Printer Friendly

Financial conglomerates and information barriers.



     A. Organizational Structure

     B. Regulatory Challenges

        1. Legal Challenge

        2. Practical Challenge

     C. Illustrations


     A. Intended Effect

     B. Actual Effect

     C. Assessment


     A. Terms

     B. Potential Benefits

     C. Limitations

        1. Scope of Application

        2. Definitional Ambiguities

        3. Reliance on Information Barriers

        4. Offsetting Costs

     D. Assessment


     A. Methodology

        1. Underlying Intuition

        2. Illustration

        3. Limitations on Use

     B. Design

        1. Mandatory Reporting of Quantitative Metrics

           a. Preferred Model to Generate Quantitative Metrics

           b. Alternative Models to Generate Metrics

        2. Imposition of Liability

        3. Scope of Application

     C. Possible Objections

        1. Unreliability of Statistical Evidence

        2. Sufficiency of Existing Protections

        3. Risk of Gaming

        4. Risk of Statistical Dueling



Financial conglomerates--firms such as JP Morgan and Goldman Sachs--adopt an organizational structure that poses fundamental regulatory challenges. The challenges are legal and practical. They arise because, under this structure, financial conglomerates act for numerous clients across a broad and diverse range of financial activities, all the while acting as principals in a similarly broad and diverse range of activities. (1) Legally, the structure subjects firms to multiple, incompatible client duties: complying with one duty leads to violating another, almost inevitably. Practically, the structure provides firms with vast reservoirs of non-public information as well as the opportunities and incentives to exploit that information, potentially harming clients and third parties.

The primary regulatory response to these challenges is the information barrier. Also known as the Chinese wall or firewall, (2) the information barrier is designed to prevent, or at least to limit, flows of non-public information within financial conglomerates. It comprises intra-firm policies and procedures to physically and electronically isolate people and records with non-public information and thus to prevent or limit the spread of that information to other parts of a firm. (3) The information barrier addresses the legal challenge by allowing a financial conglomerate to be treated conceptually as multiple distinct firms, thus providing a basis for the enterprise as a whole to discharge the otherwise incompatible duties it owes to clients. The information barrier is intended to meet the practical challenge by, in fact, constraining intra-firm information flows, thereby preventing financial conglomerates from exploiting their opportunities to use that information, such as by trading on it. The information barrier thus stands as an essential component of the financial regulatory architecture; indeed, it is probably no exaggeration to say that financial conglomerates owe their continued existence to the information barrier.

The problem for regulators, clients, and securities markets, however, is that information barriers do not function as intended. While courts and regulators permit information barriers to resolve the legal challenge of incompatible duties, empirical and other evidence indicates that information barriers fail to meet the practical challenge. It has long been suspected, and is now demonstrated, that in important contexts information barriers have failed to prevent the spread of non-public information within financial conglomerates. (4) For example, empirical studies show that information often leaks from the investment banking units of firms to trading units, where it is used for profitable trading. (5) Information barriers are thus accorded legal effect, despite often failing to have practical effect.

More than this, information barriers are routinely used publicly as a shield by financial conglomerates when they are suspected of exploiting their clients' non-public information. Especially when suspicious trading activity occurs, financial conglomerates stand behind their information barriers, insisting on their integrity in restricting information flows. In the court of public opinion, such claims may or may not be accepted, but they are rarely disproved because information flows seldom leave a trace. (6) In light of these evidential difficulties, firms have powerful incentives to rely on information barriers to defend their conduct.

Against this backdrop, the core problem for regulators lies in detecting and proving the failure of information barriers. While regulators may inspect the policies and procedures underpinning information barriers' operation, they can do little after an event to determine whether, in fact, barriers operated as intended. A firm's trading when in possession of non-public information may be so fortuitously timed as to attract suspicion, and yet plausible alternative explanations--such as the superior intellect or skill of its traders or mere coincidence--are difficult to discredit. In the absence of direct evidence, the task of detection and proof seems beyond regulators. In consequence, suspicion remains, empirical evidence mounts, and financial conglomerates continue to trumpet the integrity of their information barriers.

Consider the following example. In early 2007, Verizon hired JP Morgan to advise on its proposed multi-billion dollar acquisition of Rural Cellular Corporation (RCC). (7) Like other financial conglomerates, JP Morgan's investment banking unit advises on such merger and acquisition transactions. The firm also had a trading operation that traded for the firm's own account as well as for firm clients. (8) At the time of its engagement, JP Morgan owned no stock in RCC. (9) However, while its investment bankers were counseling Verizon on the still-secret acquisition, JP Morgan's traders began buying stock in RCC. (10) In July that year, when Verizon publicly announced its intended acquisition of RCC, RCC's price jumped 34%. JP Morgan earned a generous return on the stake in RCC it had by then accumulated. JP Morgan's trading showed uncanny prescience, but was it the result of an information leak from the firm's bankers to its traders--and thus the failure of information barriers--or benign factors, such as coincidence, superior intellect, or trading skill?

In a front-page story the following year, The Wall Street Journal (the Journal) inquired into the transaction. (11) When questioned by Journal reporters, JP Morgan adopted the routine pose of standing behind its information barriers, claiming that these barriers had prevented information about the deal from leaking to its traders. (12) Naturally, the story had no direct evidence to the contrary, but it did disclose numerous other transactions--by a range of financial conglomerates--involving similar conduct. These firms also made fortuitously timed investments in target companies while their investment bankers were simultaneously advising the acquirers. (13) Other financial conglomerates questioned in the article stood behind their information barriers. (14) The story referred to empirical evidence showing that information barriers systematically failed in these circumstances, but no evidence could pinpoint failures by any particular firm. Regulators claimed to be investigating, (15) but have taken no public action.

This Article discusses the phenomenon of failing information barriers, explains why it occurs, and proposes a regulatory solution. By examining information barriers, the Article contributes to the long-running debate in legal and financial economic literature on the merits of information barriers in addressing the regulatory challenges of financial conglomeration. (16) To date, legal scholars have questioned the practical effectiveness of information barriers and suggested denying them legal effect, but they have not considered how to detect and prove the failure of information barriers and thereby provide incentives for firms to ensure their effectiveness. (17) Business and economic scholars have focused on establishing the existence of the phenomenon of failing information barriers, but apparently without considering regulatory reforms. (18) This Article contributes to these strands of literature primarily by explaining why information barriers fail and suggesting how regulators can address those failures.

The Article argues that both market and regulatory factors explain the failure of information barriers. (19) Market factors include the inadequacy of client discipline arising from agency costs within client corporations and the incidence of harm on third parties. The regulatory explanations stem from the difficulty of detecting and proving the ineffectiveness of information barriers, which is largely the result of the nature of information: its flow rarely leaves a trace, especially where those individuals involved collaborate and seek to avoid detection.

The Article also evaluates the so-called Volcker Rule's potential influence on information barriers' effectiveness. A core plank of the Dodd-Frank Wall Street Reform and Financial Protection Act (Dodd-Frank Act), (20) the Volcker Rule bans certain financial institutions from engaging in proprietary trading. By assessing this issue, the Article extends recent scholarship on the Volcker Rule, which has focused on the rule's implications for the financial stability of financial institutions. (21) Although primarily intended to promote financial stability, the rule's prohibition on proprietary trading will significantly reduce opportunities and incentives for financial conglomerates to use nonpublic client information in violation of information barriers, especially considering that proprietary trading was a key driver of financial conglomerates' revenues. (22) In defending the eponymously named rule, Paul Volcker, a former Chairman of the Federal Reserve, asserted that he is not "so naive as to think that, even with the best efforts of boards and management, so-called information barriers can remain impermeable against the pressures to seek maximum profit and personal remuneration." (23) Despite its force, the Article argues the Volcker Rule is inadequate to deal with the phenomenon of failing information barriers.

The proposed strategy would both strengthen market discipline and allow regulators to detect and prove the failure of information barriers. In doing so, it would provide incentives for firms to bolster their information barriers and deter the use of non-public client information in violation of those barriers. Drawing on recent literature in forensic finance,* (24) the proposal relies on the capacity of statistical analysis to rule out benign explanations--such as coincidence, superior intellect, or trading skill--that might otherwise explain financial conglomerates' fortuitous trading results. Credibly ruling out these explanations as implausible would, in particular contexts, allow regulators and markets to infer that a firm's superior trading results were the result of informational advantage--and thus the failure of its information barriers. To avoid this prospect, financial conglomerates would be led to buttress their information barriers.

The methodology involves calculating various trading returns for a given financial conglomerate in a particular period and then comparing those returns. The methodology reflects the underlying intuition that, barring some plausible explanation (such as coincidence or skill), the return a firm earns from trades on securities about which it possesses non-public client information will not diverge significantly from various benchmark returns earned without non-public information--provided the firm's information barriers are effective. But if a firm systematically over-performs only when it (or any of its business units) possesses non-public information and competing explanations for that performance are ruled out, one can infer the failure of the firm's information barriers. The methodology will apply to particular types of "information events," instances when the financial conglomerate holds an information advantage, such as when it is engaged to advise on a merger and acquisition (M&A) deal or lends funds to a company.

The primary trading return calculated is the amount earned by a financial conglomerate over a period of, say, 12 months from trading in stocks when the firm (or, more specifically, a unit within it) possessed non-public information. That return would be compared against three benchmark returns calculated for the same time period, making adjustments to each return for market-wide movements. If the firm earned abnormal positive returns only for stock about which it possessed non-public information, only at times it possessed such information, and only when comparable firms failed to replicate that performance, important inferences would follow. One could rule out the possibilities that the abnormal return resulted from the intellect or skills of the firm's traders, from special features of the stocks traded, or from mere coincidence. Prima facie, the only plausible inference would be the firm's use of non-public information and, correspondingly, the failure of its information barriers. (25)

While the proposed use of statistical analysis to identify failing information barriers is novel, the use of such analysis in other fields to provide legal proof is not. (26) The law often turns to statistical analysis to provide proof in contexts where direct evidence of wrongdoing is seldom available. For example, courts and regulators rely on such analysis to help prove systemic employment discrimination, racial polarization in voting, the existence of forgeries, and the efficacy of treatments in clinical trials. (27) Statistical analysis may provide legal proof of wrongdoing.

The strategy would have dual prongs. The first would compel financial conglomerates to publicly report quantitative metrics--comprising the financial return and benchmarks referred to above--from which the failure of information barriers could be inferred statistically. By mandating such disclosure, the proposal would harness market forces, creating incentives for clients to discipline financial conglomerates' wrongdoing. The second prong would empower regulators to fine financial conglomerates that fail to rebut an inference that their information barriers failed. By imposing potential liability, the proposal recognizes the inadequacy of market forces to constrain financial conglomerates, stemming from the third-party harm that failing information barriers may cause. (28) The strategy would thus address the reasons for failing information barriers and undermine the credibility of any unjustified denials issued by financial conglomerates.

Of course, the strategy has limitations. It would not verify whether information barriers failed in a specific transaction; while financial conglomerate-specific, the required analysis necessarily applies across numerous transactions for a given type of information event. Moreover, it would not detect all systemic failures of information barriers because it may be applied only to those information events where financial conglomerates are known to possess non-public information. It might give rise to "statistical dueling" among experts and even "gaming" by financial conglomerates themselves. This Article discusses these and other potential limitations of the proposal, explaining why the proposal nevertheless holds promise in addressing the regulatory challenges of financial conglomeration.

This Article proceeds in four parts. Part II describes the organizational structure of financial conglomerates and the legal and practical regulatory challenges that the structure produces. Part III considers the phenomenon of failing information barriers and analyzes the factors explaining its existence. Part IV evaluates the newly adopted Volcker Rule as a measure for reducing the harms caused by failing information barriers. The proposed strategy for detecting and proving failing information barriers is described and possible objections discussed in Part V. A brief conclusion follows.


A. Organizational Structure

The term financial conglomerate applies to firms according to their organizational structure, rather than their legal designation. Financial conglomerates are diversified financial institutions comprising a network of subsidiaries that fall under the control of a holding company. (29) They engage in a broad and diverse range of financial activities, often including banking, securities, and asset management. In doing so, they act as both agents and principals, often simultaneously in a single transaction. From a functional perspective, they lend and borrow money, advise clients, and invest funds for themselves and for clients. (30) They diversify geographically as well as along product and service lines. They typically operate on a vast scale. Financial conglomerates thus engage in broad-ranging financial activities, act in multiple legal capacities, and serve numerous clients.

Financial conglomerates' typical range of activities is summarized in Figure 1. (31) They undertake securities activities, which divide generally into investment banking and trading. In their investment banking operations, they act for corporate clients on strategic transactions such as mergers and acquisitions, and securities offerings. (32) In their sales and trading operations, firms may act as brokers, executing trades and providing incidental advice. (33) They may also act as dealers, for example, by buying and selling securities on their own accounts from clients (an activity often referred to as market-making) (34) or by trading in financial instruments on their own accounts to profit from short-term price movements (an activity known as proprietary trading). Since the adoption of the Volcker Rule, certain financial institutions have been banned from engaging in proprietary trading. (35)

Financial conglomerates also typically engage in banking and asset management. As banks, they accept deposits and make loans. As asset managers, they manage pools of funds for outsider investors and themselves and invest according to particular strategies (as in hedge funds and private equity funds) or in particular asset classes (such as real estate or fixed income). (36)

Major financial conglomerates are typically controlled by a bank holding company that qualifies as a financial holding company. Prior to the passage of the Gramm-Leach-Bliley Act of 1999, (38) which partially repealed the Glass-Steagall Act, (39) financial conglomerates were either designated as bank holding companies (generally acting as commercial banks) or broker-dealer firms (generally acting as investment banks). The Gramm-Leach-Bliley Act allowed bank holding companies to engage in a broader range of financial activities if they qualified as financial holding companies. (40) Although investment banks continued to hold a distinct legal status after the passage of the Gramm-Leach-Bliley Act, during the financial crisis of 2007-09 the major investment banks converted to bank holding companies, were acquired by such companies, or declared bankruptcy. (41) Today, most major financial conglomerates are controlled by bank holding companies that qualify as financial holding companies. (42)

B. Regulatory Challenges

The organizational structure of financial conglomerates produces significant regulatory challenges, both legal and practical. (43)

1. Legal Challenge

The legal challenge of financial conglomeration arises from the operation of agency law, the common law doctrine regulating relations between agents and the actors (principals) on whose behalf and subject to whose control agents act. (44) The principles of agency law, which developed before the existence of financial conglomerates, pay no heed to the possibility of large, multi-function firms serving numerous clients and often simultaneously acting as both agents and principals. (45) The legal challenge generally arises because of the inflexibility of agency law when faced with financial conglomerates' organizational structure and the attribution of information in possession of one part of a firm to the firm itself. (46) It gives rise to conflicts of duty and conflicts of interest.

When applied to the financial conglomerate, agency law may impose fundamentally incompatible duties on the firm. By acting as an agent for clients, the firm becomes a repository of vast amounts of non-public information (largely garnered from these clients). (47) As explained above, the financial conglomerate can act as an agent in its investment banking, trading, and asset management activities. (48) As an agent, the firm is duty-bound to disclose material information in its possession to those clients. (49) Simultaneously, however, the financial conglomerate is duty-bound to keep confidential non-public information disclosed by its clients. (50) Routinely, information divulged by one client is material to the firm's representation of another client. According to agency law, the firm is obliged to disclose information in its possession to that other client. However, to satisfy a duty of disclosure to the latter client is to violate a duty of confidence to the former client. (51) The existence of incompatible duties occurs many times over in the financial conglomerate, considering that the firm will often represent many thousands of clients at any one time, (52) owing manifold duties in doing so. (53) The financial conglomerate thus becomes tangled in a web of potentially incompatible, or conflicting, duties.

The overlay of federal securities law further complicates the financial conglomerate's position. Under anti-fraud provisions, firms owe public investors a duty to abstain from trading using material, non-public information. (54) Thus, to disclose information in discharge of a duty of disclosure to one client may be to violate statutory anti-fraud provisions, exposing the firm to public enforcement action. (55)

The classic case of incompatible, or conflicting, duties occurs when a financial conglomerate acquires non-public information from an investment-banking client while also advising a retail brokerage client on acquiring stock in that investment-banking client. (56) Disclosing to the retail client that the investment banking client will shortly announce material information, such as an M&A deal or securities offering, would violate a duty of confidence to the latter client and violate anti-fraud provisions of federal securities laws, and yet, failing to do so may violate a duty of disclosure to the former client. The firm is thus stuck between a rock and a hard place: it must determine which duty to breach or, perhaps ultimately, which client to cease serving or which function to stop providing.

In the absence of information barriers, some courts have been unwilling to bend the unyielding application of agency law. (57) In Black v. Shearson Hammill & Co., (58) a California state court refused to relieve a broker of his duty of disclosure to his brokerage client even though that duty conflicted with a duty of confidence owed to another party. The court explained:

   [W]e have been given no sufficient reason for permitting a person

   to avoid one fiduciary obligation by accepting another which

   conflicts with it.... The [fiduciary's] conflict in duties is the

   classic problem encountered by one who serves two masters. It

   should not be resolved by weighing the conflicting duties; it

   should be avoided in advance ... or terminated when it appears.


In addition to facing conflicting duties, the financial conglomerate may face a conflict of interest. Such a conflict arises between the firm's self-interest and its duty to a client, rather than between multiple client duties. A conflict of interest arises where a firm owes a duty of loyalty to a client and yet has inconsistent financial incentives, perhaps arising from a financial position it has taken. (60) The quintessential case involves a firm taking a position through its proprietary trading operations that is inconsistent with its duty of loyalty to a client; for instance, a firm's proprietary traders might buy a significant stake in a company that the firm's client is proposing to acquire. (61)

Conflicts of interest and of duty are an inevitable feature of financial conglomerates. (62) As early as 1963, when financial conglomerates were growing in size, the Securities and Exchange Commission (SEC) recognized the "multifarious possibilities of conflict of obligation or interest in matters large and small." (63) Firms themselves also recognize the legal challenge. Morgan Stanley's former chief executive officer, Philip Purcell, explained the predicament as follows: "In our business, we are surrounded by conflicts--not just conflicts between our own interests and those of our clients, but between different parts of our firms, and between the clients in one part of the firm and the clients in another." (64)

Financial conglomerates adopt measures (other than information barriers) to avoid owing conflicting duties, although their effectiveness is unsettled. (65) For instance, in their contracts with clients in various contexts, such as mergers and acquisitions and securities underwriting, financial conglomerates routinely disclaim the existence of agency relationships and fiduciary duties. (66) They also use generalized advanced disclosure provisions, seeking to avoid liability by effectively obtaining the informed consent of their clients. They may also separately incorporate units within their firm, and thereby seek to prevent attribution of information from one unit to another, as a means of avoiding liability under agency principles. Although widely used, the legitimacy of these devices lacks scholarly consensus and is rarely judicially adjudicated. (67)

Ultimately, the financial conglomerate cannot adopt the structure it does--of simultaneously providing numerous products and services to multiple clients and acting in multiple legal capacities--without risking violating its agency and other duties. In fact, courts have acknowledged that agency doctrine makes it "exceeding difficult" for a single firm to combine the activities of investment banking and trading. (68) They have also acknowledged that, without information barriers, agency doctrine would "ultimately" lead to the "disaggregation of commercial and investment banking functions from asset management." (69)

Yet the disaggregation of financial conglomerates may impose harms. (70) The economies of scale and scope that financial conglomerates can produce would be lost, (71) as would the certification effect financial conglomerates provide as underwriters. (72) That outcome would be socially undesirable in cases where those benefits exceed the costs associated with financial conglomeration. These costs are considered next.

2. Practical Challenge

The practical challenge of financial conglomeration refers to the task of restraining the misuse of non-public information, that is, of restraining the use of non-public information garnered from a client in violation of a firm's information barriers. (73) As a repository of vast amounts of non-public client information, a financial conglomerate has opportunities and incentives to further its self-interest by using that information, potentially harming clients and third parties.

The harm to clients and third parties may occur through two channels. (74) First, a client's non-public information may be used by a firm in its trading or investing decisions or leaked to another client that trades on the information. Such conduct may harm an M&A client by increasing the price the client pays to acquire a company or other asset. It might harm a brokerage client, particularly where the firm actively trades against the position of the client, by increasing the price the client pays or decreasing the price it receives. (75)

Third-party harm may also result. This claim deserves elaboration because it lies at the core of a protracted academic debate on harms related to insider trading. (76) The focus here is not on insider trading per se, but on the potential harms arising from financial conglomerates trading on information garnered from their clients. Such trading by so-called outsiders (rather than conventional insiders like directors and officers) may cause market-wide harm by lowering market liquidity, increasing trading costs, raising the cost of equity capital, and increasing volatility. (77) It may also diminish the accuracy of stock prices by adding uncertainty to the extent to which the market should discount share prices to account for the possibility that investors will trade with counterparties with an informational advantage. (78) While scholars contest the desirability of regulating insider trading, virtually no one contests regulating trading by outsiders using non-public information from their clients. (79)

The second channel through which the misuse of non-public information may impose harm concerns flows of such information in the reverse direction, that is, from traders toward advisers within a financial conglomerate. Aware of her firm's financial stake in the stock in which her client plans to invest, an adviser faces incentives to skew her advice, exacerbating agency costs with that client--that is, the costs arising from the (economic) agent's divergence of interests from those of her principal. (80) As an example, consider the 2011 acquisition of El Paso by Kinder Morgan, in which the target company (El Paso) appointed Goldman Sachs as its M&A adviser. (81) If Goldman's advisers were aware of their own firm's significant stake (worth $4 billion) in the acquirer, they would have had incentives to skew their advice to El Paso, encouraging it to complete the sale on terms more favorable to Kinder Morgan than otherwise. (82) Skewed advice harms clients by leading to inferior client outcomes and leading cautious clients, fearful of receiving biased advice, to vigilantly monitor their financial conglomerates by, for example, hiring multiple advisers or sparingly divulging non-public information to their advisers. (83) In short, the practical challenge for regulators concerns how, in fact, to prevent intra-firm flows of non-public information that may create client and third-party harms.

C. Illustrations

To illustrate the twin regulatory challenges, consider again Verizon's 2007 acquisition of RCC. (84) As adviser to Verizon, JP Morgan was duty-bound to keep confidential its client' s non-public information, including the proposed acquisition of RCC. (85) However, JP Morgan also acted as agent to multiple clients. It served clients through its asset-management activities, which are among the largest in the world. (86) It also served clients through its trading activities. The firm thus owed multiple, incompatible duties.

JP Morgan also faced opportunities to use non-public information to benefit itself at the expense of Verizon and third parties. If the non-public information about Verizon's proposed acquisition of RCC in possession of its M&A advisers spread, the firm's asset-management or trading units could have exploited it. Any trading in RCC could have increased RCC's price, making Verizon's acquisition more expensive. Similarly, if its M&A advisers were aware of the firm's holdings in RCC, that information could have swayed their advice to Verizon, magnifying the divergence of interests between the firm and its client, thus increasing agency costs.

Consider another example, in which the Australian logistics company Toll Holdings engaged Citigroup as its M&A adviser for its proposed acquisition of Patrick Corporation. (87) The business day before Toll publicly announced its acquisition of Patrick, Citigroup acquired a significant stake--on its own account--in Patrick Corporation. When the transaction was announced, Citigroup's stake spiked in value.

Under traditional agency law analysis, Citigroup's duties were similar to those of JP Morgan above. (88) As adviser to Toll Holdings it was duty-bound to protect the confidentiality of its client's proposal to acquire Patrick. Simultaneously, as an adviser to multiple other clients, it was obliged to disclose all material information in its possession; the information about Toll's acquisition would have been material to some of those other clients.

Separately, Citigroup faced an opportunity to act in its self-interest by using nonpublic information in possession of its M&A advisers to take a stake in Patrick. In fact, it did so, acquiring one million shares in Patrick the business day prior to its client's public announcement of its bid for Patrick. (89) To the extent that trading increased the market price of Patrick, leading Toll to pay more for Patrick than otherwise, Citigroup's conduct harmed its client. Moreover, if Citigroup's M&A advisers knew of their firm's stake in Patrick, they faced opportunities and incentives to skew their advice to Toll, encouraging it to overpay for Patrick because doing so would benefit their firm.

As this discussion suggests, the practical and legal challenges intertwine. The legal challenge leads, ultimately, to the disaggregation of the financial conglomerate and to the resulting loss of any economies of scope and scale that conglomeration produces. The practical challenge stems from the harms financial conglomeration creates, including direct client harms, agency costs, and third-party harms. If poorly handled, the legal challenge diminishes the benefits of conglomeration (by requiring the disaggregation of firms), while the practical challenge magnifies the costs of conglomeration (by failing effectively to restrain the misuse of non-public information).


The regulatory response to the twin regulatory challenges of financial conglomeration is the information barrier. Required by provisions of federal securities law, (90) this device consists of policies and procedures within a firm to prevent, or at least to limit, non-public information in possession of individuals in one part of the firm from flowing to--and being used by--those in another. (91) Protocols vary by firm. (92) The device is regarded as taking its original name (the Chinese wall) from the Great Wall of China because of that wall's perceived impermeability to invading hordes. (93) In practice, the information barrier typically consists of the physical separation of workspaces, confidentiality undertakings, education, and training, as well as the separation of data storage, retrieval, and communication systems. (94) Its operation is often monitored by individuals within the firm who are regarded as "straddling" or standing "above" the barrier. (95)

A. Intended Effect

Although initially adopted to avert insider trading, (96) the information barrier was soon employed to address the legal challenge produced by the organizational structure of financial conglomerates. By constraining information flows between units of a firm, the information barrier would conceptually carve the firm into multiple, distinct firms, making it possible for the financial conglomerate to reconcile its otherwise irreconcilable duties. For instance, the duty of disclosure would require the firm's broker to disclose all material information in possession of the firm's brokerage unit, rather than in possession of the firm as a whole. (97) The firm could thus discharge its duty of disclosure to a brokerage client, or avoid violating that duty, while respecting its duty of confidence to an investment-banking client that had divulged material, non-public information to another unit of the firm.

Information barriers are also intended to address the practical challenge of financial conglomeration. (98) By preventing non-public information from flowing from one part of the firm to others, the information barrier would prevent a firm from acting on incentives and exploiting opportunities to further its self-interest at the expense of its clients and third parties. By design, therefore, the information barrier would prevent the eventual disaggregation of the financial conglomerate, permitting the economies of scope and scale that conglomeration may provide as well as diminishing the client and third-party harms that intra-firm flows of non-public information may produce.

B. Actual Effect

Consider now the extent to which information barriers address the legal and practical challenges of financial conglomeration. The legal question here is not whether courts or regulators have legitimized information barriers, such as by encouraging their use or requiring their adoption. The SEC began encouraging the use of information barriers in the 1970s, (99) and they are required in a broad range of financial contexts, such as when financial conglomerates act as broker-dealers. (100) The question is also not whether, in any context, financial conglomerates may rely on information barriers as a defense to wrongdoing. Beginning in the 1980s, information barriers began serving as a legal defense to insider trading in the context of pending tender offers, for example. (101) Rather, the question concerns whether the information barrier resolves the legal challenge of financial conglomeration, such as by providing a firm with a defense to breaching a duty it owes under agency law.

Since financial conglomerates exist, one can surmise that the legal challenge of financial conglomeration has been largely overcome. In fact, according to SEC policy statements and judicial opinions, information barriers do, generally speaking, allow firms to satisfy the incompatible agency duties they would otherwise owe. The SEC has endorsed information barriers, in combination with so-called "restricted lists," as satisfying the legal challenge of incompatible duties. (102) "Restricted lists" are lists kept by financial conglomerates identifying the securities in which they will not invest on their own account or on which they will advise. (103) The SEC took this position in an amicus curiae brief it submitted in the 1974 Second Circuit case of Slade v. Shearson, Hammill & Co. (104) Remarkably, that brief stands as the first and only statement by the SEC on the legal effect of information barriers in serving as a defense to conflicts of interest or of duty. (105)

Congress and courts have been clearer in recognizing the effect of information barriers in resolving the legal challenge of financial conglomeration. In 1999, Congress all but eliminated the risk that the legal challenge might lead to the complete disaggregation of financial conglomerates. That year, it passed the Gramm-Leach-Bliley Act, permitting commercial banks to engage in activities that were previously within the domain of investment banking. (106) The legislation allowed financial conglomerates, if qualified as financial holding companies, to engage in commercial banking, investment banking, asset management, and insurance, among other activities. (107) Its purpose was "[t]o reduce and, to the maximum extent practicable, to eliminate the legal barriers preventing affiliation among depository institutions, securities firms, insurance companies, and other financial service providers." (108) Although the statute does not address the resolution of incompatible duties, its express approval of financial conglomeration provides support for the legitimacy of measures such as information barriers.

Relying on this legislation, the most recent and authoritative judicial decision to consider the legal effect of information barriers held them effective in resolving the legal challenge of financial conglomeration. In Board of Trustees of AFTRA Retirement Fund v. JP Morgan Chase Bank, N.A. (109) the District Court for the Southern District of New York considered the legal effect of information barriers in a motion for partial summary judgment. (110) JP Morgan had invested in securities of Sigma Finance, Inc. ("Sigma") on behalf of its asset management clients, while simultaneously investing on its own account in Sigma. When Sigma became insolvent, JP Morgan profited on its own investment--in the amount of $1.9 billion--while the asset management clients lost their investments.

The firm's asset management clients claimed that JP Morgan violated its duty of disclosure as well as its duty to avoid conflicts of interest. Placing primary importance on the curative effect of JP Morgan's information barriers and Congress' decision to pass the Gramm-Leach-Bliley Act, the court dismissed both claims. Rejecting an earlier case suggesting a contrary result, the court explained that the court in the earlier decision "could not have predicted the scale of modern-day multi-service financial institutions, the scope of services they would be permitted to offer simultaneously, and the information barriers they would be required by law to erect in order to avoid liability for insider trading and related conflicts-of-interest." (111) The information barrier was not alleged to have "leaked," the court said; rather, it ensured that individuals on either side of it--those investing for asset management clients on one side and those investing for the firm on the other--made "independent decisions and shared no non-public information about Sigma." (112) In addition to preventing the spread of non-public information, the information barrier prevented the financial conglomerate's own commercial relationships from "influencing the advice and decisions made by the fiduciary in its fiduciary capacity." (113) Treating the information barrier as conceptually carving the financial conglomerate into multiple, distinct firms, the court reasoned that the firm violated no duty to avoid conflicts of interest--because it faced no conflict of interest.

To support its decision, the court also referred to the consequences that would follow from denying information barriers the effect of satisfying the legal challenge of financial conglomeration. Those consequences might include "(1) a substantial increase in the cost of virtually all financial services transactions, (2) severe restrictions on the availability of [financial] products and services ... and, ultimately, (3) the disaggregation" of the financial conglomerate. (114)

The court also rejected the claim that JP Morgan, in failing to disclose its own internal prediction that Sigma would fail, had violated its duty of disclosure to its asset management clients. (115) The court indicated that JP Morgan may have violated its duty of disclosure regarding its internal prediction about Sigma if the asset management employees knew that prediction. (116) Although the court left this issue for trial, its approach suggests that no duty to disclose information arose for information in other parts of the firm--thus confirming the view of information barriers as conceptually disaggregating the firm. The case settled before further proceedings were held, leaving the court's decision as the most recent and authoritative statement on the legal effect of information barriers.

In sum, the court allowed JP Morgan's information barriers to resolve the legal challenge of financial conglomeration. In concluding that JP Morgan had not breached its duties, the court had regard only to non-public information in the possession of the particular unit serving the client in question, rather than to such information in possession of the firm as a whole. In performing its asset-management activities, the firm neither breached its duty of loyalty (because it had no information about the other unit's investing activities) nor its duty of disclosure (because, at least on the available evidence, it lacked the information in question). The apparent integrity of the barriers and the congressional policy reflected in the Gramm-Leach-Bliley Act were influential, and ultimately led to the legal effectiveness of the firm's information barriers.

In stark contrast to their legal effect, information barriers have failed to address the practical challenge of financial conglomeration. Empirical evidence confirms what scholars and others have long suspected--that information barriers fail in important contexts. (117) In particular, information barriers have failed to prevent financial conglomerates from trading using non-public client information. They have failed in the M&A context when firms traded in the stock of target corporations while advising the bidders in these transactions. (118) They have also failed in the lending context when firms traded in the stock of companies about which they possessed non-public information as lenders to those companies. (119) Financial conglomerates have also been shown to trade using non-public client information they received in a range of other contexts, including in the roles as brokers or dealers. (120) According to the empirical evidence, financial conglomerates trade using such non-public information not only when trading on their own accounts, but also when trading for, or on behalf of, clients. (121)

According to the empirical evidence, information barriers have also failed to prevent financial conglomerates from providing skewed advice to their M&A clients. Financial conglomerates have acted disloyally to their M&A clients by providing skewed, disadvantageous advice to those clients. (122) More specifically, where the financial conglomerate advises the bidder in an M&A deal and also holds a stake in the target, its client is more likely to pay a higher premium for the target relative to deals in which its adviser is not so conflicted. Where the adviser is conflicted, its client is more likely to complete the deal than otherwise (consistent with the adviser's self-interest) and is more likely to overpay for the target (also consistent with the adviser's self-interest) compared to deals in which the adviser faces no conflict of interest. (123)

From time to time, other evidence emerges of failing information barriers in financial conglomerates. The research analyst scandals at the turn of the century, which resulted in a $1.2 billion "global settlement" in 2003 among financial regulators and ten of the country's largest financial conglomerates, (124) provide an example. The scandals arose when, keen to attract investment banking business, investment bankers allegedly pressured their research analyst colleagues to issue glowing reports about potential and existing investment banking clients, despite the existence of information barriers designed to prevent the practice. As part of the global settlement, financial conglomerates undertook to bolster the information barriers between their investment bankers and research analysts. (125) Some have suggested that information barriers fail to prevent the misuse of non-public information in trading credit default swaps. (126) More recently, the effectiveness of the information barriers was questioned by the Delaware Court of Chancery in a high-profile M&A deal in which a financial conglomerate performed multiple roles. (127)

Although courts and regulators permit information barriers to resolve the legal challenge of financial conglomeration, information barriers systematically fail to address the practical challenge: in important contexts, they fail to prevent the spread and use of non-public information. The reasons information barriers fail to achieve their intended practical effect are considered next.

C. Assessment

At first glance, the practical ineffectiveness of information barriers seems surprising. Because clients of financial conglomerates may be harmed by information leaks, they have powerful incentives to ensure information barriers' effectiveness. Moreover, clients are often commercially sophisticated and well advised and thus presumed to be capable of protecting their interests from exploitation by wayward advisers. Legal rules also provide incentives for effective information barriers: federal securities laws require financial conglomerates acting as broker-dealers and investment advisers to have in place information barriers that are "reasonably designed ... to prevent the misuse ... of material, nonpublic information"; (128) and anti-fraud provisions prohibit individuals from trading on material, non-public information, including information garnered from a client in another part of the firm. (129)

Several factors nevertheless explain the failures of information barriers. To begin, market discipline provides insufficient incentives to the extent that failing information barriers harm third parties; as explained above, we have reason to believe that third-party harm does result. (130) Market discipline will also be inadequate where financial conglomerates' clients do not fend for themselves. In the M&A context, for example, corporate clients vary in their levels of sophistication; (131) for many, such transactions are one-off events in which their managers have little experience. (132) Accordingly, while some clients may be aware of and successfully avert the risk of skewed advice or misuse of their information, others may be less savvy.

Agency costs also help to explain the inadequacy of market discipline to ensure effective information barriers. As is well known, managers' interests may not coincide with those of the firms they manage. (133) Accordingly, some senior managers of financial conglomerates' clients may not bargain as fiercely with financial conglomerates as they might otherwise. Financial conglomerates have even sought to exacerbate the agency costs of corporate managers by offering personal incentives to managers to make decisions favoring financial conglomerates, possibly at the expense of the corporations they manage. (134)

The difficulty of detecting and proving failing information barriers also helps explain the evidence. Problems of detection arise because the spread of information, particularly non-public information, rarely leaves a trace. Indeed, trading on non-public information has been described as "the kind of crime where you don't leave fingerprints." (135) Martin Lipton and Robert Mazur regard "[t]he difficulty of discovering the misuse of non-public information [as] the greatest shortcoming of the Chinese Wall approach." (136) Detection may be getting even harder as trading speeds and volumes have increased. (137) Instead of detecting the apparent violation of information barriers, regulators focus on detecting "irregular" or "suspicious" trading activity (138)--that is, fortuitously timed trades--from which they infer the use of non-public information and perhaps also the failure of information barriers.

Even if the misuse of non-public information could be detected, it would pose challenges for regulators. In particular, suspicious trading may be the result of benign rationales, including coincidence or the superior trading skill or intellect of the traders involved, rather than failing information barriers. Disproving these benign explanations can be extraordinarily difficult, especially because direct evidence of information flows is seldom available, and regulators must therefore rely on circumstantial evidence to prove wrongdoing. (139) Such evidence is less likely than direct evidence to discredit benign explanations for suspicious trading.

Regulators openly acknowledge the difficulty of proving the misuse of non-public information. (140) Discussing insider trading in particular, an SEC Director of Enforcement described the obstacles as follows:

   [Insider trading cases] are unquestionably among the most difficult

   cases we are called upon to prove, and despite careful and

   time-consuming investigations, we may not be able to establish all

   of the facts necessary to support an insider trading charge....

   The real challenge is to establish that a particular individual was

   in possession of material non-public information and in fact traded

   on it in breach of a duty, and to establish those facts based on

   admissible evidence that can withstand challenge at trial.

      Piecing together an insider trading case can be a complex and

   painstaking process. It is rare to find a "smoking gun;" virtually

   all insider trading cases hinge on circumstantial evidence. It is

   quite common for insider traders to come up with alternative

   rationales for their trading--rationales that the staff must refute

   with inferences drawn from the timing of trades, the movement of

   funds and other facts and circumstances. (141)

With the exception of recent convictions involving hedge fund SAC Capital Advisors, regulators' enforcement records against financial conglomerates reflect the difficulties regulators face in detecting the misuse of non-public information. (142) It is not uncommon for the SEC to open an investigation of suspicious trading activity, but never bring charges. For example, according to a recent report of the Government Accountability Office, the SEC suspected that some firms "were leaking [non-public] customer order information," but proof "was difficult [to establish], even after an extensive multi-year, data-intensive examination." (143) The SEC eventually closed the investigations without filing charges. (144) The same difficulties might explain the SEC's failure to commence proceedings against any of the firms named in The Wall Street Journal article about the Verizon-RCC merger referred to above. (145) The recent, unprecedented uptick in insider trading prosecutions have not involved failing information barriers. Prosecutors' success has been attributed to the use of wiretaps, (146) which typically require advance judicial approval and are less suitable for proving wrongdoing suspected after the event. (147)

Financial conglomerates commonly rely on the integrity of their information barriers when questioned about suspicious trading. This is unsurprising considering the difficulties regulators face in detecting and proving failing information barriers. The response involves, in effect, claiming that information barriers prevented the firm's left hand (the trading or other unit engaging in suspicious conduct) from possessing the nonpublic information acquired by the firm's right hand (the investment banking or other "client-facing" unit). The financial conglomerates interviewed in The Wall Street Journal article about the Verizon-RCC merger stood behind the integrity of their information barriers. (148) Goldman Sachs employed the practice when questioned by the Financial Times regarding another transaction. It explained through a spokesperson: "All investment banks have Chinese walls in place precisely to avoid and manage the type of conflict to which your article alludes. We are highly confident of the effectiveness of our Chinese walls." (149)

In an earlier transaction, Morgan Stanley even took out a full-page advertisement in The Wall Street Journal when it was suspected of revealing non-public information garnered in the course of advising an M&A client. The advertisement claimed that the firm followed "the well-known industry practice of maintaining a 'Chinese Wall' between the merger and acquisition department and arbitrage [trading] departments." (150)

The difficulties of detecting and proving failing information barriers have led regulators to attempt other methods of enforcing information barriers; although the empirical evidence casts doubt on the effectiveness of these methods. (151) Both the SEC and the Financial Industry Regulatory Authority (FINRA)--the regulator primarily responsible for regulating broker-dealers--have, from time to time, conducted reviews of financial conglomerates' information barriers. Rather than scrutinizing their effectiveness during the course of transactions by reviewing firms' trading records to determine the barriers' practical effect, these regulators have instead focused on the particular policies and procedures that constitute information barriers. (152)

As this section has explained, information barriers fail to have their intended practical effect. This result is attributable to weaknesses in market discipline, the existence of third-party harm, and difficulties of detection and proof, which produce inadequate incentives for financial conglomerates to ensure the effectiveness of their information barriers. Despite evidence of their practical ineffectiveness, information barriers receive legal effect and are often relied upon to defend against suspected wrongdoing.


Before proposing a strategy to respond to the phenomenon of failing information barriers, this Article considers the likely effect of the so-called Volcker Rule, a core plank of the Dodd-Frank Act, which was Congress' response to the Financial Crisis. (153) The Volcker Rule bans a "banking entity" from proprietary trading and from freely affiliating with private equity and hedge funds. (154) Because the definition of banking entity encompasses financial conglomerates with commercial banking operations, (155) the Volcker Rule applies to many of the largest and most important financial conglomerates. Despite its primary focus on questions of financial stability, the Volcker Rule is relevant to the question of the magnitude of the harms arising from failing information barriers. (156) In the Supporting Information accompanying the final regulations implementing the Volcker Rule, regulators claim the rule addresses the risk that:

   a banking entity may acquire substantial amounts of nonpublic

   information about the financial condition of a particular company

   or issuer through its lending, underwriting, investment advisory or

   other activities which, if improperly transmitted to and used in

   trading operations, would permit the banking entity to use such

   information to its customers' ... disadvantage. (157)

This Part argues that the Volcker Rule will significantly reduce proprietary trading by financial conglomerates and therefore reduce both the opportunities and incentives for firms to misuse non-public information. However, doctrinal ambiguities will cabin the rule's benefits, making further regulatory intervention necessary to address the harms arising from failing information barriers.

A. Terms

The Volker Rule's ban on proprietary trading is broad ranging. Proprietary trading, as defined, encompasses trading as principal in a financial instrument with a short-term trading horizon--that is, "principally for the purpose of selling in the near term (or otherwise with the intent to resell in order to profit from short-term price movements)." (158) The ban is subject to exemptions for certain "client-oriented" activities, including underwriting, market-making-related activities and risk-mitigating hedging activities. (159) These exemptions, however, are ineffective to lift the ban if the trading or other activity in question gives rise to a "material conflict of interest" involving the banking entity and a client. (160)

The Volcker Rule also bans banking entities from investing in and sponsoring hedge and private equity funds, subject to certain exemptions. (161) As with the ban on proprietary trading, these exemptions only lift the ban if the activities in question give rise to no "material conflict of interest." (162)

B. Potential Benefits

The Volcker Rule can be expected to mitigate the harms of failing information barriers by reducing both the opportunities and incentives financial conglomerates face to misuse non-public information. First, the Volcker Rule can be expected to reduce opportunities for firms to misuse non-public client information by reducing their trading--a primary means by which firms exploit non-public information. According to the newly formed Financial Stability Oversight Council (FSOC), the Volcker Rule "directly cuts off a banking entity's opportunity to profit from conflicted proprietary trading, thereby protecting customers from financial injury arising from such conflicts." (163) As some indication of the level of proprietary trading in which some financial conglomerates engaged, Goldman Sachs is estimated to have earned $3.7 billion annually from that form of trading. (164) The provision has already forced firms to sell or wind down certain parts of their business operations. (165)

By banning proprietary trading in particular, the Volcker Rule eliminates the form of trading to which information barriers are most vulnerable. When information barriers fail, the financial conglomerate has far greater incentives to trade on that information as a principal than as an agent, since as a principal it benefits to the full extent of any return earned.

The Volcker Rule's proprietary trading ban will also likely reduce firms' incentives to skew their advice to clients--another form of harm from failing information barriers. More specifically, where--because of leakages in information barriers--a firm's employees are aware that their firm holds a stake (or proposes to hold a stake) in certain financial instruments, the firm's advisers have incentives to skew their advice to clients to influence clients' trading in those instruments. As explained above, (166) in the M&A context, systematic empirical evidence indicates that financial conglomerates act on these incentives.

The Volcker Rule's restrictions on investments in hedge funds and private equity funds will also reduce financial conglomerates' incentives to misuse non-public information using funds that they sponsor. By reducing a financial conglomerate's investment in a fund, the rule will reduce the firm' s incentives to earn outsized returns, in turn diminishing its incentives to leak non-public information for use by that fund.

C. Limitations

Several factors limit the benefits likely to flow from the Volcker Rule.

1. Scope of Application

First, the Volcker Rule's narrow scope of application diminishes its potential effectiveness in mitigating the harms from failing information barriers. The rule applies only to "banking entities," namely, those financial conglomerates with a commercial bank as an affiliate. (167) While major financial conglomerates satisfy the definition, the required definitional link to banking serves to exempt other important types of financial conglomerates from the Volcker Rule's scope of application.

The Volcker Rule has no application to major independent private equity firms, such as KKR & Co. and the Blackstone Group. In recent years, these firms, which are structured as financial conglomerates, have diversified beyond their core private equity operations into other financial activities, including advising on mergers and acquisitions, underwriting securities offerings, offering asset management services, and even engaging in proprietary trading (but not banking). (168) These firms are susceptible to the legal and practical challenges arising from financial conglomeration and yet fall beyond the reach of the Volcker Rule.

The Volcker Rule may even exacerbate the pressures on private equity firms to use non-public information in violation of information barriers. These firms have been major buyers of the proprietary trading units sold by financial conglomerates seeking to comply with the Volcker Rule. (169) The sale by financial conglomerates of their proprietary trading units has been described as a "garage sale" at which "private equity firms are showing up to buy." (170) Perversely, therefore, while reducing the harms of failing information barriers at some financial conglomerates, the Volcker Rule may well be magnifying them at others. (171)

The narrow range of activities to which the Volcker Rule applies also reduces its potential benefits. The rule fully permits principal trading with a longer-term trading horizon (because such trading falls outside the definition of "proprietary trading")--despite the fact that this form of trading creates similar incentives and opportunities as proprietary trading for firms to breach their information barriers. The Volcker Rule also does nothing to restrict trading on behalf of clients (172)--despite empirical evidence demonstrating that, when information barriers fail, financial conglomerates also misuse non-public information via this form of trading. While the undesirable incentives created by such client-oriented trading are weaker than those created by proprietary trading, they may nevertheless be significant.

2. Definitional Ambiguities

In addition to its narrow application, the Volcker Rule suffers from ambiguities in its definition of proprietary trading. The rule fails to define short-term trading horizon or how to identify or establish the required intention--problems that have already exposed the rule to criticism. (173) Moreover, the ban is subject to certain exemptions--such as client-oriented activities--which themselves raise definitional concerns. The definition of "market-making activities," for example, is ambiguous and difficult to distinguish from proprietary trading. The market-making exemption is coupled with the requirement that any trading be "designed not to exceed the reasonably expected near term demands of clients"--another concept requiring clarification. (174) The exemption is also subject to the proviso that it not result in any "material conflict of interest" (175)--yet another potentially ambiguous concept. These definitional ambiguities will likely hamper the rule's effectiveness in reducing the harms arising from failing information barriers.

3. Reliance on Information Barriers

A further deficiency of the Volcker Rule is the decision by regulators to permit financial conglomerates to rely on information barriers to establish that no "material conflict of interest" exists. (176) In determining whether a material conflict of interest arises, regulators will give effect to information barriers, unless "the banking entity knows or should reasonably know that, notwithstanding [its] establishment of information barriers, the conflict of interest may involve or result in a materially adverse effect on a client, customer, or counterparty." (177) In their Supplementary Information accompanying the final Volcker Rule regulations, regulators assert that they "continue to believe that information barriers can be an effective means of addressing conflicts of interest that may arise through, for example, the spread of information among trading desks engaged in different trading activities that may result in potentially inappropriate informational advantage." (178) Regulators decided not to specify when information barriers would be ineffective, they explain, because "banking entities are better positioned to determine when information barriers may be effective given their trading activities and business structure." (179) Regulators' willingness to rely on information barriers may give greater breadth to the exemptions, and thus narrow the scope of the ban on proprietary trading, diminishing the likelihood of the Volcker Rule's potential benefits materializing. This concern is heightened by regulators' failure to even specify what amounts to effective information barriers. (180)

4. Offsetting Costs

The Volcker Rule may also create costs. In particular, its ban on propriety trading eliminates the benefits that flow from that activity. Professor Charles Calomiris argues that banning proprietary trading will "hobble" firms' "abilities to serve their clients" because they would then be unable to develop and acquire the skills and knowledge necessary to serve their clients. (181) More specifically, "prohibiting banks from proprietary trading would force banks to rely on relatively uninformed people to advise and execute for clients." (182)

Despite the concern, these potential costs of the Volcker Rule are unlikely to materialize. Although financial conglomerates may no longer trade on their own accounts, they may continue to trade on behalf of clients. Any skill and knowledge developed and acquired in the former (principal) capacity can surely be developed in the latter (agency) capacity. While a trader may not have as high-powered incentives when trading on behalf of clients as when trading for her firm's own account, (183) she nevertheless engages in the same activity, acquiring similar if not identical skills. In any case, Professor Calomiris fails to explain how the skills gained from proprietary trading are deployed for clients' benefit, if as firms claim, their information barriers separate proprietary traders from those traders advising clients. (184) In short, financial conglomerates' clients are unlikely to receive inferior advice or service in the wake of the Volcker Rule.

The Volcker Rule is also unlikely to increase financial conglomerates' risk of financial failure by depriving them of the income-diversifying effects of proprietary trading--another concern raised by critics. (185) Whether the rule's ban would produce this outcome turns on the effect of proprietary trading on the riskiness of the stream of income generated by proprietary trading and its covariance with the earnings stream of the financial conglomerate's other operations. (186) In the absence of some theory or explanation as to how income from proprietary trading may co-vary with that from a financial conglomerate's other operations to reduce overall firm risk--for example, by co-varying negatively--the claim lacks weight. (187)

D. Assessment

Although primarily directed to promoting financial stability, the Volcker Rule will likely reduce the harms associated with failing information barriers. The magnitude of the rule's potential benefits, though impossible to estimate precisely, will likely more than offset any costs the rule imposes. Prior to the Volcker Rule's imposition, proprietary trading was a key driver of financial conglomerates' revenue, (188) creating powerful incentives that operated in opposition to measures such as information barriers designed to limit informational advantages to traders. To the extent the rule removes these incentives, it will have desirable effects.

However, the rule's effect will turn on its scope of application. Definitional ambiguities promise to reduce the rule's benefits, as do the potential breadth of its exemptions. While difficult to predict, these limits are likely to substantially diminish the Volcker Rule's potential benefits in mitigating the harms of failing information barriers.

Even after the Volcker Rule is fully implemented, important sources of harm from failing information barriers will remain. An important category of financial conglomerate--major private equity firms--will be unaffected by the rule, and may even have greater incentives to use non-public information in violation of information barriers than before. Those firms to which the rule applies will continue to have opportunities to exploit non-public client information in trading on behalf of clients and in making long-term investments for their own accounts. Importantly, despite the rule's undoubted promise, regulatory attention is required to address the phenomenon of failing information barriers.


In view of the failure of information barriers to address the challenges of financial conglomeration and the limitations of the Volcker Rule, this Article proposes a regulatory reform to improve the practical effectiveness of information barriers. The reform combines statistical inference with the force of market discipline and would supplement the Volcker Rule. This Part explains the methodology the proposal employs, discusses its design, and then contends with possible objections.

A. Methodology

1. Underlying Intuition

The proposed strategy employs statistical analysis to detect and prove the misuse by financial conglomerates of non-public information. The statistical analysis recommended in this Article has been used in a recent strand of financial economic literature. (189) Applying the statistical analysis to individual financial conglomerates would allow inferences to be drawn as to a firm' s use of non-public client information in its trading activities. More specifically, the strategy would allow inferences as to whether a given financial conglomerate's trading results can be plausibly explained only as the result of failing information barriers. otherwise put, the strategy would provide evidence as to whether a firm's suspiciously high trading returns earned while in possession of (apparently "walled off") non-public information were the result not of benign factors, including skill and coincidence, but of informational advantage--and thus the use of that non-public information in violation of information barriers.

The strategy reflects a simple but powerful insight: barring some plausible alternative explanation, the return a firm earns from trades on stock of companies about which it possesses non-public client information will not diverge significantly from various benchmark returns earned from trades on stock of companies about which it lacks non-public information--provided information barriers are effective. The strategy permits plausible alternative explanations to be ruled out. To do so, it compares the return earned by a firm in possession of non-public information with various benchmark returns. Importantly, the contexts in which the former return is calculated are those in which the firm is duty-bound not to use that information, such as those in which it received nonpublic information from a client. To account for market movements, the strategy would use abnormal returns, rather than absolute returns, consistent with the literature. (190)

As Ivashina and Sun explain, the comparisons with benchmarks would allow three alternative explanations to be ruled out. (191) These explanations are superior intellect or trading skill; peculiarities of the companies in question; and coincidence. The corresponding benchmarks are: first, the relevant firm's trading return from the rest of its portfolio (that is, on trading stocks of companies about which it lacked non-public client information); second, the trading return a group of comparable firms earned when trading stocks about which the relevant firm possessed non-public information; and third, the trading return the relevant firm earned in stocks about which it had non-public information--but at times it lacked such information.

The methodology would be applied as follows. For a particular period (such as one year), calculate a financial conglomerate's (Conglomerate X's) annualized abnormal return on stock trades of companies about which it possessed non-public client information. The abnormal return represents the amount by which the actual return exceeds the expected return, which in turn is calculated using one of several broadly accepted models of estimation. (192) Of course, a positive abnormal return would be consistent with failing information barriers. This first return would then be compared with the three benchmarks referred to above, calculated for the same time period. (193)

Because financial conglomerates receive non-public information from clients in particular contexts, or for particular information events, the methodology would apply to contexts where information barriers are intended to prevent the spread of non-public information from one unit of the firm to its trading unit. These information events or contexts include when financial conglomerates advise on mergers and acquisitions or make bank loans. For each information event, the basic return would be calculated and compared with the various benchmarks. For example, the basic return would be calculated for Conglomerate X when possessing information garnered as an M&A adviser; (194) that return would then be compared with benchmarks calculated for the same period. The basic return would also be calculated for other contexts where information barriers operate, such as where Conglomerate X possessed non-public information garnered from making loans to customers (195); that return would also then be compared with the various benchmarks calculated for the same period.

Assume, for example, that the first return is positive and significant, while each of the three benchmarks is not statistically different from zero. One could dismiss the alternative explanation of superior intellect or trading skill because the firm failed to achieve above-market performance with the rest of its portfolio. Similarly, one could dismiss the possibility that the firm's superior return stemmed from peculiarities of the companies in question because comparable financial conglomerates failed to achieve above-market performance when trading in the same stocks (about which the relevant firm held non-public information). Finally, one could rule out coincidence as a possible explanation because the firm failed to achieve above-market performance when trading in the same stock but at times it lacked non-public information. (196) The only plausible explanation for the above-market returns would be the use of non-public information in the firm's possession and thus the failure of its information barriers. The various comparisons may be considered hypotheses, as Table 1 depicts.

The methodology would also allow firms to demonstrate the integrity--rather than the failure--of their information barriers. Firms might dispel suspicion by showing that they do not earn above-market returns when trading stock about which they possess nonpublic information. Those that do earn above-market returns might also dispel suspicion by, for example, showing that their returns may plausibly be accounted for by superior intellect or trading skill.

2. Illustration

The methodology may be illustrated using the JP Morgan-Verizon example above. (197) Assume that in 2007 JP Morgan earned a positive annualized abnormal return of five percent on stock trades in companies about which it possessed non-public client information. This basic return would include JP Morgan's trades in the stock of RCC while it was advising Verizon (because at this time it held non-public information about RCC) as well as JP Morgan's trades in stock of other companies about which its M&A advisers possessed non-public information. JP Morgan's superior trading return (five percent) is generally consistent with the conclusion that the firm's information barriers failed in 2007, but is not convincing evidence of that conclusion because benign explanations may well exist.

Assume further that JP Morgan failed to earn a positive annualized abnormal return on stock trades for the rest of its portfolio during the same period. (198) Assume also that comparable firms failed to earn a positive annualized abnormal return when trading in stocks about which JP Morgan possessed non-public information. Assume further that JP Morgan failed to earn a positive annualized abnormal return when trading in stocks about which it had non-public information--but at times it lacked such information. (Such a calculation would include JP Morgan's trades in RCC before or after it had non-public information about RCC.) Put differently, in a given time period, JP Morgan's superior trading performance occurred only when it traded in stocks about which it possessed nonpublic information, only at times when it possessed such information, and only when comparable financial conglomerates trading in the same stock failed to replicate that superior performance. Each of the three hypotheses is true.

Important inferences would be drawn from these hypothetical results. First, JP Morgan's superior return in 2007 could not plausibly be the result of the intellect or skill of its traders because they failed to achieve a superior return when the firm lacked nonpublic information. Second, JP Morgan's superior return could not plausibly be the result of special features of the stocks it traded because comparable firms failed to replicate the superior return when they traded the same stocks. Third, JP Morgan's superior return could not plausibly be due to the coincidence that the stocks about which the firm possessed non-public information were those stocks in which its traders had special insights because the firm failed to earn a superior return trading in those stocks at times it lacked non-public information. The only plausible remaining explanation for JP Morgan's superior trading performance for stocks about which it possessed non-public information--that is, the statistical inference--is the use of that information and, thus, the failure of the firm's information barriers.

3. Limitations on Use

The methodology requires the calculation of various trading returns, which in turn requires data of the financial conglomerate's underlying trading activities. Such data, however, is rarely publicly available. Studies that employ the methodology described above instead rely on certain trading data disclosed by financial conglomerates under Section 13(f) of the Securities Exchange Act of 1934 (the Exchange Act), which compels a broad category of firms, including financial conglomerates, to disclose on a quarterly basis their holdings of certain financial instruments over which they have investment discretion. (199) Reliance on Section 13(f) data provides a narrow snapshot of the composition of financial conglomerates' trading portfolios, with the consequence that tests relying on such publicly available data can report only the lower bound of any informed trading. (200) The trading snapshot revealed by Section 13(f) data is narrow because that statute exempts from disclosure trading data on bonds, credit default swaps, and short positions in stocks and options. (201) Moreover, the quarterly nature of the data makes it impossible to determine whether trading positions were taken and closed out within a single quarter or whether particular trades occurred before or after the receipt of non-public information. (202) To overcome these potential limitations, the proposed strategy would rely on trading records of firms, rather than on publicly available 13(f) data. (203)

The methodology also requires data about so-called information events. These are situations in which financial conglomerates possess non-public information on which they might profitably trade. Although not always possible to know when a financial conglomerate possesses non-public information, it is possible in certain contexts to identify and verify the date on which it will almost certainly possess such information. Specifically, a firm will possess non-public information on the day it is engaged to advise on an M&A transaction or agrees to lend to a borrower (or to amend its loan to a borrower). (204) Outside these contexts, information events may not be readily observable.

B. Design

The proposed regulatory strategy has two prongs. The first would require financial conglomerates to publicly report the results of testing the hypotheses described above. These results are referred to as quantitative metrics. Under the second prong, a financial conglomerate would face liability if it failed to rebut an inference of the misuse of nonpublic information in its trading activities. Each prong is considered in turn.

1. Mandatory Reporting of Quantitative Metrics

The required public reporting of quantitative metrics, that is, the results of testing the hypotheses, would allow outsiders--and clients, in particular--to draw adverse inferences about the effectiveness of financial conglomerates' information barriers and to discipline firms accordingly. The public reporting of quantitative metrics would allow clients to distinguish more reliably among financial conglomerates on the basis of their client loyalty and thereby reduce the adverse selection problem. (205) Such disclosure would also serve to sharpen the crude constraints that firm reputation has on firms' conduct. (206) To avoid such market discipline, financial conglomerates would have incentives to take greater precautions to prevent non-public information leaking to their traders, leading firms to strengthen their information barriers.

a. Preferred Model to Generate Quantitative Metrics

The preferred model for generating quantitative metrics involves compelling financial conglomerates to engage external, regulator-approved experts who would operate under a peer-review system to generate quantitative metrics. Under the proposal, financial conglomerates would be compelled to disclose--on a confidential basis--their non-public trading records to these experts as well as information about information events, that is, particular contexts in which they acquire non-public client information, such as transactions in which they advise a principal in an M&A deal or lend funds to a borrower. Financial conglomerates would report the quantitative metrics in their SEC filings and thus face liability for material misstatements or omissions.

This preferred model of generating quantitative metrics would offer several benefits. It could be integrated with existing and newly introduced practices at financial conglomerates to ensure the effectiveness of information barriers. Financial conglomerates are already required to monitor trading in securities on their so-called "watch lists" (207) and to investigate suspicious trading activity in possible violation of their information barriers (208)--although nothing suggests they generate the quantitative metrics suggested here. Financial conglomerates could employ the methodology as part of their own internal controls. The use of peer review would ensure, as much as possible, the use of good research design. Moreover, the use of regulator-approved external experts is not without precedent. The SEC relied on such experts to estimate damages resulting from the late-trading of mutual funds. (209)

The model also stands some chance of attracting financial conglomerates' support because of its parallels with existing rules. For instance, the Volcker Rule will likely require firms to collect, analyze and report data--known in the regulations as quantitative metrics--to assist regulators in identifying impermissible proprietary trading. (210) As part of the 2003 "global settlement," financial conglomerates agreed to generate aggregate data about their accuracy in evaluating stocks and thus provide evidence regarding the independence of their analysts. (211) By enlisting financial conglomerates in the regulatory process, firms may "buy-in" to the strategy and be less likely to seek to evade it. (212)

b. Alternative Models to Generate Metrics

The preferred model for generating quantitative metrics--involving the self-reporting by firms of quantitative metrics calculated by external, regulator-approved experts--avoids many of the defects afflicting the alternatives. To begin, the preferred model trumps the alternative of requiring financial conglomerates simply to publicly disclose their trading data (rather than to report the quantitative metrics derived from that data). The public disclosure of trading records would allow scholars, financial media, and others (rather than regulator-approved experts under the preferred model), to test the hypotheses and generate the quantitative metrics. However, such a model would risk disclosing confidential information about financial conglomerates' trading strategies, potentially diminishing firms' incentives to innovate and exposing them to predatory trading by competitors. (213) Because clients are beneficiaries of financial conglomerates' trading strategies, such a model could harm them. While techniques such as delaying disclosure (214) could be adopted to avert these risks, the preferred model of having experts generate the required metrics (using non-public, privately disclosed, firm data) avoids the risks altogether.

The alternative of requiring the public disclosure of trading data would also suffer from reliance on scholars and others to generate quantitative metrics. Using Section 13(f) data, scholars have shown that information barriers systematically fail, but they have refused to identify those firms implicated. (215) Unless non-compliant firms are named, the proposed strategy would be unlikely to effectively harness market discipline to strengthen information barriers because comparisons among firms could not be made.

A second alternative model would involve rating agencies (or their equivalent) generating the quantitative metrics, possibly assigning ratings to individual financial conglomerates. The model would offer cost savings by collectivizing the costs of analysis and also overcome the competitive disadvantage firms may suffer from divulging their confidential trading records. However, the model would face the funding problem afflicting all credit rating agencies. To fund ratings, firms must either charge the users of those ratings or the firms being rated, neither of which is an appealing option. Users will rarely pay (because of the ease with which information spreads), and if the firms themselves must pay, legitimate doubts arise about the independence of the ratings produced. (216) Moreover, the conduct of credit rating agencies in the lead-up to the financial crisis does not inspire confidence in the ratings model. (217)

A third alternative model would involve a regulator, such as the SEC, generating the quantitative metrics. Assuming the SEC had the willingness and expertise to undertake the analysis, (218) this alternative would further burden a resource-constrained regulator already encumbered by Dodd-Frank Act rulemaking. It nevertheless offers the closest alternative for generating metrics to the preferred model above.

In sum, the proposed strategy would require financial conglomerates to engage regulator-approved outside experts to calculate quantitative metrics under the review of their peers in a manner analogous to that used by prominent academic journals. They would do so using data confidentially disclosed by financial conglomerates. Financial conglomerates would then be compelled to report those quantitative metrics--possibly including brief interpretations of them--in their regulatory filings. Such disclosures would likely harness the disciplinary power of market forces, providing incentives for financial conglomerates to buttress their information barriers to avert adverse inferences being drawn. The proposed strategy would be cost-effective, imposing relatively modest costs to generate quantitative metrics, especially relative to the benefits and when considered against the expense for regulators of closely monitoring information barriers. It would seek to ensure that alternative explanations--other than failing information barriers for suspicious trading activities--are accounted for.

2. Imposition of Liability

The proposed strategy's second prong involves fining financial conglomerates whose information barriers were demonstrated to have failed. Liability would be desirable given the limits of market discipline. (219) Under this prong, an adverse inference drawn from the quantitative metrics would give rise to a prima facie case of wrongdoing. Any financial conglomerate against which an adverse inference was drawn (based on quantitative metrics) would be given the opportunity to rebut the inference, such as by showing that the trading at issue was the result of client instructions. The strategy would thus heed the Supreme Court's caution in employment discrimination about the use of statistics--that, although "competent in proving [wrongdoing]," they are "not irrefutable; they come in infinite variety and, like any other kind of evidence, they may be rebutted." (220)

Several insights from optimal deterrence theory would be incorporated into the strategy's design. (221) First, the fine imposed for failing information barriers should not necessarily be set so high as to prevent information barriers from failing. A wrongdoer would be optimally deterred if it faced a sanction equal to the social cost of its wrong. Under that condition, the wrongdoer would internalize the social cost it creates and thus be led to take optimal precautions to deter that wrong. (222) Such a regime would lead it to minimize (rather than necessarily prevent) the social harm it creates. (223)

In setting the magnitude of the fine, regard should also be given to the probability of detecting and proving failing information barriers. The lower the probability, the higher the sanction financial conglomerates should face for demonstrated failures of information barriers. (224) According to optimal deterrence theory, where the probability of sanction is less than one, the sanction' s magnitude should be set such that its expected value equals the social cost of the wrong. (225) Under that condition, a wrongdoer would expect to face the social cost of its wrong--and thus be led to minimize both that social cost and the costs of the precautions the wrongdoer takes. (226)

Calibrating the optimal fine is inevitably difficult (as in any context), and somewhat speculative assessments of the various parameters would be required. Nevertheless, the magnitude of the fine should be set to lead firms to internalize the costs they impose, subject to some adjustment to reflect the impossibility of detecting every wrong.

Finally, the proposal would not initially provide a private right of action based on the reported quantitative metrics for corporate clients or others. While a private right of action would powerfully supplement SEC enforcement in creating incentives for firms to buttress their information barriers, this Article favors the more modest approach of permitting only public enforcement, at least until the need for supplemental private enforcement were demonstrated. (227)

3. Scope of Application

The proposed strategy would be applied to financial conglomerates. Accordingly, it would apply more broadly than the Volcker Rule, capturing not only commercial banks, but also major private equity firms and other financial institutions without commercial banking units that nevertheless employ a conglomerate structure. (228)

C. Possible Objections

The proposed strategy may invite several objections regarding the limitations of statistical evidence and the existing availability of the proposed strategy. This Part considers the legitimacy of these objections.

1. Unreliability of Statistical Evidence

Skeptics of the proposal would be right to assert that statistical inference is not foolproof in detecting and proving wrongdoing. In no circumstances may statistical analysis establish conclusively that any outperformance by a financial conglomerate trading stocks in companies about which it possesses non-public information was the result of ineffective information barriers. As is well known, statistical analysis cannot establish causation. (229) It operates negatively by allowing researchers to reject the likelihood of plausible alternative explanations. (230) It provides information only as to the probability that a difference is due to chance. (231)

This objection in no way defeats the proposal. To begin, while statistical inference cannot provide conclusive proof or direct evidence of causation, it may nevertheless provide legal proof. As Judge Posner has explained, "All evidence is probabilistic--statistical evidence merely explicitly so, because nothing with which the law deals is metaphysically certain." (232) To rely on statistical evidence is simply to accept that such evidence, like other evidence, has the "tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence." (233) Moreover, the approach employed in statistical analysis--of arriving at a conclusion by ruling out plausible alternative explanations--is consistent with judicial fact-finding. (234)

Second, in a range of contexts where direct evidence of wrongdoing is seldom available, courts do make statistical inferences to prove wrongdoing. In the employment discrimination context, for example, courts use statistical inference to prove systemic disparate treatment in violation of Title VII of the Civil Rights Act of 1964. (235) Courts recognize the suitability of such inference given the usual absence of direct evidence--because employers rarely admit to wrongful discrimination or leave an evidentiary trail of such conduct. (236) In this context, as well as in many others, (237) courts accept the value of statistical inference to provide evidence of wrongdoing that is otherwise difficult to detect and prove. (238)

Third, statistical analysis has a record of detecting corporate wrongdoing. Statistical analysis first uncovered the practice of options backdating, a practice in which corporations grant options to executives and retroactively set a lower exercise price for the options to increase their value. (239) Statistical analysis also uncovered insider trading by officers and directors in violation of Section 16(b) of the Exchange Act. (240) The practices of "late trading" by mutual funds and "spinning" hot initial public offerings were also first detected by statistical analysis. (241)

The proposed strategy also would not rely uncritically on statistical evidence to infer wrongdoing. Under it, statistical inference would, at most, give rise to a prima facie case of failing information barriers, and financial conglomerates would then be permitted to rebut the inference. Accordingly, the proposal does not treat statistical evidence as dispositive.

Some concerns about the strategy will inevitably remain. Although similar issues exist in proving employment discrimination, (242) the calculation of abnormal returns and the selection of comparable corporations (for control purposes) must be performed with care. The strategy would not verify whether information barriers failed in a specific transaction; while firm-specific, the required analysis necessarily applies across repeated transactions. It would not detect all systemic failures of information barriers because it may be applied only to particular contexts where financial conglomerates are known to possess non-public information.

Nevertheless, the strategy offers promise in detecting wrongdoing in a field of activity where direct evidence is not available and benign explanations exist. It would apply on a firm-specific basis, allowing inferences as to whether a particular firm's information barriers are effective--a more useful measure than one showing the failure of information barriers in a particular deal. While it could not detect failing information barriers in contexts where information events are not readily available, the strategy nevertheless applies in those contexts--particularly M&A deals--where information barriers have been shown to fail. Most importantly, the strategy allows benign alternative rationales for failing information barriers--namely, traders' superior intellect or skill or mere coincidence--to be assessed and, if appropriate, ruled out. The fact that statistical analysis cannot provide conclusive proof provides no legitimate basis for rejecting a strategy incorporating such analysis.

2. Sufficiency of Existing Protections

Skeptics may assert that regulators can already apply the suggested methodology without the need for further regulatory intervention. While nothing would seem to prevent regulators from performing the tests proposed in this Article, nothing suggests they have done so. (243) The SEC, already resource-constrained, is currently burdened with writing rules required under the Dodd-Frank Act. (244) Moreover, policymaking roles at the SEC are dominated by lawyers, (245) who may be less likely than other professionals--such as economists--to appreciate the value of statistical inference in deterring and catching wrongdoing. The SEC may nevertheless be expected to perform the oversight role envisaged here of approving experts and enforcing the liability regime. In fact, the proposed strategy incorporates something of a market check on the diligence of the SEC by requiring the public reporting of data showing whether a financial conglomerate's information barriers are failing. (246) Any omission by the SEC to pursue a matter would invite public scrutiny.

A related concern--that clients can already protect themselves, perhaps by using the suggested empirical techniques--also fails to defeat the proposal. As explained above, empirical evidence suggests limits to market discipline, which is unsurprising given the third-party nature of some harm arising from the misuse of non-public information and the agency costs experienced by corporate clients. (247)

3. Risk of Gaming

Gaming, that is, strategic behavior to prevent the suggested strategy from detecting wrongdoing, might take the form of firms providing inaccurate or incomplete data to the external experts or trying in some other way to disguise their activities. However, some information events would be difficult to game; for instance, M&A deals and bank loans are easily identifiable and it would generally be clear when a firm possessed non-public information from these activities. (248) Moreover, the proposed strategy employs safeguards, including having regulators approve experts and subjecting firms to liability under anti-fraud provisions of the federal securities laws for misstating their quantitative metrics (by requiring them to report the metrics in their SEC filings). The recently adopted whistleblower bounty scheme might also diminish the prospect that any gaming would remain hidden. (249) Nevertheless, as with any regulatory strategy, the risk of gaming cannot be ruled out, and would need to be addressed as problems arise.

4. Risk of Statistical Dueling

Skeptics may also fear the risk of "statistical dueling" (250)--that disputes might descend into battles between experts employed by the various stakeholders involved. Views among experts may differ, creating disputes. However, the risk of statistical dueling significantly hampering the proposal's effectiveness would seem no greater than in the various other legal contexts in which statistical evidence is used. (251) Moreover, as with any novel development, something of a learning period may be involved after the proposal's initial adoption, as processes are fine-tuned. (252) In addition, where the SEC seeks to impose liability, the adversarial approach would test opposing views about the statistical evidence just as it tests differing perceptions as to non-statistical evidence. Ultimately, the adversarial process, the proposed peer-review process, and the transparency of enforcement actions promise to ensure that the proposal' s required statistical analyses comport with good social science practice.


This Article examines the use of information barriers to address the legal and practical challenges of financial conglomeration. Although information barriers are accorded legal effect, they often fail to address the practical challenge of effectively restricting intra-firm flows of non-public information. This Article argues that limits on market discipline and evidential obstacles to detecting and proving the misuse of nonpublic information account for the failures of information barriers. The Volcker Rule can be expected to reduce the pressures within firms that lead to breaches of barriers, but definitional ambiguities and its limited scope of application hamper the rule's effect.

This Article proposes a regulatory solution. The proposal applies the insight that trading on non-public information in violation of information barriers may be detected and proved by statistical inference. Using financial conglomerates' non-public trading data, the proposal would compel financial conglomerates to report quantitative metrics from which plausible inferences could be drawn as to the failure of their information barriers. The reported metrics would attract both market discipline and potential legal liability, harnessing forces that would create powerful incentives for financial conglomerates to buttress their information barriers. The proposed strategy would thereby address the practical challenge of financial conglomeration, diminishing the client and third-party harms arising from the misuse of non-public client information, while still allowing the economies and scope and scale that may flow from financial conglomeration. The proposal would also undermine financial conglomerates' practice of reflexively responding to accusations of wrongdoing by expressing confidence in the integrity of their information barriers. Under the proposal, they would need to prove as much. Only then will information barriers more effectively address the fundamental regulatory challenges of financial conglomeration.

(1.) See Law Commission, Fiduciary Duties and Regulatory Rules, 1992, Consultation Paper 124, at 12-18 (U.K.) (describing the organizational structure of financial conglomerates).

(2.) See, e.g., Nat'l Ass'n of Sec. Dealers & N.Y. Stock Exch., Joint Memorandum on Chinese Wall Policies and Procedures, Notice to Members No. 91-45 (1991); In re Barclays PLC et al., CFTC Docket No. 12-25, at 34 (June 27, 2012), available at documents/legalpleading/enfbarclaysorder062712.pdf. The term "information barrier" is now more commonly used. See generally Sec. & Exch. Comm'n, Staff Summary Report on Examinations of Information Barriers: Broker-Dealer Practices Under Section 15(g) of the Securities Exchange Act of 1934, at 45 (Sept. 27, 2012), available at (discussing broker-dealers' use of information barriers).

(3.) See infra Part III.

(4.) See infra Part III.B (discussing the extent to which information barriers resolve the legal and practical challenges of financial conglomeration).

(5.) See infra notes 117-123 and accompanying text.

(6.) Trading on non-public information has been described as "the kind of crime where you don't leave fingerprints." See Michael C. Jensen, The Financiers: The World of the Great Wall Street Investment Banking Houses 305 (1976). The claim is not that the use of non-public information is rarely detected and proved, but more specifically that the breach of information barriers resulting in such activity is rarely detected and proved. See infra notes 135-147 and accompanying text.

(7.) Mark Maremont & Susanne Craig, Trading in Deal Stocks Triggers Look at Banks, Wall St. J., Jan. 14, 2008, at A1; see also Part II.C (discussing Verizon's acquisition of RCC).

(8.) Several years after the RCC transaction, Congress imposed limits on financial conglomerates' ability to trade on their own accounts. See infra Part IV (discussing the so-called Volcker Rule).

(9.) More specifically, the particular units of JP Morgan that subsequently acquired stock in RCC had not previously owned RCC shares. Maremont & Craig, supra note 7. The article does not disclose whether other units of the firm had previously owned stock in RCC. Id.

(10.) Through a spokesperson, JP Morgan claimed its purchases in RCC were made on behalf of its clients. Id. The article does not report whether those clients were entities affiliated with JP Morgan or not. Id.

(11.) Id.

(12.) According to the article, JP Morgan explained the circumstances as a "chance" occurrence, stating, "one arm of a bank will buy a stock without knowing that another arm is advising on a deal." Maremont & Craig, supra note 7.

(13.) Id.

(14.) Regarding its involvement in a deal, Citigroup responded that it has "no reason to believe any breach occurred" in its information barriers. Id. Regarding another deal, Morgan Stanley asserted that it has "strict policies" (namely, information barriers) "to prevent the spread of non-public information" and did not believe they had been violated. Id. (internal quotation marks omitted).

(15.) Id.

(16.) The legal literature is voluminous. Examples include: ANDREW CROCKETT ET AL., CONFLICTS OF INTEREST IN THE FINANCIAL SERVICES INDUSTRY: WHAT SHOULD WE DO ABOUT THEM? (2003); Arthur B. Laby, Resolving Conflicts of Duty in Fiduciary Relationships, 54 Am. U. L. Rev. 75 (2004); Martin Lipton & Robert B. Mazur, The Chinese Wall Solution to the Conflict Problems of Securities Firms, 50 N.Y.U. L. Rev. 459 (1975); HARRY MCVEA, FINANCIAL CONGLOMERATES AND THE CHINESE WALL: REGULATING CONFLICTS OF INTEREST (1993); Note, Conflicting Duties of Brokerage Firms, 88 HARV. L. REV. 396 (1974); Norman S. Poser, Chinese Wall or Emperor's New Clothes? Regulating Conflicts of Interest of Securities Firms in the US. and the U.K., 9 MICH. Y.B. INT'L LEGAL STUD. 91, 93 (1988); JOSEPH E. STIGLITZ, THE ROARING NINETIES: A NEW HISTORY OF THE WORLD'S MOST PROSPEROUS DECADE 140-69 (2003); Larry L. Varn, The Multi-Service Securities Firm and the Chinese Wall: A New Look in the Light of the Federal Securities Code, 63 NEB. L. Rev. 197 (1984).

Recent financial economic literature on the effectiveness of information barriers includes: Andriy Bodnaruk et al., Investment Banks as Insiders and the Market for Corporate Control, 22 Rev. Fin. Stud. 4989 (2009); Victoria Ivashina & Zheng Sun, Institutional Stock Trading on Loan Market Information, 100 J. FIN. ECON. 284 (2011); Massimo Massa & Zahid Rehman, Information Flows within Financial Conglomerates: Evidence from the Banks-Mutual Funds Relation, 89 J. Fin. Econ. 288 (2008); H. Nejat Seyhun, Insider Trading and the Effectiveness of Chinese Walls in Securities Firms, 4 J.L. Econ. & Pol'Y 369, 369-73, 386-89 (2008).

(17.) See, e.g., Lipton & Mazur, supra note 16, at 499-510; Poser, supra note 16, at 138-44; Varn, supra note 16, at 267-82 (discussing the limits of information barriers and suggesting circumstances when they would be denied legal effect).

(18.) See supra note 14 (presenting empirical evidence of failing information barriers in various contexts); see also Part III.B (discussing the extent to which information barriers resolve the legal and practical challenges of financial conglomeration).

(19.) See infra Part III.C (assessing the effect of information barriers).

(20.) Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. No. 111-203 (2010).

(21.) See, e.g., charles W. Calomiris, The Volcker Rule: Unworkable and Unwise, Economic Policies for the 21st Century, e21 (Mar. 25, 2010), commentary/03_25_2010_Volcker.pdf; Hal. S. Scott, Prepared Written Testimony Before the Comm. on Banking, Housing, and Urban Affairs of the U.S. Senate (Feb. 4, 2010), available at scott_commentary_on_volcker_proposals_201002.pdf; Charles K. Whitehead, The Volcker Rule and Evolving Financial Markets, 1 Harv. Bus. L. Rev. 39 (2011).

(22.) See infra Part IV.B (discussing the potential benefits of the Volcker Rule in limiting financial conglomerates' misuse of non-public client information).

(23.) Paul A. Volcker, Statement before the Comm. on Banking, Housing, and Urban Affairs of the U.S. Senate (Feb. 2, 2010), available at _on_financial_reforms_201002.pdf.

(24.) See infra note 189 and accompanying text (discussing "forensic finance" literature).

(25.) See infra Part V (discussing a proposed regulatory strategy to address failing information barriers).

(26.) See Hillary A. Sale, Disappearing Without a Trace: Sections 11 and 12(a)(2) of the 1933 Securities Act, 75 Wash. L. Rev. 429, 83-88 (2000) (discussing the increased use of statistical analysis to provide evidence in civil litigation).

(27.) See infra note 237 and accompanying text (describing the use of statistical analysis by courts and regulators to provide proof).

(28.) See infra notes 75-82 and accompanying text (discussing the third-party harm caused by failing information barriers).

(29.) The subsidiaries of major financial conglomerates often number in the thousands. See Dafna Avraham et al., A Structural View of U.S. Bank Holding Companies, Fed. Reserve Bank OF N.Y. Econ. Policy Rev., July 2012, at 66, available at ("Today, the four most complex firms ... each have more than 2,000 subsidiaries, and two have more than 3,000 subsidiaries.").

(30.) Brokers typically act for "customers," whereas other units within financial conglomerates, including investment banking, act for "clients." For consistency, this Article uses the latter term to encompass those actors typically referred to as either clients or customers.

(31.) See, e.g., Morgan Stanley, Annual Report on Form 10-K 2-6 (2013), available at (discussing the financial conglomerate's business segments); and JP Morgan Chase & Co., 2010 Annual Report 67 (2011), available at 23b/2010_JPMC_AnnualReport_.pdf (depicting the financial conglomerate's business segments graphically). Financial conglomerates that were, until recently, investment banks give less emphasis to their banking activities. See, e.g., Goldman Sachs & Co., 2013 Annual Report 29 (2013), available at (describing Goldman's business segments). Figure 1 refers to the primary functional segments or business lines within financial conglomerates. Some firms further divide their segments, such as into consumer and wholesale segments, and also blur some segments, such as including corporate lending as a securities-related activity.

(32.) In addition to M&A deals and securities offerings, strategic transactions include divestitures, restructurings, and spin-offs. See Goldman Sachs & Co., supra note 31 (describing investment banking services). For convenience, strategic transactions generally are referred to simply as M&A transactions.

(33.) See Sec. & Exch. Comm'n, Study on Investment Advisers and Broker-Dealers: As Required by Section 913 of the Dodd-Frank Wall Street Reform and Consumer Protection Act 8-11 (Jan. 2011) [hereafter SEC Investment Adviser and Broker-Dealer Study], available at (describing the activities in which broker-dealers engage); see also Securities Exchange Act § 3(a)(4), 15 U.S.C. § 78c(a)(4) (2012) (defining a "broker" as a "person engaged in the business of effecting securities transactions for [others]").

(34.) Securities Exchange Act § 3(b)(5), 15 U.S.C. § 78c(a)(5) (2012) (defining a "dealer" as a "person engaged in the business of buying and selling securities ... for [his] own account"); Fin. Stability Oversight Council, Study & Recommendations on Prohibitions on Proprietary Trading & Certain Relationships with Hedge Funds & Private Equity Funds 18-19 (Jan. 2011) [hereinafter FSOC Study], available at 201%2018%2011%20rg.pdf (describing "market-making").

(35.) See infra Part IV.A (describing the Volcker Rule's prohibition on proprietary trading).

(36.) Hedge funds and private equity funds are investment vehicles that pool together money from various investors--typically institutions and high net-worth individuals--to invest in financial instruments or real assets. See Nora Jordan et al., Advising Private Funds: A Comprehensive Guide to Representing Hedge Funds, Private Equity Funds and Their Advisers 1 (2010) (defining investment funds generally). Their investment strategies differ--hedge funds invest with a short-term horizon in liquid securities and private equity funds typically invest with a longer-term horizon in businesses, often to acquire control of them.

(37.) Asset management activities are also often referred to as "investment management" or "funds management" activities.

(38.) Financial Services Modernization Act (Gramm-Leach-Bliley Act), Pub. L. No. 106-02, 113 Stat. 1338 (1999) (codified in various sections of 12 and 15 U.S.C.).

(39.) Banking Act of 1933 (Glass-Steagall Act), Pub. L. No. 73-66, 48 Stat. 162 (1933) (codified as amended in various sections of 12 U.S.C.), repealed in part by Gramm-Leach-Bliley Act, Pub. L. No. 106-02, 113 Stat. 1338.

(40.) To qualify as a financial holding company, a bank holding company must meet three criteria, including being well-capitalized and well-managed. See Gramm-Leach-Bliley Act § 103(a) (codified at 12 U.S.C. § 1843(l) (2012)) (requiring banks engaged in holding company activities to well capitalize all subsidiary depository institutions, well manage all their depository institutions, and file all appropriate declarations and certifications). Financial holding companies are permitted to engage in activities that are "financial in nature" or incidental or complementary to such activities. See id. (codified at 12 U.S.C. § 1843(k) (2012)) (permitting financial holding companies to engage in financial and certain related activities).

(41.) Fin. Crisis Inquiry Comm'n, The Financial Crisis Inquiry Report: Final Report of the National Commission on the Causes of the Financial and Economic Crisis in the United States 362-63 (2011) [hereinafter FCIC Report].

(42.) See Dafna Avraham et al., A Structural View of U.S. Bank Holding Companies, Fed. Res. Bank OF N.Y. Econ. Pol'y Rev., July 2012, at 65, 65 ("Large banking organizations in the United States are generally organized according to a bank holding company (BHC) structure.").

(43.) The dual challenges identified in this Article are analogous to the two purposes information barriers may serve. See Norman S. Poser, supra note 16 ("What has not been generally appreciated is that the Chinese Wall may have two very different purposes.... on the one hand, its purpose may be only prophylactic: to prevent inside information ... from being misused.... on the other hand, its purpose may also be legal: to provide a defense to the firm against liability....").

(44.) The Restatement (Third) of Agency defines agency as "the fiduciary relationship that arises when one person (a 'principal') manifests assent to another person (an 'agent') that the agent shall act on the principal's behalf and subject to the principal's control, and the agent manifests assent or otherwise consents so to act." Restatement (Third) of Agency § 1.01 (2006).

(45.) See, e.g., Ross Cranston, Insider Dealing--Informational Imbalances and Financial Businesses, in European Insider Dealing 203, 210 (Klaus J. Hopt & Eddy Wymeersch eds., 1991) ("English common law developed at a time when organisations were relatively simple. its application to the multifunctional financial business is not an easy matter."); see also Restatement (Third) of Agency 6 (2006) ("Doctrines within the common law of agency are formulated without regard to whether a person is an individual or a legal or commercial entity or other legally recognized nonindividual person, including an organization.").

(46.) An extensive literature identifies the legal challenge arising from the clash between organizational structure and agency doctrine. See generally Leonard Chazen, Reinforcing The Chinese Wall: A Response, 51 N.Y.U. L. Rev. 552 (1984) (discussing the conflicting financial institution duties of confidentiality and investment disclosure); Lipton, supra note 16 (discussing inherent conflicting duties within financial institutions); Martin Lipton & Robert B. Mazur, The Chinese Wall: A Reply to Chazen, 51 N.Y.U. L. Rev. 579 (1976) (arguing that Chinese Wall provisions that fail to disallow security recommendations and restricted lists for tainted securities are inadequate protections against the conflicting duties of brokerage firms); Bevis Longstreth, Fiduciaries, Capital Markets and Regulation: The Current Challenge, 7 Ann. Rev. BANKING L. 237, 248-49 (1988) (discussing the conflicting duties arising from individuals in a firm possessing material non-public information); Note, Conflicting Duties of Brokerage Firms, supra note 16 (discussing the dynamics behind the conflicting duties faced by brokerage firms as well as potential means for relieving these conflicts); Varn, supra note 16 (discussing the conflicting brokerage firm duties of client confidentiality of non-public information and disclosure of relevant investment information and arguing that policies put in place to alleviate this conflict should serve as a rebuttal to the presumption that knowledge of an individual member of the firm is equivalent to knowledge of the firm).

(47.) See, e.g., Goldman Sachs & Co., supra note 32, at 177 ("We [Goldman Sachs] regularly receive confidential information as part of our normal client relationships."); Victor Brudney & Robert Charles Clark, A New Look at Corporate Opportunities, 94 Harv. L. Rev. 997, 1039-41 (1981) (discussing the access to inside information of investment professionals); William R. McLucas et al., Settlement of Insider Trading Cases with

the SEC, 48 Bus. Law. 79, 101 (1992-93) (referring to the opportunities for investment bankers to profit from the non-public information that regularly comes into their possession); A.C. Pritchard, United States v. O'Hagan: Agency Law and Justice Powell's Legacy for the Law of Insider Trading, 78 B.U. L. Rev. 13, 47 (1998) ("The securities markets provide an especially tempting venue for a faithless agent to breach his fiduciary duty to his principal and misappropriate confidential information.").

(48.) In acting as a broker-dealer, a financial conglomerate may be an agent of its client (and thus owe fiduciary duties). See SEC Investment Adviser and Broker-Dealer Study, supra note 33, at 50-73 (discussing the fiduciary duties broker-dealers may owe); Norman S. Poser & James A. Fanto, Broker-Dealer Law and Regulation § 16 (4th ed. 2011) (discussing the fiduciary duties of broker-dealers). The relationship between a financial conglomerate as M&A adviser and its client may be fiduciary, depending on the features of the relationship. See generally Andrew F. Tuch, Investment Bankers as Fiduciaries: Implications for Conflicts of Interest, 29 Melb. U. L. Rev. 478 (2005) (examining whether M&A advisers owe fiduciary duties to their clients). However, it is common in their engagement letters with M&A clients for financial conglomerates to disclaim the existence of fiduciary duties.

Similarly, financial conglomerates may owe fiduciary duties to their underwriting clients; although, again, this will depend on a factual assessment, including the contractual contractual relations between the parties. See EBC I, Inc. v. Goldman Sachs & Co., 936 N.Y.S.2d 92, 100-01 (N.Y. App. Div. 2011) (describing the question as to whether a fiduciary relationship existed as factual and finding against the existence of a fiduciary relationship on the immediate facts); see generally Andrew F. Tuch, Securities Underwriters in Public Capital Markets: The Existence, Parameters and Consequences of the Fiduciary Obligation to Avoid Conflicts, 7 J. Corp. L. Stud. 51 (2007) (considering the circumstances in which underwriters owe fiduciary duties to their issuer clients).

(49.) The duty of disclosure arises as a matter of agency law. See Restatement (Second) of Agency § 381 (1958); Restatement (Third) of Agency § 8.01 (2006); see also Black v. Shearson, Hammill & Co., 72 Cal. Rptr. 157, 160-61 (Cal. Ct. App. 1968) (imposing a duty of disclosure on a financial conglomerate). The so-called "shingle theory" of federal securities regulation may also provide a basis for financial conglomerates when acting as broker-dealers to disclose adverse information in their possession to clients. See Varn, supra note 16, at 209-11 (discussing the duty of agents to give principals relevant information).

(50.) The duty of confidence arises as a matter of agency law. Restatement (Third) of Agency § 8.05 (2006). Engagement letters between financial conglomerates and investment banking clients also typically require financial conglomerates to protect the confidentiality of any non-public client information they receive.

(51.) See Poser, supra note 16 at 94-103 (discussing situations "in which a strict interpretation of an agent's duty of loyalty would make it difficult if not impossible for a multi-service securities firm to pursue its normal activities."); Varn, supra note 16, at 200-15 (discussing "the problem of conflicting duties"). In fact, the variety of fact patterns in which agency duties may arise makes it difficult to generalize when conflicts of duties arise. Indeed, noting the "multifarious possibilities of conflict of obligation or interest" that arise in financial conglomerates, the SEC warned of the dangers of generalizing "as to the problems presented or possible remedies." Report of Special Study of Securities Markets of the Securities and Exchange Commission, H.R. Doc. No. 88-95, pt. 5, at 65 (1963) [hereinafter Special Study of the SEC].

(52.) in its brokerage and asset management activities in particular, a financial conglomerate will act for many thousands of clients. See, e.g., Goldman Sachs & Co., supra note 32, at 5 (discussing the firm's "diverse" client base, described as "numbering 2,000 institutional clients and third-party distributors, and over 25,000 private wealth management accounts").

(53.) Duties may arise under agency law or statute. As an example of the latter, the firm, as an asset manager, must typically avoid--and not simply manage--conflicts with the interests of its asset-management clients. The duties a financial conglomerate may owe to its asset management clients may be governed by the Employee Retirement Income Security Act (ERISA), which imposes a duty on advisers to avoid conflicts of interest. See AFTRA Ret. Fund v. JP Morgan Chase Bank, N.A., 806 F. Supp. 2d 662, 681 (S.D.N.Y. 2011) (noting that ERISA Regulations require agents to disclose material conflicts of interest).

(54.) Section 10(b) of the Securities Exchange Act of 1934 and associated Rule 10b-5 prohibit trading in securities on the basis of material, non-public information. 15 U.S.C. § 78j(b) (2012); 17 C.F.R. § 240.10-5 (2013). Legal doctrine requires a person in possession of non-public information--such as the financial conglomerate--to either disclose the information or abstain from trading on it. See Sec. & Exch. Comm'n v. Tex. Gulf Sulphur Co., 401 F.2d 833, 849-50 (2d Cir. 1968) (holding that an insider holding material information that could affect investors' desires to buy, sell, or hold securities must disclose such information or refrain from trading), cert. denied, 394 U.S. 976 (1969).

(55.) See Varn, supra note 16, at 207-08 (pointing out that courts take a rather broad stance when imposing liability for disclosure or use of insider information). However, because information is not automatically attributed from an individual to her firm, violation of Rule 10b-5 will not follow as a matter of course. For a more in-depth discussion, see Donald C. Langevoort, Insider Trading: Regulation, Enforcement and Prevention § 12:3 (2013).

(56.) See Varn, supra note 16, at 202-03 (discussing the "classic case" of incompatible duties for financial conglomerates).

(57.) On occasion, courts have failed to admit the existence of incompatible duties by refusing to concede that a financial conglomerate owes a duty of disclosure. See, e.g., Cotton v. Merrill, Lynch, Pierce, Fenner & Smith, Inc., 699 F. Supp. 251, 256-57 (N.D. Okla. 1988) (deciding that a financial conglomerate serving as M&A advisor in a deal did not breach its duty of disclosure to a brokerage client by failing to disclose information about the proposed M&A deal to that brokerage client). In the main, however, scholars accept the risks of financial conglomerates owing incompatible duties. See supra note 16 and accompanying text (noting the abundance of legal literature about incompatible duties).

(58.) 72 Cal. Rptr. 157 (Cal. Ct. App. 1968).

(59.) Id. at 161.

(60.) It is typically only when one party owes a duty of loyalty--such as that imposed by agency law--that a conflict of interest becomes problematic. See Paul Nelson, Capital Markets Law and Compliance: The Implications of MiFID 146-50 (2008) (explaining that a conflict of interest arises only where there is "some special reason," such as the existence of a fiduciary relationship, "under which one party's interest has to be subordinated, in his own conduct, to the interest of the other party").

(61.) According to Paul Volcker testifying before Congress, "When the bank [or financial conglomerate] itself is a 'customer,' i.e., it is trading for its own account, it will almost inevitably find itself, consciously or inadvertently, acting at cross purposes to the interests of an unrelated commercial customer of a bank." S. Rep. No. 111-176, at 91 (2010).

(62.) See id. (providing Paul Volcker's description of conflicts of interest as "inherent in the participation of [financial conglomerates] in proprietary or private investment activity"); Inst. of Bankers, Conflicts of Interest in the Changing Financial World xv (R. M. Goode ed., 1986) (discussing the "inescapable" existence of conflicts of interest in financial conglomerates); Law Commission, supra note 1, at 12-18, 61-63 (discussing potential sources from which conflicts of interest and of duty arise in financial conglomerates and how to manage them); Poser, supra note 16, at 95 ("[C]onflicts of interest and of duty ... are endemic to a multi-service firm."); Stiglitz, supra note 16, at 158 ("The problems [of conflicts of interest] are endemic to the banking industry, and have long been recognized.").

(63.) Special Study of the SEC, supra note 51.

(64.) Philip Augar, The Greed Merchants: How the Investment Banks Played the Free Market Game 23 (2005) (quoting Philip Purcell, Chief Exec. Officer, Morgan Stanley, Speech at Securities Industry Association 2003 Annual Meeting Conference: Reconnecting with America (Nov. 7, 2003)).

(65.) Contractual techniques to modify or exclude fiduciary duties, while commonly employed, have doubtful effects. See, e.g., Law Commission, supra note 1, at 80-88 (discussing uncertainty as to the effect of contractual terms seeking to modify or exclude fiduciary duties); Restatement (Third) of Agency § 1.02 cmt. b (2006) (agreements by parties negatively characterizing the relationship as one not of agency are not determinative of the status of a relationship). A financial conglomerate may also avoid owing multiple incompatible duties by separately incorporating business units to prevent information within one unit (and entity) from being attributed to another. See LAW COMMISSION, FIDUCIARY DUTIES AND REGULATORY RULES: REPORT ON A REFERENCE UNDER SECTION 3(1)(e) of the Law Commissions Act 1965, 1995, Cm. 236, at 102-03 (U.K.) (distinguishing the use of information barriers between different departments of the same company from the use of information barriers between different companies in the same group of companies). But see Australian Sec. & Inv. Comm 'n v. Citigroup Global Mkts. Australia Pty Ltd (No. 4) (2007), 160 FCA 35, 75-77, 82-84 (Austl.) (holding that a contractual disclaimer of fiduciary duties may be effective to exclude a fiduciary relationship arising between the contracting parties).

(66.) This claim is based on interviews with market professionals.

(67.) For an apparent instance of the lack of scholarly consensus, compare Restatement (Third) of Agency § 8.06 (2006) (permitting conduct that would otherwise violate the duty of loyalty if informed consent is provided), with Restatement (Second) of Agency § 387 (1958) (making the agent's duty of loyalty subject to the agreement of the parties, a more easily satisfied requirement than informed consent).

(68.) Slade v. Shearson, Hammill & Co., 517 F.2d 398, 400 (2d Cir. 1974).

(69.) Bd. of Trustees of AFTRA Ret. Fund v. JP Morgan Chase Bank, 806 F. Supp. 2d. 662, 690 (S.D.N.Y. 2011). For more a more detailed discussion of the case, see infra notes 109-16 and accompanying text.

(70.) Disaggregation of the type required to avoid conflicts of duty and of interest would require, at a minimum, segregating the functions of broker and dealer. Such a proposal was considered and rejected during the New Deal era. See generally Sec. & Exch. Comm'n, Report on the Feasibility and Advisability of the Complete Segregation of the Functions of Dealer and Broker (1936), available at 1930/1936_0620_SECSegregationReport.pdf (emphasizing the need for "flexible control and evolutionary development" under the current scheme). Various proposals for disaggregation have been advanced. For a review and recent reform proposal, see generally Arthur A. Wilmarth, Jr., Narrow Banking: An Overdue Reform That Could Solve the Too-Big-to-Fail Problem and Align U.S. and U.K. Regulation of Financial Conglomerations (Part 1), 31 BANKING & Fin. Services POL'Y Rep. 1 (2012), available at

(71.) See Crockett et al., supra note 16, at 6-9 (discussing the costs and benefits of disaggregating financial conglomerates); Ingo Walter, Conflicts of Interest and Market Discipline Among Financial Services Firms 9-10 (N.Y. Univ. Working Paper No. 2451/27263, 2003), available at (last visited Feb. 27, 2014) (discussing the benefits of financial conglomeration, including economies of scope). The benefits of financial conglomeration have been contested. See, e.g., Arthur E. Wilmarth, Jr., How Should We Respond to the Growing Risks of Financial Conglomerates?, in FINANCIAL MODERNIZATION After Gramm-Leach-Bliley 70-90 (Patricia C. McCoy ed., 2002) (questioning the economic benefits of financial conglomeration).

(72.) The certification effect concerns the extent to which the identity of an underwriter--and its relationship to the issuer of securities--influences investors' perceptions of the investment quality of those securities. See Tiago Duarte-Silva, The Market for Certification by External Parties: Evidence from Underwriting and Banking Relationships, 98 J. Fin. ECON. 568, 581 (2010) (showing that underwriters that have lending relationships with issuers provide stronger certifications than other underwriters); Randall S. Kroszner & Raghuram G. Rajan, Is the Glass-Steagall Act Justified? A Study of the U.S. Experience with Universal Banking Before 1933, 84 Am. Econ. Rev. 810, 819-20 (1994) (finding that securities underwritten by bank-affiliated underwriters defaulted less often than securities underwritten by unaffiliated underwriters); Manju Puri, Commercial Banks in Investment Banking: Conflict of Interest or Certification Role?, 40 J. Fin. Econ. 373, 375-76 (1996) (finding that bank-underwritten securities were perceived to be higher quality by investors than securities underwritten by non-bank underwriters).

(73.) The focus here is on the use of non-public information in trading activities.

(74.) In theory, the misuse of non-public information may benefit clients on some occasions. For example, a financial conglomerate advising a target corporation on an M&A deal may use such information from its client to trade in the stock of the target, increasing the price at which the target is ultimately sold. Using such information disclosed by some clients, a financial conglomerate may benefit other firm clients. Using such information might also improve a financial conglomerate's ability to "certify" the securities of its underwriting clients. As to the certification effect, see supra note 72 and accompanying text. Of course, these benefits may come at the cost of even greater third-party harm. As to third-party harm, see infra note 77 and accompanying text. Despite the theoretical possibility that the misuse of non-public information will provide benefits, this Article takes as given the desirability of ensuring information barriers' effectiveness (and thus the undesirability of the misuse of non-public information). The author is unaware of any claims, by regulators, scholars, or financial conglomerates suggesting that the misuse of non-public information be permitted in any circumstances.

(75.) See FSOC Study, supra note 34, at 48 (discussing the possibility of a broker-dealer harming its client by actively trading against the position of that client).

(76.) For an overview of the debate, see Stephen M. Bainbridge, The Law and Economics of Insider Trading: A Comprehensive Primer 61-78 (2001) (unpublished manuscript), available at (providing an overview of the debate surrounding insider trading); Laura Nyantung Beny, Insider Trading Laws and Stock Markets Around the World: An Empirical Contribution to the Theoretical Law and Economics Debate, 32 J. CORP. L. 237, 241-56 (2007) (providing empirical evidence concerning insider trading); Langevoort, supra note 55, §§ 1: 1-1:6.

(77.) Viral Acharya & Timothy Johnson, Insider Trading in Credit Derivatives, 84 J. Fin. Econ. 110, 114 (2007) (internal citations omitted) ("Insider trading has been the focus of a large body of research in equity markets which has found that insider trading lowers liquidity and increases trading costs, raises the cost of equity capital, and increases volatility."). For empirical evidence across countries that is consistent with the view that insider trading imposes market-wide harms, see generally Beny, supra note 76.

(78.) Diminished share price accuracy is likely to lead to inefficiency by "reduc[ing] the likelihood that resources will be allocated to implement the most promising real investment projects available in the economy; by decreasing the feasibility of executive compensation based on securities prices; and by "undermin[ing] the market for corporate control." Merritt B. Fox, Insider Trading in a Globalizing Market: Who Should Regulate What?, 55 Law & Contemp. Probs. 263, 282-83 (1992).

(79.) As is well-known, scholars contest whether trading on non-public information by "insiders" is harmful. See, e.g., Henry G. Manne, Insider Trading and the Stock Market 78-86 (1966) (discussing the market effects of a rule permitting insider trading); Bainbridge, supra note 76, at 61-82 (discussing the reasons for federal regulation of insider trading and analyzing different arguments about whether or not insider trading should, in fact, be regulated); Beny, supra note 76, at 241-56 (reviewing the debate over the effects of insider trading). The phenomenon under consideration in this Article, however, refers to "outsider trading," which is trading by financial conglomerates on the non-public information of their clients. No apparent contention exists among scholars that such trading should be banned. Moreover, the claim that "insider trading" increases the accuracy of stock prices (by increasing the speed with which they reflect all available information) is inapposite in the context of outsider trading because trading by financial conglomerates is not conspicuously informed trading and, in consequence, may not prompt derivative trading sufficient to increase stock price accuracy. Observers may be unable to distinguish between (informed) trading on a firm's own account and (uninformed) trading on behalf of a client, especially if financial conglomerates disguise the signal their trading sends. As to the possibility of disguising signals, see Pritchard, supra note 47, at 52.

(80.) See Michael C. Jensen & William H. Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, 3 J. Fin. Econ. 305, 308-10, 357 (1976) (defining and discussing agency costs).

(81.) In re El Paso Corp. S'holder Litig., 41 A.3d 432, 445 (Del. Ch. 2011).

(82.) Aware of the possibility of skewed incentives, procedures were put in place to prevent Goldman's advisers from advising El Paso on its sale to Kinder Morgan; instead, they were to advise only on an alternative proposed spin-off of one of El Paso's business units. The Delaware Court of Chancery questioned the effectiveness of these procedures. Id. at 440 ("Goldman still played an important role in advising the Board [of El Paso] by suggesting that the Board should avoid causing Kinder Morgan to go hostile and by presenting information about the value of pursuing the spin-off instead of the Kinder Morgan deal.").

(83.) The practice of investment banking clients hiring multiple financial conglomerates as advisers is common. See, e.g., Vipal Monga & Suzanne Stevens, Nice Work If You Can Get It, The Deal Pipeline, Sept. 11, 2009, available at

(84.) See supra notes 7-13 and accompanying text (discussing Verizon's acquisition of RCC).

(85.) The analysis in these examples excludes the potential effect of contractual disclaimers or the existence of multiple legal entities within a financial conglomerate. As to the doubtful effect of these measures in overcoming the legal challenges of financial conglomeration, see supra note 65 and accompanying text.

(86.) See Asset Management, J.P. Morgan, (last visited May 30, 2014) ("As one of the largest asset and wealth managers in the world, with assets under supervision of $2.2 trillion and assets under management of $1.5 trillion (as of September 30, 2013), we provide global market insights and a range of investment capabilities that few other firms can match.").

(87.) Australian Sec. and Inv. Comm'n v. Citigroup Global Mkts. Australia Pty Ltd (No. 4) (2007), 160 FCR 35, 43 (Austl.).

(88.) The Toll Holdings-Patrick Corporation transaction occurred under Australian statutory and common law. Although Australian and U.S. legal principles are common in many respects, the analysis above assumes the transaction was subject to U.S. law.

(89.) Australian Sec. and Inv. Comm'n, 160 FCR at 62-65.

(90.) The relevant provisions reflect a piecemeal approach to regulation. As broker-dealers, financial conglomerates must "establish, maintain and enforce" information barriers. Securities Exchange Act § 15(g), 15 U.S.C. § 78o(g) (2012). Registered investment advisers are subject to an identical provision. See Investment Advisers Act § 204A, 15 U.S.C. § 80b-4A. Broker-dealers and investment advisers may face civil penalties for "knowingly or recklessly fail[ing] to establish, maintain, or enforce" these required policies or procedures. Securities Exchange Act § 21A(b)(1)(B), 15 U.S.C. § 78U-1(b)(1)(B). In addition, the Federal Deposit Insurance Corporation (FDIC), the primary federal regulator for state-chartered banks that do not join the Federal Reserve System, requires banks to have in place information barriers when these banks simultaneously provide asset management and commercial banking services. See Fed. Deposit Ins. Corp., Trust Examination Manual §§ D-D.1 (2005), available at trustmanual/section_8/section_viii.html#mpi ("Policies and procedures should be established prohibiting the use of material, nonpublic information, when deciding whether to buy or sell securities."); see also FSOC Study, supra note 34, at 49-50 (describing the requirement for financial conglomerates to have in place information barriers).

(91.) See Tender Offers, Securities Act Release No. 6239, 45 Fed. Reg. 60,410, 60,415 (Sept. 12, 1980) (codified as amended at 17 C.F.R. § 240.14e-3) (describing the effectiveness and risks associated with using information barriers and other policies and procedures to prevent the use of material, nonpublic information relating to a tender offer); Nat'l Ass'n of Sec. Dealers & N.Y. Stock Exch., supra note 2 (specifying the requirements for memorialization, review, supervision, and education in an information barrier system).

(92.) See Langevoort, supra note 55, § 12:3 ("The actual rules vary from firm to firm.").

(93.) P.G. Willis, Chinese Walls: Myth, Metaphor and Reality--Living with Fiduciary Duties in Resources Relationships, in AMPLA Yearbook 573, 574 (2002).

(94.) Self-regulatory organizations have described in broad terms the "minimum elements" for "adequate" information barriers. See, e.g., Nat'l Ass'n of Sec. Dealers & N.Y. Stock Exch., supra note 2 (explaining the "minimum elements" for "adequate" information barriers).

(95.) See infra note 112 (referring to these individuals as "wall straddlers").

(96.) Poser, supra note 16, at 103 ("In the U.S., the concept of the Chinese Wall was created as a preventive measure to control the specific problem of misuse of insider information by multi-service securities firms.").

(97.) As an indication of the intended effect of information barriers, in Slade v. Shearson, Hammill & Co., 18 Fed. R. Serv. 2d 265 (S.D.N.Y. 1974), the financial conglomerate, Shearson, argued that its information barrier prevented information in possession of its investment banking unit from being disclosed to its brokerage unit, thus relieving its brokerage unit of the duty to disclose it. Id. at 1.

(98.) In doing so, the information barrier is intended to serve the prophylactic purpose of preventing the use of non-public information. See Poser, supra note 16, at 93.

(99.) See In re Merrill Lynch, Pierce, Fenner & Smith, Inc., Exchange Act Release No. 8459, [1967-69 Transfer Binder] Fed. Sec. L. Rep (CCH) ¶ 77,629, at 83,350 (Nov. 25, 1968) (accepting an offer of settlement that required the firm to implement information barriers).

(100.) See supra note 90 (identifying various regulations requiring the use of information barriers).

(101.) See 17 C.F.R. § 240.14e-3 (2012) (permitting information barriers to serve as a defense to the offence under Rule 14e-3 of trading on non-public information in the context of tender offers). The use of information barriers may serve as a defense to insider trading in other contexts as well. See id. § 240.10b5-1 (permitting firms to use information barriers to demonstrate that a purchase or sale of securities was not "on the basis of' material non-public information for purposes of insider trading liability). See also Poser, supra note 16, at 93 ("Only in recent years has the [Chinese] wall been used for legal purposes, and then only as a defense against liability under a specific SEC rule outlawing insider trading in connection with a pending tender offer.").

(102.) Slade v. Shearson, Hammill & Co., Inc., 517 F.2d 398, 403 (2d Cir. 1974). The district court certified a question to the Second Circuit, although the case settled before the Second Circuit determined the matter. Slade v. Shearson, Hammill & Co., 42 CIV. 4779, 1974 WL 376 (S.D.N.Y. Mar. 18, 1974) (mem.).

(103.) Lipton & Mazur, supra note 16, at 467-68. Some firms also adopt "watch lists," which list the securities in which firms' compliance departments will monitor trading with the purpose of detecting suspicious trading by the firm or its employees. Gorman, supra note 16, at 494-95.

(104.) Id.

(105.) Cf. supra note 101 and accompanying text.

(106.) Financial Services Modernization Act (Gramm-Leach-Bliley Act), Pub. L. No. 106-02, 113 Stat. 1338 (1999) (codified in scattered sections of 12 and 15 U.S.C.).

(107.) See supra notes 38-42 and accompanying text (discussing the qualification of bank holding companies and financial holding companies).

(108.) H.R. Rep. No. 106-74, pt. 1, at 2 (1999).

(109.) 806 F. Supp. 2d 662 (S.D.N.Y. 2011).

(110.) Before this decision, the most authoritative statement of the legal effect of information barriers was provided by the district court decision in Slade v. Shearson, Hammill & Co., 18 Fed. R. Serv. 2d 265 (S.D.N.Y. 1974). See also Poser, supra note 16, at 108 (stating that the district court's decision stood as "the most authoritative statement of law on the subject"). In Slade, the district court rejected the financial conglomerate's reliance on its information barriers as a defense to its alleged breach of the duty of disclosure to a brokerage client. Poser, supra note 16, at 107.

(111.) Bd. of Trustees, 806 F. Supp. 2d at 688.

(112.) Id. at 688-89 & n.149 (citation omitted) (internal quotation marks omitted). The court noted that some individuals at JP Morgan--whom it referred to as "wall straddlers"--possessed information from both sides of the information barrier. Id. at 689. These individuals' presence did not compromise the information barriers' legal effectiveness.

(113.) Id. at 689 (citation omitted) (internal quotation marks omitted).

(114.) Id. at 690.

(115.) The investment management clients also alleged that the firm violated its duty to disclose its conflicts of interest. Bd. of Trustees, 806 F. Supp. 2d at 693. However, the court reasoned that JP Morgan owed no such duty because no conflict of interest arose. Id. at 694.

(116.) The court explained that an information barrier cannot circumvent a duty to disclose information by "those employees on whom [investors] reasonably rely for important information and guidance--especially when the evidence strongly suggests that those employees were actually kept in the loop about and aware of the risks materializing in one of their fiduciary clients' [sic] largest investments." Id at 695-96 (citations omitted) (internal quotation marks omitted). By inference, JP Morgan's information barrier would have protected it from disclosing its prediction about Sigma if only its personnel on the (lending) side of the information barrier knew that prediction, rather than also those on the asset management side.

(117.) See, e.g., Poser, supra note 16, at 93 (referring to Chinese walls at some firms as "nothing but a convenient fiction aimed at avoiding liability for market abuses"); Stiglitz, supra note 16, at 149 ("[Economists have long been skeptical of Chinese walls...."); Volcker, supra note 23 (describing as "naive" the view that "Chinese Walls can remain impermeable against the pressures to seek maximum profit and personal remuneration").

(118.) See Bodnaruk et al., supra note 16, at 5024 (finding results consistent with financial conglomerates violating information barriers by exploiting non-public information from M&A clients in their trading activities); Narasimhan Jegadeesh & Yue Tang, Institutional Trades Around Takeover Announcements: Skills vs. Inside Information 23-24 (Dec. 2010) (unpublished manuscript), available at sol3/papers.cfm?abstract_id=1568859 (finding that financial conglomerates use non-public information garnered from their M&A clients for the benefit of some of their brokerage clients). See also Seyhun, supra note 16 (concluding that non-public information garnered by financial conglomerates through having their employees serve on corporate boards is used by financial conglomerates in their trading operations and interpreting that evidence as supporting the notion of information barriers as "porous and ineffective"). In this M&A context, advisers know the identity of the proposed bidder and target companies prior to the public disclosure of this information--information allowing firms to profit by taking stakes in the target company prior to its price increasing at the time the deal is announced.

(119.) See generally Ivashina & Sun, supra note 16 (finding that institutional investors, including financial conglomerates, trade using non-public information garnered from participating in discussions by borrowers to amend their loan agreements); Massa & Rehman, supra note 16 (finding that financial conglomerates' asset managers trade using non-public information from the firms' lending divisions). In this context, financial conglomerates will possess non-public information about the borrower, including information about the likelihood of default; this information may enable firms to trade against the borrower in equities or derivatives markets, and to profit on the borrower's expected reduction in value.

(120.) See Acharya & Johnson, supra note 77, at 116 (finding that financial conglomerates make use of material, non-public information--garnered from lending to corporate borrowers--in their trading in credit default swaps--financial instruments allowing parties to speculate on the occurrence of a particular "credit event" in an underlying security).

(121.) See, e.g., Massa & Rehman, supra note 16 (finding that mutual funds condition their decisions on the lending activities of their affiliated banks); Sergey Barabanov et al., Underwriters and the Broken Chinese Wall: Institutional Holdings and Post-IPO Securities Litigation, J. Fin. RESEARCH (forthcoming), available at (finding evidence consistent with financial conglomerates trading using nonpublic information garnered from their roles as lead securities underwriters).

(122.) Bodnaruk et al., supra note 16, at 4990 (finding that M&A advisers taint their advice to bidders, leading to inferior deal outcomes for their clients relative to clients of un-conflicted financial conglomerates).

(123.) Id. at 5016-24. Other evidence suggests that corporations with conflicted M&A advisers may not experience inferior deal outcomes. See, e.g., Charles W. Calomiris & Donna M. Hitscherich, Banker Fees and Acquisition Premia for Targets in Cash Tender Offers: Challenges to the Popular Wisdom on Banker Conflicts, J. Emp. Leg. Stud. 909, 937 (2007); Charles W. Calomiris & Hal J. Singer, How Often Do "Conflict of Interests" in the Investment Banking Industry Arise During Hostile Takeovers 4 (Feb. 26, 2004) (unpublished manuscript), available at However, these studies cast no doubt on the evidence reported above since the conflicts of interest facing the financial conglomerates in these studies did not stem from the financial stakes firms held in the corporations involved.

(124.) See Press Release, Fin. Indus. Regulatory Auth., Ten of Nation's Top Investment Firms Settle Enforcement Actions Involving Conflicts of Interest Between Research and Investment Banking (Apr. 28, 2003), available at (announcing regulatory action against financial conglomerates arising from conflicts of interest between research and investment banking).

(125.) Id. (discussing reforms imposed on financial conglomerates, including the physical separation of research and investment banking departments "to prevent the flow of information between the two groups").

(126.) For discussion of "endemic" insider trading in credit default swap markets and the possible use of information barriers to "blunt suggestions of insider trading," see Yesha Yadav, Insider Trading in the Derivatives Market (and What it Means for Everyone Else) 31-32 (Vanderbilt Law and Econ. Research Paper No. 13-24, 2013), available at

(127.) See In re El Paso Corp. S'holder Litig., 41 A.3d 432, 440-42 (Del. Ch. 2012) (questioning the effectiveness of the information barriers of a financial conglomerate that provided M&A advice to the target company while also holding a stake in the acquirer).

(128.) Securities Exchange Act § 15(g), 15 U.S.C. § 78o(g) (2012). See also supra note 90 (explaining that broker-dealers and advisers must have information barriers according to the Securities Exchange Act and the FDIC).

(129.) The SEC prosecutes trading on the basis of material, non-public information under the general antifraud provisions of the federal securities laws, most commonly Section 10(b) of the Securities Exchange Act and the associated Rule 10b-5.

(130.) See supra notes 76-79 and accompanying text (discussing third-party harms caused by the use of nonpublic information, such as lowered market liquidity and diminished accuracy of stock prices).

(131.) Although sparse, some evidence suggests that financial conglomerates regard their corporate clients as falling on a spectrum of sophistication, with some materially more sophisticated than others. See, e.g., Wall Street and the Financial Crisis: The Role of Investment Banks, Hearing Before the Permanent Subcomm. on Investigations of the Comm. on Homeland Sec. and Gov't Affairs, 111th Cong. 28-29 (referring to an e-mail by a Goldman Sachs employee distinguishing among institutional clients on the basis of their sophistication).

(132.) Former Delaware Chancellor William T. Allen recognized that clients typically have "little or no experience in the sale of a public company.... Naturally, they turn for guidance to their specialist advisors who will typically have had a great deal of relevant experience." William T. Allen, Independent Directors in MBO Transactions: Are They Fact or Fantasy?, 45 BUS. Law. 2055, 2061 (1989-90).

(133.) A voluminous literature stemming from the seminal work of professors Adolf Berle and Gardiner Means considers the lack of coincidence between the interests of corporate managers and shareholders. See generally Adolf A. Berle, Jr. & Gardiner C. Means, The Modern Corporation and Private Property (1932) (explaining the divergence of interests between corporate management and widely dispersed shareholders). For important recent contributions and assessments of strands of the literature, see generally William W. Bratton, Berle and Means Reconsidered at the Century's Turn, J. Corp. L. 737 (2001); Jennifer G. Hill, Then and Now: Professor Berle and the Unpredictable Shareholder, 33 SEATTLE U. L. Rev. 1005 (2010).

(134.) For example, financial conglomerates offered stock to corporate executives in "hot" initial public offerings, apparently to induce those executives to direct business to financial conglomerates. See Therese H. Maynard, Spinning in a Hot IPO--Breach of Fiduciary Duty or Business as Usual?, 43 WM. & Mary L. Rev. 2023, 2026 (2002) (discussing how the CEO of a computer software firm received shares of Pixar's hot IPO in the hope that he would "direct his company's future investment banking business to" the IPO's lead underwriter). Some financial conglomerates also offer something of a job-placement network to help corporate executives with future employment, a service that may encourage executives to make decisions in the interests of financial conglomerates, rather than in those of the firm. See Ivo Welch, Corporate Finance: An Introduction 874 (2009) (discussing factors influencing corporate managers' selection of financial advisers, including financial conglomerates' ability to help with job placements).

(135.) Jensen, supra note 6. In sentencing Rajat Gupta for insider trading, Judge Jed Rakoff asserted that "[i]nsider trading is an easy crime to commit but a difficult crime to catch." Kara Scannell, Gupta Sentenced to Two Years in Jail, Fin. Times (Oct. 24, 2012), See also Insider Trading: Tipping the Scales, The ECONOMIST, Oct. 15, 2011, at 83 ("Insider trading is tough to prove, and it can take years to gather enough evidence to bring charges.").

(136.) See Lipton & Mazur, supra note 16, at 494 ("The difficulty of discovering misuse of inside information is, of course, the greatest shortcoming of the Chinese Wall approach.").

(137.) Insider Trading: Tipping the Scales, supra note 135 ("Detection [of insider trading] is also getting harder. Investors are able to trade more quickly and in greater volumes than ever before.").

(138.) See Ben Protess & Michael J. De La Merced, S.E.C., Suspecting Insider Trading, Freezes Account Over Heinz Buyout, N.Y. Times Dealbook (Feb. 15, 2013, 8:52 PM) (quoting the head of the SEC's market abuse unit as saying, "Irregular and highly suspicious options trading immediately in front of a merger or acquisition announcement is a serious red flag that traders may be improperly acting on confidential nonpublic information.").

(139.) See Melissa Maleske, An Insider Falls, Inside Counsel, Aug. 2012, at 28, 29 (discussing the use of circumstantial evidence to prove insider trading cases and the difficulty of proving "a circumstantial case"). The primary exception to the use of circumstantial evidence involves wire-tap evidence, which proved instrumental in the high profile convictions of hedge fund manager Raj Rajaratnam and former corporate director Rajat Gupta. See Chad Bray & Reed Albergotti, Gupta Jurors Hear Wiretaps, Wall St. J., May 25, 2012, at C3 (referring to prosecutors' view of wiretapped phone conversations as "the most important pieces of evidence that Mr. Gupta shared confidential information ... with Mr. Rajaratnam"). However, wire-tap evidence must be planned in advance because of the need for judicial approval.

(140.) Margaret Cole, Director of Enforcement, U.K. Fin. Servs. Auth., Address at the London School of Economics: Insider Dealing in the City (Mar. 17, 2007), available at Communication/Speeches/2007/0317_mc.shtml (discussing the difficulty of proving insider trading); Linda Chatman Thomsen, Testimony Concerning Insider Trading: Before the U.S. Senate Comm. on the Judiciary, Sec. & Exch. Comm'n (Oct. 5, 2006), (describing several cases where nonpublic information was allegedly misused).

(141.) Thomsen, supra note 140.

(142.) Regulators achieved apparently spectacular success in prosecuting individuals associated with SAC Capital Advisors for insider trading or related offenses. According to reports, the multi-year investigation heated up only when federal prosecutors received permission from a judge to wiretap the phone of one prominent target. Michael Rothfield et al., Trading Charges Reach SAC, Wall St. J, Nov. 21, 2012, at A1. Cooperating witnesses also gathered evidence using body wires. These effective investigative techniques require advance suspicions. They would seem less suitable when regulatory interest is triggered by suspicious trading activity. See Maleske, supra note 139 (discussing the advantages of wiretap evidence over circumstantial evidence in proving insider trading cases).

(143.) U.S. Gov't Accountability Office, GAO-11-529, Proprietary Trading: Regulators Will Need More Comprehensive Information to Fully Monitor Compliance with New Restrictions When Implemented 33 (2011), available at (discussing the practice of front-running) [hereinafter GAO Proprietary Trading Report].

(144.) Id.

(145.) See Maremont & Craig, supra note 7 (discussing cases involving the spread of nonpublic information). In one high profile instance, the SEC detected the apparent use of non-public information by (and the failure of information barriers at) a financial conglomerate. The conduct allegedly occurred between 2003 and 2005. However, as is often the case, the firm settled the action, neither admitting nor denying liability. See Merrill Lynch, Pierce, Fenner & Smith Inc., Exchange Act Release No. 63760, 100 SEC Docket 957, 2011 WL 231575, at *1-2, 5-6 (Mar. 25, 2011) (finding that the firm violated Section 15(g) of the Exchange Act, instituting administrative and cease-and-desist proceedings, and imposing remedial sanctions).

(146.) See Dennis K. Berman, The Galleon Legacy: White-Collar Wiretaps, WALL St. J., May 12, 2011, at C1 (discussing the use of wiretaps as a new shift in regulators' prosecution of insider trading); Michael Rothfeld, For U.S., an OldPlaybook in the Gupta Case, Wall St. J., Oct. 28, 2011, at B1 (asserting that recent "insider-trading convictions or guilty pleas in the past two years [have been] based largely on wiretaps that proved defendants leaked confidential information"); Michael Rothfeld et al., Fund Titan Found Guilty, Wall St. J., May 12, 2011, at A1 (referring to juror comments that wiretaps of Mr. Rajaratnam "were the deciding factor" in his conviction); Jenny Strasburg & Reed Albergotti, Insider Targets Expanding, Wall St. J. (Feb. 28, 2012), (referring to the use of wiretaps and other investigative methods once confined to terrorist and drug cases and the cooperation of witnesses who have been enlisted to secretly record conversations).

(147.) One would also expect the effectiveness of this form of evidence to decline as wrongdoers appreciate the risk of their conversations being recorded.

(148.) Maremont & Craig, supra note 7.

(149.) Lucas van Praag, Goldman Highly Confident in its Chinese Walls, Fin. TIMES (Apr. 26, 2010),; see also Patrick Jenkins, Goldman Had Two Roles in Lloyds Deal, Fin. Times (Apr. 23, 2010), 33752ad4-4e6f-11df-b48d-00144feab49a.html#axzz2w3clFcy0 ("Goldman bankers stress that its Chinese walls bar underwriters from knowing anything about the investment activity of its proprietary traders....").

(150.) Warren A. Law, Wall Street and the Public Interest, in WALL STREET AND REGULATION 157, 167 (Samuel L. Hayes, III ed., 1987).

(151.) See supra Part III.B (discussing the use of information barriers).

(152.) In an industry review of information barriers in 1990, the SEC focused on matters such as the training of employees, policies on employee trading restrictions, and the physical separation of staff and record-keeping. Sec. & Exch. Comm'n, Broker-Dealer Policies and Procedures Designed to Segment the Flow and Prevent the Misuse of Material Non-Public Information 5-9 (1990), available at It did not review firms' trading records. It concluded that the National Association of Securities Dealers (NASD) and the New York Stock Exchange (NYSE) (now superseded by FINRA) oversee information barriers using an examinations program. Id at 18.

(153.) The Dodd-Frank Act amended the Bank Holding Company Act of 1956 ("BHCA") by introducing section 13. Dodd-Frank Act § 619, 12 U.S.C. § 1851 (2012). Section 13 constitutes the so-called Volcker Rule. It requires the "appropriate Federal banking agencies," the SEC, and the Commodity Futures Trading Commission (CFTC) to issue regulations to implement the Volcker Rule. BHCA § 13(b)(2); 12 U.S.C. § 1851(b)(2). On December 10, 2013, the Board of Governors of the Federal Reserve System, the Office of the Comptroller of the Currency, the FDIC, the SEC, and the CFTC finalized regulations. Prohibitions and Restrictions on Proprietary Trading and Certain Interests In, and Relationships With, Hedge Funds and Private Equity Funds, 79 Fed. Reg. 5535 (Jan. 31, 2014) (to be codified at 12 C.F.R. pts 44, 248, 255, 351) [hereinafter Volcker Rule Implementing Regulations].

(154.) See BHCA § 13(a)(1), 12 U.S.C. § 1851(a)(1) (listing banned activities, including engaging in proprietary trading and acquiring an interest in a hedge fund or private equity fund).

(155.) More specifically, the term captures any entity that is, or is affiliated with, a federally insured depository institution (such as a bank or thrift). BHCA § 13(h)(1), 12 U.S.C. § 1851(h)(1). The Volcker Rule thus applies to Goldman Sachs and Morgan Stanley, firms that converted to bank holding companies during the Financial Crisis.

(156.) The Volcker Rule required the Financial Stability Oversight Council to study and make recommendations on implementing the rule with the purpose of achieving certain objectives. BHCA § 13(b)(1), 12 U.S.C. § 1851(b)(1). These objectives include "promoting] and enhancing] the safety and soundness of banking entities" and "reduc[ing] conflicts of interest between the self-interest" of those institutions covered by the rule and "the interests of the customers of such entities and companies." Id. The inclusion of the latter objective suggests Congress' awareness of the potential for the Volcker Rule to address the legal and practical challenges of financial conglomeration.

(157.) Volcker Rule Implementing Regulations, 79 Fed. Reg. at 5659.

(158.) BHCA § 13(h)(6), 12 U.S.C. § 1851(h)(6). The Rule achieves this result by defining proprietary trading as "engaging as a principal for the trading account of the banking entity ... in any transaction to purchase or sell, or otherwise acquire or dispose of, any security ... or financial instrument" and then defining "trading account" as "any account used for acquiring or taking positions in [such] securities and instruments" for the purpose expressed above. BHCA § 13(h)(4), (6), 12 U.S.C. § 1851(h)(4), (6).

(159.) "Market-making-related activities" are permitted "to the extent that ... [they] are designed not to exceed the reasonably expected near term demands of clients, customers, or counterparties." BHCA § 13(d)(1)(B), 12 U.S.C. § 1851(d)(1)(B).

(160.) BHCA § 13(d)(2)(A)(i), 12 U.S.C. § 1851(d)(2)(A)(i). The proviso refers to conflicts of interest in terms of their effect on "clients, customers, or counterparties," not simply clients. Id.

(161.) BHCA § 13(a)(1)(B), 12 U.S.C. § 1851(a)(1)(B). The statute uses the expression "[acquiring or retaining] any equity, partnership, or other partnership interest" in place of "investing." Id. More specifically, a financial conglomerate may continue to sponsor hedge and private equity funds, with limited exceptions, but may only make de minimis investments in such sponsored funds. The relevant permitted activity is described as "organizing and offering a private equity or hedge fund"--an activity which includes sponsoring such a fund, provided the fund does not share the banking entities name or a variant of it. BHCA § 13(d)(1)(G), 12 U.S.C. § 1851(d)(1)(G). For this activity to be permissible, the fund must be "organized and offered only in connection with the provision of bona fide trust, fiduciary, or investment advisory services and only to persons that are customers of such services of the banking entity." BHCA § 13(d)(1)(G)(ii), 12 U.S.C. § 1852(d)(1)(G)(ii). The banking entity may "not acquire or retain an equity interest, partnership interest, or other ownership interest in the fund[]," with the exception of seed investments (to establish the fund) or de minimis investments in such a fund (but apparently not in a third-party fund). BHCA § 13(d)(1)(G)(iii), 12 U.S.C. § 1852(d)(1)(G)(iii).

(162.) Supra note 160 and accompanying text.

(163.) FSOC Study, supra note 34, at 49 ("In imposing a prohibition on proprietary trading, the Volcker Rule directly cuts off a banking entity's opportunity to profit from conflicted proprietary trading, thereby protecting customers from financial injury arising from such conflicts.").

(164.) Lauren Tara LaCapra, Goldman Lobbying Hard To Weaken Volcker Rule, REUTERS (May 4, 2011), ("The [Volcker] rule ... will cost Goldman at least $3.7 billion in annual revenue, by one [research analyst] estimate. And billions more could be at stake if regulations now being drawn up are extra-tough.").

(165.) E.g., Paritosh Bansal, BofA Spins Off Its Last Large Private Equity Fund, Reuters (June 2, 2011), (explaining Bank of America's effort to comply with the Volcker Rule by spinning off its private equity fund as a new entity--North Cove Partners); Donal Griffin, Citigroup Said to Shut $400 Million Proprietary Fund as Ahmed Has New Role, Bloomberg (June 2, 2011), (noting Citigroup shut down a $400 million hedge fund to comply with the Volcker Rule).

(166.) See supra Part III.B (explaining how information barriers fail to prevent financial conglomerates from trading on non-public client information).

(167.) Supra note 160.

(168.) See, e.g., Steve Eder & Megan Davies, Goldman Proprietary Traders Jump to KKR, REUTERS (Oct. 21, 2010), (describing the decision of KKR, another prominent private equity firm, to build up its proprietary trading activities); Our Businesses, Blackstone, (last visited Feb. 27, 2014) (describing the activities of Blackstone, a prominent private equity firm, which include financial advisory services).

(169.) For instance, KKR purchased the New York proprietary trading desk of Goldman Sachs. Peter Lattman, Goldman's Proprietary Trading Desk to Join K.K.R., N.Y. TIMES DEALBOOK (Oct. 21, 2010, 7:29 PM), blogs&_r=0 ("Even mighty Goldman Sachs is not immune. Its proprietary trading desk in New York ... is heading to the private equity giant Kohlberg Kravis Roberts as Goldman winds down the operation to comply with new federal regulations.").

(170.) Id. ("Wall Street is holding a garage sale, and private equity firms are showing up to buy. The banks are offloading businesses to comply with new rules....").

(171.) The Volcker Rule does include a provision for extending its reach to non-banking entities, but only if these entities are systemically important. See BHCA § 13(b)(2), 12 U.S.C. § 1851(b)(2) (2012) (imposing limitations on proprietary trading and fund sponsorship and investment activities by certain non-bank financial companies). This possible extension of the Rule--justified on the grounds of promoting financial stability rather than mitigating conflicts of interest--fails to overcome the limitation associated with the Rule's narrow reach.

(172.) One of the activities permitted as an exception of the Volcker Rule's general prohibition is transacting "on behalf of customers." BHCA § 13(d)(1)(D), 12 U.S.C. § 1851(d)(1)(D). The exception is subject to the proviso that it not give rise to material conflicts of interest, and might thus be interpreted to subject trading on behalf of clients to regulation under the Volcker Rule. BHCA § 13(d)(2)(A)(i), 12 U.S.C. § 1851(d)(2)(A)(i). However, an interpretation that regulates agency activities of banks must fail. Analytically, activities permitted as exceptions to the general prohibition on proprietary trading must be conducted in a principal capacity; trading on behalf of clients is not conducted in such a capacity and thus is outside the scope of the general prohibition.

(173.) Reflecting the difficulty of identifying proprietary trading, the Government Accountability Office was unable to study proprietary trading at subject firms, other than such trading undertaken in separately designated business units. GAO Proprietary Trading Report, supra note 143, at 14. See also FSOC Study, supra note 34, at 27-47 (discussing at length the considerations involved in identifying proprietary trading for purposes of the Volcker Rule).

(174.) BHCA § 13(d)(1)(B), 12 U.S.C. § 1851(d)(1)(B) (2012); see also FSOC Study, supra note 34, at 28-29.

(175.) BHCA § 13(d)(2)(A)(i), 12 U.S.C. § 1851(d)(2)(A)(i).

(176.) See supra note 160 and accompanying text.

(177.) Volcker Rule Implementing Regulations, 79 Fed. Reg. 5535, 5796 (Jan. 31, 2014) (to be codified at 12 C.F.R. pts 44, 248, 255, 351).

(178.) Id. at 5664. The regulations do not cite any empirical evidence to support the claimed effectiveness of information barriers.

(179.) Id.

(180.) Instead, according to the final regulations implementing the Volcker Rule, regulators require information barriers to be "memorialized in written policies and procedures ... that are reasonably designed ... to prevent the conflict of interest from involving or resulting in a materially adverse effect on a client, customer, or counterparty." Id at 5660.

(181.) Calomiris, supra note 21, at 2. Although not precisely articulating the nature of this knowledge and skill, Professor Calomoris' point seems to be that bankers' ability to advise their clients would be diminished if they were deprived of the ability to engage in proprietary trading. Professor Calomiris argues that:

   The source of ... [clients'] confidence [in banks] is simple: the

   person sitting at the bank trading desk is not a bureaucrat, but a

   highly paid financial expert. In order for the bank to afford to

   pay such a person, he or she must be creating value for the bank

   related to his or her knowledge of the markets, and the obvious way

   such an individual creates that value for a bank is by using

   knowledge of the markets to take positions for the bank.


(182.) Id.

(183.) Traders acting on behalf of clients, rather than trading on their firm's own account, are typically remunerated on the basis of funds under management as well as trading profits. Proprietary traders, in comparison, typically earn only a percentage of their trading profits. See Peter Muller, Proprietary Trading: Truth and Fiction, 1 Quantitative Fin. 6, 6-7 (2001) (discussing the compensation incentives of various types of traders). The difference in forms of compensation is regarded as providing higher powered incentives to proprietary traders. Id. (arguing that the interest of proprietary traders are more closely aligned with the financial conglomerates on whose account they trade than are the interests of traders who have delegated fiduciary responsibility).

(184.) The point here is not that information barriers prevent skills gained from proprietary trading from spreading to client-focused traders, but that client-focused traders may develop the relevant skills by virtue of the activities they perform.

(185.) See Calomiris, supra note 21, at 3 (highlighting risks of failure in large firms).

(186.) See Robert C. Clark, The Regulation of Financial Holding Companies, 92 Harv. L. Rev. 787, 815 (1979) (discussing how the combination of banking and non-banking businesses within a financial conglomerate would affect the financial stability of the enterprise as a whole); Robert C. Clark, The Soundness of Financial Intermediaries, 86 Yale L.J. 1, 51-60 (1976) (discussing the application of portfolio theory to financial conglomerates); Howell E. Jackson, The Expanding Obligations of Financial Holding Companies, 107 Harv. L. Rev. 507, 584 n.278 (1994) (discussing how regulation banning financial conglomerates from engaging in particular activities may affect the overall financial stability of the enterprise).

(187.) The benefits of income diversification have been considered in the context of the Glass-Steagall Act--with some scholars claiming that the combination of certain non-banking activities (banned by Glass-Steagall) with banking activities would promote bank stability. See Daniel R. Fischel et al., The Regulation of Banks and Bank Holding Companies, 73 Va. L. Rev. 301, 320 (1987) (discussing whether the provision of non-banking activities increases the incidence of bank failure). However, the issue does not seem to have empirical support in the context of proprietary trading.

(188.) See Steve Rosenbush, Investment Banks Jockey for Position, BLOOMBERG BUSINESSWEEK (June 29, 2006), market-and-financial-advice (explaining that investment banks' revenue is increasingly coming from non-traditional banking activities, specifically from increased trading); Emily Thornton et al., Inside Wall Street's Culture of Risk, BLOOMBERG BUSINESSWEEK Mag. (June 11, 2006), (discussing the fact that investment banks are taking bigger risks to produce revenue). For example, in 2005 Goldman Sachs earned 15% of its total revenue from traditional investment banking activities, compared with 66% from trading and principal investments. Goldman Sachs Group, Inc., Annual Report (Form 10-K) 5 (Feb. 7, 2006); see also Goldman Sachs: Behind the Brass Plate, The Economist, Apr. 29, 2006, at 77 (discussing the importance of proprietary trading to Goldman Sachs' revenue).

(189.) Referred to as "forensic finance," this strand of literature provides evidence of unlawful or unethical practices in the financial services industry. See Jay R. Ritter, Forensic Finance, 22 J. Econ. Persp. 127, 127-28 (2008) (describing studies in economics and finance that led to regulatory enforcement actions in various contexts); see also Ivashina & Sun, supra note 16, at 286 (describing "the so-called 'forensic finance' literature" as providing "micro-evidence of unethical or illegal practices in finance"). Examples of the literature include Bodnaruk et al., supra note 16; Ivashina & Sun, supra note 16; Massa & Rehman, supra note 16.

(190.) See Ivashina & Sun, supra note 16, at 291-92 (discussing the methodology for calculating abnormal returns).

(191.) These alternative explanations are based on those discussed by Ivashina & Sun, supra note 16, at 290-91.

(192.) For instance, Ivashina and Sun calculate expected returns "using a (i) one-factor market model ... and (ii) a four-factor market model which also includes Fama and French ... factors and the Cahart ... momentum factor." See Ivashina & Sun, supra note 16, at 291-94 (citing Mark M. Carhart, On Persistence in Mutual Fund Performance, 52 J. Fin. 57 (1997); Eugene F. Fama & Kenneth R. French, Common Risk Factors in the Returns on Stocks and Bonds, 33 J. Fin. Econ. 3 (1993)).

(193.) These analytical steps mirror those typically deployed to provide evidence of discrimination in systemic disparate treatment cases. See Ramona L. Paetzold & Steven L. Willborn, The Statistics of Discrimination: Using Statistical Evidence in Discrimination Cases 28-32 (2012) (describing the analytical techniques generally used in these types of cases).

(194.) Bodnaruk and his co-authors study the M&A context, focusing on financial conglomerates advising clients bidding to acquire another company and thus having non-public information about the identity of target companies in such transactions. See Bodnaruk et al., supra note 16, at 49-90.

(195.) Ivashina and Sun study financial conglomerates that garner information about customers when negotiating amendments to loan agreements with those customers. See Ivashina & Sun, supra note 16, at 291-92.

(196.) The particular explanation ruled out is that the companies the financial conglomerate advises (about which it received non-public information) are those about which it happened to have special insights that lead to superior trading performance.

(197.) See supra notes 7-13 and accompanying text.

(198.) More specifically, for each benchmark the assumption is that the (annualized abnormal) return is not statistically different from zero, not that the return is precisely zero.

(199.) See Securities Exchange Act § 13(f), 15 U.S.C. § 78m(f) (2012); 17 C.F.R. § 240.13f-1 (2013). Under Section 13(f) of the Securities Exchange Act, financial conglomerates disclose their trading records on a quarterly basis. 17 C.F.R. § 240.13f-1(a)(1). Section 13(f) applies to any entity that invests in, buys, or sells securities for its own account or that exercises investment discretion over the accounts of clients. 15 U.S.C. § 78m(f)(1). The provision applies to institutional investment managers with investment discretion of $100 million or more. Division of Investment Management: Frequently Asked Questions About Form 13F, SEC. & Exchange Commission (Oct. 2013), [hereinafter Division of Investment Management]. Because financial conglomerates typically perform these functions, they are generally required to comply with the provision. Banks and bank holding companies that exercise investment discretion over $100 million or more in securities are required to file Form 13F, even though they are excluded from the definition of investment adviser in Section 202(a)(11) of the Investment Advisers Act. Id.

(200.) See Ivashina & Sun, supra note 16, at 285 (describing limitations with using Section 13(f) data and referring to findings using such data as representing the lower bound of any trading on non-public information).

(201.) Division of Investment Management, supra note 199 (referring specifically to answers to questions 41, 42 & 43). Ivashina and Sun, who use 13(f) data in their study, observe that non-public information "could also be used to trade in bonds or credit default swaps (CDS), but we only observe stock positions of individual investors" and that "besides information on stock holdings which are available from the CDA/Spectrum 13f database, there is limited investor-specific information for other securities." See Ivashina & Sun, supra note 16, at 288.

(202.) See Ivashina & Sun, supra note 16, at 285 ("We acknowledge that the quarterly frequency of institutional stock holdings drawn from SEC 13(f) filings imposes a limitation on the analysis."); id. at 290 (acknowledging that the authors cannot "observe full composition of the investors' portfolios (bonds or short positions)"); see also Guohua Li, Informed Institutional Trading Around Corporate Merger and Acquisition Announcements, 6 J. Trading 35, 36 (2011) (discussing the difficulty of capturing "quick entry-and-exit executions" when using 13(f) data).

(203.) Studies relying on Section 13(f) nevertheless draw valuable inferences. See, e.g., Ivashina & Sun, supra note 16, at 290, 294, 296 (claiming strong evidence of the misuse of non-public information, despite relying on quarterly 13(f) data).

(204.) In the study by Ivanisha and Sun, the relevant information event was the renegotiation by a company of a loan with its lending syndicate. See id at 286-87 (positing that lenders use such information to make trades).

(205.) The publication of quantitative metrics could thus reduce the "lemons" discount for some firms--or, fearing such a discount, firms could strengthen their information barriers. This is a reference to the "lemons problem" identified by George A. Akerlof, The Market for "Lemons": Quality Uncertainty and the Market Mechanism, 84 Q. J. Econ. 488, 489-90 (1970).

(206.) See Andrew F. Tuch, Multiple Gatekeepers, 96 Va. L. Rev. 1583, 1613-14 (2010) (discussing the limits of firm reputation in deterring undesirable conduct). Cf. Fischel et al., supra note 187, at 325 (asserting that reputation will constrain the willingness of financial conglomerates to exploit conflicts of interest).

(207.) "Watch lists" are lists that financial conglomerates and other firms maintain of companies about which the firm has, or expects to have, material, non-public information. See Theodore A. Levine et al., Multiservice Securities Firms: Coping with Conflicts in a Tender Offer Context, 23 WAKE FOREST L. Rev. 41, 65 (1988) (describing "Watch List" or "Grey List" procedures used by financial services firms as a response to the possibility of conflicts of interest); Maremont & Craig, supra note 7 ("Banks are supposed to keep a closely guarded 'watch list' or 'gray list' of potential deals on which they're advising, and to monitor trading in the affected securities.").

(208.) See Maremont & Craig, supra note 7 (explaining various instances where the investigations identified no impropriety).

(209.) See Ritter, supra note 189, at 131 ("After the late trading scandal came to light in September 2003, many of the large mutual fund families that knowingly permitted the late trading were forced to hire independent consultants, approved by the [SEC], to estimate damages and oversee reforms.").

(210.) FSOC Study, supra note 34, at 27-47 (recommending that financial conglomerates provide to regulators quantitative trading metrics designed to assist regulators in identifying impermissible proprietary trading activity); see also GAO Proprietary Trading Report, supra note 143, at 37-40 (referring to the need for firms and regulators to monitor "metrics" that could indicate whether impermissible proprietary trading occurs).

(211.) See Press Release, Sec. & Exch. Comm'n, Ten of Nation's Top Investment Firms Settle Enforcement Actions Involving Conflicts of Interest Between Research and Investment Banking (Apr. 28, 2003), available at (explaining the terms of the settlement).

(212.) As legal anthropologist Annelise Riles has observed, where regulatory reforms have succeeded, it has been because they have enrolled the targets of regulation in the purposes of regulation, encouraging the targets to take responsibility for the regulatory problem. Annelise Riles, Collateral Knowledge: Legal Reasoning in the Global Financial Markets 227 (2011). Where financial conglomerates do not "buy-in" to new regulation, they are less likely to comply with its terms. Id. at 230-39. Although industry acceptance should not determine regulatory reform, the factor deserves consideration.

(213.) See, e.g., Martin N. Bailey et al., Squam Lake Working Grp. on Fin. Regulation, A New Information Infrastructure for Financial Markets 5 (Feb. 2009) (unpublished manuscript), available at (proposing that large financial institutions release financial data, but cautioning that "it is important to protect proprietary business models and incentives to innovate" and that "[p]ublic disclosure of a firm's sensitivities also raises concerns about predatory trading by competitors"); Robert P. Bartlett, Making Banks Transparent, 65 Vand. L. Rev. 293, 298-99 (2012) (discussing concerns about revelation of a bank's proprietary investment strategies).

(214.) Bailey et al., supra note 213 (suggesting public disclosure of financial information be delayed to mitigate risks associated with dulling incentives to innovate and predatory trading by competitors).

(215.) For instance, the article by Bodnaruk et al., supra note 16, showing the systematic failure of information barriers, attracted widespread media attention. The authors declined to identify the financial conglomerates whose information barriers apparently failed--although in interviews, they revealed that eight major (unidentified) firms had been implicated. See David Weidner, There's No Force to Battle in This 'Dark Side', MarketWatch (Jan. 14, 2008), ("Unfortunately, the professors didn't name names. They did say that eight of the 10 biggest banks named in the global research settlement seemed to be trading ahead of deals."). Scholars, it seems, would prefer to focus their efforts on targeting systematic practices, rather than risk making an adversary of a named perpetrator, especially one vociferously denying the allegations.

Studies relying on Section 13(f) data may be criticized on the basis that such data presents an incomplete picture of firms' trading activities. See supra note 202 and accompanying text. However, a plausible interpretation is that these studies report only the lower bound of any trading on non-public information. See Ivashina & Sun, supra note 16, at 285 (describing limitations with using Section 13(f) data and referring to findings as representing the lower bound of any trading on non-public information). Moreover, conclusions as to the failure of information barriers in Part III.B rely on studies using other data to detect failing information barriers. See, e.g., Seyhun, supra note 16 (using data disclosed by companies under Section 16 of the Exchange Act to conclude that information barriers are "porous and ineffective"). Such conclusions also rely on anecdotal evidence.

(216.) No viable business model exists to allow credit rating agencies to profit from providing ratings, despite the value they hold for clients, without introducing an entirely new set of conflicts of interest associated with rating agencies' funding. Credit rating agencies, such as Moody's and Standard & Poor's, are paid directly by the institutions whose securities they rate. While this "issuer pays" business model impairs the independence of rating agencies' ratings, no credible alternative business model exists. See Erik Sirri, Investment Banks, Scope, and Unavoidable Conflicts of Interest, Fed. Res. Bank Atlanta Econ. Rev., Fourth Quarter 2004, at 23, 33 ("Rating agencies may prefer to receive payment from investors in lieu of, or in addition to, payment from issuers, but no credible model exists for doing so."). Moreover, rating agencies are likely to lose significant value from producing ratings because of their inability to prevent paying users from disclosing those ratings to non-paying users. John C. Coffee, Jr., Ratings Reform: The Good, The Bad, and The Ugly, 1 Harv. Bus. L. Rev. 231, 255 (2011) [hereinafter Coffee, Ratings Reform] ("Because the rating agency cannot effectively prevent the communication of its ratings to non-paying investors once it discloses it ratings to its clients, it cannot capture the full value of the financial information it creates."). Similarly, in the context under consideration, any agency that analyzed the data and provided information would be unlikely to capture the value of the information it produced because the agency could not effectively exclude free-riders from acquiring the information. However, the use of a rating agency paid by financial conglomerates themselves would introduce the same range of conflicts of interest that afflict credit rating agencies. For a discussion of the conflicts of interest afflicting credit rating agencies, see John C. Coffee Jr., Gatekeepers: The Professions and Corporate Governance 283-314 (2006). Various measures could be adopted to guard against these conflicts; one, a suggestion from Professor John Coffee, would be to permit financial conglomerates to pay for a rating, but not to select the rater. Coffee, Ratings Reform, supra. This would be tantamount to a regulator-selected analyst. In the absence of any requirement that financial conglomerates be rated, it is unlikely any financial conglomerate would pay for the service for the same reason they do not already disclose the data to clients.

(217.) See FCIC Report, supra note 41, at 118-22, 146-55 (referring to mistaken ratings of securities by credit rating agencies).

(218.) The SEC recently signaled strong interest in employing statistical analysis to detect and approve wrongdoing with the announcement of the establishment of a Center for Risk and Quantitative Analytics, a unit of the Division of Enforcement. Press Release, Sec. & Exch. Comm'n, SEC Announces Enforcement Initiatives to Combat Financial Reporting and Microcap Fraud and Enhance Risk Analysis (July 2, 2013), available at

(219.) See supra notes 130-36 and accompanying text.

(220.) Int'l Bhd. of Teamsters v. United States, 431 U.S. 324, 339-40 (1977) (referring to "employment discrimination" rather than to "wrongdoing").

(221.) As to optimal deterrence theory generally and the objective of reducing the harm associated with wrongdoing, see Gary S. Becker, Crime and Punishment: An Economic Approach, 76 J. POL. Econ. 169, 176, 209 (1968); Guido Calabresi, The Costs of Accidents: A Legal and Economic Analysis 26 (1970); Steven Shavell, Foundations of Economic Analysis of Law 1-3 (2004).

(222.) See Shavell, supra note 221, at 473-514 (discussing the deterrence of undesirable acts through the use of monetary sanctions imposed by the state).

(223.) Id. at 178 (stating in the context of unilateral accidents that "[t]he social goal here will be minimization of the sum of the costs of care and of expected accident losses" and referring to this sum as "total social costs").

(224.) Id. at 479-80 (discussing the optimal probability and magnitude of sanctions where enforcement occurs); id at 473 ("The principal problems for society that are studied are the choice of the level of enforcement effort--which determines the probability of penalizing parties--and the choice of the magnitude of sanctions, so as to maximize social welfare.").

(225.) Id. at 482-83. This condition applies for the risk-neutral wrongdoer. However, actors vary in their tolerance for risk and thus will not be equally deterred by different combinations of probability and magnitude of sanction with the same expected value. Shavell, supra note 221, at 479. The risk-averse actor, one whose experience of disutility of sanctions increases out of proportion to their size, will be more deterred by a combination with a higher sanction than one with a lower sanction and the same expected value. The reverse is true of the risk lover: holding the expected sanction constant, a risk-loving individual will be more deterred by a higher probability of sanction than a higher magnitude sanction.

(226.) Id. at 474-75, 477-78.

(227.) This approach is favored because of the risk that a private right of action would create excessive incentives for financial conglomerates to bolster their information barriers and because it is probably more feasible to permit private enforcement (in the event that incentives prove inadequate) than to eliminate such a right once it has been permitted (if incentives prove excessive).

(228.) As to the scope of the Volcker Rule, see supra Part IV.C. 1.

(229.) See id. at 70-71 (discussing limits on causal inferences that may be drawn from statistical analysis). For a recent illustration of the limits of statistical evidence to provide causal inferences, see Lawrence Summers, Lessons Can Be Learned From Reinhart-Rogoff Error, WASH. POST (May 5, 2003), (referring to "the chestnut that correlation does not establish causation").

(230.) Paetzold & Willborn, supra note 193, at 71-72.

(231.) Paetzold & Willborn, supra note 193, at 69-71 (discussing the reasons for the inability of statistical analysis to establish conclusively any observed disparity).

(232.) United States v. Veysey, 334 F.3d 600, 605-06 (7th Cir. 2003) (internal citations omitted) (internal quotation marks omitted); see also Braker v. Bridgestone/Firestone Co., 966 F. Supp. 874, 876 (W.D. Mo. 1996) ("Not all commentators have agreed on this approach.... This reluctance to recognize statistical probability as a basis for a verdict has been eroded in recent times when there are insurmountable obstacles to producing particularized proof.").

(233.) Fed. R. Evid. 401(a).

(234.) In the context of Rule 10b-5 actions, for example, Judge Posner endorsed the same analytical approach to determine whether a plaintiff stated "'with particularity facts giving rise to a strong inference that the defendant acted with the required state of mind.'" Makor Issues & Rights, Ltd. v. Tellabs Inc., 513 F.3d 702, 705, 711 (7th Cir. 2008) (quoting 15 U.S.C. § 78u-4(b)(2) (2012)) ("As more and more alternatives to a given explanation are ruled out, the probability of that explanation's being the correct one rises.").

(235.) Title VII of the Civil Rights Act of 1974 bars private employers from discriminating based on race, color, religion, sex, or national origin. 42 U.S.C. §§ 2000e-1 to 2000e-17 (2012). Courts have long recognized the importance of statistics to prove discrimination in cases under Title VII, particularly in cases of systemic disparate treatment. See, e.g., Paetzold & Willborn, supra note 193, at ix ("[Statistical analysis has become a central means of proving discrimination in a wide variety of settings."); Charles A. Sullivan & Lauren M. Walter, Employment Discrimination 175-207 (2009) (discussing the use of statistical inferences in establishing systemic disparate treatment in employment discrimination cases). See also Int'l Bhd. of Teamsters v. United States, 431 U.S. 324, 339 (1977) (quoting Mayor of Philadelphia v. Educational Equality League, 415 U.S. 605, 620 (1974)) ("'Statistical analyses have served and will continue to serve an important role' in cases in which the existence of discrimination is a disputed issue."); Alabama v. United States, 304 F.2d 583, 586 (5th Cir. 1962), aff'd per curiam, 371 U.S. 37 ("[Statistics often tell much, and Courts listen.").

(236.) The use of statistical methods to prove discrimination has increased as discriminatory intent has become more subtle and difficult to detect. See Paetzold & Willborn, supra note 193, at ix-x ("In Title VII cases, the value of statistical evidence has increased over time because, as the more blatant forms of discrimination became rarer, the power of statistical analysis to detect and quantify less obvious patterns of behavior became more important."). See also Thomas v. Eastman Kodak Co., 183 F.3d 38, 58 n.12 (1st Cir. 1999) (quoting Hodgens v. General Dynamics Corp., 144 F.3d 151, 171 n.8 (1st Cir. 1998)) (stating, in reference to the use of circumstantial evidence, "This method of proving a Title VII claim is all the more important now ... since 'smoking gun' evidence is 'rarely found in today's sophisticated employment world.'"); Reid v. Lockheed Martin Aeronautics Co., 205 F.R.D. 655, 660 (N.D. Ga. 2001) ("In contrast to the early days of Title VII, it is now more uncommon to find an employer that overtly encourages wholesale discrimination on the basis of race; race discrimination today comes in more subtle forms.").

Scholars also advocate for the use of statistical evidence. For example, Hillary Sale argues for the use of statistical evidence to allow shareholders, other than the original shareholders, to obtain remedies under Sections 11 and 12(a)(2) of the Securities Act (creating liability for disclosure errors). Sale, supra note 26. Drawing parallels with the use of statistical evidence to allow plaintiffs to prove causation in toxic tort cases, Professor Sale argues for the use of such evidence to allow shareholders to trace the source of the securities they hold and thereby obtain remedies when otherwise the purposes for which the statute was enacted would be defeated. Id. at 483-87.

(237.) Courts and regulators employ statistical inference in other areas of substantive law, including to prove--or to disprove--factual claims concerning racial polarization in voting, existence of forgeries, and efficacy of treatments in clinical trials. See David P. Leonard et al., The New Wigmore: A Treatise on Evidence § 12.1 (2013) ("Statistical assessments are prominent in many kinds of cases, ranging from antitrust to voting rights. They have been employed to prove--or disprove--factual claims such as the presence of illegal discrimination, racial polarization in voting, the identities of criminals, the existence of forgeries, the causes of trends in sales or prices, and the quantum of damages caused by illegal conduct."); see generally E9 Statistical Principles for Clinical Trials, 63 Fed. Reg. 49,583 (Sep. 16, 1998) (acknowledging the importance of statistics in clinical trials and setting forth guidelines for such testing).

(238.) For example, an employer's decision to, say, employ a white worker over a black worker may appear suspicious. Repeated employment decisions would heighten suspicions but still may be explained as coincidence or on the basis of other benign rationales. Statistical inference helps to credibly rule out competing inferences by considering, for example, precisely how likely it is that all 100 of the firm's recently hired workers are white in an employment pool that is 50% black.

(239.) See Stephen J. Choi et al., Scandal Enforcement at the SEC: The Arc of the Option Backdating Investigations 6-8 (N.Y. Univ. Law & Econ., Working Paper No. 11-20, 2013), available at (discussing the history of the option backdating scandal). See also Ritter, supra note 189, at 131-37 (discussing the role of so-called forensic finance in uncovering options backdating). See Maremont & Craig, supra note 7 ("Academic studies played a major role in uncovering the recent stock-option backdating scandal.").

(240.) See Exchange Act § 16(b), 15 U.S.C. § 78p(b) (2012). The provision requires directors, executive officers, and the holders of ten percent of a class of equity securities of public companies to disclose all personal trading in the issuer's stock within two business days of the transaction. Exchange Act § 16(a), 15 U.S.C. § 78p(a). Prior to changes instituted by the Sarbanes-Oxley Act, Section 16(a) required disclosure of changes in beneficial ownership by insiders within ten days after the close of the month in which the change occurred. Using the disclosed data, interested observers generate quantitative metrics from which they draw inferences about whether directors and other insiders trade using non-public information. A strand of empirical literature studies insider trading using data disclosed under Section 16(a). See, e.g., Josef Lakonishok & Inmoo Lee, Are Insider Trades Informative?, 14 Rev. Fin. Stud. 79, 83-84 (2001) (discussing the use of disclosures made by insiders pursuant to Section 16(a)); Ji-Chai Lin & John S. Howe, Insider Trading in the OTCMarket, 45 J. Fin. 1273, 1274-76 (1990) (using data obtained from disclosures demanded of investors by Section 16(a)); see also James D. Cox et al., Securities Regulation: Cases and Materials 944 (7th ed. 2013) (noting that disclosures required by Section 16(a) are intended in part "to discourage high-level insider trading by assuring its publicity"). Section 16(a) of the Exchange Act has also served as a model for a proposed regulatory strategy designed to hold financial firms accountable for trading to benefit from the short-term market effect of analyst reports. See Jill E. Fisch, The Analyst as Fiduciary: A Misguided Quest for Analyst Independence? 48-49 (Sept. 2006) (unpublished manuscript), available at

(241.) Ritter, supra note 189 (providing examples of so-called forensic finance).

(242.) See Sullivan & Walter, supra note 235, at 171-72 ("[Statistical evidence is subject to its own constraints, including the use of appropriate comparison groups and statistical significance....").

(243.) In a recent report, the staff of the SEC reviewed features of information barriers the highlighted practiced it believed to be effective; none suggest the use of strategies relying on statistical analysis. See Sec. & Exch. Comm'n, supra note 2, at 7, 24-42.

(244.) See, e.g., Ben Protess, Dodd-Frank Inches Along, N.Y. Times Dealbook (Sept. 6, 2011, 1:29 PM), (describing the rule-making burden on regulators arising from the Dodd-Frank Act and describing regulators as "badly behind schedule"); Bailey et al., supra note 213, at 5 (arguing in favor of public dissemination of financial data because "[regulatory capacity is limited: despite the talented individuals with sound intentions throughout regulatory agencies, the inherent complexity of financial markets means potential problems can be difficult to recognize and respond to").

(245.) Cox et al., supra note 240, at 14 (referring to the "dominance of lawyers in policymaking roles at the SEC").

(246.) The market check would also combat the possibility of the SEC lacking prosecutorial zeal, which claim has been leveled at the SEC recently. See, e.g., Matt Taibbi, Is the SEC Covering Up Wall Street Crimes?, Rolling Stone (Aug. 17, 2011), (referring to "a recent series of damaging critiques of the SEC made by reporters, watchdog groups and members of Congress, all of which seem to indicate that top federal regulators spend more time lunching, schmoozing and job-interviewing with Wall Street crooks than they do catching them"); Stavros Gadinis, The SEC and the Financial Industry: Evidence from Enforcement Against Broker-Dealers, 67 BUS. Law. 679, 725-26 (2012) (providing evidence of the so-called revolving-door phenomenon, involving SEC staff joining the firms they previously regulated).

(247.) See supra Part III.B-C (discussing the actual effect and assessment of information barriers).

(248.) In the case of a merger, for instance, a financial conglomerate could be presumed to hold non-public information at the time it was engaged to advise.

(249.) Introduced by Section 922 of the Dodd-Frank Act, the whistleblower bounty scheme may increase incentives for employees to "whistle blow" on gaming by financial conglomerates. See generally Sec. & Exch. Comm'n, Annual Report on the Dodd-Frank Whistleblower Program, Fiscal Year 2012 (2012), (summarizing whistleblower activities in 2012); Amanda M. Rose, Better Bounty Hunting: How the SEC's New Whistleblower Program Changes the Securities Fraud Class Action Debate, Nw U. L. Rev. (forthcoming 2014), available at (outlining the whistleblower regime).

(250.) Caridad v. Metro-North Commuter R.R., 191 F.3d 283, 292 (2d Cir. 1999).

(251.) See supra note 237 and accompanying text (describing the various contexts in which courts and regulators use statistical analysis to provide proof).

(252.) The use of statistical analyses in discrimination cases also seems to have experienced problems initially as parties became increasingly familiar with their limits. See, e.g., Thomas J. Campbell, Regression Analysis in Title VII Cases: Minimum Standards, Comparable Worth, and Other Issues Where Law and Statistics Meet, 36 Stan. L. Rev. 1299 passim (1984) (discussing two misuses of regression analysis in early discrimination cases); William T. Bielby & Pamela Coukos, "Statistical Dueling" with Unconventional Weapons: What Courts Should Know about Experts in Employment Discrimination Class Actions, 56 Emory L. J. 1563, 1574-76 (2007) (criticizing certain statistical approaches used at the class certification phase of Title VII cases).

Andrew F. Tuch, Associate Professor, Washington University School of Law. For helpful comments and suggestions, I thank Bill Alford, Adam Badawi, Lucian Bebchuk, Andriy Bodnaruk, Lisa Bernstein, Nuncio D'Angelo, Deborah DeMott, Stavros Gadinis, Scott Hirst, Victoria Ivashina, Howell Jackson, Louis Kaplow, Pauline Kim, Reinier Kraakman, Josh Lerner, Geoffrey Miller, Noam Noked, Russell Osgood, Larry Ribstein, Laura Rosenbury, Adam Rosenzweig, Hillary Sale, Roy Shapira, Steve Shavell, Holger Spamann, and Peter Wiedenbeck, as well as participants at various seminars and workshops. For assistance in understanding market and regulatory practices, I thank Philip Augar, Carole Schiffman, Erik Sirri, and Barry Volpert. For research assistance, I thank Amanda Berk, Michael Geiselhart, and Anna Tkacheva. Financial support was generously provided by Washington University School of Law and the John M. Olin Center for Law, Economics, and Business at Harvard Law School. All errors are entirely my own.

Table 1. Summary of the explanation--other than the use of non-

public information ("NPI")--that is ruled out when a given

hypothesis is true.

Hypothesis                            Alternative Explanation Ruled


1. Firm' s return from trading        Superior intellect or trading

stock about which it possesses        skill by the firm

NPI > Firm's return for the rest

of its portfolio

2. Firm's return from trading         Peculiarities of the companies

stock about which it possesses        about which the firm possessed

NPI > Return by comparable firms      NPI

trading in stock about which the

firm possesses NPI

3. Firm's return from trading         Coincidence (or endogeneity)--

stock about which it possesses        i.e., the firm advises

NPI > Firm' s return from trading     companies (from which it

stock about which it possessed        garners non-public

NPI--but at times when it lacked      information) about which it

NPI                                   happens to have special


Figure 1: Range of activities in which financial conglomerates

typically engage.


(Commercial) Banking

* Accepting deposits

* Making loans

* Offering other

products and services,

including credit cards


Investment Banking

* M&A

* Securities


Sales and Trading

* Brokering

* Dealing

* Proprietary

trading and investing

* Research analysis

Asset Management (37)

Asset Management

* Hedge funds

* Private equity funds

* Mutual funds

* Venture capital funds

* Real estate

* Fixed income

* Infrastructure

COPYRIGHT 2014 University of Iowa Journal of Corporation Law
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2021 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Tuch, Andrew F.
Publication:The Journal of Corporation Law
Date:Mar 22, 2014
Previous Article:Killing conscience: the unintended behavioral consequences of "pay for performance."
Next Article:I'll have Mayo with my Myriad: the effect of Mayo Collaborative Services v. Prometheus Laboratories, Inc. and Association for Molecular Pathology v....

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |