suggest that stock liquidity enhances the power of exit threats and improves firm value. For example, in Edmans (2009), the manager is compensated on the stock price and can take fundamental actions to improve firm value. Stock liquidity encourages strategic traders to acquire more information on firm fundamentals and trade on it in larger volumes (or blocks). The manager is sensitive to the resulting stock price, and therefore takes actions to increase firm value and induce (informed) blockholders to stay. Liquidity thus enhances the power of blockholder exit threats and improves firm value. This theoretical prediction forms the basis of our empirical tests.

A full structural model of liquidity, blockholdings, and firm value is not only a difficult endeavor, but also confounded by the fact that the finance literature has not converged on a definitive measure of liquidity. We therefore bypass the structural approach entirely, appealing instead to exogenous liquidity shocks. We first examine two foreign financial crises, namely the Russian default crisis and the Asian financial crisis. These were unexpected exogenous events (from an individual firm’s perspective) whose duration was unknown at the time of their onset. A significant body of studies indicates that these events significantly decreased liquidity in the U.S. stock market (Acharya and Pedersen (2005); Chordia, Sarkar, and Subrahmanyam (2005)). We test how the association between firm value (Q) and blockholding shifts around the above exogenous liquidity shocks. We find that firms with larger block ownership were impacted significantly more during these crises. A one standard deviation increase in block ownership corresponds to a decrease of about 4.1% in Q during the crisis period, an economically large effect. Blockholder exit threats thus appear to be strongly operating in our sample.

Due to the foreign nature of the above crises, we have some confidence in positing that their impact on the U.S. stock market was primarily through stock liquidity. However, it is possible that these crises affected Q directly. We therefore pick a non-crisis candidate for an exogenous liquidity shock, namely the decimalization in early 2001 when the New York Stock Exchange and AMEX (and subsequently Nasdaq) started quoting and trading its listed issues in dollars and cents (decimalization) as opposed to increments of a sixteenth of a dollar. Decimalization is therefore a liquidity increasing shock. We expect and indeed find a significantly greater increase in firm value post-decimalization for firms with larger block holdings. The economic magnitude is also significant, with a one standard deviation increase in block ownership corresponding to an increase of about 7.1% in Q.

One concern with tests using liquidity shocks is that the observed results could simply be an artifact of an ongoing trend, i.e., if the block-ownership-firm-value association was trending downwards before the crises and upwards before decimalization. In such a case, the continuation of these trends would be an alternative explanation for our results. To alleviate this concern, we run falsification tests by constructing periods of “pseudo-shocks” to denote periods of equal length before the actual liquidity shocks (Russian, Asian Crisis, and Decimalization). We do not find any analogous significant effects around “pseudo-shocks”, lending further confidence to our interpretation that the effect of liquidity on the association between firm value and block ownership is not due to trends, but rather due to exit threats.

Finally, we conduct the same analysis for the recent U.S. financial crisis in 2008, and find strong results for this event as well. A one standard deviation increase in block ownership corresponds to a decrease of about 2.5% in Q during the crisis period. However, we are cautious in imputing this result solely to exit threats. The economic impact of the recent crisis on our sample of U.S. firms extended far beyond just liquidity shocks; the accompanying collapse of the housing sector and the economic downturn also adversely affected firms’ fundamentals (e.g., demand for their products). It is therefore difficult to attribute Q-based results around the U.S. financial crisis to just liquidity effects.

The above results on the role of stock liquidity in modulating the association between blockholdings and firm value do not conclusively implicate exit threats. Liquidity can improve firm valuation even when blockholders govern through intervention (i.e., “voice”). For example, Maug’s (1998) model shows that liquidity enhances voice because it allows the blockholder to buy additional shares at a price that does not reflect the benefits of intervention. Since both intervention and exit threats could be operating in the data, it is important to separate the two. One distinction of intervention theories from exit threat theories is that the sensitivity of managerial wealth to the stock price (i.e., stock option delta) plays no direct role in the intervention theories. By contrast, exit threat models predict that blockholder exit threat will be more effective in firms whose managers’ wealth is more sensitive to the stock price. Accordingly, we find that the impact of liquidity shocks on the blockholder-firm value association is far more pronounced for firms whose managers have a significant interest in the firm’s stock price.

We measure the management’s interest in the stock price through management’s equity incentives. However, management’s equity position also confers control rights, which could be valuable for deterring blockholder intervention. To test this possibility, we conduct the analysis with management blockholdings rather than management sensitivity to stock price, as it is the ownership of shares that confer control rights. The results indicate that management blockholdings do not drive our results. The insignificance of management blockholdings is further consistent with exit threat theories, because, except in cases of severe internal discord, management is unlikely to threaten itself with exit.

We then make another attempt to separate the threat of exit from intervention. We assume that blockholder intervention, while prevalent, is likely to be less effective when managers are more entrenched. Using an entrenchment index measure to proxy for this variation, we find that the results continue to hold even among firms with entrenched managers, suggesting exit rather than intervention as the source of our results. In addition, Admati and Pfleiderer (2009) propose that the effectiveness of exit threats lies not as much in promoting good investments by the manager as it does in deterring bad investments by the manager (by contrast, blockholder intervention should be equally effective in both cases). Admati and Pfleiderer (2009) suggest that the “bad” agency problem is more severe in cash-rich firms. Accordingly, we find that our liquidity shock results hold significantly more strongly for cash-rich firms, suggesting that governance via exit, while potentially concomitant with governance via intervention, is distinct from it.

We conclude by placing our study in the context of the corporate governance literature. Many studies examine the motivation for and the information content of blockholder exit, which is made possible by the law viewing blockholder sales as a material event and thus requiring all beneficial owners owning more than ten percent of a firm’s shares to report their sales on Forms 4 and 5.3 McCahery, Sautner, and Starks (2010) use survey evidence to show that institutions use exit as their primary governance mechanism. However, they only have survey evidence, not direct evidence. Turning to direct evidence, Boehmer and Kelley (2009) link blockholders to price informativeness by showing that institutional ownership improves price efficiency even after controlling for institutions’ trading behavior (i.e., actual exits). To the extent ownership carries with it an implicit threat of exit, that study’s findings correlate well with our study. However, that study does not show any links to firm value. On the other hand, Gallagher, Gardner, and Swan (2012) link blockholders to price informativeness and firm value; however, they do not study the effect of liquidity (in their study, liquidity is a measure of price informativeness so it is a dependent rather than independent variable), and neither do they use natural experiments. Fang, Noe, and Tice (2009) use natural experiments to study the effect of liquidity on firm value.

Our study makes a number of contributions over prior studies: we show that the effect is particularly strong when block ownership is higher (providing evidence that liquidity is acting through “governance through exit” rather than other channels), we perform falsification tests and also show that the effect is particularly strong for equity-aligned managers (a feature specific to exit theories). The results of our study complement Edmans, Fang, and Zur (2012), who use an exogenous shock to liquidity to identify a threat of exit. Specifically, they show that liquidity encourages blocks to form in the first place, and that conditional upon block formation, the blockholder is more likely to choose exit than voice.

The full paper is available for download here.