Complexity, Innovation and the Regulation of Modern Financial Markets

Electronic copy available at: Electronic copy available at: copy available at:

Complexity, Innovation and the Regulation of Modern Financial Markets

Dan Awrey∗

∗ University Lecturer in Law and Finance and Fellow, Linacre College, Oxford University. The author would like to thank John Armour, Blanaid Clarke, Merritt Fox, Anna Gelpern, Lawrence Glosten, Jeff Golden, Sean Griffith, Christian Johnson, Donald Langevoort, Katharina Pistor, Morgan Ricks, Colin Scott, Arthur Wilmarth and Kristin van Zwieten for their extremely helpful comments and to the organizers and participants of workshops hosted by Harvard University, Oxford University, Fordham University and University College Dublin for the opportunity to present previous drafts of this paper. The author would also like to acknowledge the generous support of both the Institute for New Economic Thinking and the Columbia University Global Finance and Law Initiative.

Electronic copy available at: Electronic copy available at: copy available at:



The intellectual origins of the global financial crisis (GFC) can be traced back to blind spots emanating from within conventional financial theory. These blind spots are distorted reflections of the perfect market assumptions underpinning the canonical theories of financial economics: modern portfolio theory; the Modigliani and Miller capital structure irrelevancy principle; the capital asset pricing model and, perhaps most importantly, the efficient market hypothesis. In the decades leading up to the GFC, these assumptions were transformed from empirically (con)testable propositions into the central articles of faith of the ideology of modern finance: the foundations of a widely held belief in the self-correcting nature of markets and their consequent optimality as mechanisms for the allocation of society’s resources. This ideology, in turn, exerted a profound influence on how we regulate financial markets and institutions.

The GFC has exposed the folly of this market fundamentalism as a driver of

public policy. It has also exposed conventional financial theory as fundamentally incomplete. Perhaps most glaringly, conventional financial theory failed to adequately account for the complexity of modern financial markets and the nature and pace of financial innovation. Utilizing three case studies drawn from the world of over-the-counter (OTC) derivatives – securitization, synthetic exchange-traded funds and collateral swaps – the objective of this paper is thus to start us down the path toward a more robust understanding of complexity, financial innovation and the regulatory challenges flowing from the interaction of these powerful market dynamics. This paper argues that while the embryonic post-crisis regulatory regimes governing OTC derivatives markets in the U.S. and Europe go some distance toward addressing the regulatory challenges stemming from complexity, they effectively disregard those generated by financial innovation.

Key words: complexity; financial innovation; financial regulation; shadow banking system; OTC derivatives; securitization; collateral swaps; synthetic exchange-traded funds; ETFs; Dodd-Frank Act; European Market Infrastructure Regulation.

Electronic copy available at: Electronic copy available at: copy available at:


TABLE OF CONTENTS I. Introduction 1 II. Toward A More Robust Theory of Complexity and its Drivers 9 III. Toward A Supply-Side Theory of Financial Innovation 31 IV. The Relationship Between Complexity and Financial Innovation: 43 Three Case Studies V. Complexity and Financial Innovation: The Regulatory Challenges 54 VI. OTC Derivatives Regulation in the Wake of the GFC: A Brave New

World 57

VII. Conclusion 78


I. Introduction

The intellectual origins of the ongoing global financial crisis (GFC) can be

traced back to shortcomings – blind spots – emanating from within conventional

financial theory. These blind spots are distorted reflections of the perfect market

assumptions underpinning the canonical theories of financial economics: modern

portfolio theory (MPT); the Modigliani and Miller (M&M) capital structure

irrelevancy principle; the capital asset pricing model (CAPM) and, perhaps most

importantly, the efficient market hypothesis (EMH).1 These theories share a common

and highly stylized view of financial markets, one characterized by, inter alia, perfect

information, the absence of transaction costs and rational market participants. Yet in

reality financial markets – and market participants – rarely (if ever) strictly conform

to these assumptions.2, 3 Information is costly and unevenly distributed; transaction

costs are pervasive and often determinative, and market participants frequently exhibit

cognitive biases and bounded rationality.4 Despite these seemingly uncontroversial

1 These theories, their centrality to the field of financial economics, and their underlying assumptions are each discussed in greater detail in Parts II and III. 2 The most notable exception arguably being public secondary markets for equity securities, where a significant body of empirical research exists to support the view that these markets generally conform to the assumptions of semi-strong form EMH. For a survey of this empirical work, see Burton Malkiel, “The Efficient Market Hypothesis and Its Critics” (2003), Centre for Economic Policy Studies Working Paper No. 91, available at and Eugene Fama, “Market Efficiency, Long-Term Returns and Behavioral Finance” (1998), 49 J. Fin. Econ. 283. Even in this context, however, it is still unrealistic – and, indeed, actually inconsistent with the operation of the arbitrage mechanism at the heart of conventional financial theory – to expect that markets will always be in equilibrium; see Sanford Grossman and Joseph Stiglitz, “On the Impossibility of Informationally Efficient Markets” (1980), 70:3 Am. Econ. Rev. 393. 3 As Ron Gilson has observed, it is not altogether clear whether the authors of these theories were initially attempting to describe real world financial markets or, alternatively, provide the basis for a research agenda which – by relaxing the perfect market assumptions – could enhance our understanding of how these markets work in practice; Ron Gilson, “Market Efficiency after the Financial Crisis: It’s Still a Matter of Information Costs” (May 2011) at 17 [working paper on file with author]. Ultimately, at least one of these authors did explicitly adopt the latter view; see Merton Miller, “The Modigliani-Miller Propositions After Thirty Years” (1988), 2 J. Econ. Perspectives 99 at 100. 4 Observing this divergence between theory and reality, Fischer Black, the former M.I.T. finance professor, Goldman Sachs executive and co-author of the Black-Scholes option pricing formula, once quipped: “Markets look a lot less efficient from the banks of the Hudson than from the banks of the Charles”; personal conversation quoted in Peter Bernstein, Against The Gods: The Remarkable Story of Risk (John Wiley & Sons, New York, 1996) at 7.


observations, however, the empirically (con)testable assumptions of conventional

financial theory have been transformed into the central articles of faith of the ideology

of modern finance: the foundations of a widely held belief in the self-correcting

nature of markets and their consequent optimality as mechanisms for the allocation of

society’s resources.5

The ideology of modern finance has exerted a profound influence on how we

regulate financial markets and institutions. Perhaps most significantly, the pervasive

belief in the social desirability of unfettered markets represented the driving force

behind the sweeping agenda of financial deregulation witnessed in many jurisdictions

in the decades leading up to the GFC.6 This market fundamentalism was grounded in

the conviction that rational and fully informed market participants – utilizing

sophisticated quantitative methods and the innovative financial instruments these

methods made possible – had effectively mastered risk. Public regulation, by

implication, was largely relegated to a supporting role: namely, the provision of

private property rights and efficient contract enforcement necessary to support private

risk-taking. Ultimately, it was this market fundamentalism which justified turning a

blind eye to the potential adverse effects of vast global current account imbalances7;

5 Simon Johnson and James Kwak, 13 Bankers: The Wall Street Takeover and the Next Financial Meltdown (Pantheon Books, New York, 2010) at 5 and 104-118. 6 See Financial Crisis Inquiry Commission (FCIC), Final Report of the National Commission on the Causes of the Financial Crisis in the United States (Public Affairs, New York, 2011) at xviii; Richard Posner, A Failure of Capitalism: The Crisis of ’08 and the Descent into Depression (Harvard University Press, Cambridge, 2009); George Cooper, The Origins of the Financial Crisis: Central Banks, Credit Bubbles and the Efficient Market Fallacy (Random House, New York, 2008); Gilson (n 3) at 2-3 and Johnson and Kwak (n 5) at 69. The term ‘deregulation’ does not entirely capture the breadth or fundamental character of this trend. Indeed, it is perhaps more accurate to say that deregulation during this period was characterized by (1) significant devolution of regulation from public to private actors, and (2) a non-interventionist stance toward the regulation of many financial markets and institutions which emerged, developed and matured during this period. 7 The influence of market fundamentalist thinking on the established wisdom underpinning the post- war push to liberalize international trade and capital flows is reflected in the comments of Stanley Fischer, former First Deputy Managing Director of the International Monetary Fund (IMF): “free capital movements facilitate a more efficient allocation of global savings, and help channel resources


acquiescing to the build-up of huge amounts of risk within the so-called ‘shadow

banking’ system8, and devolving significant responsibility for the design and

implementation of capital adequacy standards to the very financial institutions which

were ultimately subject to this micro-prudential regulation.9 At times, it appeared as

if the only question to which ‘more markets’ was not the consensus answer was:

where do we turn when markets fail?

The GFC has revealed the folly of market fundamentalism as a driver of

public policy. It has also exposed conventional financial theory as fundamentally

incomplete. Perhaps most glaringly, conventional financial theory failed to

adequately account for both the complexity of modern financial markets and the

nature and pace of financial innovation. From sub-prime mortgages, securitization

and credit default swaps (CDS) to sophisticated quantitative models for measuring

and managing risk, the footprints of complexity and innovation can be observed

throughout modern financial markets – and, importantly, at almost every significant

step along the road to the GFC.10 Complexity and innovation have combined to

into their most productive uses, thus increasing economic growth and welfare”; Stanley Fischer, “Capital Account Liberalization and the Role of the IMF”, lecture given at the IMF Annual Meetings (September 19, 1997), available at 8 The shadow banking system includes (1) non-bank financial institutions such as finance companies, structured investment vehicles, securities lenders, money market mutual funds, hedge funds and U.S. government sponsored entities, and (2) financial instruments such as repurchase agreements, asset- backed securities, collateralized debt obligations and other derivatives, insofar as these institutions and instruments perform economic functions (i.e. maturity, credit and liquidity transformation) typically associated with more ‘traditional’ banks; see Gary Gorton and Andrew Metrick, “Regulating the Shadow Banking System” (Fall 2010), Brookings Papers on Economic Activity 261 and Zoltan Pozsar, Tobias Adrian, Adam Ashcraft and Halley Boesky, “Shadow Banking”, Federal Reserve Bank of New York Staff Reports No. 458 (July 2010). 9 As most infamously epitomized by the ill-fated Consolidated Supervised Entity (CSE) Program administered by the U.S. Securities and Exchange Commission; see SEC’s Oversight of Bear Stearns and Related Entities: The Consolidated Supervised Entity Program, Report of the Securities and Exchange Commission Office of the Inspector General, Report No. 446-A (September 25, 2008). 10 And, indeed, the road to many previous financial crises. See for example, John Kenneth Galbraith, The Great Crash of 1929 (Houghton Mifflin Company, New York, 1954) at 46-50, 72-76, 80-86 and 89, describing the role of financial innovations such as margin trading and so-called ‘investment trusts’ in helping to fuel the speculative bubble which ultimately precipitated the 1929 U.S. stock market


generate significant asymmetries of information and expertise within financial

markets, thereby opening the door to suboptimal contracting and exacerbating already

pervasive agency cost problems.11 At the same time, the pace of innovation has left

financial regulators and regulation chronically behind the curve. Together,

complexity and innovation thus give rise to a host of regulatory challenges, the full

implications of which we are only just now beginning to understand.

Perhaps nowhere is the myopia of market fundamentalism more evident than

in connection with the pre-crisis regulation of over-the-counter (OTC) derivatives

markets. Over the course of the past three decades, these markets have grown from

an obscure financial backwater into a global behemoth – the $USD700 trillion gorilla

of modern financial markets. Prevailing dogma prior to the GFC viewed the

seemingly insatiable demand for many species of OTC derivatives as a rational

response to market imperfections. Supply, in turn, was a rational response to this

demand. That supply met demand within the marketplace was then generally

interpreted as being dispositive of these instruments’ private and social utility. This

viewpoint was firmly rooted in the autonomous rational actor framework

underpinning MPT, the M&M capital structure irrelevancy principle, CAPM and the

EMH. Not coincidentally, conventional financial theory also provided the rationale –

crash. More recent examples include both (1) the role of portfolio insurance in the 1987 stock market crash, and (2) the role of high frequency traders, automated execution algorithms and exchange traded funds in the so-called ‘flash crash’ of May 6, 2010; see Report of the Presidential Task Force on Market Mechanisms, submitted to the President of the United States, the Secretary of the Treasury and the Chairman of the Federal Reserve Board (January 1988) at v and Findings Regarding the Market Events of May 6, 2010, Report of the Staffs of the Commodity Futures Trading Commission and Securities and Exchange Commission to the Joint Advisory Committee on Emerging Regulatory Issues (September 30, 2010). 11 In the context of a principal-agent (or other co-operative) relationship between two or more parties, the term ‘agency costs’ refers to costs incurred by the parties in connection with the monitoring and bonding of the other parties, along with any residual (hidden) losses stemming from the misalignment of incentives as between the parties; see Michael Jensen and William Meckling, “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure” (1976), 3:4 J. of Fin. Econ. 305.


forcefully articulated by, amongst many others, U.S. Federal Reserve Board

Chairman Alan Greenspan12 – for why public regulatory intervention was not

necessary to ensure the safe and efficient operation of OTC derivatives markets. This

stance was ostensibly bolstered by the emergence of private actors such as the

International Swaps and Derivatives Association (ISDA), along with various trade

execution, clearing and settlement platforms, to provide the legal and operational

infrastructure necessary to support the development and growth of these new


OTC derivatives markets epitomize both the complexity of modern financial

markets and the nature and pace of innovation within them. For this reason, they

offer us an illuminating window into the regulatory challenges generated by the

interaction of these powerful (and yet poorly understood) market dynamics. Perhaps

not surprisingly, these challenges ultimately stem from the availability and allocation

of a single and immensely precious commodity: information. How costly is it to

acquire? Who has it? And, importantly, who doesn’t?14 As we shall see, the answers

to these and other related questions are highly instructive in terms of how we should

12 See for example, Alan Greenspan, “The Regulation of OTC Derivatives”, testimony before the House Committee on Banking and Financial Services, 105th Congress, 2nd session (July 24, 1998), stating: “professional counterparties to privately negotiated contracts also have demonstrated their ability to protect themselves from losses, from fraud, and counterparty insolvencies… Aside from the safety and soundness regulation of derivatives dealers under the banking and securities laws, regulation of derivatives transactions that are privately negotiated by professionals is unnecessary. Regulation that serves no useful purpose hinders the efficiency of markets to enlarge standards of living.” See also Alan Greenspan, “Technological Change and the Design of Bank Supervisory Policies”, remarks at the Conference on Bank Structure and Competition of the Federal Reserve Bank of Chicago (May 1, 1997); Alan Greenspan, “Government Regulation and Derivatives Contracts”, remarks to the Financial Markets Conference of the Federal Reserve Bank of Atlanta (February 21, 1997); U.S. Treasury Department Press Release, “Joint Statement by Treasury Secretary Robert E. Rubin, Federal Reserve Board Chairman Alan Greenspan and Securities and Exchange Commissioner Arthur Levitt” (May 7, 1998), and Lawrence Summers, testimony before the Senate Banking Committee (July 31, 1998). 13 See Dan Awrey “The Dynamics of OTC Derivatives Regulation: Bridging the Public-Private Divide” (2010), 11:2 Eur. Bus. Org. L. Rev. 155. 14 And, indeed, if it can be acquired, manipulated, filtered and/or analyzed within applicable temporal, cognitive, resource and/or technological constraints.


approach the regulation of OTC derivatives markets – and the broader financial

system – going forward.

The objective of this paper is to start us down the path toward a more robust

understanding of the regulatory challenges which flow from complexity and

innovation within modern financial markets. It does not, however, seek to ‘correct’

the blind spots of conventional financial theory. This is an important point. What

follows is not an indictment of the methodologies of positive economics from which

the insights of conventional financial theory have largely derived.15 Indeed, the

rigorous logic and hypothesis testing at the core of this discipline have contributed

greatly to our understanding of the economic world. At the same time, however, it

must be acknowledged that the intellectual tools of this discipline – and the

assumptions upon which they are founded – have been (at best) misconstrued and (at

worst) hijacked by those seeking to advance the cause of market fundamentalism.16 It

is in response to this pyrrhic victory of rhetoric over reality that this paper seeks to

establish a more stable and constructive equilibrium between financial theory and

how we approach financial regulation.17 Just as market fundamentalism has been

found wanting in the wake of the GFC, so too will any approach to regulation which

favors ideological purity over the rigorous and ongoing evaluation of the market

frictions and market failures which attract regulatory scrutiny and the anticipated

15 For a robust description (and defense) of these methodologies, see Milton Friedman, “The Methodology of Positive Economics” in Milton Friedman (ed.), Essays in Positive Economics (Chicago University Press, Chicago, 1966). 16 In this respect, it is irrelevant for the purposes of this paper whether policymakers were ‘true believers’ in market fundamentalism or simply utilizing it for their own ends. What is important, rather, is that this ideology influenced (either directly or indirectly) how these policymakers approached the regulation of financial markets and institutions. 17 Although certainly not a more static one.


costs and benefits of various forms of regulatory intervention.18 Put somewhat

differently, the only antidote to ideological fervor is the systematic study of how

markets – and regulation – work in practice.19

One further point of clarification is perhaps in order. This paper is not an

attempt to dissect the proximate or root causes of the GFC. Considerable scholarly

ink has already been spilled on this subject and, even then, the debate over precisely

what happened and why seems poised to rage on well into the new millennium.20

More importantly for the present purposes, however, while the crisis has undoubtedly

served to bring these issues into sharper focus, the regulatory challenges generated by

complexity and financial innovation existed prior to, and independently of, the events

and circumstances which culminated in the GFC.

The remainder of this paper proceeds as follows. Part II begins by articulating

a theoretical framework for understanding complexity which conceptualizes it as a

function of two variables: information costs and bounded rationality. It then

examines six key drivers of high information costs (and information failure) within

modern financial markets and their points of intersection with the cognitive and

18 This paper thus adopts as its normative touchstone the evaluative framework provided by welfare economics, pursuant to which ‘optimal’ or ‘efficient’ markets or regulation are understood to be those which maximize net social welfare. Reflective of the real-world limitations facing policymakers, optimal or efficient regulation will be further understood to refer to that which maximizes net social welfare within resource and technological constraints – or, cloaked in the jargon of welfare economics, the tangency between the utility possibilities frontier and the highest attainable social welfare indifference curve (i.e. the ‘constrained bliss-point’); see Per-Olov Johansson, An Introduction to Modern Welfare Economics (Cambridge University Press, Cambridge, 1991) at 28-29 and John Eatwell, Murray Milgate and Peter Newman (eds.), The New Palgrave: Allocation, Information and Markets (Macmillan, New York, 1989) at 1. For a more fulsome discussion of welfare economics and its utility (and limitations) in the domain of financial regulation, see Awrey (n 13) at 165-167. 19 This approach is reflected in Ronald Coase’s statement that “satisfactory views on policy can only come from a patient study of how, in practice, the market, firms, and government handle the problem of harmful effects.”; Ronald Coase, “The Problem of Social Cost” (1960), 3 J. of Law & Econ. 1 at 10. 20 For a very useful synopsis of this literature, see Andrew Lo, “Reading About the Financial Crisis: A 21-Book Review” (October 26, 2011), available at


temporal constraints on our ability to process information.21 Part III shifts the focus

to financial innovation and advances a theory which re-conceptualizes it as a process

of change – but not necessarily one of improvement – influenced by, inter alia, the

supply-side incentives of the principal innovators: financial intermediaries. Part IV

then examines the multifaceted and mutually reinforcing relationship between

complexity and financial innovation through the lens of three case studies drawn from

the world of OTC derivatives: securitization, synthetic exchange-traded funds (ETFs)

and collateral swaps. Leveraging these case studies, Part V seeks to identify the

regulatory challenges generated by the interaction of these powerful market dynamics.

Part VI then examines whether and to what extent the embryonic post-crisis

regulatory regimes governing OTC derivatives markets in the U.S. and Europe

effectively respond to these challenges and canvasses potential options for further

reform. Part VII concludes.

As American essayist H.L. Mencken once observed: “for every complex

problem there is an answer which is clear, simple and wrong”.22 Consistent with this

axiom, this examination fails to generate an obvious or straightforward set of

prescriptions. As in virtually all areas of public policy, tradeoffs abound. This paper

concludes, therefore, by extracting and synthesizing the common themes flowing

from this exploration of complexity and financial innovation. These themes

underscore the importance and pervasiveness of information costs, asymmetries of

information and agency cost problems within modern financial markets and, thus, the

manifest need for mechanisms which (1) subsidize the production and dissemination

of information and (2) align the incentives of both public and private actors with 21 As explored in greater detail in Part II, these drivers include technology, opacity, interconnectedness, fragmentation, regulation and reflexivity. 22 Regrettably, the author was unable to unearth the original source for this oft-cited quotation.


broader social welfare. They also highlight the nature and pace of change within

modern financial markets and the resulting desirability of regulation designed and

built with the objective of ensuring sufficient flexibility, responsiveness and

durability. Viewed in this light, while this paper does not have in mind a specific

destination, it can be understood as strongly advocating certain modes – and a general

direction – of travel.

II. Toward A More Robust Theory of Complexity and its Drivers

Modern financial markets are very, very complex. This complexity is

compounded by the nature and pace of financial innovation. But what do we mean

when we say that financial markets are ‘complex’ and ‘innovative’? And what are

the key drivers of complexity and innovation within modern financial markets? This

section (and the next) sketch out preliminary – and at this stage largely theoretical –

answers to these all-important questions.

A. An Economic Framework for Understanding Complexity

It is almost trite to observe that modern financial markets are ‘complex’.23

Curiously, however, scholars in the fields of both law and finance have expended

relatively little time or effort attempting to understand this complexity or

systematically identify its potential sources.24, 25 So what makes modern financial

23 For a small sampling of the legal academic work acknowledging the complexity of financial markets, see Steven Schwarcz, “Regulating Complexity in Financial Markets” (2009), 87 Wash. L. Rev. 211; Emilios Avgouleas, “What Future for Disclosure as a Regulatory Technique? Lessons from the Global Financial Crisis and Beyond” (2009), available at; Gregory Krohn and William Gruver, “The Complexities of the Financial Turmoil of 2007 and 2008” (2008), available at, and Steven Schwarcz, “Rethinking the Disclosure Paradigm in a World of Complexity”, (2004), U. Ill. L. Rev. 1. 24 At least part of the explanation for this lack of attention likely stems from the fact that the theoretical and empirical literature examining MPT, the M&M capital structure irrelevancy principle, CAPM and the EHM has historically focused on the public markets for equity and, to a lesser extent, debt securities. In a recent review of the literature examining the EMH, for example, 53 of the 54 cited works were primarily or exclusively concerned with its application within the context of public equity markets; see Malkiel (n 2). This of course makes perfect sense: these theories implicitly rely on the


markets complex? We can take our first tentative steps toward answering this

question by constructing a simple (and hopefully intuitive) framework which

conceptualizes complexity as a function of two variables. The first variable

encompasses the costs incurred by actors in connection with searching for, acquiring,

filtering, manipulating and analyzing information (i.e. information costs). The second

variable, then, consists of cognitive and temporal constraints on an actor’s ability to

process this information (i.e. bounded rationality).26 In many ways, this framework

brings together, renders explicit, elaborates on and formalizes intuitions previously

existence of the secondary market liquidity typically associated with public capital markets (in effect, to ensure the efficient operation of the arbitrage mechanism which moves markets toward equilibrium). What is more, it is the public nature of these markets which afford scholars access to the information necessary to measure how rapidly new information is impacted into security prices. Simultaneously, however, it must be acknowledged that this research strategy generates an inherently biased (and increasingly myopic) sample if one’s ultimate objective is to measure the informational efficiency of modern financial markets. As we shall see, the vast majority of the complexity – and thus the information costs and bounded rationality – within modern financial markets does not emanate from within the relatively transparent (and static) public markets for capital. 25 This is not to say, however, that scholars have not attempted to construct models designed to reflect the complex dynamics of modern financial markets; see for example, Robert May, Simon Levin and George Sugihara, “Ecology for Bankers” (2008), 451 Nature 893; Robert May and Nimalan Arinaminpathy, “Systemic Risk: The Dynamics of Model Banking Systems” (2010), 46 J. of Royal Society Interface 823, and Prasanna Gai, Andrew Haldane and Sujit Kapadia, “Complexity, Concentration and Contagion” (2011), 58:5 J. of Monetary Econ. [forthcoming]. Many of these models share a common methodology – first employed by Herbert Simon – which is, in effect, based on identifying similarities between financial systems, on the one hand, and physical, biological or other social systems, on the other; see Herbert Simon, “The Architecture of Complexity” (1962), 106:6 Proceedings of the American Philosophical Society 467. The obvious shortcoming of this methodology, however, is that while models drawn from other disciplines (and developed to analysis other subject matter) might mimic the complexity of financial markets (at a given moment of time), they fail to explain why financial markets are complex. This is the question at the heart of the present inquiry. 26 Bounded rationality is a semi-strong form of rationality pursuant to which economic actors are assumed to be ‘intendedly rational, but only limitedly so’; Oliver Williamson, The Economic Institutions of Capitalism (The Free Press, New York, 1985) at 45, citing Herbert Simon, Administrative Behavior, 2ed. (Macmillan, New York, 1961) at xxiv. The concept of bounded rationality is grounded in the notion that, if the mind is a scarce resource, there will exist cognitive and temporal constraints on our ability to process information. The sources and species of bounded rationality and related cognitive biases are themselves already the subject of a rich theoretical and experimental literature upon which the present inquiry does not attempt to build. For a survey of this literature, see Nicholas Barberis and Richard Thaler, “A Survey of Behavioral Finance” in George Constantinides, Milton Harris and René Stulz, (eds.), Handbook of the Economics of Finance (Elsevier, Amsterdam, 2003). See also, Daniel Kahneman, Thinking, Fast and Slow (Penguin, London, 2011).


articulated by scholars such as Ron Gilson and Reinier Kraakman27; Steven

Schwarcz28; Henry Hu29; Gary Gorton30 and Robert Bartlett.31

As a starting point, we can envision a perfectly rational and fully informed

actor. This actor incurs no information costs and processes information completely

free from the distortions of bounded rationality. In effect, the attributes of this

hypothetical actor reflect the central assumptions of conventional financial theory.

Simultaneously, we can envision a real world actor – be it a single individual or a

group of individuals working together in a firm or other organization – attempting to

understand a particular constellation of facts or state of the world: a ‘snowball’

interest rate swap; the balance sheet of a large, complex financial institution (LCFI),

or the myriad of systemic interconnections between financial markets and institutions,

for example. To fully understand this constellation of facts or state of the world, this

real world actor must invest in the acquisition, filtering, manipulation and analysis of

information.32 It may also exhibit some form and measure of bounded rationality.

The difference between our hypothetical and real world actors can be understood in

terms of their respective tolerances for complexity.

27 Ron Gilson and Reinier Kraakman, “The Mechanisms of Market Efficiency” (1984), 70:4 Virginia L. Rev. 549 and Ron Gilson and Reinier Kraakman, “The Mechanisms of Market Efficiency Twenty Years Later: The Hindsight Bias” (2004), 46 Corp. L. Commentator 173. 28 Schwarcz (n 23). 29 Henry Hu, “Misunderstood Derivatives: The Causes of Information Failure and the Promise of Regulatory Incrementalism” (1992-1993), 102 Yale L.J. 1457. 30 Gary Gorton, “The Panic of 2007”, prepared for the Federal Reserve Bank of Kansas City, Jackson Hole Conference (August 2008) at 20-34, available at 31 Robert Bartlett III, “Inefficiencies in the Information Thicket: A Case Study of Derivatives Disclosures During the Financial Crisis” (2010), 36:1 J. Corp. Law 1. This paper is also available at Subsequent pinpoint citations refer to the ssrn version of this article. 32 And, where our actor is an organization, coordination costs.


The first important insight we can draw from this framework is that an actor’s

tolerance for complexity is inherently relative.33 What one actor views as

immediately comprehensible, another may view as too complex to understand. Thus,

we can envision a second real world actor attempting to understand the same

constellation of facts or state of the world, but facing a different quantum of

information costs and/or measure (and/or kind) of bounded rationality. Ultimately,

we would expect the differences between each actor’s information costs and bounded

rationality – i.e. their relative tolerances for complexity – to be a function of several

variables. Variables specific to each actor might conceivably include, inter alia,

economies of scale in the production and/or analysis of information; technological

and/or resource constraints and, importantly, the actor’s initial position within the

constellation of facts or state of the world in question. External variables, meanwhile,

might include market structure, regulation and other institutional features which

subsidize (or impede) the free flow of information and thus level (or tilt) the

informational playing field.

We might thus predict, for example, that an LCFI acting as a market maker

within an opaque, dealer-intermediated, quote-driven market might enjoy a higher

tolerance for complexity in respect of that market than, say, a pension fund manager,

a regulator, or a law professor perched high atop his ivory tower. Put differently, we

would expect to observe clear hierarchies vis-à-vis different actors in terms of both

access to information and the resources needed to effectively process it. As we shall

see, such hierarchies abound within modern financial markets: hierarchies between

market participants; between market participants and regulators and, indeed, even

between regulators. Ultimately, this simple observation – essentially that complexity

33 Unless, of course, we assume that all actors are perfectly rational and fully informed.


is a subjective phenomenon and that, as a result, different actors may find themselves

asymmetrically exposed to its dangers and opportunities – helps explain the existence

and potential value of financial intermediaries. As explored in greater detail below, it

is also the source of many of the regulatory challenges stemming from the complexity

of modern financial markets.

The second important insight we can draw from this framework is that our

tolerance for complexity is not infinite.34 More specifically, we can envision a

frontier beyond which the combination of high information costs and bounded

rationality can be expected to render full comprehension impossible within a given

timeframe. Beyond the complexity frontier, actors will be forced to employ heuristics

as a second-best strategy for understanding a particular set of facts or state of the

world.35, 36 As we shall see, the mere acknowledgement that there may exist elements

of the financial system which are so complex as to render full comprehension a

practical impossibility has potentially profound regulatory implications.

B. Six Drivers of Complexity

Armed with this provisional framework for understanding complexity, we can

embark on an examination of the sources (or drivers) of high information costs – and

information failure – within financial markets and the points of intersection between

these costs and our own bounded rationality. Predictably, complexity itself hampers

our ability to construct anything resembling a complete account of these drivers or the

34 Unless, once again, we assume that actors are perfectly rational and fully informed. 35 This is not to suggest, of course, that actors might not also elect to employ heuristics in less complex circumstances. Ultimately, we are all satisficers. 36 There exists a more fundamental question here, although one which resides beyond the scope of this thesis, as to how to conceptualize the behavior of market participants beyond the complexity frontier. Intuitively, the autonomous rational actor model upon which conventional financial theory tends to rely would seem to possess limited explanatory power beyond the point at which high information costs and bounded rationality combine to force the use of heuristics.


various interactions between them. Nevertheless, taking a broad look across the

financial system, it is possible to identify at least six – in many respects intertwined

and overlapping – sources of complexity: technology, opacity, interconnectedness,

fragmentation, regulation and reflexivity.

As we shall see, these drivers of complexity can be broken down into three

categories: those influencing our capacity to process information; those impacting on

the availability and/or intelligibility of the information itself and, finally, those

accelerating the velocity of informational change. The lines of demarcation between

each of these categories can perhaps be clarified by drawing an analogy with

marksmanship. The first category – which includes both financial and information

technology – can be understood as relating to both the quality of the rifle and the

proficiency of the individual marksman. The second category, meanwhile, includes

drivers which – like darkness, fog, foliage or distance – obscure the visibility of the

target. Drivers falling into this category include technology (again), opacity,

interconnectedness, fragmentation and regulation. Lastly, we must somehow account

for the fact that the target itself may be in motion. Thus, we need a category and

driver – reflexivity – which reflects the inherent dynamism of modern financial

markets. Ultimately, just as we would expect each of these factors to influence the

accuracy of the marksman’s shot, so too would we expect each driver of complexity

to influence the extent to which, in practice, actors are able to understand various

constellations of facts or states of the world.

Technology. There is little doubt that advances in information technology,

telecommunications and financial theory over the course of the past half century have

made a positive (gross) contribution toward the informational efficiency of financial


markets.37 Faster and more powerful computers have enabled market participants to

employ sophisticated and data-intensive quantitative (i.e. statistical) techniques to

calculate the value of financial assets with greater precision and to better understand

and more effectively manage various risks.38 A revolution in telecommunications,

meanwhile, has made possible the almost instantaneous transmission of information

to every corner of the globe.39 Finally, breakthroughs in financial theory – perhaps

most notably the development of MPT40, CAPM41, the Black-Scholes option pricing

model (Black-Scholes)42 and their respective progeny – have given birth to a universe

37 See Robert Merton, “Financial Innovation and the Management and Regulation of Financial Institutions”, National Bureau of Economic Research Working Paper No. 5096 (April 1995) at 6. 38 Powerful computers, for example, have made possible the use of ‘value-at-risk’ (VaR) methodologies and portfolio stress testing to measure and manage the risk of institutional insolvency; Scott Frame and Lawrence White, “Empirical Studies of Financial Innovation: Lots of Talk, Little Action?” (2004), 42:1 J. of Economic Lit. 116 at 120. See also Scott Frame and Lawrence White, “Technological Change, Financial Innovation, and Diffusion in Banking”, Federal Reserve Bank of Atlanta Working Paper No. 2009-10 (March 2009) at 20-21, available at and Lawrence White, “Technological Change, Financial Innovation, and Financial Regulation in the U.S.: The Challenges for Public Policy”, Wharton Centre for Financial Institutions Working Paper 97-33 (1997) at 7, available at 39 Indeed, strong linkages between revolutions in telecommunications and finance are by no means a recent phenomenon. From the telegraph, consolidated ticker tape and electronic fund transfer, to the fax, the internet and the blackberry, the evolution of finance is intricately intertwined with the evolution of how we communicate with one another; see generally Kenneth Garbade and William Silber, “Technology, Communication, and the Performance of Financial Markets” (1978), 33 J. of Finance 819. 40 MPT flows from the premise that there is a tradeoff between risk and return. On the basis of certain assumptions, MPT prescribes, for a given level of risk (variance), how to select a portfolio with the highest possible return (or, conversely, for a given level of return, how to select a portfolio with the least risk). MPT thus makes possible the construction of an efficient frontier from which an investor can choose their desired portfolio on the basis of their individual risk preferences. One of the key insights of MPT is that an asset should not be selected on the basis of its individual risk-return characteristics, but rather with a view to the effect of its addition in terms of the overall risk-return characteristics of the investor’s portfolio; see Harry Markowitz, “Portfolio Selection” (1952), 7:1 J. of Finance 77, and Harry Markowitz, Portfolio Selection: Efficient Diversification of Investments (John Wiley & Sons, New York, 1959). 41 CAPM is used to calculate the expected rate of return on an asset to be added to a diversified portfolio on the basis of (1) the risk free rate of return; (2) the sensitivity of the asset to non- diversifiable (systemic) risk, and (3) the expected market return; see William Sharpe, “Capital Asset Prices: A Theory of Market Equilibrium Under Conditions of Risk” (1964), 19:3 J. of Finance 425, and Jack Treynor, “Toward a Theory of Market Value of Risky Assets” (1962) in Robert Korajczyk, (ed.), Asset Pricing and Portfolio Performance: Models, Strategy and Performance Metrics (Risk Books, London, 1999). 42 Black-Scholes is used to calculate the exact theoretical price of a real option; see Fischer Black and Myron Scholes, “The Pricing of Options and Corporate Liabilities” (1973), 81 J. of Pol. Econ. 637. While the original Black-Scholes model technically applied to the valuation of European options (i.e.


of new financial instruments which have been credited with, amongst other

contributions, enhancing price discovery, market liquidity and systemic resilience. In

short, there exists a strong prima facie argument that these technological

advancements have combined to significantly lower information costs within modern

financial markets.

Upon closer scrutiny, however, these technological advancements are also the

source of potentially significant information costs.43 The origins of this informational

dark side can be traced back to conceptual breakthroughs such as MPT, CAPM and

Black-Scholes, the resulting emergence of “financial science”44 within the field of

economics, and its subsequent rise to prominence within the theory and practice of

modern finance.45 The sophisticated mathematical models residing at the core of this

discipline render its theoretical underpinnings largely inaccessible to all but a

relatively small handful of academic economists, along with the so-called ‘quants’

employed by investment banks, hedge funds and other financial institutions.46 Even

in practice, the utilization of these models contemplates both information-intensive

quantitative processes and the formulation of subjective judgments on the basis of

accumulated technical expertise and experience in order to generate important input

options exercisable only at maturity), its progeny have been adapted to value far more exotic instruments. 43 This is not to suggest that these costs outweigh the informational benefits of these technological advancements. My point here is simply that the existence of these costs contributes, utilizing my definition, to the complexity of modern financial markets. 44 See generally Hu (n 29). The discipline is now generally known as financial economics. 45 For a historical survey of this rise, see generally Peter Bernstein, Capital Ideas: The Improbable Origins of Modern Wall Street (The Free Press, New York, 1992); Robert Merton, “Influence of Mathematical Models in Finance on Practice: Past, Present and Future” (1994), 347 Philosophical Transactions of the Royal Society of London 451. 46 See Richard Whitley, “The Transformation of Business Finance Into Financial Economics: The Roles of Academic Expansion and Changes in U.S. Capital Markets” (1986), 11 Acct. Org. & Soc’y 171 at 173 and Hu (n 29) at 1470.


variables.47 Developing a comprehensive understanding of financial theory and how

to utilize these models in practice thus requires an enormous upfront investment in

human capital.48 Accordingly, while advances in financial theory are largely

responsible for laying the foundations of modern (and at times more informationally

efficient) financial markets, they must simultaneously be viewed as a potentially

significant driver of information costs and, thus, complexity.49

Advances in financial theory and information technology have further

contributed to the complexity of modern financial markets by making possible the

development and wide-spread use of new and increasingly sophisticated financial

instruments. Specifically, the existence of relatively robust markets for instruments

such as OTC swaps50, asset-backed securities (ABS)51 and collateralized debt

47 The Black-Scholes option pricing model is a good example. Prior to the development of Black- Scholes, market participants seeking to determine the value of an option faced a problem: namely, they were required to accurately predict, inter alia, the probability distribution of the possible prices for the underlying asset at maturity; Hu (n 29) at 1468, citing Stephen Figlewski, “Theoretical Valuation Models” in Stephen Figlewski, William Silber and Marti Subrahmanyam, (eds.), Financial Options: From Theory to Practice (McGraw Hill, New York, 1992). Market participants were thus required to formulate subjective judgments about the state of future market conditions. A significant part of the (perceived) genius of Black-Scholes was that it enabled market participants to calculate the precise theoretical value of a European option without having to construct such a probability distribution. In reality, however, Black-Scholes simply substituted the need to predict future asset prices with the need to predict the future volatility of those prices. 48 Furthermore, as illustrated below, the nature and pace of financial innovation operates so as to demand significant ongoing investment in order to preserve the value of this human capital. 49 Hu (n 29) at 1470. 50 A swap is a series of mutual forward obligations whereby two counterparties agree to periodically exchange (or ‘swap’) cash flows over a specified period of time. The classic example of a swap is an interest rate swap pursuant to which one party – typically a borrower with fixed rate obligations – agrees to make payments at a fixed interest rate to a counterparty who in turn agrees to pay the borrower a variable (or ‘floating’) rate. The fixed rate borrower receiving a floating rate thus stands to benefit from any subsequent increase in interest rates, whereas its counterparty receiving the fixed rate under the swap will benefit from any decline. The periodic payments due under a swap are calculated with reference to what is often referred to as a ‘notional amount’. The resulting obligations are then typically netted out against one another such that only one counterparty is obligated to remit payment in any given period. 51 An ABS is a security the income stream from which is backed by a pool of (typically illiquid) underlying assets such as mortgages, automobile loans, credit card receivables or student loans.


obligations (CDOs)52 implicitly rely on two necessary, if not individually sufficient,

conditions: (1) the development of rational models for determining their intrinsic

value, and (2) the ability to meet the computational demands of these models within a

timeframe which enables market participants to profit from their use.53 Financial

theory has satisfied the first condition, advances in information technology the


The development of the “originate-and-distribute”54 mortgage lending model

provides an illustrative example. Recent years have witnessed the increasing use of

computer-generated credit scoring tools to process residential mortgage applications.

The sub-prime mortgage market in particular was (originally) predicated on the use of

sophisticated quantitative tools to assist lenders in better managing their exposure to

high-risk borrowers.55 The utilization of these tools served to enhance the

transparency of mortgage underwriting standards, thereby facilitating the

development of a deep secondary market for mortgages repackaged and distributed

via the process of securitization.56 In very broad terms, securitization is a financing

technique which transforms non-liquid assets such as mortgages and loan receivables

52 A CDO is a type of ABS typically created to hold fixed income assets such as bonds, CDS or, frequently, other ABS. 53 In the absence of the first condition, one would expect a wide divergence between bid-ask spreads, ultimately leading either to very thinly traded markets or complete market failure. In the absence of the second condition, one would expect the existence of substantial transaction costs to alter the economic incentives of potential market participants, ultimately with much the same effect. A third pre-condition for many instruments – and in particular OTC derivatives – was the development of standardized legal documentation; see Awrey (n 13) at 163. 54 Or “originate-to-distribute”, depending on your views respecting why financial intermediaries innovate; see Part IV. 55 Frame and White (2009) (n 38) at 6. 56 See John Straka, “A Shift in the Mortgage Landscape: The 1990s Move to Automated Credit Evaluations” (2000), 11:2 J. of Housing Research 207; Michael LaCour-Little, “The Evolving Role of Technology in Mortgage Finance” (2000), 11:3 J. of Housing Research 173; Susan Gates, Vanessa Perry and Peter Zorn, “Automated Underwriting in Mortgage Lending: Good News for the Underserved?” (2002), 13:2 Housing Policy Debate 369, 389, 370, and Scott and Frame (2009) (n 38) at 14-15.


into more readily alienable ABS (or MBS in the case of mortgages).57 This is

achieved by pooling assets together and then slicing, dicing and reconstituting the

associated cash flow rights into separate tranches. On the supply side, the design of

these MBS – and especially the pricing of the tranches – is itself heavily reliant on,

once again, sophisticated financial models and modern information technology.58 On

the demand side, purchasers employ the same technologies to measure and manage

the risks associated with holding these securities in their portfolios.59 At every stage

of the process, financial theory and information technology combine to facilitate the

development of new financial instruments and markets. While the acronyms may

change, this same fundamental story can been observed playing out across modern

financial markets.

So how have these developments combined to render financial markets more

complex? In the wake of the GFC, it has been widely acknowledged that even the

most (ostensibly) sophisticated counterparties failed to grasp the technical nuances of

many of the new instruments and markets made possible by the confluence of

advances in financial theory and information technology.60 Gary Gorton, for

example, has observed that many market participants did not fully appreciate how the

57 Among other implications, securitization has the effect of reducing (and potentially eliminating) lenders’ exposure to borrower default. As a corollary, it also dilutes the incentives of lenders to screen for and monitor creditor and asset quality. 58 Frederic Mishkin, “Financial Innovation and Current Trends in U.S. Financial Markets”, National Bureau of Economic Research Working Paper No. 3323 (April 1990) at 8-9, available at See also Peter Tufano, “Financial Innovation” in George Constantinides, Milton Harris and René Stultz, (eds.), Handbook of the Economics of Finance (Elsevier, Amsterdam, 2003) at 321-22. 59 David Li, for example, developed a formula known as the Gaussian copula which became widely employed prior to the GFC to evaluate the relationships between the default risks associated with various assets held within securitization structures; see Felix Salmon, “Recipe for Disaster: The Formula That Killed Wall Street”, Wired (February 23, 2009) at 1. 60 See, e.g., Containing Systemic Risk: The Road to Reform, Counterparty Risk Management Policy Group III (August 6, 2008) [the “CRMPG III Report”] at 53, observing: “there is almost universal agreement that, even with optimal disclosure in the underlying documentation, the characteristics of these instruments were not fully understood by many market participants.”


unique structure of sub-prime mortgages made the MBS and CDOs into which they

were repackaged particularly sensitive to volatility in underlying home prices.61

Along a similar vein, Joshua Coval, Jakub Jurek and Erik Stafford have demonstrated

how ratings agencies and other market participants failed to perceive both (1) how the

structure of CDOs (and CDO-squared62) amplified initial errors with respect to the

calculation of default risk on underlying assets, and (2) the systematic

interconnections between these assets.63 Advances in financial theory and

information technology have, accordingly, proven themselves to be less than perfect

tools for understanding the complex dynamics of the very instruments and markets

which they have combined to make possible.64 Put simply, technology has been

unable to keep pace with itself. The (net) contribution of technology toward the

complexity of modern financial markets must ultimately be measured by the extent of

this imperfection.

Opacity. A second significant driver of complexity is the opacity of many

financial instruments, markets and institutions. There are in essence two species of

opacity. The first stems from the simple non-availability of information within a

particular segment of the marketplace.65 Markets exhibiting this form of opacity – in

particular with respect to pricing information and the identity of counterparties – have

61 See Gorton (n 30) at 20-34. As Gorton explains, the unique structure of sub-prime mortgages (specifically their short duration, step-up rates and pre-payment penalties) effectively provided lenders with an implicit embedded option on home prices. 62 In broad terms, a CDO-squared is simply a CDO which has invested in securities issued by other CDOs. 63 See Joshua Coval, Jakub Jurek and Erik Stafford, “The Economics of Structured Finance” (2009), 23:1 J. of Econ. Perspectives 3. 64 Indeed, many of these imperfections are attributable to the unrealistic assumptions (e.g. the existence of autonomous rational actors, perfect information, liquidity) underpinning many financial models – assumptions which, not coincidentally, largely mirror those of conventional financial theory. 65 That is, the non-availability of information to a particular subset of market participants (and, potentially, regulators).


historically included those for OTC swaps, ABS, CDOs and repurchase agreements

(or ‘repos’)66, along with so-called ‘dark pools’.67 Many financial institutions also

exhibit this form of opacity: the most frequently cited example perhaps being the

historical lack of transparency surrounding the investors, holdings and trading

strategies of hedge funds.68 Even traditional commercial banks, however, manifest

opacity of this variety insofar as the marketplace does not generally possess the

borrower or asset specific information needed to accurately determine the value of

these banks’ loan books and, accordingly, the enterprise value of the lenders

themselves.69 Furthermore, while banks and other financial institutions can be

expected to possess a reasonable amount of information regarding their own

counterparties, one would at the same time expect a marked decline in the extent and

quality of the information they possess in respect of their counterparties’

counterparties (and so on down the counterparty daisy chain). Investors in ABS,

CDOs and especially CDO-squared face an analogous challenge insofar as it is often

not possible to penetrate the layers of securitization in order to evaluate the quality of

66 A repurchase agreement is essentially a sale of securities under an agreement by which equivalent securities are to be repurchased at a future date. The duration of these agreements vary from overnight to months or even years, with compensation paid to the seller either in the form of interest or as a mark-up incorporated into the repurchase price. The purchaser may also be required by the seller to post collateral. Louise Gullifer, (ed.), Goode on Legal Problems of Credit and Security (Sweet & Maxwell, London, 2008) at 250. 67 Dark pools are effectively private OTC trading platforms used to match orders internally (i.e. between clients of the same firm) and between institutional trading desks; see “Big Traders Dive Into Dark Pools”, Bloomberg Businessweek (October 3, 2007), available at 68 See Andrew Lo, “Hedge Funds, Systemic Risk, and the Financial Crisis of 2007-2008”, written testimony prepared for the U.S. House of Representatives Committee on Oversight and Government Reform (November 13, 2008) and Willa Gibson, “Is Hedge Fund Regulation Necessary?” (2000), Temple L. Rev. 681, 710. 69 See Robert Bartlett III, “Making Banks Transparent” (2011), Vanderbilt L. Rev. [forthcoming]; Donald Morgan, “Rating Banks: Risk and Uncertainty in an Opaque Industry” (2002), Am. Econ. Rev. 874. But see Mark Flannery, Simon Kwan and Mahendrarajah Nimalendran, “Market Evidence on the Opaqueness of Banking Firms’ Assets” (2004), 71 Fin. Econ. 419.


the underlying assets.70 This first species of opacity can thus be understood as giving

rise to classic asymmetries of information.

The second species of opacity stems from the dense “information thicket”71

generated by the overwhelming volume of data swirling around within modern

financial markets. This opacity is the product of information which, while publicly

available in a strictly technical sense, is extremely (if not prohibitively) costly to

acquire, filter, manipulate and/or analyze.72 The balance sheets of LCFIs exemplify

this form of opacity. The number of positions held by LCFIs, the technical

sophistication of the financial instruments used to take these positions, and the

intricate (and potentially contradictory) nature of the resulting market and

counterparty exposures render it virtually impossible to construct – in a timely

fashion – a comprehensive picture of the overall risk profile of these institutions.73, 74

Much of the explanation for the growth of this information thicket in recent years can

once again be traced back to the development of new financial instruments. As

described above, the computational demands associated with many of these

instruments are exceedingly high.75 As explained by Robert Bartlett:

70 See Gorton (n 30) at 45 and 59. See also, Howell Jackson, “Loan Level Disclosure in Securitization Transactions: A Problem with Three Dimensions”, Harvard Public Law Working Paper No. 10-40 (July 27, 2010), available at 71 See Bartlett (n 31). 72 See Schwarcz (2009) (n 23) at 222. 73 As arguably evidenced by the fact that, in retrospect, the pre-GFC CDS spreads on LCFIs reflected significant under-pricing of the default risks associated with these institutions (the primary counter- argument being that the low spreads reflected the so-called ‘too-big-too-fail’ subsidy). In fact, CDS spreads within the financial services sector suggested that risks were at historically low levels; see Financial Services Authority (FSA), The Turner Review: A Regulatory Response to the Global Banking Crisis (2009) [the “Turner Review”] at 46, available at 74 The information thicket surrounding LCFIs is exacerbated by the existence of the first species of opacity insofar as, for example, GAAP only mandates that positions be reported in the aggregate. 75 See Schwarcz (n 23) at 13 and Gorton (n 30) at 48-49. See also Warren Buffett, Letter to Shareholders (May 2, 2009), available at and CRMPG III Report (n 60).


“Valuing even a single CDO investment – let alone a portfolio of such investments – requires a multi-faceted analysis of a considerable amount of both legal and financial data, ranging from an estimation of the default and prepayment risks of hundreds (potentially thousands) of underlying assets, analysis of the particular overcollateralization and subordination provisions attached to particular tranches of CDO securities, and an assessment of potential counterparty risk of the CDO’s various hedge counterparties.”76

Furthermore, insofar as these instruments facilitate the reconstitution and

redistribution of risk within the financial system (often via transactions within

relatively opaque markets), they obscure the location, nature and extent of the

ultimate exposures.77 Like the first species of opacity, the information thicket

manifests the potential to generate acute asymmetries of information. Unlike the first,

however, this second species of opacity thus raises the additional and rather sobering

prospect that information may become altogether “lost”.78

Robert Bartlett’s event study involving Ambac Financial provides a

compelling illustration of how the information thicket may result in the loss of

information.79 Ambac was and is a large, publicly-listed monoline insurance

company which, prior to the GFC, was active in the business of insuring multi-sector

CDOs. As a result of the confluence of (1) statutory accounting rules mandating

disclosure by monoline insurers of their largest exposures, and (2) European

regulatory requirements mandating disclosure of large volumes of legal and financial

documentation in respect of insured CDOs, it is possible to construct a relatively

complete picture of Ambac’s exposures and, accordingly, its financial health.80 In

2008, a number of CDOs insured by Ambac experienced multi-notch credit rating

76 Bartlett (n 31) at 4. 77 See Schwarcz (2004) (n 23) at 10 and 13. 78 In the sense of being unknown to anyone; Gorton (n 30) at 45. 79 See Bartlett (n 31). 80 See Ibid. at 5, 8-12.


downgrades. Bartlett’s analysis of the abnormal returns surrounding the

announcement of each of these downgrades revealed no significant reaction in

Ambac’s stock price, short-selling data or the CDS spreads on its senior debt

securities.81 The subsequent disclosure of these downgrades within Ambac’s

quarterly earnings announcement, however, was associated with significant one-day

abnormal returns.82 Bartlett attributes this inefficiency to the low salience of

individual CDOs within Ambac’s portfolio and the logistical challenges of processing

CDO disclosures.83 In effect, however, the density of the information thicket

overwhelmed the powerful incentives possessed by market participants to seek out

and exploit such informational inefficiencies.

Interconnectedness. The ongoing process of market liberalization – aided by

advances in telecommunications84 – has sparked a pronounced trend toward greater

globalization and integration of financial markets and institutions. This process has

generated complex linkages within and between these markets and institutions and,

importantly, the real economies they support. Financial institutions are connected to

one another via their (increasingly complex) counterparty arrangements.85 The

balance sheets of these institutions, meanwhile, are connected to markets – and via

81 See Ibid. at 23-35. 82 See Ibid. at 28. Using a single factor market model, Bartlett reports a one-day abnormal return of – 43%. 83 See Ibid. at 1, 7, 48-49. 84 See Mishkin (n 58) at 10. 85 And, indeed, their counterparties’ counterparty arrangements; See Ricardo Caballero and Alp Simsek, “Complexity and Financial Panics”, National Bureau of Economic Research Working Paper 14997 (May 2009) at 2, available at Furthermore, the widespread use of collateral in connection with many of these arrangements can generate linkages between the relevant counterparties (and markets) and prices within the markets for the collateral assets. During the GFC, for example, decreases in the value of senior tranches of sub-prime MBS held as collateral in the repo market triggered what eventually became the complete paralysis of this market; See Zachary Gubler, “Instruments, Institutions and The Modern Process of Financial Innovation”, (2011) Delaware J. of Corp. L. [forthcoming] at 20-21, available at


markets to the balance sheets of other financial institutions – through mark-to-market

accounting methods.86 These balance sheet linkages in turn generate systemic

feedback effects between asset values, leverage and liquidity.87 At an even higher

macro level, household savings patterns in China88 are linked to global asset values

via the resulting demand for (primarily U.S.) government securities, the consequent

reduction in yields on these securities, and the incorporation of these lower yields as a

proxy for the real risk-free rate into the discount rates used in asset pricing models.89

These are but a small sampling of the myriad of intricate, constantly evolving

and often undetected interconnections which shape modern financial markets. While

we have arguably come some distance in identifying and understanding the dynamics

of some of these interconnections90, the acquisition, analysis and ongoing monitoring

of markets and institutions which this entails comes at a high (informational) cost.

Put differently, these interconnections make it more costly to identify and monitor

86 Mark-to-market or ‘fair value’ accounting refers to the practice, reflected in Generally Accepted Accounting Principles (GAAP) and International Financial Reporting Standards (IFRS), of accounting for the value of an asset on the basis of its current market price, the market price of similar assets or, if neither is available, another metric of ‘fair’ value. 87 The basic (spiral) pattern of these effects can be summarized as follows: (1) rising asset values inflate bank balance sheets, allowing them to extend greater leverage; (2) the resulting expansion of credit stimulates demand for assets and liquidity; and (3) increased demand for assets and liquidity has the effect of inflating prices while simultaneously reducing the liquidity premium on the assets. These effects operate in reverse in an environment of falling asset prices; Tobias Adrian and Hyun Song Shin, “Liquidity and Financial Cycles”, presentation to the 6th BIS Annual Conference (June 2007) and IMF, “Assessing the Systemic Implications of Financial Linkages” in Global Financial Stability Report, Volume 2 (April 2009), available at 88 Or, more precisely, China’s resulting current account surplus (combined with its managed exchange rate regime). 89 See Turner Review (n 73) at 11-13. This has a double-barreled effect in terms of stimulating demand: (1) lower yields on U.S. government securities reduce real interest rates (thereby making it cheaper to employ leverage to purchase assets), and (2) the incorporation of lower yields into discount rates reduces risk premiums (thereby making the assets themselves cheaper). 90 For an overview of some of the tools used to evaluate systemic linkages within the financial system (including the network approach, co-risk models, distress dependence matrices and default intensity models), see IMF (n 87). For a critique of these tools, see Steven Schwarcz, “Systemic Risk” (2008), 97 Geo. L. J. 193 at 206.


potential sources of risk within the financial system.91 What is more, the sheer

number of these linkages, their intricacy and their rapid evolution suggest that our

ability to identify and understand them will ultimately be constrained by bounded

rationality. It is perhaps not surprising, therefore, that many of these interconnections

are only revealed (or their importance fully understood) at the point at which they

become channels for the transmission of financial shocks. Ultimately,

interconnectedness represents a significant source of opacity – and thus complexity –

within modern financial markets.

Fragmentation. One of the most striking features of many of the transactions

which exemplify modern financial markets is the extent to which they result in the

fragmentation of economic interests. The archetypal example of this is securitization.

As Kate Judge explains, by repackaging underlying assets such as mortgages into

ABS, repackaging ABS into CDOs, and CDOs into CDO-squared, securitization

transforms what was initially, in many instances, a bilateral relationship into a

complex web involving potentially hundreds of dispersed counterparties.92 Judge has

coined the term “fragmentation nodes”93 to describe this category of transactions.

Each successive fragmentation node attenuates the informational and economic

relationship between counterparties and the underlying assets in which they have,

ultimately, invested.94 This attenuation has the double-barreled effect of (1)

increasing information and coordination costs for counterparties and (2) diluting their

incentives to coordinate their activities and/or invest in the acquisition of

91 See Avgouleas (n 23) at 22. Indeed, as explored in greater detail in Parts V and VI, OTC derivatives offer a compelling example of such interconnectedness and how costly it can be to monitor. 92 See Kate Judge, “Fragmentation Nodes: A Study in Financial Innovation, Complexity and Systemic Risk” (2011), Stanford L. Rev. [forthcoming] at 104-105, 127 and 139. 93 See Ibid. at 105. 94 See Ibid.


information.95 Like interconnectedness, fragmentation thus represents a potentially

significant driver of opacity within modern financial markets.96

Regulation. The complexity of modern financial markets is further

compounded by the complexity of the regulatory regimes which govern them. This

regulatory complexity manifests both substantive and structural elements.

Substantive regulatory complexity stems from what U.S. Senator Charles Schumer

and New York Major Michael Bloomberg, speaking in reference to the U.S.

regulatory landscape, have characterized as the “thicket of complicated rules”97 which

have built up over time within many regulatory regimes. The recently enacted Dodd-

Frank Wall Street Reform and Consumer Protection Act98, to take one example, runs

to 848 pages, is estimated to require up to 243 new federal regulations99 and is

believed by many – no doubt speaking with a touch of hyperbole – to manifest a

“trillion unintended consequences”.100 This comes on top of the substantial pre-

existing edifice of federal securities laws, regulations and jurisprudence governing

U.S. financial markets. Synthesizing this regulation – to say nothing of staying

abreast of new regulatory developments – represents no small challenge for either

market participants or financial regulators.

Structural regulatory complexity, meanwhile, stems from the disconnect

between the increasingly globalized and integrated structure of many financial

95 See Ibid. at 104. 96 See Ibid. at 105. 97 McKinsey & Co., Sustaining New York’s and the US’ Global Financial Services Leadership (City of New York and United States Senate, 2007) [the “Bloomberg Report”] at ii. 98 Pub. L. No. 111-203, 124 Stat. 1376, 1675–82, 1762–84 (2010) [the “Dodd-Frank Act”]. 99 This estimate was made by New York law firm Davis Polk & Wardwell; see “The Uncertainty Principle”, The Wall Street Journal (July 15, 2010). 100 “A Trillion Unintended Consequences”, The Wall Street Journal (July 7, 2010).


markets and institutions, on the one hand, and the fragmentation exhibited within and

between many regulatory regimes, on the other.101 In the U.S., for example, federal

responsibility for financial regulation is currently divided between a cacophony of

regulators including the Federal Reserve Board, Financial Stability Oversight Council

(FSOC), Securities and Exchange Commission (SEC), Commodity Futures Trading

Commission (CFTC), Federal Deposit Insurance Corporation (FDIC), Financial

Industry Regulatory Authority (FINRA), Office of the Comptroller of the Currency

(OCC), Federal Housing Financing Agency (FHFA) and Consumer Financial

Protection Bureau (CFPB).102 A similar degree of regulatory fragmentation can be

observed within the E.U., where the new European Systemic Risk Board, European

Banking Authority, European Securities and Market Authority and European

Institutional and Occupational Pensions Authority must coordinate their activities

both with each other and with national supervisors in each of the bloc’s 27 member

states.103 This regulatory fragmentation results in higher information costs for both

market participants (seeking to understand and comply with regulation) and

regulators (seeking to coordinate their activities).104 What is more, the inevitable

gaps generated by this fragmentation open the door to regulatory arbitrage.105 As we

101 See Merton (n 37) at 31. 102 Compounding this fragmentation, many segments of the U.S. financial services industry are also highly regulated at the state level. 103 For an overview of the new structure of financial supervision in the E.U., see See also Eilis Ferran, “Understanding the New Institutional Architecture of E.U. Financial Market Supervision” (2011), available at 104 Dan Awrey, “The FSA, Integrated Regulation and the Curious Case of OTC Derivatives” (2010), 13:1 U. Penn. J. of Bus. L. 101. 105 As examined in greater detail in Part IV, the term ‘regulatory arbitrage’ refers to transactions or strategies designed to exploit gaps or differences within or between regulatory regimes, ultimately with the intention of either reducing costs or capturing profits; see Frank Partnoy, “Financial Derivatives and the Costs of Regulatory Arbitrage” (1996-1997), 22 J. Corp. L. 211 at 211, n. 1.


shall see, these gaps can also provide the stimulus for financial innovation and, as a

result, contribute still further to the complexity of modern financial markets.

Reflexivity. Complexity does not exist independently of the observer.106 It is

observers, after all, who incur information costs and who are inevitably constrained

by bounded rationality. Yet we are not simply passive observers within financial

markets: we are participants. Economists develop theories of market behavior which

in turn influence the very behavior of market participants which economists seek to

understand.107 Asset values affect our perception of risk, which affects the

availability of credit, which affects asset values.108 Regulators introduce rules

designed to constrain the behavior of market participants, incentivizing market

participants to find ways of circumventing these constraints, thereby necessitating

further regulatory intervention.109 The interactions between the cognitive perceptions

of market participants and regulators, the actions predicated on these perceptions and

the impact of these actions within markets generate complex and often self-

reinforcing feedback loops. George Soros has characterized the interference created

by these feedback loops as “reflexivity”.110 As Soros explains:

“In situations that have thinking participants, there is a two-way interaction between the participants’ thinking and the situation in which they participate. On the one hand, participants seek to understand reality; on the other, they seek to bring about a desired outcome. The two functions work in opposite directions: in the cognitive function reality is the given; in the participating function, the participants’ understanding is the constant. The two can interfere with each other by rendering

106 A fact which is reflected in the framework for understanding complexity set out above. 107 See Donald Mackenzie, An Engine, Not A Camera (MIT Press, Cambridge, 2006). 108 To clarify, asset values affect our perception of risk (and thus the availability of credit) primarily by impacting the value of the collateral pledged and received in connection with the extension of credit. 109 Edward Kane has characterized this interaction as the “regulatory dialectic”; Edward Kane, “Technology and the Regulation of Financial Markets” in Anthony Saunders and Lawrence White, (eds.), Technology and the Regulation of Financial Markets: Securities, Futures and Banking (Lexington Books, Lexington, 1986) at 187-193. 110 George Soros, The Alchemy of Finance (John Wiley & Sons, New Jersey, 2003) at 2.


what is supposed to be given, contingent… Reflexivity renders the participants’ understanding imperfect…”111

Further explaining:

“The imperfection I am concerned with arises because we are participants. When we act as outside observers we can make statements that do or do not correspond to the facts without altering the facts; when we act as participants, our actions alter the situation we seek to understand.”112

The incursion of information costs with a view to better understanding the complex

dynamics of financial markets (whether in search of knowledge or profit or as a

means of achieving regulatory ends) will thus invariably alter these dynamics, thereby

demanding the incursion of further information costs.113 It is a game without end.

Furthermore, our location within the object of study – indeed, ultimately, as the object

of study – would, intuitively, seem likely to magnify the extent of our bounded

rationality. Accordingly, while many economists have tended to shy away from the

utilization of concepts such as reflexivity, any systematic attempt to understand the

drivers of complexity within modern financial markets must somehow account for

this uniquely human element.

Technology, opacity, interconnectedness, fragmentation, regulation and

reflexivity together generate significant information costs and set us on a collision

course with our own bounded rationality. In the process, they drive financial markets

toward – and potentially beyond – the complexity frontier: often leading these

markets to function in very different ways from those posited by conventional

financial theory. Indeed, this process is in many ways the defining feature of what I

111 Ibid. at 2. 112 Ibid. [emphasis added]. 113 See Schwarcz (2009) (n 23) at 238.


have characterized as modern financial markets. Yet this is only one half of the story.

To more fully appreciate the regulatory challenges posed within modern financial

markets we must also examine the unique nature of financial innovation and,

ultimately, the important relationship between complexity and innovation. In many

respects, this examination boils down to a single question: who benefits from the

complexity of modern financial markets?

III. Toward a Supply-Side Theory of Financial Innovation

The word ‘innovation’ brings to mind products and processes – the printing

press, indoor plumbing, penicillin, the designated hitter, etc. – which have

unequivocally made the world a better place. Economists, however, employ the term

in a somewhat more expansive (and, on the surface at least, less normative) fashion to

describe unanticipated shocks to the economy.114 Yet beneath this veneer of

academic objectivity there survives a marked tendency within the literature to view

these unanticipated shocks as being more in the nature of “unforecastable

improvements”.115 This view seems likely to have been influenced by Joseph

Schumpeter’s conception of innovation as the catalyst of the ‘Creative Destruction’

which fuels growth within capitalist economies.116 As Schumpeter explains:

“The fundamental impulse that sets and keeps the capitalist engine in motion comes from the new consumers, goods, the new methods of production or transportation,

114 Along with the responses of economic actors to these shocks; Tufano (n 58) at 310. 115 Merton Miller, “Financial Innovation: The Last 20 Years and the Next” (1986), 21:4 J. of Fin. & Quant. Anal. 459 at 460 [emphasis added]. See also Frame and White (2004) (n 38) at 5 (“Profit- seeking enterprises and individuals are constantly seeking new and improved products, processes, and organizational structures that will reduce their costs of production, better satisfy customer demands, and yield greater profits… When successful, the result is an innovation.” [emphasis added]); Frame and White (2009) (n 38) at 4 (“we define financial innovation as something new that reduces costs, reduces risks, or provides an improved product/service/instrument that better satisfies financial system participants’ demands.”), and Merton (n 37) at 6 (“Looking at financial innovations… one sees them as the force driving the global financial system towards its goal of greater economic efficiency.”). 116 See Joseph Schumpeter, Capitalism, Socialism and Democracy (Harper & Row, New York, 1975) [orig. pub. 1942] at 119.


the new markets, the new forms of industrial organization that capitalist enterprise creates.”117


“The opening up of new markets, foreign and domestic, and the organizational development from the craft shop and factory to such concerns as U.S. Steel illustrate the same process of industrial mutation – if I may use the biological term – that incessantly revolutionizes the economic structure from within, incessantly destroying the old one, incessantly creating a new one. This process of Creative Destruction is the essential fact about capitalism.”118

While Schumpeter himself may not necessarily have espoused this view, it is not

difficult to see how one might interpret his analysis as equating innovation – in the

form of new goods, methods of production or forms of industrial organization – with

progress. Indeed, Schumpeter’s utilization of biological terminology is evocative of

a Darwinian survival of the fittest. As will soon become apparent, however, the

welfare implications of financial innovation are not nearly so straightforward.119 This

indeterminacy points to the desirability of a more cautious, less value-laden

understanding of financial innovation as an ongoing process of experimentation

whereby new institutions, instruments, techniques and markets are (or are perceived

to be) created.120 Ultimately, framing our understanding of financial innovation as

simply a process of (perceived) change – and not necessarily one of improvement –

has profound implications in terms of the way we look at modern financial markets.

117 Ibid. at 82-83. 118 Ibid. at 83. 119 See generally Robert Litan, “In Defence of Much, But Not All, Financial Innovation”, The Brookings Institution (February 17, 2010), available at; James Van Horne, “Of Financial Innovations and Excesses” (1985), 40:3 J. of Finance 621; Tufano (n 58) at 327-329. 120 See Tufano (n 58) at 309 and Gubler (n 85).


A. The Conventional View: Financial Innovation as a Demand-Side Response to Market Imperfections We know relatively little about what stimulates financial innovation. The

dominant economic view, grounded in Proposition I of the M&M capital structure

irrelevancy principle121, envisions financial innovation as a rational demand-side

response to market imperfections.122 These imperfections – many of which are

themselves the products of exogenous changes to the economic environment123 –

include, inter alia, regulation and taxes124; incomplete markets125; transaction

costs126; asymmetries of information and the ensuing agency costs127, and other

inefficiencies which constrain the ability of market participants to maximize their

utility functions. Following this view, these imperfections generate demand for

financial innovations which promise, amongst other things, greater choice, lower

121 The M&M capital structure irrelevancy principle advances, on the basis of certain assumptions, that the value of a firm is independent of its capital structure (i.e. its mix of equity, debt and other capital); see Franco Modigliani and Merton Miller, “The Cost of Capital, Corporation Finance and the Theory of Investment” (1958), 48:3 Am. Econ. Rev. 261. The assumptions underlying the M&M principle include, inter alia, the absence of (1) information costs (and thus asymmetries of information and agency cost problems); (2) bankruptcy costs and (3) taxes. In a world where these assumptions held true, the M&M principle would suggest that there should be no demand for financial innovation (at least in terms of security design). 122 See Tufano (n 58) at 313-314. For more recent work in which the dominance of this demand-side view is evident, see Nicola Gennaioli, Andrei Shleifer and Robert Vishny, “Financial Innovation and Financial Fragility” (2010), Fondazione Eni Enrico Mattei Working Paper 114.2010, available at 123 See Mishkin (n 58) at 1. 124 See, e.g., Frame and White (2004) (n 38) at 9; Mishkin (n 58) at 11; Kane (n 109); Miller (n 115), and Van Horne (n 119) at 623-624. 125 See, e.g., Darrell Duffie and Rohit Rahi, “Financial Market Innovation and Security Design: An Introduction” (1985), 65:1 J. of Econ. Theory 1; Tufano (n 58) at 314, and Van Horne (n 119). 126 See, e.g., Robert Merton, “On the Application of the Continuous Time Theory of Finance to Financial Intermediation and Insurance” (1989), 14 Geneva Papers on Risk and Insurance 225. 127 See, e.g., Tufano (n 58) at 315. For a survey, see Milton Harris and Artur Raviv, “The Design of Securities” (1989), 24:2 J. of Fin. Econ. 55, and Franklin Allen and Douglas Gale, Financial Innovation and Risk Sharing (MIT Press, Cambridge, 1994) at 144-147.


Figure 1.1: Innovation in an M&M World

costs, enhanced liquidity and/or more effective risk management.128 Figure 1.1

depicts the relationship between issuers and investors in an M&M world.

Viewed in this light, for example, the extreme interest rate volatility of the 1970s and

early 1980s lead to innovations such as adjustable rate mortgages, variable-rate

certificates of deposit, financial futures and interest rate swaps129; U.S. regulatory

constraints on the remuneration arrangements, eligible investors and trading strategies

of registered investment companies and advisers spurred the development of hedge

funds, and the thirst for yield on fixed income investments in the low interest rate

environment of the 2000s stimulated demand for, inter alia, new forms of CDOs and

synthetic CDOs domiciled in tax efficient jurisdictions such as Ireland and the

Cayman Islands.130

However, while this demand-side story is important, it paints a fundamentally

incomplete picture. First, it is deeply rooted in the Schumpeterian paradigm in which

128 See Tufano (n 58) at 313-314, citing Robert Merton, “Operation and Regulation in Financial Intermediation: A Functional Perspective” in Peter Englund, (ed.), Operation and Regulation of Financial Markets (Economic Council, Stockholm, 1993); Robert Merton, “A Functional Perspective of Financial Intermediation” (1995), 24:2 Financial Management 23, and Bank for International Settlements (BIS), Recent Innovations in International Banking (BIS, Basel, 1986). 129 See Hu (n 29) at 1466; Mishkin (n 58) at 2-5, and Van Horne (n 119) at 622-623. 130 See Adair Turner, “The Financial Crisis and the Future of Financial Regulation”, speech at The Economist’s Inaugural City Lecture (January 21, 2009), available at, explaining that a reduction in medium and long-term real risk free rates “had driven among investors a ferocious search for yield – a desire among any investor who wishes to invest in bond-like instruments to gain as much as possible spread above the risk-free rate, to offset at least partially the declining risk-free rate”.

Issuers Investors

(Innovative) Economic Claims



Figure 1.2: Innovation in a World with Financial Intermediaries

the intersection of supply and demand are too frequently viewed as being dispositive

of an innovation’s private and social utility. Second, and more importantly, it fails to

adequately account for the incentives of the institutions at the center of the market for

financial innovation: it ignores the role of financial intermediaries.

B. The Supply-side View: Financial Intermediaries as a Driver of Innovation

Curiously, the supply-side dynamics of financial innovation have been largely

overlooked by both academics and policymakers. So who are the primary suppliers

of financial innovation and what are their incentives to innovate? The suppliers are,

by and large, financial intermediaries such as commercial and investment banks,

securities dealers, investment funds and insurance companies. At first glance, the

incentives of these intermediaries might appear relatively straightforward: profit.131

In a competitive environment, however, one would expect these profits to rapidly

erode as imitators enter the marketplace, attract market share and drive down

margins.132 One would further expect the rate of this profit erosion – and thus the

131 See Mishkin (n 58) at 1. 132 See Van Horne (n 119) at 622. What little empirical evidence exists on this front (at least with respect to financial innovation) is inconclusive and not altogether relevant to the present inquiry. In a widely cited empirical study of financial innovations from 1976 to 1984, Peter Tufano found that financial intermediaries did not charge higher prices in the brief ‘monopoly’ period before imitations appeared and, in the long-run, charged lower prices than rivals offering imitative products. Tufano did

Issuers Investors

Economic Claims


Financial Intermediary


Market Access




inclination of financial institutions to innovate – to be a function of the diffusion

speed of the innovation.

We would thus expect the incentives of potential innovators to be relatively

muted in the absence of some means of preventing imitators from freely appropriating

the innovation. This is the traditional economic justification – articulated by

Schumpeter and others – for the extension of intellectual property rights to

innovators.133 By granting innovators a temporary monopoly on the fruits of their

invention, these rights provide the economic incentives (i.e. rents) necessary to spur

innovation. The problem, of course, is that intellectual property rights do not extend

to the vast majority of financial innovations.134 JPMorgan cannot patent a CDO

structure.135 Goldman Sachs cannot copyright the acronym ‘CDS’. It is perhaps

find, however, that innovating banks captured a larger share of underwriting business for the relevant products than did imitators; Peter Tufano, “Financial Innovation and First Mover Advantages” (1989), 25 J. of Fin. Econ. 213. In a more recent study, Kenneth Carrow found an inverse relationship between the number of imitators and the size of underwriting spreads; Kenneth Carrow, “Evidence of Early Mover Advantages in Underwriting Spreads” (1999), 15:1 J. of Fin. Services Research 37. Neither study, however, is particularly illuminating or immediately relevant insofar as (1) their research was focused exclusively on innovations within markets for publicly-traded securities, and (2) neither researcher looked beyond underwriting spreads to examine other potential benefits – the informational advantages associated with market-making or reputational effects, for example – derived from being an innovator. 133 See, e.g., Kenneth Arrow, “Economic Welfare and the Allocation of Resources for Invention” in National Bureau of Economic Research, The Rate and Direction of Inventive Activity: Economic and Social Factors (Princeton University Press, Princeton, 1962); Avinash Dixit and Joseph Stiglitz, “Monopolistic Competition and Optimum Product Diversity” (1977), 67 Am. Econ. Rev. 297, and Jean Tirole, The Theory of Industrial Organization (MIT Press, Cambridge, 1988). 134 Outside the limited scope of business method patents; see the Federal Circuit Court of Appeals decision in State Street Bank v. Signature Financial, 47 U.S.P.Q. 2d 1596 (Fed. Cir. 1998) [“State Street”]. However, one would expect such patents to be of limited practical application in the context of financial innovation insofar as the application process contemplates public disclosure as a precondition to protection. More specifically, it is likely that financial intermediaries will in many instances find such disclosure unpalatable for strategic reasons. This intuition finds empirical support in the form of studies finding that the decision in State Street did not have an appreciable impact on the number of patent applications filed by financial firms; see Robert Hunt, “Business Method Patents and U.S. Financial Services”, Federal Reserve Bank of Philadelphia Working Paper No. 08-10 (May 2008). See also Steven Pokotilow and Ian DiBernardo, “Protection for Financial Indices, ETFs and Other Products” (2006), 263:63 N.Y.LJ., available at for a discussion of the limits on intellectual property rights in financial indices and ETFs in the U.S. 135 For an inside look at the development of CDOs by JPMorgan Chase & Co., see Gillian Tett, Fool’s Gold: How the Bold Dream of a Small Tribe at J.P. Morgan Was Corrupted by Wall Street Greed and Unleashed a Catastrophe (Free Press, New York, 2009).


unsurprising, therefore, that the diffusion rates of many financial innovations are

exceptionally high.136 As a corollary, we would expect to observe relatively little

innovation. Yet this is precisely the opposite of what we often see occurring within

modern financial markets. This observation suggests that we need to develop a better

understanding of why financial intermediaries innovate.

The key insight is derived from understanding that financial intermediaries

possess at least three very different incentives to innovate. First, as previously

acknowledged, they innovate in response to the emergence of ‘genuine’ demand

within the marketplace. Second, they often possess their own incentives stemming

from, for example, the desire to mitigate the impact of various regulatory

requirements. A prime example of this, examined in greater detail in Part IV, is the

use (and adaptation) of securitization techniques by banks to circumvent capital

adequacy requirements. Third, financial intermediaries possess supply-side

incentives to design and implement strategies with the intention of recreating the

monopolistic conditions – usually afforded by the protection of intellectual property

rights – which allow for the ongoing extraction of rents. There are at least two such

strategies and, together, they help reveal the multifaceted relationship between

complexity and financial innovation.

The first strategy involves artificially accelerating the pace of innovation.137

Financial intermediaries engage in this strategy for the purpose of achieving product

differentiation138 – not only vis-à-vis the innovations of their competitors but,

136 See Hu (n 29) at 1484. Although, as we shall see, this diffusion is in many cases limited to a relatively small group of financial intermediaries. 137 See Ibid. at 1479, and Henry Hu, “New Financial Products, the Modern Process of Financial Innovation, and the Puzzle of Shareholder Welfare” (1991), 69 Texas L. Rev. 1273 at 1275. 138 See Tufano (n 58) at 309.


crucially, between previous generations of their own innovations. In this respect, this

strategy is broadly analogous to the short-term ‘planned obsolescence’ through

innovation observed within, inter alia, the fashion, consumer electronics, software

and academic textbook industries.139 This strategy does not necessarily rely on the

existence of any natural demand in the marketplace, nor on the innovation itself being

‘new’ in any material respect. Rather, it can theoretically be premised on little more

than, for example, capitalizing on investor short-termism, other behavioral factors, or

simply tapping the instinctive human desire for the ‘next new thing’.140 The practical

effect of this strategy is to reset the diffusion clock141 – in essence creating more

(albeit shorter) monopoly-like periods – thereby enabling intermediaries to extract

greater rents from their innovations.142 Importantly, this strategy also manifests the

potential to generate what U.K. FSA Chairman Adair Turner has characterized as

“socially useless”143 over-innovation.

The second strategy employed by financial intermediaries in response to the

appropriability problem is to embrace complexity as an integral component of their

business models. More specifically, many financial intermediaries have harnessed

139 Very briefly, planned obsolescence is a strategy pursuant to which producers intentionally design products which are no longer functional and/or fashionable beyond a certain limited period of time. For a timely real world example of this strategy, readers might look to Apple’s relatively frequent releases of new versions of its iPhone and iPad products (and, concomitantly, the overwhelming demand for these products even amongst customers owning previous generations of them). See generally Drew Fudenberg and Jean Tirole, “Upgrades, Tradeins and Buybacks” (1998), 29 Rand J. Econ. 235; Michael Waldman, “Planned Obsolescence and the R&D Decision” (1996), 27 Rand J. Econ. 583, and Michael Waldman, “A New Perspective on Planned Obsolescence” (1993), 108 Q. J. Econ. 273. See also Glenn Ellison and Drew Fudenberg, “The Neo-Luddite’s Lament: Excessive Upgrades in the Software Industry” (2000), 31 Rand J. Econ. 253 and Laurence Miller, Jr., “On Killing Off the Market for Used Textbooks and the Relationship Between Markets for New and Secondhand Goods” (1974), 82 J. Pol. Econ. 612. 140 See Van Horne (n 119) at 626. Or, in the case of academic textbooks, having a captive audience. 141 Who, after all, would want to imitate previous innovations now viewed as being outmoded? 142 Primarily in the form of higher underwriting spreads. 143 “Financial Services Authority Chairman Backs Tax on ‘Socially Useless’ Banks”, The Guardian (August 27, 2009).


technology (and especially financial theory) to develop – and move an increasingly

large proportion of their business activities into – new and relatively opaque

institutions, instruments and markets.144 They have also lobbied fiercely against

regulatory reforms which would seek to achieve, amongst other objectives, a leveling

of the informational playing field.145 Interestingly, this confluence of technology and

opacity has not necessarily been utilized, as one might predict, to thwart imitators and

thereby slow the diffusion rate of innovation.146 Indeed, small groups of financial

intermediaries have often collaborated in the development of new financial

instruments, markets and institutions.147 The resulting complexity has instead often

been used by intermediaries as a group to prevent the commoditization of many

financial innovations, ultimately forestalling the redistribution of rents from

innovators to consumers which one might otherwise expect to take place over time.148

Within more arcane and opaque markets, these rents flow not only from higher

underwriting spreads but also the informational advantages derived from the role

financial intermediaries play as market-makers.149 It is in their quest to maximize and

exploit their comparative informational advantage that financial intermediaries have

thus driven us toward – and beyond – the complexity frontier.

144 This of course makes perfect sense given the expectation of higher profit margins within such markets. 145 See Gary Rivlin, “The Billion Dollar Bank Heist”, Newsweek (July 11, 2011), available at; Edwar Wyatt and Eric Lichtblau, “A Finance Overhall Fight Draws a Swarm of Lobbyists”, The New York Times (April 19, 2010), available at, and Brady Dennis and Steven Mufson, “Bankers Lobby Against Financial Regulatory Overhaul”, The Washington Post (March 19, 2010), available at 146 The most notable exception to this likely being a financial institution’s investment strategies, where opacity is employed specifically with a view to preventing imitation. 147 See Awrey (n 13) for an exploration of how financial intermediaries and other private actors – and ISDA in particular – have collaborated in the development of OTC derivatives markets. 148 And, simultaneously, preventing a potentially costly innovation ‘arms race’ between competing financial intermediaries. 149 Including, inter alia, (1) pricing and counterparty information, and (2) lower search costs for underwriting opportunities. See Part IV for a discussion of the market-making role played by financial intermediaries within OTC derivatives markets.


This, of course, begs an important question: why would consumers of financial

innovation – upon learning of the existence and potential use of these strategies – not

take appropriate countermeasures? More specifically, why would rational and fully

informed consumers not (1) apply a ‘lemons’ discount; (2) insist on the utilization of

costly contracting mechanisms designed to reveal information about the quality of the

innovation, or (3) refuse to transact with financial intermediaries which they

suspected of engaging in these strategies?150 As a preliminary matter, one might

observe that these consumers’ lower tolerance for complexity would impede this

learning process.151 However, while this would almost certainly be true on one level,

the relevant question simply becomes: why would consumers – or competing

financial intermediaries – with a higher tolerance for complexity not share the fruits

of their knowledge with less sophisticated consumers? Why, in other words, would

this information not ultimately find its way into the broader marketplace?

There are a number of potential explanations for this type of market failure.

The ‘shrouding’ model developed by Xavier Gabaix and David Laisbon, for example,

demonstrates how ‘shrouding’ – i.e. the process by which producers hide information

from consumers respecting high priced add-ons – can flourish even in highly

competitive markets.152 Gabaix and Laisbon’s model proceeds on the basis of a

distinction between ‘sophisticated’ and ‘myopic’ consumers.153 Using examples

150 Ultimately dis-incentivizing their use. For a theoretical discussion of the so-called ‘lemons’ (i.e. adverse selection) problem, see George Akerlof, “The Market for Lemons” (1970), 84:3 Q. J. Econ. 488. 151 Indeed, one would expect that artificially accelerating the pace of innovation would itself impede this process. 152 See Xavier Gabaix and David Laibson, “Shrouded Attributes, Consumer Myopia, and Information Suppression in Competitive Markets” (2006), 121:2 Quarterly J. of Econ. 505. 153 And the existence of both in the marketplace; ibid. at 510.


drawn from the banking154, hospitality155 and office product industries156, Gabaix and

Laisbon then illustrate how producers utilize marketing strategies which obscure

high-priced add-ons (often in the ‘fine print’) with the objective of exploiting myopic

customers who, by definition, fail to recognize that the proverbial wool is being

pulled over their eyes. Sophisticated customers – who can see through the shrouding

– then exploit the marketing schemes designed to target myopic customers by, for

example, opting out of the add-ons. The result is an equilibrium in which producers,

competitors offering close substitutes157, and sophisticated consumers158 have no

incentive to ‘de-bias’ myopic customers by revealing the existence or true cost of the

add-ons.159 Gabaix and Laisbon further observe that, over the long run, shrouding

may be sustained by, inter alia, the entrance of new myopic customers; the

development of new shrouding techniques or, importantly, new rounds of


Second, even where these strategies are transparent to the marketplace, there

remains the fundamental issue of market access. For example, as we will examine in

greater detail in Part IV, the dealer intermediated structure of OTC derivatives

markets – combined with the economies of scale associated with market making161 –

154 Where various ATM, minimum balance and other fees are often shrouded; ibid. at 506. 155 Where hotels, for example, shroud add-ons such as parking, telecommunications and room service charges; ibid. at 507-508. 156 Where printer manufacturers, for example, often advertise low prices for inkjet printers, but not the (far higher) cost of patented ink cartridges; ibid at 506. 157 Who risk de-biasing their own consumers. 158 Who can be understood as receiving a subsidy from the marketing strategies designed to exploit myopic consumers; ibid. at 509-510. 159 Ibid. 160 Ibid. at 522-523. 161 More specifically: (1) the informational benefits derived from access to a larger proportion of overall trading activity (i.e. deal flow) and (2) the hedging benefits derived from being able to trade


has resulted in the concentration of trading activity within a small oligopoly of

financial intermediaries. What is more, virtually all of these intermediaries are

LCFIs. Market participants looking to utilize OTC derivatives have thus historically

enjoyed a limited menu of counterparty options outside these powerful and opaque

institutions. This in turn is likely to have diluted the impact of any market discipline

which might have otherwise been brought to bear on those intermediaries who engage

in strategies designed to extract rents from their higher tolerance for complexity.

All of this is not to suggest that this nascent supply-side theory of financial

innovation fully encapsulates the incentives – or explains the behavior – of all

financial intermediaries, in all markets, at all times. Demand-side factors are clearly

important. Nor am I suggesting that financial intermediaries have engaged in some

sort a grand conspiracy to make financial markets more complex. What I am

suggesting, however, is that re-conceptualizing financial innovation as a process of

change influenced by the incentives of innovators – who have the most to gain and

possess a comparative informational advantage – can enhance our understanding of

the complex and rapidly evolving dynamics within modern financial markets. What

is more, re-conceptualizing financial innovation in this light serves to illuminate the

regulatory challenges stemming from the interaction of complexity and innovation.

We will turn our attention to these challenges in a moment. First, however, it is

important to unpack the multifaceted relationship between complexity and financial


with a larger number of counterparties, looking to take a larger (and more diverse) number of exposures.


IV. The Relationship between Complexity and Financial Innovation: Three Case Studies

As may already be apparent, complexity and financial innovation are mutually

reinforcing dynamics. This symbiosis can be observed across at least four

dimensions. First, as described above, complexity can be utilized by financial

intermediaries for the purpose of preventing the commoditization of an innovation.

Second, financial intermediaries which enjoy a higher tolerance for complexity

relative to other market participants (and regulators) can exploit this advantage – i.e.

extract rents – by offering ‘innovative’ products and services which their clients may

not fully understand. Third, newer and more innovative financial instruments

invariably demand the incursion of high (initial) information costs on the part of both

market participants and regulators. What is more, these instruments often (1) trade

within less developed and more opaque markets and (2) generate unanticipated and

undetected interconnections within and between financial markets and institutions,

thereby exacerbating complexity. Finally, insofar as financial innovation is employed

as a reflexive response to changes in the prevailing regulatory environment, both this

innovation and the regulation which spawned it can be viewed as contributing to the

complexity of modern financial markets.

A. Complexity and Financial Innovation within OTC Derivatives Markets

There exists no shortage of potential case studies illustrating various

dimensions of the relationship between complexity and financial innovation. Three

particularly compelling examples, however, are securitization, synthetic ETFs and

collateral swaps. It should come as no surprise that all three of these case studies are


drawn from the world of OTC derivatives.162 OTC derivatives markets have long

been recognized at hotbeds of financial innovation.163 Perhaps more importantly,

however, the dealer-intermediated microstructure which characterizes these markets

has bestowed upon OTC derivatives dealers a distinct informational advantage –

especially in terms of pricing and deal flow – vis-à-vis their clients, other market

participants and regulators.

The defining feature of this microstructure is the fact that dealers perform an

explicit market-making role: structuring derivatives instruments and marketing them

to clients on the basis that they are willing to take either side of the transaction.164

These dealers then typically look to eliminate the resulting exposures by seeking out

and entering into offsetting transactions with other clients or, in many cases, other

OTC derivatives dealers.165 Dealers are thus central – indeed, essential – to the

operation of OTC derivatives markets: representing not only the primary source of

innovation, but also of market access, information and liquidity.166 This reality is

reflected in the concentration of trading activity within these markets. As of June

2010, for example, the fourteen largest OTC derivatives dealers (the so-called ‘G14’) 162 Nor that they are drawn from the vast, opaque and intricately interconnected plumbing of the shadow banking system. 163 See, e.g., Darrell Duffie, Ada Li and Theo Lubke, “Policy Perspectives on OTC Derivatives Market Infrastructure”, Federal Reserve Bank of New York Staff Report No. 424 (March 2010) at 10, available at; René Stulz, “Over-the-Counter Derivatives Markets Act of 2009”, testimony to the House Financial Services Committee (October 7, 2009) at 5; Darrell Duffie and Henry Hu, “Competing for a Share of Global Derivatives Markets: Trends and Policy Choices for the United States”, Stanford University Rock Center for Corporate Governance Working Paper No. 50 (June 3, 2008) at 3, available at, and Dan Awrey, “Regulating Financial Innovation: A More Principles-Based Proposal?” (2011), 5:2 Brook. J. of Corp., Fin. & Comm. L. 274. 164 This description is most apt in respect of swaps markets. The circumstance is somewhat more complicated in respect of many securitization markets, where dealers can also perform a role more closely resembling that of an underwriter in a traditional securities offering. Ultimately, the dealer’s role will generally hinge on how bespoke the instrument is to the needs of a particular client or clients. 165 See Deutche Börse Group, The Global Derivatives Market: An Introduction (April 2008) at 17, available at Subject to applicable regulatory constraints, dealers can also engage in so-called ‘proprietary’ trading for their own account. 166 See Duffie et. al. (n 163) at 10.


were responsible for approximately 82% of the global swaps market.167 This

microstructure has historically deprived the marketplace of objective and transparent

market-access and pricing mechanisms. To put it bluntly, OTC derivatives markets

bear almost no resemblance to the perfect markets of conventional financial theory.

The information costs (and information failure) generated by this

microstructure are compounded by, inter alia, the opacity of the LCFIs, hedge funds

and many other counterparties which utilize OTC derivatives168; the fragmentation

which many OTC derivatives engender169 and, in many cases, the sophisticated

technical aspects of the instruments themselves.170 Furthermore, as amply illustrated

by the GFC, the widespread use of OTC derivatives strengthens and expands the

intricate web of interconnections within and between financial markets and

institutions. Collectively, these attributes epitomize the complexity of modern

financial markets. They also render securitization, synthetic ETFs and collateral

swaps uniquely illuminating case studies in terms of both the relationship between

complexity and financial innovation and, ultimately, the regulatory challenges posed

by the interaction of these powerful market dynamics.

167 ISDA, “Concentration of OTC Derivatives Among Major Dealers”, ISDA Research Note, Issue 4 (2010), available at Broken down by instrument, the G14 held 82% of the total outstanding notional amount of interest rate derivatives, 90% of CDS, and 86% of equity derivatives; ibid. 168 See discussion at 19-20. Indeed, the fact that the identify of counterparties to OTC derivatives matters cuts against the grain of conventional financial theory. 169 See discussion at 24-25. 170 It is certainly the case that many OTC derivatives are (at least from an economic perspective) relatively straightforward to understand and use. It would take a small upfront investment to familiarize oneself with, for example, the basic structure and potential uses of a single currency interest rate of foreign exchange swap. At the same time however, the derivatives universe is populated by a diverse array of far from complex instruments. For a comprehensive description of the technical aspects of many of these instruments, see Satyajit Das, The Swaps and Financial Derivatives Library: Products, Pricing, Applications and Risk Management, 3rd ed. (John Wiley & Sons, New York, 2005) and Richard Flavell, Swaps and Other Derivatives, 2nd ed. (John Wiley & Sons, New York, 2009).


B. Three Case Studies in Complexity and Financial Innovation

Securitization.171 The case study which has to this point garnered the most

scholarly attention is undoubtedly securitization.172 As described in Part II,

securitization is a process whereby the cash flows associated with non-liquid assets

are pooled together, restructured and sold as securities. Most structured finance

vehicles are, in effect, a form of credit derivative.173 The first ABS was issued by the

U.S. Government National Mortgage Association (Ginnie Mae) in 1970.174 This

nascent ABS market initially revolved around the issuance of residential MBS by

U.S. government sponsored enterprises (GSEs) such as Ginnie Mae, the Federal

National Mortgage Association (Fannie Mae) and the Federal Home Loan Mortgage

Corporation (Freddie Mac).175 Between 1970 and 2010, annual issuances within this

so-called ‘agency’ MBS market grew from approximately $USD452 million to over

$USD1.9 trillion.176 As of June 30, 2011, the outstanding amount of U.S. mortgage-

related securities stood at approximately $USD7 trillion.177

171 Some might object, perhaps justifiably, to the assertion that securitization vehicles constitute OTC derivatives. Ultimately, however, while there are important economic (and legal) distinctions between securitization vehicles and other species of derivatives (e.g. swaps, options, etc.), they do ultimately fall within the generic – if somewhat overbroad – definition of a derivative as a financial contract the value or expected performance of which is linked to another, underlying, asset or assets. 172 See for example Schwarz (2004 and 2009) (n 23); Gorton (n 30); Bartlett (n 31); Jackson (n 70); Gubler (n 85) and Judge (n 92). 173 Essentially because the obligations of the issuers of these securities to make periodic payments to the holders are contingent upon the (non-)performance of the underlying assets (as measured by their ability to generate the expected cash flows). 174 Shelagh Heffernan, Modern Banking (John Wiley & Sons, Chichester, 2005) at 46. 175 Prohibited by law from originating mortgages, the GSEs would acquire mortgages from private lenders, securitize them and then guarantee the income streams generated by the resulting MBS; ibid. at 47. 176 Securities Industry and Financial Markets Association (SIFMA), U.S. Mortgage-Related Securities Issuance (June 13, 2011), available at 177 SIFMA, U.S. Mortgage-Related Securities Outstanding (August 1, 2011), available at www.


Observing this success, private sector financial institutions – primarily large

commercial and investment banks – began structuring and distributing ‘private label’

ABS in the mid-1980s.178 Notably, the timing of this move roughly corresponded

with the completion of the 1988 Basel Capital Accord (Basel I). These financial

institutions employed the structures developed by the GSEs in connection with

residential mortgages and quickly adapted them to securitize cash flows derived from

a far broader range of underlying assets including, inter alia: commercial mortgages;

home equity and student loans; automobile, aircraft and equipment leases; credit card

receivables; corporate debt; swaps, and even other securitizations.179 Between 1985

and 2011, the outstanding amount of non-mortgage-related ABS issued in the U.S.

and Europe grew over 1800% – from an estimated $USD1.2 billion to over $USD2.2


The emergence and precipitous growth of both agency and private label

securitization markets – to say nothing of the markets for CDOs and CDO-squared –

is attributable to a complex bundle of supply-side, demand-side and other incentives.

The agency ABS market, for example, grew at least in part out of a desire on the part

of the U.S. federal government to expand home ownership, essentially as a means of

ameliorating rising economic inequality.181 Investors, meanwhile, flocked to ABS,

CDOs and other securitizations in search of both (1) higher yields182 and (2)

diversified exposure to, inter alia, the U.S. residential and commercial property

178 SIFMA (n 176) and Heffernan (n 174) at 47. 179 See SIFMA, U.S. Asset-Backed Securities Outstanding (July 5, 2011), available at 180 Ibid. and SIFMA, Europe Structured Finance Outstanding (May 25, 2011), available at 181 See Raghuram Rajan, Fault Lines: How Hidden Fractures Still Threaten the World Economy (Princeton University Press, Princeton, 2010) at 21-45 and FCIC (n 6) at 38-42. 182 Turner (n 130).


sectors.183 Ultimately, however, much of this growth is attributable to the supply-side

incentives of the commercial and investment banks which structured and sold these

securities. As a preliminary matter, financial institutions sponsoring securitized

offerings earned sizable fees in connection with these transactions. What is more,

securitization enabled originators to shift the market, liquidity, interest rate and other

risks associated with the underlying assets off their balance sheets. Most importantly,

however, securitization enabled banks to secure relief from capital adequacy

requirements184, thus freeing up capital for reinvestment.185 Viewed in this light, the

supply-side incentives come front and centre: the more assets a bank could repackage

and sell via securitization, the more capital it could deploy toward new investments,

and the more assets it would have to fuel the securitization machine. Introduce CDOs

and CDO-squared into this mix – and thus the ability to make new assets out of thin

air – and it is little wonder that securitization markets witnessed such exponential

growth in the decades leading up to the GFC.

The complexity generated by the constant stream of new innovation within

ABS, CDO and other securitization markets is well documented. As both Gorton and

Coval et. al. observe, many of the most (ostensibly) sophisticated institutional

investors failed to fully grasp the complex technical aspects of both mortgage-backed

183 FCIC (n 6) at 43. 184 While a detailed examination of capital adequacy requirements is well beyond the scope of this paper, these requirements – and specifically those articulated under Basel I, II and III – prescribe, inter alia, that banks and certain other classes of financial institution maintain a specified ratio of capital to risk-weighted assets. Insofar as many securitization vehicles attract a lower risk weighting than the underlying assets under these requirements, financial institutions will ceteris paribus be required to hold a lower amount of capital and, accordingly, will be incentivized to repackage and sell these assets via securitization. 185 Viral Acharya, Phillipp Schnabel and Gustavo Suarez, “Securitization Without Risk Transfer” (August 8, 2011), available at; Financial Crisis Inquiry Commission, “Overview on Derivatives”, Preliminary Staff Report (June 29, 2010) at 6, and Alan Greenspan, “The Role of Capital in Optimal Banking Supervision and Regulation” (1998), Federal Reserve Bank of New York Policy Review 163 at 165-166.


ABS and the more complex CDOs into which they were repackaged.186 Along the

same vein, the structure of many of these instruments undermined the ability of both

underwriters and investors to effectively screen for and monitor asset and creditor

quality.187 These informational problems became more acute with each successive

fragmentation node.188 Ultimately, these factors combined to obscure from view the

enormous risks building within this market.

Synthetic ETFs. A second (and considerably less notorious) case study

illustrating the relationship between complexity and financial innovation is the

burgeoning market for synthetic ETFs. ETFs are exchange-traded investment funds

designed to replicate the value of a portfolio of assets (e.g. the FTSE, S&P 500 or

MSCI Emerging Markets Index).189 ETFs are generally regarded as low cost and

liquid vehicles for investors seeking portfolio diversification.190 Their economic

rationale is thus very much grounded in MPT. Introduced in the early 1990s, plain

vanilla ETFs physically replicate the reference portfolio by purchasing the underlying

assets.191 Synthetic ETFs, in contrast, are a more recent innovation designed to

replicate the reference portfolio through the use of OTC derivatives.192

186 Gorton (n 30) at 20-34 and Coval et. al. (n 63). 187 Gorton (n 30) at 45 and 59 and Jackson (n 70). 188 Judge (n 92) at 3. 189 The investment firm BlackRock estimates that there are now in excess of 2,700 ETFs worldwide, replicating various portfolios of public equity and debt securities, across virtually every conceivable investment style, country and region; see “Too Much of a Good Thing”, The Economist (June 25, 2011). 190 IMF, Global Financial Stability Report (April 2011) at 68, available at; Financial Stability Board (FSB), “Potential Financial Stability Issues Arising From Recent Trends in Exchange- Traded Funds (ETFs)” (April 12, 2011) at 1, available at; Bank of England, Record of the Interim Financial Policy Committee Meeting of June 16, 2011 (June 24, 2011) at 8, and ibid. 191 BIS, “Market Structures and Systemic Risks of Exchange-Traded Funds”, BIS Working Paper No. 343 (April 2011) at 4, available at 192 Ibid. and FSB (n 190) at 2.


While there exist a number of ways to structure a synthetic ETF, perhaps the

most common technique involves the sponsor of the fund entering into a total return

swap193 with a financial intermediary.194 There are two components – or ‘legs’ – of

this swap. In the first leg, the ETF sponsor contracts with the financial intermediary

to receive the total return on the reference portfolio in exchange for cash equal to the

notional amount of the swap.195 In return, the financial intermediary transfers a

portfolio of collateral to the ETF sponsor. Importantly, the collateral assets are often

unrelated to those which the synthetic ETF has been designed to replicate.196 The

second leg of the swap then involves the transfer of the total return on the collateral

package back to the financial intermediary.197

Synthetic ETFs have thus far proven especially popular in Europe and Asia.198

The growing demand for these derivatives has been stoked by institutional investors

in search of higher returns in less liquid fixed income and emerging markets where

physical replication of the reference portfolio would almost certainly prove

193 A prototypical total return swap (or TRS) involves swapping cash flows calculated with reference to a floating rate of interest for those derived from the total return (i.e. all capital gains and interest/dividend income) on a given asset or portfolio of assets; BIS (n 191) at 5. 194 This structure is commonly referred to as the ‘unfunded swap structure’; ibid. This is in contrast to the ‘funded swap structure’ which, in a nutshell, involves the ETF sponsor buying a structured note secured by a collateral pledge from a financial intermediary. Notably, in the funded swap structure, the financial intermediary posts eligible collateral into a ring-fenced custodial account. Accordingly, unlike the unfunded swap structure, the ETF sponsor is not the beneficial owner of the collateral assets; see ibid. at 6 for further details. 195 Ibid. This has the benefit of transferring the tracking risk in the reference portfolio to the swap counterparty. 196 Ibid. and The Economist (n 189). For ETFs domiciled in the E.U., for example, the Undertakings for Collective Investments in Transferrable Securities (UCITS) Directive 88/220/EEC (as amended) only prescribes that the collateral assets be selected from among certain prescribed classes of equity or debt securities; see UCITS Directive, Arts. 22 and 23 and FSB (n 190) at 4 for further details. 197 BIS (n 191) at 5. 198 FSB (n 190) at 3. Synthetic ETFs are less popular in the U.S. owing to regulatory constraints imposed under the Investment Company Act of 1940, codified at 15 U.S.C. §80a (1940) [the “ICA”]; see IMF (n 190) at 68. Notably, in March 2010 the SEC announced that it was conducting a review of the use of derivatives by ETFs; see SEC, “SEC Staff Evaluating the Use of Derivatives by Funds”, Press Release 2010-45 (March 2010), available at


prohibitively expensive.199 At least some of the impetus for the development of

synthetic ETFs, however, stems from the desire on the part of the financial

intermediaries acting as swap counterparties to remove less liquid collateral from

their balance sheets – ultimately with a view to enhancing their liquidity profile,

lowering securities warehousing costs and, once again, obtaining relief from

regulatory capital requirements.200 In the extreme – and in particular where the

financial intermediary is affiliated with the fund sponsor – synthetic ETFs can thus be

utilized as a “dumping ground”201 for lower quality assets.202 This in turn serves to

highlight the fact that these instruments expose investors to both (1) counterparty

credit risk in connection with the swap itself and (2) market and liquidity risk in

connection with the swap collateral.203 Accordingly, while synthetic ETFs are

themselves exchange-traded (and thus highly regulated204) instruments, their

complexity and risk profile more closely resemble the OTC derivatives which reside

at the core of this increasingly popular investment fund structure.

The complexity associated with synthetic ETFs stems primarily from the

opacity of the underlying swaps and, more specifically, their collateral packages.

This opacity is illustrated by a recent exercise conducted by the BIS involving a

199 The Economist (n 189) and BIS (n 191) at 1. These increased costs are attributable to, inter alia, the wider bid-ask spreads typically encountered within these markets; BIS (n 191) at 4. 200 Ibid. at 1 and 8-10; FSB (n 190) at 2, and Bank of England (n 190) at 8. In effect, synthetic swaps can thus be utilized to perform the same economic function (i.e. liquidity transformation) as collateral swaps (see below). 201 The Economist (n 189). 202 IMF (n 190) at 71-72. 203 Ibid. and BIS (n 191) at 8-9. What is more, these risks are likely to be exacerbated during periods of market turmoil. 204 As previously mentioned, these instruments are subject to the ICA in the U.S. and the UCITS Directive in the E.U., along with the rules of the exchange on which they trade.


widely traded synthetic ETF replicating the MSCI Emerging Markets Index.205 With

the assistance of the fund sponsor, the BIS was able to determine that the collateral

package for this fund contained over 1000 securities, consisting largely of Japanese

equities and unrated U.S. corporate bonds.206 In the end, however, the BIS found that

a more detailed breakdown of the assets in the collateral package was “not readily

available”207 and that obtaining this information “would be a cumbersome

process”.208 It is also worth noting that the geographic dispersion of the assets within

the collateral package bears little relation to the emerging market portfolio the fund is

designed to replicate. The BIS exercise thus reinforces the concern that investors in

synthetic ETFs may be operating with less than perfect information respecting the

risks to which they are ultimately exposed.

Collateral swaps.209 Our final case study is the emerging market for so-called

‘collateral swaps’. A collateral swap is essentially a form of secured lending whereby

one counterparty transfers relatively liquid assets to another in exchange for a pledge

of less liquid collateral.210 In a typical collateral swap, a bank holding a portfolio of

ABS or other securitizations will transfer these assets to a pension fund or insurance

company which, in exchange for a periodic fee, will deliver a portfolio of more liquid

205 BIS (n 191) at 9-10. This fund utilizes the ‘funded’ swap structure. 206 Ibid. 207 Ibid. 208 Ibid. 209 As with securitization, there is a very legitimate argument that, despite their name, collateral swaps should not be categorized as OTC derivatives. Indeed, insofar as these instruments are structured as long-dated repo contracts, they bear little similarity with more traditional swaps (i.e. a series of forward agreements). 210 For this reason, these transactions are often referred to within collateral management circles as ‘liquidity transfers’. In effect, collateral swaps are economically quite similar to a long-dated repo arrangement.


collateral such as high-grade government or corporate bonds.211 The pension fund or

insurer thereby receives a higher yield on its (ostensibly) safe investments, while the

bank obtains access to a portfolio of liquid assets which it can then re-pledge to

obtain funding from central banks and other sources which, in the wake of the GFC,

have been less willing to accept ABS and other securitizations as eligible collateral.212

The development of collateral swaps is thus, in effect, an innovative response to both

the post-crisis funding constraints on banks and the need to satisfy new liquidity

requirements soon to be imposed under Basel III.213

Collateral swaps contribute to the complexity of modern financial markets in

at least three ways. First, the collateral swap market is extremely opaque. Nobody

knows with any certainty, for example, how big this market is, who the major players

are, or the size of the aggregate exposures. As a result, it is exceedingly difficult to

ascertain the nature and extent of the attendant risks.214 Second, given the identity of

the counterparties, collateral swaps seem destined to strengthen the interconnections

between banking markets, on the one hand, and insurance and pension funds, on the

other. Finally, as described above, collateral swaps are a reflexive response to

changes in the post-crisis market and regulatory environment.

211 Jennifer Hughes, “Concern Mounts Over Rise of Collateral Swaps”, The Financial Times (June 30, 2011), available at; Izabella Kaminska, “The Privatization of Liquidity Ops”, The Financial Times (December 17, 2010), available at; Izabella Kaminska “It’s Stock Lending Jim, But Not As You Know It”, The Financial Times (October 28, 2010), available at, and Aaron Wollner, “Funding Needs Drive Banks to ‘Borrow’ Liquidity from Insurers and Pensions Funds”, Life & Pension Risk (October 28, 2010), available at 212 Ibid. 213 See Ibid. In effect, the counterparties to collateral swaps are arbitraging differences in the capital adequacy regimes applicable to banks, on the one hand, and pension funds and insurance companies, on the other. 214 Bank of England (n 190) at 8.


Taken together, securitization, synthetic ETFs and collateral swaps exemplify

both the complexity of modern financial markets and the nature and pace of financial

innovation. The salient question thus becomes: what are the regulatory challenges

flowing from the interaction of these ubiquitous forces?

V. Complexity and Financial Innovation: The Regulatory Challenges

As amply illustrated by our three case studies, complexity and financial

innovation together generate a host of regulatory challenges. Sophisticated new

instruments, derived from esoteric financial theory, structured in ways which obscure

the attendant risks, and traded in opaque dealer-intermediated markets by opaque

financial institutions raise clear investor protection issues. Paramount amongst these

are the potential for both (1) uninformed (suboptimal) contracting215, and (2) fraud,

misconduct and other opportunistic behavior on the party of financial intermediaries.

The potential for suboptimal contracting in turn raises the prospect of both

overinvestment and excess leverage leading, ultimately, to the build-up of systemic


Simultaneously, opacity and the pace of innovation also render it more

difficult for regulators to effectively police financial markets and – in conjunction

with interconnectedness and fragmentation – to locate and monitor potential risks.

Meanwhile, the vast array of intricate, evolving and often undetected interconnections

within and between markets and institutions – themselves often the byproducts of

financial innovation – foment systemic fragility and manifest the potential to become

215 As Milton Friedman observed, optimal contracting necessitates that the actions of counterparties are both voluntary and informed; Milton Friedman, Capitalism and Freedom (University of Chicago Press, Chicago, 1962) at 13. Accordingly, where counterparties face high information costs, asymmetries of information and the resulting agency costs problems, there is reason to question the private (and social) optimality of the contracts into which they enter.


channels for the transmission of contagion during periods of market distress.216

Reflexivity contributes still further to this fragility insofar as its self-reinforcing

feedback effects drive the formation of asset bubbles.217

Financial innovation itself represents yet another source of systemic

vulnerability. Newer, less liquid and highly concentrated markets frequently lack the

legal, operational and/or risk management infrastructure necessary to withstand

financial shocks.218 Compounding matters, the appropriability of financial innovation

dilutes the incentives of market participants to invest in the development of such

infrastructure.219 In the end, financial regulators face the decidedly daunting prospect

of mounting effective responses to these (and other) challenges as, all the while, the

forces of regulatory arbitrage – often in the guise of financial innovation – shift the

ground beneath their feet.

Lurking in the background is one final regulatory challenge: welfare

indeterminacy. Regulators cannot directly observe the preferences of their

constituents, nor do they have any practical means of aggregating these preferences

into a social welfare function.220 Simultaneously, they possess imperfect knowledge

216 Essentially, these interconnections exacerbate informational problems during periods of market distress as financial institutions seek to determine the sources and scope of their potential exposures. Where the informational costs are too great, the resulting uncertainty can lead to panic and the mass withdrawal of liquidity from the financial system; Schwarcz (2009) (n 23); Gorton (n 30), and Caballero and Simsek (n 85). What is more, these interconnections may result in the transmission of financial shocks faster than regulators are able to address them; Schwarcz (2009) (n 23) at 215, citing W. Brian Arthur, “Complexity and the Economy”, Science (April 2, 1999). 217 See Soros (n 110) at 23. 218 Gubler (n 85) at 15. 219 Hu (n 29) at 1482. Indeed, such under-investment is part of a broader issue stemming from the fact that financial stability is, in effect, a public good. 220 Indeed, many critics of welfare economics have gone so far as to suggest that the concept of social welfare is both logically incoherent and inherently contested; see Timothy Besley, Principled Agents? The Political Economy of Good Government (Oxford University Press, Oxford, 2006) at 21. Perhaps most notably, the assumption that the aggregation of individual utilities or preferences into a social welfare function is in fact possible has been challenged by Kenneth Arrow; see Kenneth Arrow, “A


of (exogenous) future events and the (endogenous) welfare consequences of their

policy choices.221 These blind spots limit the ability of regulators to evaluate the net

welfare effects of, inter alia, (1) existing financial institutions, instruments and

markets; (2) existing regulation; (3) financial innovation, or (4) contemplated

regulatory intervention. It is impossible to know with any real certainty, for example,

whether the net social costs of taxpayer funded bailouts for the financial institutions

at the epicenter of the GFC exceed those which would have resulted from the

economic turmoil which these bailouts likely averted222; whether the systemic

benefits flowing from the implementation of the Basel III capital adequacy

framework will outweigh any attendant costs in terms of lost economic growth223, or

whether the benefits of OTC derivatives stemming from more complete markets,

enhanced price discovery and improved market liquidity exceed the costs arising

from inefficient contracting, opportunistic behavior and potential systemic risks.

What is certain, however, is that this welfare indeterminacy represents a significant

regulatory challenge.

Difficulty in the Concept of Social Welfare” (1950), 58:4 J. of Pol. Econ. 328. Arrow argued that the task of aggregating individual preferences is “plagued by the difficulties of interpersonal comparison.”; Ibid. at 329. Under certain specified conditions, Arrow illustrated that a rational paradox could result from the aggregation of the preferences of as few as two individuals faced with as few as three potential states, thus precluding the construction of a social welfare function. For a discussion of the unrealistic nature of many of the assumptions underpinning Arrow’s analysis, see Awrey (n 13) at 52. 221 Indeed, we do not even know with certainty which future events are exogenous and which are endogenous. Furthermore, even if we could determine the net welfare effects of a given policy choice at a particular moment in time, there is no guarantee that it would be representative of the net effects at any other moment. 222 Although this has not stopped scholars from attempting to quantify these costs; see Pietro Veronesi and Luigi Zingales, “Paulson’s Gift” (2009), available at 223 Although, once again, this has not stopped various observers from attempting to quantify these costs; see for example Patrick Slovik and Boris Cournède, “Macroeconomic Impact of Basel III”, OECD Economics Department Working Paper No. 844 (February 2011), available at www.oecd-; Douglas Elliott, “Basel III, the Banks, and the Economy”, The Brookings Institution (July 2010), available at; The Institute of International Finance, “Interim Cumulative Effect Report” (June 2010), available at, and Douglas Elliott, “Quantifying the Effects on Lending of Increased Capital Requirements”, The Brookings Institution (September 2009), available at


The common theme running through this inventory of regulatory challenges is

the existence of pervasive, acute and often deeply entrenched asymmetries of

information and expertise within modern financial markets. These twin asymmetries

– exacerbated, if not always caused, by complexity and financial innovation – can be

observed both within the marketplace itself and, importantly, between market

participants and regulators. These asymmetries have combined to make the entire

financial system increasingly reliant on a relatively small oligopoly of intermediaries

which serve as the repositories and purveyors of this information and expertise. As

made all too clear by the economic turmoil unleashed by the GFC, the nature and

extent of this reliance has generated what can fairly be described as the mother of all

agency cost problems.

VI. OTC Derivatives Regulation in the Wake of the GFC: A Brave New World

Prior to the GFC, the approach adopted toward OTC derivatives regulation in

jurisdictions such as the U.S. and U.K. – which account for the vast majority of global

trading activity224 – can perhaps best be described as ‘non-interventionist’.225 Swaps

markets effectively (if not at all times legally) fell outside the perimeter of securities 224 As of April 2010, for example, these two jurisdictions accounted for roughly 70% of global turnover in OTC interest rate derivatives and 55% of the global turnover in OTC foreign exchange derivatives; BIS, Triennial Central Bank Survey of Foreign Exchange and Derivatives Markets Activity in April 2010 (Sept. 2010) at 5-6, available at Other jurisdictions with a meaningful share of global turnover in these instruments include Japan (6% of OTC foreign exchange derivatives and 3% of OTC interest rate derivatives), Singapore (5% and 3%) and Switzerland (5% and 3%); ibid. While reliable comparable data for equity, credit and commodity-linked derivatives is more difficult to come by, the available data suggests a similar (if not greater) degree of geographic concentration within these market segments; Duffie and Hu (n 163) at 12-16. 225 It is worth briefly noting that a handful of observers have suggested that, despite appearances, the U.S. Federal Reserve Board and other federal banking regulators actually played a robust oversight role in respect of OTC derivatives; see for example, Schuyler Henderson, Henderson on Derivatives, 2nd ed. (LexisNexis, London, 2010). Ultimately, however, these observers downplay (or altogether ignore) the myriad of ways in which these regulators systematically relaxed the regulatory rules surrounding these instruments in the decades leading up to the GFC. For a survey of these actions, see Awrey (n 104); Saule Omarova, “The Quiet Metamorphosis: How Derivatives Changed the ‘Business of Banking’” (2009), 63 U. Miami L. Rev. 1041, and Saule Omarova, “From Gramm-Leach-Bliley to Dodd- Frank: The Unfulfilled Promise of Section 23A of the Federal Reserve Act” (2011), 89 North Carolina L. Rev. 1683. More fundamentally, these observers seemingly fail to appreciate the rather obvious point that U.S. banking regulators do not enjoy jurisdiction over global markets.


and futures regulation in both jurisdictions.226 ABS, CDOs and other securitizations,

meanwhile, were frequently offered under exemptions from the prospectus,

registration and other requirements imposed under applicable securities laws.227 This

non-interventionist approach was shaped by the prevailing free market ideology

which viewed market participants as invariably best positioned to address the risks

arising in connection with OTC derivatives.228 It was also influenced by mounting

competitive pressures within the increasingly global market for investment banking

services.229 Ultimately, however, this approach effectively disregarded the risks and

regulatory challenges generated by complexity and financial innovation. The

$USD700 trillion dollar question thus becomes: what lessons, if any, have

policymakers taken away from the GFC?

The frenzied and destructive events of March-September 2008 spurred

policymakers on both sides of the Atlantic to fundamentally reevaluate their

approaches toward the regulation of OTC derivatives markets.230 This ‘rethink’ was

226 See Awrey (n 104) for a detailed description of the pre-crisis regulatory treatment of swaps in both the U.S. and U.K. 227 In the U.S., for example, exemptions could be obtained under sections 3(a)(2) and 4(2) of the Securities Act of 1933, codified at 15 U.S.C. § 77a (1933) [the “Securities Act”] and sections 3(c)(1) and 3(c)(7) of the ICA. Very briefly, Section 3(a)(2) provides an exemption for securities issued by federally regulated banks and savings and loan associations. Section 4(2) provides an exemption for transactions not involving a public offering of securities. Section 3(c)(1) provides an exemption where the beneficial holders of outstanding securities number less than 100 at any time. Section 3(c)(7), meanwhile, provides an exemption where the issuer does not make a public offering and the securities are owned by certain qualified purchasers (i.e. those meeting a prescribed income or asset test). The SEC would subsequently expand the available exemptions through the promulgation of Rule 144A under the Securities Act (adopted in 1990) and Rule 3a-7 under the ICA (adopted in 1992). 228 The influence of this ideology is most clearly visible in connection with the Congressional hearings leading up to the enactment of the Commodity Futures Modernization Act of 2000, Pub. Law No. 106- 554, 114 Stat. 2763 (2000) [“CFMA”] which, inter alia, prohibited federal securities and futures regulators from regulating OTC derivatives markets. 229 See Duffie and Hu (n 163) and the Bloomberg Report (n 97). 230 This shift began (modestly enough) in March 2008 – in the immediate aftermath of the Bear Stearns bailout – when the CFTC and SEC entered into a mutual cooperation agreement with a view to enhancing coordination and facilitating the review of new derivatives instruments; see CFTC Press Release 5468-08, “CFTC, SEC Sign Agreement to Enhance Coordination, Facilitate Review of New Derivatives Products” (March 11, 2008), available at Then, in November, the CFTC,


motivated by two principal observations. First, when the chips were down, the size,

technological sophistication, opacity, interconnectedness and fragmentation of OTC

derivatives markets – in short, their complexity – meant that nobody knew with any

certainty where or how big the counterparty credit (and thus systemic) risks were.

Second, bilateral risk management – i.e. privately negotiated collateral and netting

arrangements – had not effectively mitigated these risks. Manmohan Singh, for

example, has estimated that as of 2008 bilateral swap markets were under-

collateralized by as much as $USD2 trillion.231 Perhaps most importantly, prevailing

market practice dictated that intra-dealer exposures were often entirely


On March 4, 2009, the European Commission announced its commitment to

implement reforms designed to increase transparency and reduce systemic risk within

OTC derivatives markets.233 This commitment would eventually be met in the form

of the E.U. Regulation on OTC Derivatives, Central Counterparties and Trade

Repositories (or EMIR234), adopted on September 15, 2010.235 The U.S. Treasury

SEC and Federal Reserve Board entered into a memorandum of understanding to establish a framework for consultation and information sharing on regulatory issues related to central counterparties for CDS contracts; see _2000s.html. Shortly thereafter, the CFTC announced that the CME had certified a proposal to clear CDS through the CME’s clearing facilities; see CFTC Press Release 5592-08, “CFTC Announces that CME Has Certified a Proposal to Clear Credit Default Swaps” (December 23, 2008), available at 231 As measured by derivatives payables; see Manmohan Singh, “Collateral, Netting and Systemic Risk within OTC Derivatives Markets”, IMF Working Paper 10/99 (2010), available at See also Manmohan Singh and James Aitken, “Counterparty Risk, Impact on Collateral Flows and Role for Central Counterparties”, IMF Working Paper 09/173 (2009) and Miguel Segoviano Basurto and Manmohan Singh, “Counterparty Risk in the Over-The-Counter Derivatives Market”, IMF Working Paper 08/258 (2008), both available at 232 Singh (2010) (n 231) at 7. 233 See European Commission, “Driving European Recovery”, COM(2009) 114 (March 4, 2009) at 7, available at See also European Commission, “Ensuring Efficient, Safe and Sound Derivatives Markets: Future Policy Actions”, COMM(2009) 563 (October 20, 2009); European Commission, “Ensuring Efficient, Safe and Sound Derivatives Markets”, COMM(2009) 332 (July 3, 2009), and European Commission Staff Working Paper Accompanying COMM(2009) 332 (July 3, 2009) SEC(2009) 905, all available at 234 Which stands for the ‘European Market Infrastructure Directive’.


Department, meanwhile, was also eager to signal its enthusiasm for a new approach:

unveiling the draft Over-the-Counter Derivatives Markets Act in August 2009.236

These reforms would ultimately be enacted in July 2010 as part of the Dodd-Frank

Wall Street Reform and Consumer Protection Act.237

A. The U.S. Regulatory Response

The Obama Administration has characterized the objectives of the new U.S.

regime as to: (1) guard against the build-up of systemic risk; (2) promote transparency

and efficiency; (3) thwart market manipulation, fraud, insider trading and other abuse,

and (4) prevent inappropriate marketing to unsophisticated counterparties.238 Title

VII of the Dodd-Frank Act employs four primary mechanisms in pursuit of these

objectives.239 First, it confers upon the CFTC and SEC the authority to mandate that

financial instruments falling within the definition of either a “swap” or “security-

based swap”240 be centrally cleared through CFTC-regulated derivatives clearing

organizations or SEC-regulated securities clearing agencies (collectively, CCPs).241

235 SEC(2010) 1058 and 1059 (September 15, 2010). EMIR is not scheduled to come into full force and effect until December 31, 2012. The U.K. is obligated under E.U. law to implement EMIR. 236 See U.S. Department of the Treasury Press Release TG-261, “Administration’s Regulatory Reform Agenda Reaches New Milestone: Final Piece of Legislative Language Delivered to Capitol Hill” (August 11, 2009), including the proposed text of the Over-the-Counter Derivatives Markets Act of 2009. 237 While Title VII of the Dodd-Frank Act (governing OTC derivatives) technically came into force on July 16, 2011, the effective date of the vast majority of the contemplated reforms has been delayed pending the completion of the requisite rulemaking process. Each of these reforms will take effect 60 days following the publication of the relevant final rule; Dodd-Frank Act, s. 754. 238 Treasury Department (n 236). 239 Not including (1) the ‘push out’ of (most) derivatives activities conducted by federally insured banks to separate non-bank affiliates; see Dodd-Frank Act, s. 716 or (2) the so-called ‘Volcker Rule’ limiting the proprietary trading activities of bank holding companies; ibid., s. 619. 240 Taken together, the definitions of swap and security-based swap encompass the vast majority of OTC derivatives instruments; see ibid., s. 721 and 761. That said, the dividing line between swaps and security-based swaps is not altogether clear under the Dodd-Frank Act, especially with respect to swaps based on a portfolio of assets, such as those which often form the subject matter of structured finance transactions. 241 Dodd-Frank Act, s. 723 and 763. Unless otherwise indicated, all subsequent references to “swap” shall, for the purposes of this description of the operative provisions of Title VII of the Dodd-Frank


In very broad terms, CCPs interpose themselves between the counterparties to

bilateral OTC transactions, effectively assuming the obligations of each party to the

other.242 The principle advantage of centralized clearing and settlement through CCPs

is the potential mitigation of both counterparty credit and systemic risk via the (1)

multilateral netting of exposures243; (2) collateralization of residual net exposures244;

(3) enforcement of robust risk management standards245, and (4) mutualization of

losses resulting from the failure a clearing member.246 Simultaneously, of course,

CCPs concentrate counterparty credit – and thus systemic – risk.

The Dodd-Frank Act contemplates an exemption from the clearing

requirement if one of the counterparties (1) is not a “financial entity”; (2) is using the Act, be construed so as to include a “security-based swap”. The process for determining whether a particular group, category, type or class of swap be will subject to the central clearing and exchange- trading requirements can be initiated by either a CCP or the relevant regulator; ibid., s. 723(a)(3). CCPs are required to submit to the CFTC or SEC, as applicable, “any group, category, type, or class of [security-based] swap” it intends to accept for clearing and provide notice of this submission to its members; ibid. In reviewing a submission, the CFTC or SEC will determine whether the submission is consistent with the core principles of the relevant CCP; ibid. The relevant regulator is also required to take into account the following factors: (1) “the existence of significant outstanding notional exposures, trading liquidity, and adequate pricing data”; (2) “the availability of a rule framework, capacity, operational expertise and resources, and credit support infrastructure to clear the contract on terms that are consistent with the material terms and trading conventions on which the contract is then traded”; (3) “the effect on the mitigation of systemic risk, taking into account the size of the market for such contract and the resources of the CCP available to clear the contract”; (4) “the effect on competition, including appropriate fees and charges applied to clearing”, and (5) “the existence of reasonable legal certainty in the event of the insolvency of the relevant CCP or one or more of its clearing members with regard to the treatment of customer and swap counterparty positions, funds, and property.”; ibid. 242 Duffie et. al (n 163) at 5. As Duffie and his co-authors explain, a “CCP stands between two original counterparties as the seller to the original buyer, and as the buyer to the original seller.”; ibid. See also BIS and the Technical Committee of the International Organization of Securities Commissions (IOSCO), “Guidance on the Application of the 2004 CPSS-IOSCO Recommendations for Central Counterparties to OTC Derivatives CCPs”, Consultative Report (May 2010) at 1, available at 243 Multilateral netting involves eliminating offsetting or redundant positions via, inter alia, the utilization of portfolio compression or so-called ‘tear up’ procedures. 244 Effectively creating a first loss position which serves as a capital buffer in the event of counterparty default. 245 By prescribing rules respecting, for example, capital; initial and variation margin; collateral; position portability; segregation of client assets, and stress testing. 246 See IMF, “Global Financial Stability Report: Meeting New Challenges to Stability and Building a Safer System” (April 2010) at 97, available at and BIS, “New Developments in Clearing and Settlement Arrangements for OTC Derivatives”, Committee on Payment and Settlement Systems Publication No. 77 (March 2007), available at


instrument to “hedge or mitigate commercial risk”, and (3) provides prescribed

information to the relevant regulator respecting how it meets its financial obligations

in connection with bilaterally cleared swaps.247 For the purposes of this commercial

end-user exemption, a financial entity includes a swap dealer248, major swap

participant249, and certain other identified classes of financial institution.250 In order

to incentivize greater utilization of centrally-cleared derivatives, it is likely that the

new regime will ultimately impose higher capital and margin requirements on both

swap dealers and major swap participants in connection with bilaterally cleared


Second, the Dodd-Frank Act gives regulators the authority to require that any

swap subject to the clearing requirement also trade on a regulated board of trade,

247 Dodd-Frank Act, s. 723(a)(3). The non-financial or hedging counterparty retains the option to require that the instrument be centrally cleared; ibid. 248 Section 721(a) of the Dodd-Frank Act and section 3(a)(71) of the Exchange Act define a swap dealer as: “any person who—(i) holds itself out as a dealer in [security-based] swaps; (ii) makes a market in [security-based] swaps; (iii) regularly enters into [security-based] swaps . . . ; or (iv) engages in any activity causing the person to be commonly known in the trade as a dealer or market maker in [security-based] swaps”. This definition does not include a person who enters into swaps for their own account (or in a fiduciary capacity), but does not do so as part of a regular business; ibid. 249 Section 721(a) and 761(a) of the Dodd-Frank Act define a major swap participant as: “any person who is not a [security-based] swap dealer and—(i) maintains a substantial [net] position in swaps for any of the major swap categories as determined by the [relevant regulator], excluding (I) positions held for hedging or mitigating commercial risk. . . (ii) whose outstanding swaps create substantial counterparty exposure that could have serious adverse effects on the financial stability of the United States banking system or financial markets.” The definition also includes a financial institution falling under the definition of financial entity as set out in the Dodd-Frank Act that is (1) highly leveraged; (2) not subject to capital requirements, and (3) maintains a substantial net position in outstanding swaps for any of the major swap categories as determined by the relevant regulator; ibid. The definition of a “substantial position” is left to be defined by the relevant regulators; ibid. 250 See Dodd-Frank Act, s. 723(a)(3). 251 See Treasury Department (n 236). Ultimately, however, the Dodd-Frank Act only mandates that the CFTC, SEC, and federal banking regulators, as applicable, set minimum capital and margin requirements; Dodd-Frank Act, s. 731 and 764. See CFTC, Proposed Rules Respecting Margin Requirements for Uncleared Swaps for Swap Dealers and Major Swap Participants, 76 Fed. Reg. 23,732-23,749 (April 28, 2011) and CFTC, Notice of Proposed Rulemaking Respecting Capital Requirements of Swap Dealers and Major Swap Participants, 76 Fed. Reg. 27,802-27,841 (May 12, 2011), both available at


exchange, or alternative swap execution facility.252 This execution requirement will

not apply, however, where (1) no board of trade, exchange or swap execution facility

makes the swap available to trade or (2) one of the counterparties to the swap falls

within the commercial end-user exemption to the clearing requirement.253 Where

swaps are subject to this execution requirement, the expectation is that this will

enhance price discovery, promote greater market transparency and curb opportunities

for market abuse.

Third, the Dodd-Frank Act requires all swap dealers254, major swap

participants255, CCPs256, swap execution facilities257 and swap data repositories

(SDRs)258 to register with the SEC, CFTC, and/or federal banking regulators. Once

registered, swap dealers and major swap participants are subject to, inter alia, capital;

margin; reporting; recordkeeping, and business conduct requirements.259 CCPs

registered with the CFTC, swap execution facilities and SDRs, meanwhile, are

required to (1) comply with a set of ‘core principles’ and other requirements and (2)

design, implement, monitor, and enforce technical regulation in furtherance of these

252 Dodd-Frank Act, s. 723 and 763. Section 721(a) defines a swap execution facility as “a trading system or platform in which multiple participants have the ability to execute or trade swaps by accepting bids and offers made by multiple participants in the facility or system”. 253 Ibid. 254 Ibid., s. 731 and 764. 255 Ibid. 256 Ibid., s. 725. 257 Ibid., s. 733 and 763. 258 Ibid., s. 728 and 763. An SDR is a centralized registry that maintains a database of transaction records. SDRs may also manage trade life-cycle events and downstream trade processing services; see BIS and IOSCO (n 242) at 1. 259 Ibid., s. 731 and 764. The capital and margin requirements will only apply in respect of bilaterally cleared swaps. The corresponding requirements for centrally cleared swaps will be set by the relevant CCP; ibid. Section 737 also contemplates that the CFTC may set position limits (excluding bona fide hedges) for swaps that perform or affect a significant price discovery function with respect to registered entities; see also, CFTC, Proposed Rules Respecting Position Limits for Derivatives, 76 Fed. Reg. 4752-4777 (January 26, 2011), available at


principles.260 While the Dodd-Frank Act does not articulate a similar set of core

principles for CCPs registered with the SEC, it does mandate that the two agencies

adopt consistent and comparable rules governing these registrants.261

Finally, the Dodd-Frank Act imposes extensive recordkeeping and reporting

requirements on these new registrants. Swap counterparties are required to report all

centrally and bilaterally cleared swaps to an SDR.262 SDRs, CCPs and swap

execution facilities are then obligated to provide granular counterparty and transaction

information to the relevant regulators.263 These regulators are, in turn, required to

publically disseminate anonymized transaction and pricing data on a “real time”

basis.264 This public reporting requirement is explicitly designed to enhance price

discovery.265 More broadly, these requirements are designed to leverage the

centralization of transaction data within SDRs, CCPs, swap execution facilities and

other institutions with a view to generating greater market transparency and, as a

consequence, enabling regulators to more effectively monitor the location, nature and

extent of potential systemic risks.266

The Dodd-Frank Act carves up jurisdiction over bilateral OTC derivatives on

the basis of a distinction between (1) contracts for the sale of a commodity for future

delivery and swaps (subject to CFTC jurisdiction) and (2) security-based swaps 260 Ibid., s. 725, 728, 733 and 763. 261 Ibid., s. 712(a)(7). 262 Ibid., s. 727, 729 and 766. These provisions set out rules respecting which counterparty is required to report the swap. In the circumstance where no SDR will accept the swap, it must be reported directly to the relevant regulator; ibid., s. 729 and 766. Notably, this reporting obligation also applies to swaps entered into prior to the enactment of the Dodd-Frank Act; ibid. 263 Ibid., s. 725, 728 and 733. 264 Ibid., s. 727. For the purpose of these requirements, reporting on a “real time” basis refers to reporting within a time frame which is “technologically practicable”; ibid. 265 Ibid. 266 See IMF (n 246) at 105–106.


(subject to SEC jurisdiction).267 Simultaneously, however, it mandates consistency

and comparability between SEC and CFTC rules and regulations governing

functionally or economically similar products and registrants.268 To this end, the SEC

and CFTC have been handed joint responsibility for fleshing out the innumerable

technical details of the new regime.269 The two agencies are thus currently engaged in

the monumental task of issuing proposed and final rules respecting, inter alia, the

process by which regulators determine whether a swap will be subject to the clearing

requirement270; risk management and business conduct standards for CCPs, SDRs,

swap dealers and major swap participants271; margin and capital requirements for

swap dealers and major swap participants272, and ownership limitations and

267 Dodd-Frank Act, s. 712, 722 and 761–763. 268 Ibid., s. 712(a). 269 Ibid., s. 712(d)(1). Including the definitions of swap; security-based swap; swap dealer; security- based swap dealer; major swap participant; major security-based swap participant, and eligible contract participant; ibid. The Obama Administration has requested and received a joint plan for harmonizing the regulation of OTC derivatives markets; see CFTC and SEC, Joint Report of the SEC and the CFTC on Harmonization of Regulation (October 16, 2009), available at 270 See CFTC, Final Rule Respecting the Process for Review of Swaps for Mandatory Clearing, 76 Fed. Reg. 44,464-44,475 (July 26, 2011), available at and SEC Release No. 34-63557, Proposed Rule Respecting the Process for Submissions for Review of Security-Based Swaps for Mandatory Clearing and Notice Filing Requirements and Notice Filing Requirements for Clearing Agencies; Technical Amendments to Rule 19b-4 and Form 19b-4 Applicable to All Self-Regulatory Organizations, 75 Fed. Reg. 82,490-82,533 (December 15, 2010), available at 271 See for example, CFTC, Final Rule Respecting Derivatives Clearing Organization General Provisions and Core Principles, 76 Fed. Reg. 69,334-69,480 (November 8, 2011); CFTC, Notice of Proposed Rulemaking Respecting Information Management Requirements for Derivatives Clearing Organizations, 75 Fed. Reg. 78,185-78,197 (December 15, 2010); CFTC, Proposed Rules Respecting Business Conduct Standards for Swap Dealers and Major Swap Participants With Counterparties, 75 Fed. Reg. 80,638-80,663 (December 22, 2010); CFTC, Notice of Proposed Rulemaking Respecting Swap Data Repositories, 75 Fed Reg. 80,898-80,945 (December 23, 2010), and CFTC, Notice of Proposed Rulemaking Respecting Core Principles and Other Requirements for Swap Execution Facilities, 76 Fed. Reg. 1214-1259 (January 7, 2011), all available at See also, SEC Release 34-63347, Proposed Rule Respecting Security-Based Swap Data Repository Registration, Duties and Core Principles, 75 Fed. Reg. 77,306-77,377 (November 19, 2010); SEC Release 34-63845, Proposed Rule and Interpretation Respecting Registration and Regulation of Security-Based Swap Execution Facilities, 76 Fed. Reg. 10,948-11,070 (February 2, 2011), and SEC Release 34-64766, Proposed Rules Respecting Business Conduct Standards for Security-Based Swap Dealers and Major Security-Based Swap Participants, 76 Fed. Reg. 42,396-42,459 (June 29, 2011), each available at 272 Ibid.


governance requirements for CCPs, designated contract markets, exchanges and swap

execution facilities.273

The Dodd-Frank Act also seeks to enhance the regulation of ABS and other

securitizations – including, importantly, those offered under exemptions from the

prospectus and registration requirements under the Securities Act.274 First, it requires

issuers of ABS and other securitizations to disclose information respecting the quality

of the assets backing each tranche or class of security.275 Where necessary for

investors to perform independent due diligence, issuers must also disclose more

detailed asset or loan-level data.276 Second, it requires “securitizers”277 to disclose

fulfilled and unfulfilled repurchase requests across all trusts aggregated by the

securitizer.278 Third, it compels credit rating agencies to include information in their

rating reports respecting the representations, warranties and enforcement mechanisms

available to investors in connection with a securitization and, importantly, how these 273 See CFTC, Notice of Proposed Rulemaking Respecting Governance Requirements for Derivatives Clearing Organizations, Designated Contract Markets, and Swap Execution Facilities; Additional Requirements Regarding the Mitigation of Conflicts of Interest, 76 Fed. Reg. 722-737 (January 6, 2011), available at; see also SEC Release 34-64017, Proposed Rule Respecting Clearing Agency Standards for Operation and Governance, 76 Fed. Reg. 14,472-14,539 (March 3, 2011) and SEC Release 34-64018, Ownership Limitations and Governance Requirements for Security-Based Swap Clearing Agencies, Security-Based Swap Execution Facilities, and National Securities Exchanges with Respect to Security-Based Swaps under Regulation MC, 76 Fed. Reg. 12,645-12,648 (March 3, 2011), both available at 274 Section 943 of the Dodd-Frank Act introduced s. 3(a)(77) of the Exchange Act, which defines an “asset-backed security” as a fixed income or other security collateralized by any type of self-liquidating financial asset that allows the holder of the security to receive payments which depend primarily on the cash flows from that asset. Notably, the definition expressly includes both CDOs and CDO-squared. 275 Ibid., s. 942(b). The Dodd-Frank Act then requires the SEC to adopt regulations prescribing the specific format and content of these disclosures; ibid. 276 Ibid. 277 The Dodd-Frank Act defines a securitizer as (1) an issuer of an ABS or other securitization or (2) a person who organizes and initiates an ABS transaction by selling or transferring assets, either directly or indirectly, to the issuer; ibid., s. 941(b). 278 Ibid., s. 943(2). The Dodd-Frank Act characterizes the objective of this provision as to make it easier for investors to identify asset originators with clear underwriting deficiencies; ibid. This obligation only applies, however, where the transaction documentation contains a covenant to repurchase an asset; see SEC Rule 15Ga-1 and SEC Release 33-9175, Final Rules Respecting Disclosure for Asset-Backed Securities Required By Section 943 of the Wall Street Reform and Consumer Protection Act, 76 Fed. Reg. 4489-4515 (January 26, 2011), available at


provisions differ from other offerings of similar securities.279 Finally, it imposes risk

retention requirements on securitizers: mandating that, in certain prescribed

circumstances280, they maintain at least 5% of the credit risk in connection with any

assets they sell into a securitization.281, 282 As with the new regime governing swaps,

the securitization provisions of the Dodd-Frank Act contemplate substantial post-

enactment rulemaking.283

B. The European Regulatory Response

The scope and substantive requirements of the new European regime are

broadly consistent with Title VII of the Dodd-Frank Act.284 EMIR mandates that all

“eligible”285 OTC derivatives between “financial counterparties”286 be cleared and

settled through a CCP.287 This mandatory clearing requirement also applies to non-

financial counterparties whose derivatives positions – excluding those objectively

279 Dodd-Frank Act, s. 943(1) and SEC Rule 15Ga-1. 280 Specifically, the risk retention requirements may be reduced where the underwriting standards employed by the originator indicate that those assets manifest less credit risk. In addition, these requirements do not apply in respect of ABS collateralized exclusively by certain “qualified residential mortgages”; ibid., s. 941(b). 281 Ibid. 282 These risk retention requirements must also be viewed in conjunction with Basel III which, when effective, will impose more conservative capital requirements in respect of some securitization exposures. For an overview of these requirements, see Standard & Poors, “Tougher Capital Requirements Under Basel III Could Raise the Costs of Securitization” (November 17, 2010), available at 283 The OCC, Federal Reserve Board, FDIC and SEC are responsible for promulgating regulation in respect of the risk retention requirements. The SEC, meanwhile, is responsible for adopting regulation in respect of the disclosure requirements. 284 Although, as will be explored in greater detail below, there is considerable scope for substantive divergence. 285 Much like the new U.S. regime, EMIR establishes a process for determining whether an instrument is eligible for centralized clearing. This process can unfold in one of two ways. The first way is a ‘bottom-up’ process, pursuant to which a CCP applies to the European Securities and Markets Authority (ESMA) for a determination; Art 4(1). The second ‘top-down’ process involves ESMA, in conjunction with the European Systemic Risk Board (ESRB), determining that a contract should be subject to the mandatory clearing requirement; Art. 4(5). 286 A financial counterparty is defined as including a bank; investment bank; insurance company; UCITS fund; pension fund, or alternative investment fund manager; ibid., Art. 2(6). 287 Ibid., Art . 3.


linked to the counterparty’s commercial activities – exceed a prescribed threshold.288,

289 Both financial and non-financial counterparties entering into OTC derivatives not

subject to the mandatory clearing requirement, meanwhile, are required to hold

“appropriate and proportionate”290 capital and ensure that they have put in place

appropriate procedures and arrangements to “measure, monitor and mitigate

operational and credit risk”.291

EMIR also establishes a uniform authorization requirement for CCPs.292

While these CCPs will continue to be registered and supervised at the national level,

the Regulation empowers ESMA to develop technical standards and to ensure the

uniform and objective application of these standards across the E.U.293 To this end, it

imposes organizational and conduct of business requirements on CCPs respecting,

inter alia, initial capital294; governance295; ownership296; access297; transparency298;

288 There are actually two thresholds: an information threshold and a clearing threshold. Non-financial counterparties exceeding the information threshold are required to report the details of any OTC derivatives instrument to a trade repository; ibid., Arts. 6(1) and 7(1). Non-financial counterparties exceeding the clearing threshold are subject to the mandatory clearing requirement; ibid., Art. 7(2). Instruments that are objectively ascertained to be linked to a non-financial counterparty’s commercial activities will not be taken into account in determining whether the counterparty has exceeded the clearing threshold; ibid., Art. 3(4). ESMA and the ESRB have been handed primary responsibility for articulating the substance of both thresholds no later than June 30, 2012; ibid., Art. 7(3). 289 It is not clear on the face of this provision how transactions between a financial and non-financial counterparty not exceeding either the information or clearing tests would be treated. If EMIR is to be consistent with Title VII of the Dodd-Frank Act, however, such transactions should be exempt. 290 EMIR, Art. 8(1). The European Commission is empowered under EMIR to adopt technical regulation specifying the amount of capital necessary to comply with Art. 8(1); ibid., Art. 8(2). 291 Ibid. 292 Ibid., Art . 10. CCPs, derivatives exchanges and alternative execution facilities are already subject to E.U. regulation under MiFID. The E.U. has launched a consultation which is seeking to, inter alia, determine how MiFID should be updated to reflect emerging trends in this area; see E.U. Press Release IP/10/1677, “Financial Services: Improving European Rules for a More Robust Framework for All Financial Actors and Instruments” (December 8, 2010), available at 293 See Explanatory Memorandum accompanying EMIR at 9. 294 All CCPs are required to have permanent, available and separate capital of at least EUR 5 million; EMIR, Art. 12(1). 295 Ibid., Arts. 24, 25, 26 and 31. These governance requirements contemplate, amongst many other matters: (1) clear separation between the reporting lines for risk management and other operations; (2) remuneration policies designed to support sound risk management; (3) frequent and independent


outsourcing299; segregation300; position portability301, and interoperability.302 It also

imposes prudential requirements respecting, inter alia, margin and collateral

mechanisms303; permitted investments304; default waterfalls, funds and other

procedures305, and risk modeling, stress testing and back testing.306

Lastly, EMIR requires all “trade repositories”307 (TRs) to register with

ESMA.308 It then subjects this new class of registrants to organizational and

operational requirements respecting, inter alia, governance309; access310; information

safeguarding311; transparency312, and data availability.313 Financial counterparties,

audits, and (4) the establishment of an independent risk committee to advise the board of directors on any arrangements that may impact the risk management of the CCP. 296 Ibid., Art. 28. 297 Ibid., Art. 35. Most importantly, CCPs must establish non-discriminatory, transparent and objective criteria for ensuring fair and open access to the CCP. 298 Ibid., Art. 36. Notably, in certain prescribed circumstances, these requirements empower national regulatory authorities to refuse authorization and/or “take other appropriate measures” in response to issues surrounding the identity, influence or holdings of a CCP’s owners. 299 Ibid., Art. 33. 300 Ibid., Art. 37. 301 Ibid. 302 Ibid., Arts. 48, 49 and 50. 303 Ibid., Arts. 39 and 43. 304 Ibid., Art. 44. These requirements are designed to ensure that a CCP will only invest in highly liquid assets to which it enjoys prompt and non-discriminatory access. 305 Ibid., Arts. 40, 42 and 45. These requirements prescribe, inter alia, (1) that a CCP shall maintain a fund to cover losses arising from the default of a clearing member; (2) the order in which the financial resources of a CCP shall be deployed in the event of default, and (3) that a CCP shall have in place procedures to be followed in various default scenarios. 306 Ibid., Art. 46. Specifically, a CCP must regularly review its models and parameters and subject its models to rigorous and frequent stress tests to evaluate their resilience in extreme but plausible market conditions. It must also perform back-tests to evaluate the reliability of the methodology adopted. The results of these tests must be reported to the relevant national authority. 307 Ibid., Art. 1(2). TRs are the E.U. equivalent of SDRs under the Dodd-Frank Act. 308 Ibid., Art. 51. 309 Ibid., Art. 64(1)-(4). 310 Ibid., Art. 64(5). 311 Ibid., Art. 66. 312 Ibid., Art. 67.


along with non-financial counterparties whose derivatives positions exceed a

prescribed information threshold, are required to report all OTC derivatives

transactions to a registered TR.314 TRs are in turn required to make this information

available to both ESMA and the relevant national authorities and to publicly disclose

aggregate derivatives positions broken down by class.315

C. The Post-Crisis Regulatory Response: A Preliminary Assessment

On the surface, the Dodd-Frank Act and EMIR represent a wholesale shift in

terms of the regulation of OTC derivatives markets. But how far do these reforms go

in responding to the risks and regulatory challenges stemming from complexity and

financial innovation? On at least one level, these reforms can be viewed as holding

out considerable promise. The Dodd-Frank Act and EMIR both introduce

mechanisms designed to subsidize the production and dissemination of information

for use by both market participants and regulators. CCPs, for example, can be

understood as simplifying the complex and constantly evolving network of bilateral

derivatives exposures – theoretically making it less costly for end-users, dealers and

regulators to evaluate counterparty credit risk in connection with centrally cleared

swaps.316 SDRs and TRs, meanwhile, will serve as important nodes for the

aggregation and dissemination of derivatives trading data in respect of both centrally

and bilaterally cleared instruments.317 The enhanced disclosure requirements for ABS

313 Ibid. 314 Ibid., Arts. 6 and 7(1). 315 Ibid., Art. 67. 316 In effect by transforming a complex ‘web’ of exposures into a simpler ‘hub and spoke’ network; see Gai et. al. (n 25) at 22-23. 317 Although this will ultimately depend on the type and format (and thus usability) of the information which must be made available to regulators and the public. For a discussion of the relevant issues in this regard, see CPSS/IOSCO, “Report on OTC Derivatives Data and Reporting Requirements: Final Report” (January 2012), available at


and other securitizations under the Dodd-Frank Act are, similarly, a step in the right


Simultaneously, however, considerable work remains to be done to shine a

more powerful light on some of the murkier corners of the global financial system.

Almost two years after the enactment of the Dodd-Frank Act, the Office of Financial

Research – the new federal agency charged with the task of improving the quality of

financial information available to U.S. policymakers – has yet to produce any

meaningful research or market data.318 More fundamentally, finalizing the legislative

frameworks governing CCPs, SDRs/TRs and other major market participants has

been an extremely slow – and often opaque – process in many jurisdictions.319

Indeed, the projected timeframes for full implementation of these reforms in the U.S.,

Europe and elsewhere (originally slated for December 2012) are now far from

clear.320 Compounding matters, uneven implementation across jurisdictions – in

terms of both timing and substantive content – may actually serve to increase

information costs.321 While progress has been measurable, we are thus still some

distance from realizing the objective of meaningfully reducing information costs

within OTC derivatives markets and, ultimately, leveling the informational playing


318 Having produced only one working paper – a survey of existing quantitative measures of systemic risk – and no actual financial data as of April 2012; see 319 For an overview of the status of these reforms, see FSB, “Overview of Progress in the Implementation of the G20 Recommendations for Strengthening Financial Stability”, Report to G20 Leaders (November 4, 2011) at 2 and 16-18 and FSB, “OTC Derivatives Market Reforms”, 2nd Progress Report on Implementation (October 11, 2011), both available at 320 Ibid. 321 And, of course, raises the prospect of regulatory arbitrage.


Moreover, while timely and comprehensive access to information is

undoubtedly a necessary condition for both optimal private contracting and effective

public oversight, it is by no means sufficient. As soberly illustrated by the collapse of

the U.S. MBS market in 2007-2008322, the subsequent run in the repo market at the

epicentre of Lehman’s demise323, and Robert Bartlett’s event study involving the

derivatives disclosures of Ambac Financial324, the sheer volume of information

available within modern financial markets – combined with the rapid pace of change

– can overwhelm the powerful incentives of even the most sophisticated market

participants. Regulators, likewise, have struggled with what is, in effect, information

overload. As we have seen, this dense “information thicket”325 is rendered even more

impenetrable by other drivers of complexity including, inter alia, technology,

interconnectedness, fragmentation, regulation and reflexivity. Viewed from this

perspective, the marginal benefits of simply generating more information may be

very limited indeed.

One intuitively appealing potential policy response – especially if we believe

that the complexity of modern financial markets contributes to market failure and

other socially suboptimal outcomes – is to enhance the resources, incentives and

expertise of public regulators.326 Thus, for example, we can take steps to ensure that

front-line supervisory agencies such as the SEC and CFTC are better funded and, as a

322 See Gary Gorton, “The Subprime Panic” (2009), 15:1 European Financial Management 10 and Gorton (n 30). 323 See Gary Gorton and Andrew Metrick, “Securitized Banking and the Run on Repo”, NBER Working Paper No. w15223 (August 2009); Gary Gorton, “Slapped in the Face by the Invisible Hand: Banking and the Panic of 2007”, Working Paper (May 2009), and Gary Gorton, “Information, Liquidity and the (Ongoing) Panic of 2007”, NBER Working Paper No. w14649 (January 2009), each available at 324 Bartlett (n 31). 325 Ibid. 326 At least insofar as the anticipated benefits exceed the marginal costs.


corollary, that their sources of funding are sufficiently insulated from undue political

interference.327 We can also re-examine how we compensate supervisory personnel

with a view to both attracting and retaining top talent and better aligning their private

incentives with the pursuit of public regulatory objectives.328

Lamentably, the trajectory of financial regulation in many jurisdictions

appears to be heading in something of the opposite direction. The CFTC’s budget, for

example, has been under almost constant threat from Congressional Republicans since

the Dodd-Frank Act expanded the agency’s mandate to include (joint) oversight of

OTC derivatives markets.329 Moreover, while financial sector compensation practices

have figured prominently in the post-crisis debate330, relatively little time or attention

has been paid to how we compensate the public regulators which oversee this vast,

powerful and socially important industry.331 Given the enormity of the stakes, there

exists a strong case for re-evaluating these (and other) decisions in terms of their

likely impact on both the capacity and incentives of public regulators to effectively

monitor modern financial markets.

327 Admittedly, this is more of a problem in the U.S. (where regulators such as the SEC and CFTC rely on Congress for funding) than in the U.K. (where funding is derived principally from industry levies). 328 Frederick Tung and Todd Henderson, for example, have proposed a compensation structure for bank supervisors which, inter alia, links their compensation to the value of equity and debt in the banks they oversee; see Frederick Tung and Todd Henderson, “Pay for Regulator Performance”, Working Paper (January 16, 2012), available at For a discussion of some of the potential pitfalls of this particular proposal, see ibid. at 61-70. 329 See Shahien Nashiripour, “Tight Budget Set for US Markets Regulator”, The Financial Times (November 16, 2011), available at 330 For a small sampling of this research, see Lucian Bebchuk and Holger Spamann, “Regulating Bankers’ Pay” (2010), 98:2 Georgetown L. J. 247; Lucian Bebchuk, Alma Cohen and Holger Spamann, “The Wages of Failure: Executive Compensation at Bear Stearns and Lehman 2000-2008” (2010), 27 Yale. J. on Reg. 257, and Rüdiger Fahlenbrach and René Stulz, “Bank CEO Incentives and the Credit Crisis” (2011), 99:1 J. of Fin. Econ. 11. 331 With the notable exception of Tung and Henderson (n 328).


Another potential response is to simplify some of the more complex elements

of modern financial markets. David Scharfstein and Adi Sunderam, for example,

have identified a number of potential options for reducing complexity within the U.S.

residential mortgage and MBS markets.332 These options include: (1) limiting the

availability of (or altogether prohibiting) mortgages with ‘risky’ characteristics such

as high loan-to-value ratios, self-financed down-payment assistance, adjustable rates

or negative amortization; (2) prohibiting the securitization of such risky mortgages;

(3) simplifying the capital structures which can be used in connection with

securitization vehicles, and (4) prohibiting the re-securitization of junior tranches of

MBS into CDOs.333 An analogous set of measures for bilaterally cleared swaps

might conceivably include: (1) restrictions on the types of swaps available to non-

financial end-users334; (2) mandating higher (and higher quality) collateral335, and (3)

prohibiting the re-hypothecation of pledged collateral.

Theoretically, regulatory intervention of this kind would serve at least two

purposes.336 First, restrictions on the availability of risky mortgages or more complex

swaps would insulate those with lower tolerances for complexity from the negative

consequences of both their own suboptimal decision-making and the sharp practices

of more sophisticated financial intermediaries. Second, by simplifying the arcane

plumbing of these markets, such measures would reduce information costs for both

332 David Scharfstein and Adi Sunderam, “The Economics of Housing Finance Reform: Privatizing, Reguating and Backstopping and Mortage Markets” (February 2011), available at 333 Ibid. at 40-45. 334 Articulating a clear (and yet un-arbitragable) boundary between permitted and prohibited swaps would obviously be an important and difficult task in connection with the implementation of any such proposal. 335 Indeed, regulators in both the U.S. and Europe have signalled their desire to impose such requirements in connection with the implementation of the Dodd-Frank Act and EMIR, respectively; see n 251, 290 and 291. 336 In addition to the enhanced financial stability posited by Scharfstein and Sunderam (n 332).


market participants (investing in swaps, MBS and CDOs and posting or receiving

these instruments as collateral)337 and regulators (attempting to identify, monitor and

respond to the attendant risks). Ultimately, of course, the welfare implications of

Sharfstein and Sunderam’s and other similar proposals are difficult to evaluate: while

they may serve to reduce information (and agency) costs and promote greater

financial stability, one might also expect them to have an adverse impact on both the

ability of counterparties to effectively hedge risk338 and, more broadly, the flow of

credit to the real economy. Indeed, as we have seen, this welfare indeterminacy is

itself an important contributing factor to the complexity of modern financial markets.

While the Dodd-Frank Act and EMIR can be seen as representing (at the very

least) a marginal improvement over the pre-crisis status quo in terms of responding to

the regulatory challenges generated by complexity, these same regimes effectively

disregard the challenges arising from the nature and pace of financial innovation.

Indeed, in at least one respect, these nascent regimes may actually incentivize socially

suboptimal over-innovation. Specifically, the newly created regulatory dichotomy

between centrally and bilaterally cleared swaps generates two distinct payoff

structures for market participants. This, in turn, invites financial innovation – or,

perhaps more accurately, “faux customization”339 – motivated by the desire to avoid

the marginal costs associated with central clearing. Ultimately, there are any number

of reasons why dealers or other counterparties might find it more advantageous to

utilize bilateral instruments (even after accounting for higher margin and capital

requirements). Post-crisis constraints on the supply of high quality collateral, for 337 Thus potentially ameliorating the adverse selection problem which triggered the run or repo; see Gorton and Metrick (n 323). 338 At least in the case of bilaterally cleared swaps. 339 Sean Griffith, “Governing Systemic Risk: Toward a Governance Structure for Derivatives Clearinghouses” (2012), 61 Emory L. J. 1153 at 1197.


example, have increased the opportunity costs of central clearing relative to the (often

under-collateralized340) bilateral market.341 Along a similar vein, moving

standardized instruments on to CCPs will require dealers to unbundle netted positions

involving both standardized and non-standardized instruments.342 In the end, these

collateral and netting benefits may prove very substantial indeed.

The prospect of faux customization is rendered even more acute by virtue of

the fact that, at present, OTC derivatives dealers enjoy effective control over the

CCPs which, in the vast majority of cases, will make the initial determinations in

terms of a swap’s eligibility for central clearing.343 As Sean Griffith explains: “major

dealers have an incentive to exert governance control to keep clearing eligible

products off of clearinghouses so that they can continue to trade in the higher margin

bilateral market.”344 Importantly in this regard, neither the Dodd-Frank Act nor

EMIR mandate regulatory review of a CCP’s decision that an instrument is ineligible

for central clearing. Compounding matters, one might expect regulators to be

reluctant to overturn a CCP’s eligibility determination out of concern that forcing

instruments on to CCPs could exacerbate systemic risk.345 Indeed, this reluctance

might be reinforced by asymmetries of information and expertise vis-à-vis regulators

and CCPs. In this respect, there appears to be ample room for improvement in terms

340 It is at present unclear whether the margin requirements contemplated under either the Dodd-Frank Act or EMIR would eliminate this arbitrage opportunity. 341 Manmohan Singh and James Aitken, “Deleveraging post Lehman – Evidence from Reduced Rehypothecation”, IMF Working Paper 09/42 (2009), available at; Singh (n 231) at 3-4, and Tracy Alloway, “Financial System Creaks as Loan Lubricant Dries Up”, The Financial Times (November 28, 2011), available at 342 Singh (n 231) at 4 and 8. 343 Both directly through their equity interests in CCPs and indirectly through their ability to re-route order flow; see Sean Griffith, “Incentive Problems in Derivatives Trading: Towards a New Governance Structure for Clearinghouses” (June 1, 2010) [working paper on file with author] at 24-25. 344 Ibid. at 23. 345 Ibid.


of how the Dodd-Frank Act and EMIR address these information and incentive


One possible way of addressing the problem of faux customization would be

to impose a targeted anti-arbitrage rule (or TAAR) on swap dealers and other market

participants. The primary thrust of a TAAR would be to mandate that market

participants obtain regulatory approval as a pre-condition to entering into any ‘new’

or ‘innovative’ species of bilateral swap.346 In order to obtain this approval, the

market participant(s) submitting the application would need to demonstrate that the

innovation responded to a legitimate economic need and not simply the desire to

avoid central-clearing requirements. To minimize the duplication of effort and

expense, the relevant regulatory authority could then issue ‘blanket’ orders

authorizing all other market participants within their jurisdiction to trade in the new


A well designed TAAR would offer two potential benefits. First, it would

alter the anticipated payoffs from regulatory arbitrage: in effect deterring financial

innovation not motivated by a legitimate economic rationale.347 Second, it would

provide an incentive for risk adverse market participants to bring new bilateral

instruments to the attention of regulators with a view to obtaining ‘pre-clearance’ for

their prospective use. A TAAR would thus manifest potentially significant

informational benefits – bringing new innovations within the perimeter of regulation

346 What precisely constituted a ‘new’ or ‘innovative’ swap would of course need to be fleshed out. Here, however, the definition of innovation introduced in Part III – focusing as it does on change as opposed to improvement – would arguably provide a very useful starting point. 347 The question of how to distinguish between faux customization and economically ‘legitimate’ innovation would of course be of central importance in terms of the operation of a TAAR. The key for the present purposes, however, is that the burden of proof in this regard would be on the market participant(s) making the application.


more rapidly than would otherwise be the case – while simultaneously reducing the

deleterious systemic effects of regulatory arbitrage.348

Ultimately, the objective of this paper is not to exhaustively canvas the ways

in which regulators might better respond to the challenges posed by complexity and

financial innovation. Rather, it has been to punctuate the fact that by simply

acknowledging the complexity of modern financial markets and the nature and pace

of financial innovation we can potentially gain a more complete and nuanced

understanding of the problems we face and, hopefully, how we might go about

addressing them. In this respect, this paper should be understood as aspiring to build

the foundations for a broader research agenda examining complexity, innovation and

the regulation of modern financial markets.

VII. Conclusion

Complexity and innovation define modern financial markets. Together, they

also generate a host of pressing regulatory challenges. As we have seen, these

challenges stem from high information costs, deeply entrenched asymmetries of

information and expertise, and the acute agency cost problems these asymmetries

generate. These challenges underscore the necessity (if not sufficiency) of

mechanisms such as CCPs and SDRs/TRs which subsidize the production and

dissemination of information as a means of promoting both more efficient private

contracting and more effective public oversight. They also potentially justify more

radical regulatory intervention with a view to reducing complexity within some of the

348 Ultimately, of course, further analysis is required to ascertain both the feasibility and desirability of a TARR. For a critical analysis of the prospective costs and benefits of a General Anti-Avoidance Rule (or GAAR) in the tax context, see the Aaronson Report, “A Study to Consider Whether a General Anti- avoidance Rule Should be Introduced into the U.K. Tax System” (November 11, 2011), available at


more arcane corners of the global financial system. Simultaneously, these challenges

point to the desirability of regulation capable of responding to the inherent dynamism

of modern financial markets and, more specifically, the nature and pace of financial

innovation. Here, measures such as a well designed TAAR for bilateral swap markets

could potentially help reveal valuable information and deter socially questionable

forms of innovation. In the end, while recent regulatory reforms under the Dodd-

Frank Act and EMIR have arguably gone some distance in addressing these

challenges, considerably more work thus remains to be done before modern financial

markets begin to resemble the perfect markets envisioned by conventional financial