In the recent Sotheby’s case, the Delaware Court of Chancery permitted the creation of a two-tier defensive poison pill that provided for different treatment of activist and passive acquirers of stock in the art auction house. At about the same time, Valeant Pharmaceuticals and hedge fund Pershing Square Capital Management made headlines when they teamed up to bid on Allergan Inc., a smaller pharmaceutical company. Allergan’s board of directors defended against the joint bid by instituting a one-year poison pill. Pershing Square, run by activist investor Bill Ackman, then filed a lawsuit in Delaware, claiming that the structure of the poison pill may limit the ability of shareholders to call a special meeting. The lawsuit quickly settled, but the broader story has now been front-page business news for nearly two months.
These two recent corporate dramas have led to many discussions about corporate governance in mainstream news outlets and in the legal blogosphere. Importantly, these cases come in the midst of a heated stretch in the long-standing debate over corporate governance as it relates to activist shareholders. These cases have thus provided new fuel for the fire.
The debate over activist shareholders has focused on Williams Act disclosures, short term-ism, poison pills, and other related issues in corporate governance. Many academics have tended to argue for more shareholder influence over corporate affairs. As Delaware Chief Justice Leo Strine highlights in a 2014 essay in the Columbia Law Review, however, a fervent reliance on empirical studies to “prove” any position neglects the complexity of the issues at hand.
Empirical studies are extremely valuable to academics, practitioners, and judges, as they ground our thinking and can inspire action. But the underlying assumptions of any work can threaten its prescience, and the drawbacks of certain empirical studies tend to get lost in the noise of a clamorous debate. This blog post highlights three fundamental concerns with empiric-based legal storytelling on both sides of the activist debates:
(1) Averages are simplifications;
(2) Outliers require careful treatment; and
(3) “Long” and “short” are relative, malleable terms.
(1) Averages are simplifications. By definition, averages are not descriptive with respect to variation across data points (e.g. range, standard deviation, etc.). Information is lost when experts rely on the “average” impact of any given cause.
A business school professor provided one of my classes with a simple example: A beverage company performs market research to drive the launch of a new tea product. If 50% of customers ask for hot tea, and the other 50% ask for cold tea, the company should not develop lukewarm tea. We would expect those sales to be tepid, at best.
The clearest examples of the potential misuse of averages in governance debates are the articles that seek to demonstrate the impact of different governance events on firm value. For example, a study might find that shareholder value grows on average when activists make proposals. But such a sample could include several firms with marginal gains in value, and one firm whose reputation – and with it, shareholder value – was destroyed by an activist campaign.
In fact, this problem is demonstrated crisply by one frequently cited study related to the share price impact of 13D filings. In this study, there was a 14.3% average abnormal return for target firms 12 months after a hedge fund’s 13D filing for an accumulation of 5% or more of a firm’s shares. But the median firm saw no material gains, and many firms had steeply negative abnormal returns. As Professor John C. Coffee Jr. describes it, “the target firm stalked by the activist fund faces an outcome that may either be a feast or a famine.” Such a scenario warrants a nuanced weighing of outsized returns against corporate instability. Unfortunately, the current debate doesn’t seem to strike that tone.
(2) Outliers require careful treatment. Just as means and medians can gloss over nuance, the way research examines (or fails to examine) outliers can also lead to relatively superficial analysis. Basic statistical methods – and the desire to easily articulate results – often dismiss or otherwise ignore extreme cases as anomalous.
We can define outliers, here, both quantitatively (as in the “feast or famine” example, above) and qualitatively (the Valient-Allergan example represents an outlier in the sense that the activist behaviors are pushing up against legal boundaries). But when it comes to our legal system, the unusual cases tend to be precisely the ones at the heart of a debate. People instinctively respond differently to corporate transactions that generate outsized returns for certain shareholders, change the culture and trajectory of family-owned or iconic brands, transform innovative companies into cash cows, or otherwise impact the social and economic landscape in a way that run-of-the-mill M&A deals simple don’t.
Outliers, here, are thus not a mere statistical challenge. Rather, they may represent the battleground itself. Most empirical studies do not – or cannot – properly incorporate such quantitatively and qualitatively extreme cases.
(3) “Long” and “short” are relative, malleable terms. Short-termism – the concept that some shareholders take a myopic view of company performance – is ever-present as a talking point on all sides of the activist debate. Harvard Law Professor Lucian Bebchuk has implied that long term can mean anything from two-and-a-half (FN 40) to five years (see generally this 2013 study). Harvard Business School Professor Robert Pozen recently argued that the “war” on corporate short-termism is misdirected, pointing to oft-cited studies that hedge funds on average hold stocks for one or two years. Presumably, Pozen thinks that a one- or two-year period is not short-term. On the other side of the debate, Martin Lipton of Wachtell, Lipton, Rosen, & Katz has also referenced numerous studies with time horizons between two and five years. But how long is long for securities analysis?
One could argue that the debate is anchored downward because of the presence of high-frequency traders acting on the order of milliseconds. Compared to holding a stock for fractions of a second, holding an asset for multiple quarters or years appears quite long-term. At the other extreme, Joe Tsai, co-founder of the Chinese e-commerce giant Alibaba, wrote in a blog post related to Alibaba’s IPO that the firm required a consolidated ownership structure because their goal was to thrive for over one hundred years.
The answer probably lies somewhere between those poles. Then-Vice Chancellor, now-Delaware Chief Justice, Leo Strine has criticized Lucian Bebchuk’s view of the “long-term” as “myopic” and as “sitcom-length rather than motion-picture-length“ (see FN 40). It seems intuitive that great firms, which have survived decades of competition, seek strategies that will be accretive over a longer timeframe than one or two or three years. A pharmaceutical CEO, for example, regularly makes decisions that will not impact customers for upwards of 10 years (i.e. the time it takes for a drug to go from discovery to FDA approval).
Academics might respond, in turn, that for practical purposes it is too difficult to develop airtight quantitative studies over a horizon beyond a few years; there are simply too many variables, too much “noise,” or too many disparate data sets to piece together. This may be true, but methodological challenges should mean that we treat these studies with a grain of salt, not that we change our normative expectations.
In sum, fresh eyes reveal that high-level debates can overlook basic methodological questions. Far from a novel concept, we must remember that numbers don’t tell the whole story: quantitative rigor does not obviate human judgment, nor does it capture the ephemeral ways that market events impact society. The academy, the regulators, and the courts all embrace empirical, quantitative studies because they provide useful evidence. And yet, there is still an important place for qualitative analysis that affixes values alongside value.
What is the question we are trying to answer? How does a particular study’s methodology contribute to answering that question? How does a study’s assumptions strengthen or weaken its explanatory power? By asking basic questions such as these, we are reminded that transparency is invaluable not only for functioning capital markets, but for constructive scholarship and debate.