Executive and non-executive directors with access to private and price-sensitive information have, as insiders, an advantage over other investors when trading in the shares of their firm. Even if insiders trade after, rather than before, the release of price-sensitive information on, for example, earnings or dividends, their trades may integrate non-disclosed information about the firm’s prospects. Indeed, these trades may also reflect broader information that is not firm-specific, such as data on peer companies and competitors, and on industry trends that insiders may hold given their direct or indirect connections to other corporate boards. This information is likely to generate gains for those trading on it.
In our recent paper entitled 'Insider Trading and Networked Directors', we apply network analysis to map the corporate connections of executive and non-executive directors, and to capture insiders’ access to non-firm-specific information. By means of network centrality measures (we focus on eigenvector centrality), we examine the relation between director networks and insider trading in the UK. We analyze the share transactions carried out by CEOs and chairmen, as well as other executive and non-executive directors. We focus on how director centrality within director networks affects the following two fundamental aspects of insider trading: (1) the market reaction around the insider trading announcement day; and (2) the frequency, value, and profitability of insider trading. We distinguish between insider transactions of the following types: purchases and sales, transactions by the various insiders (eg, the CEO or the chairman), as well as routine versus non-routine, ie, opportunistic trades. Since this study focuses on the information content of insider trading, we are particularly interested in opportunistic trades.
We find that directors with stronger connections—the ones who hold board positions in several firms—possess more non-firm-specific information as their insider trades trigger significantly higher abnormal returns. Importantly, this positive relationship between network characteristics and the market reaction is only observed for opportunistic purchase transactions (and not for routine trades). Our results are robust to a number of further analyses that include alternative network measures (degree, closeness, and betweenness) and insider trading measures (transaction sequences, clustered transactions, alternative event windows). They are also robust to the use of different subsamples (option-related transactions, unique daily transactions), analyses that control for important corporate events (eg, transactions close to M&A announcements), and to those that address endogeneity concerns. Moreover, when directors, who sit as executives or non-executives on the boards of multiple companies, carry out a sequence of trades in the shares of these companies, the market reacts stronger to the later transactions than to the initial transaction. Surprisingly, well-connected directors (who have several board positions in other firms) trade less frequently and for smaller total values annually. However, these directors make greater trading profits than the less well-connected directors.
This study makes the following four major contributions to the literature. First, it systematically examines the relation between insider trading and director networks captured by graph-theoretical measures. Second, our results imply that the gains from insider trading are not only determined by access to firm-specific information as documented by the existing literature, but also by access to information that is not firm-specific, such as information on peer companies and the broader industry. As such, networks enable directors to collect that type of information, and ultimately improve their trading performance. Third, while most studies on insider trading focus on trading performance, we also investigate the trading patterns (trading frequency, sequenced transactions, clustered transactions, and the direction of sequenced transactions). We find that better-connected insiders, whose transactions induce strong positive market reactions at the announcement, are not necessarily the most active traders. As they have an informational advantage, they purchase more selectively, ie less frequently and for lower share values than the less well-connected directors. Finally, our analysis uses the partitioning of insider transactions into opportunistic and routine trades as proposed by Cohen et al. We extend Cohen et al’s partitioning by distinguishing transaction sequences from single transactions. As stated above, we find that the timing of a trade within a sequence of trades matters.
Marc Goergen is Professor of Finance at the IE Business School and a research member of the European Corporate Governance Institute (ECGI).
Luc Renneboog is Professor of Corporate Finance at the Department of Finance, Tilburg University, and a research member of the European Corporate Governance Institute (ECGI).
Yang Zhao is Lecturer in Banking and Finance at the Newcastle University Business School.