Profiting from connections: Do politicians receive stock tips from brokerage houses?

Abstract

This study investigates whether brokerage houses appear to provide stock tips to politicians. Our results indicate that trades by politicians who are politically connected to the brokerage house where the trade is executed are more profitable. Our estimates suggest that these connected trades earn an incremental 0.3% over a five-day window relative to the politician's average profitability. Given the average number of trades our sample politicians execute in a year, the 0.3% return per trade translates to an incremental $3,411 in trading profits each year. We provide additional support by investigating the frequency and differential profitability of politicians' trades immediately before the brokerage house issues a revised recommendation, as well as during a period when Goldman, Sachs & Co. was sanctioned for providing stock tips to high priority clients. Additional tests suggest that brokerages may provide stock tips to politicians in exchange for favorable legislative outcomes or political information.

Introduction

This study investigates whether brokerage houses appear to provide stock tips to politicians. Prior research focuses on the incentives that brokerage houses face to provide institutional clients with privileged access to in-house research in exchange for trading commissions earned from those institutional investors (Irvine et al., 2007; Christophe et al., 2010; Kadan et al., 2018). In this study, we are interested in traders that are unlikely to generate significant trading volume for brokerage houses but potentially can provide a different type of benefit: political favors granted by members of Congress. The prior work on politicians' stock portfolios has focused on superior profitability stemming from their use of “political intelligence” obtained through their governmental position (Ziobrowski et al. 2004, 2011). By contrast, we focus on politicians’ preferential access to broker-supplied information.¹

We maintain that brokerage houses face strong incentives to share information with select politicians and that this practice complements their overall political strategies. Researchers generally rely on more observable political strategies (e.g., campaign contributions) to study the economic benefits that accrue to politically active organizations but recognize that many organizations have a menu of less observable political tactics available to them (Cooper et al., 2010; Faccio 2006; Faccio et al., 2006). We expect that providing stock tips allows brokers to further cultivate their relationships with members of Congress, and unlike campaign contributions, results in a direct financial benefit to legislators.

We investigate this question by focusing on politically connected brokerage houses, where a political connection is identified when a politician receives any campaign contributions during the year from a Political Action Committee (PAC) sponsored by the brokerage house.² If brokerages tip politicians as a means of reinforcing connections, we predict that a politician's trades executed through brokerage houses connected to the politician will be more profitable than a politician's trades executed through brokerage houses without these political connections. Our findings indicate that, on average, the excess return in the 5-day window following the trade is approximately 30 basis points higher for connected trades than non-connected trades. Given that each year the typical politician in our sample executes approximately 103 connected trades at an average dollar value of $10,800 per trade, the 0.3% return per trade translates to an incremental $3,411 in trading profits per year. This finding is robust to including controls for alternative sources of information, firm characteristics, and politician, firm, and year fixed effects. We document that this increased profitability disappears in the period surrounding the Stop Trading on Congressional Knowledge (STOCK) Act's passage when politicians faced heightened public scrutiny of their trading activities. However, connected trades are incrementally profitable both prior to as well as after the period surrounding the STOCK Act's passage. We provide additional support for our inferences by investigating the frequency and differential profitability of trades executed by politicians immediately before the brokerage house issues a revised recommendation, as well as during a period when Goldman, Sachs & Co. was sanctioned for providing stock tips to high priority clients.

In additional analyses, we investigate two potential political favors brokerages could receive for providing stock tips. One potential favor is access to “political intelligence” (i.e., political information, see Christensen et al., 2017). A second potential favor is the opportunity to influence legislative outcomes for the brokerage house industry and/or regulatory enforcement of brokerage houses (e.g., Brown et al., 2015; Cohen and Malloy 2014; Cooper et al., 2010; Correia 2014; Mehta et al., 2020; Yu and Yu 2011). To provide some initial insights on these two potential political favors, we investigate whether cross-sectional variation in the profitability of politicians' trades is consistent with brokerages providing stock tips to politicians in a position to provide these benefits. Specifically, we examine whether the effect of political connections on the profitability of trades varies systematically with the politicians’ access to political information and ability to influence legislative and regulatory outcomes for brokerages. Generally, we expect that brokerage houses are more likely to favor politicians in more powerful positions, as they have greater insights into upcoming policy changes and potentially more influence over policy developments (Christensen et al., 2017). Using a coarse measure of political power, we find limited evidence that powerful politicians execute even more profitable trades when connected to the brokerage house through which the trade is executed. Given this limited evidence, we perform two additional tests to better understand the potential political favors brokerages could receive from providing stock tips.

First, we focus on politicians that can influence more directly legislative and regulatory outcomes for the brokerage house. Specifically, if brokerages obtain legislative or regulatory benefits from tipping politicians, we expect brokerage houses to target those members of Congress that have direct oversight over the legislation and regulation specific to the banking sector. Consistent with this prediction, we find evidence that the profitability of connected trades is greater when the politician serves on the House financial services committee or the Senate banking committee. These findings raise questions for future research about whether brokerages provide stock tips in exchange for other non-monetary benefits outside of the political arena.

Second, we perform additional analysis where we investigate a possible “quid pro quo” information exchange between brokerages and connected politicians. For this analysis, we focus on the setting of U.S. comprehensive healthcare reform because provisions were modified frequently throughout the debate, and early access to private political information would result in significant benefits to connected market participants (Schweizer 2011). Christensen et al. (2017) use this setting and document that brokerage houses that were connected to politicians throughout the healthcare debate issued more profitable recommendations on healthcare stocks prior to public hearings and roll call votes (i.e., during private information windows). If brokerage houses provide connected politicians with stock tips in exchange for healthcare policy news, we expect connected politicians to execute more profitable trades for stocks unaffected by healthcare reform (i.e., non-healthcare stocks). During the comprehensive healthcare reform period, we find that politicians connected to brokerage houses are more likely to trade in non-healthcare stocks than politicians not connected to brokerage houses and that these trades are more profitable. These results suggest brokerage houses that were likely to have received material information from politicians about the progress of comprehensive healthcare reform also provided these politicians with value-relevant information about unaffected stocks. Taken together, our results suggest that one motivation to providing stock tips to politicians is the granting of political favors, namely reduced regulatory oversight and scrutiny as well as access to inside policy information (e.g., “political intelligence”). It is important to note, however, that these tests are indirect. Specifically, our tests are unable to pinpoint this exchange of stock tips for political favors, and thus an important caveat is that our indirect evidence is merely suggestive of a quid pro quo exchange.

This study contributes to several streams of the academic literature. First, we contribute to a line of research that studies the timing and profitability of trades in advance of sell-side research. These studies document the returns that accrue to brokers directly through market making activities (Juergens and Lindsey 2009), the privileged access to sell-side research that some investors enjoy by attending analyst-hosted investor conferences (Markov et al., 2017), as well as the profitability of trades for institutional clients with privileged access to analysts’ research (Irvine et al., 2007; Christophe et al., 2010; Kadan et al., 2018). We document evidence consistent with one class of individuals receiving specialized services from brokerage houses. While the trading volume generated by institutional clients provides a reasonable incentive for analysts to invest in information generation and selective dissemination, it is unlikely that the volume generated from Congressional trading activity provides similar incentives. We contribute to this line of research by proposing two non-monetary benefits, namely access to private political information and reduced regulatory oversight, for providing privileged access to sell-side research. In addition, our evidence on the trading behavior before recommendation revisions suggests that existing regulations and safeguards in place at brokerages may be insufficient to prevent insider trading.

Second, we contribute to research that investigates repeated information exchange and relationship building. Prior studies have focused on the relationship between firm management, brokerage houses, analysts and/or investors. These studies examine various mechanisms through which relationships are formed and the effects of these relationships on the flow of information and trading profits (e.g., Bushee and Noe 2000; Guiso et al., 2004; Hong et al., 2004; Hong et al., 2005; Gruber 2005; Cohen et al., 2008; 2010; Mayew 2008). We contribute to this literature by identifying stock tips as a potential mechanism beyond campaign contributions that brokerages use to build and maintain their relationships with politicians.³

Third, we contribute to the literature that studies the profitability of Congressional trading behavior. Ziobrowski et al. (2004) and Ziobrowski et al. (2011) document superior profitability of trades by members of the Senate and House, respectively, presumably due to private information gained during their work in Congress. Their evidence coincided with heightened public scrutiny over Congressional trading which culminated in the passage of the Stop Trading on Congressional Knowledge Act in early 2012. Although the STOCK Act was designed to eliminate the trading benefits associated with being a member of Congress, our findings raise the possibility that politicians are still able to personally benefit from their access to private political information by sharing it with outside parties in exchange for other information. Notably, we find that the disclosure of the brokerage account at which a trade is executed drops by half following the STOCK Act, making detection of connected trades more difficult in more recent years. Based on our findings, we recommend that regulators consider requiring politicians to provide additional disclosures for their trading activities, specifically the brokerage house through which each trade is executed and whether the politician receives campaign support from that brokerage house. While campaign contributions are reported to the Federal Election Commission, including this information in politicians' Personal Financial Disclosures would reduce information processing costs. Combined, these additional disclosures would provide regulators increased transparency on politicians’ trading activities in an effort to prevent politicians from profiting from their positions in Congress.

The paper proceeds as follows. Section 2 outlines the background for our study and motivates our research question. Section 3 discusses the data and sample used to test our hypotheses. Our main analysis for the overall profitability of Congressional trades is contained in Section 4, followed by additional analyses on trading before information events and the types of political favors received in exchange for stock tips in Sections 5 and 6, respectively. Section 7 concludes.