Erik F. Gerding
Draft of September 19, 2009
Disclosure 2.0:
Leveraging Technology to Address “Complexity” and
Information Failures in the Financial Crisis
Erik F. Gerding
Associate Professor
University of New Mexico School of Law
This article advocates leveraging advances in computer software, information systems,
and the Open Source movement to enhance securities disclosure and remedy some of the
information asymmetries that exacerbated the current financial crisis.
This article responds to a critique of mandatory securities disclosure by several legal
scholars (Troy Paredes and Steven Schwarcz) that disclosure overloads investors with too much
information and fails to help investors analyze the complexity of modern financial instruments
and markets.
However, because the financial crisis stemmed in large measure from information
failures, it would be counterproductive to dilute securities disclosure for failing to stop the crisis.
Instead, disclosure must be radically enhanced to enable investors to better analyze:



intricate financial instruments, such as asset-backed securities and derivatives;
the models used to price these instruments, set risk management policies, and
guide trading strategies; and
the complex market interactions of these instruments and trading strategies.
Various technologies can revolutionize disclosure including:





“tagging” assets that underlay asset-backed securities and derivatives to allow
investors to trace which assets affect which financial instruments;
access to underlying data that is aggregated in financial statements;
“open source” risk models;
real-time financial disclosure; and
interactive disclosure that allows users to change certain assumptions or
accounting methods to see how financial disclosure would change.
Employing these technologies poses certain risks and costs, which this article analyzes.
The article argues that an “open source” approach to technologically-enhanced disclosure can
mitigate agency costs and advocates experimental testing of disclosure effectiveness.
1
Erik F. Gerding
Draft of September 19, 2009
1.
Introduction
Recent legal scholarship has continued the tradition of critiquing mandatory disclosure,
but with a modern twist. Rather than argue that disclosure is superfluous, several scholars have
contended that disclosure is either counterproductive given the cognitive limitations of investors
or hopelessly ineffective because the complexity of modern financial instruments prevents
investors from adequately understanding those instruments. Professor Troy Paredes (now a
commissioner of the Securities Exchange Commission (SEC)) has put forth the first half of this
argument; he posits that required securities disclosure overloads investors with too much
information and, therefore, disclosure requirements should be scaled back.1 Professor Steven
Schwarcz has articulated the second half of the argument; he argues securities disclosure should
be deemphasized because disclosure cannot help investors understand the complexity of assetbacked securities and derivatives.2 This article labels these two arguments as the “complexity
critique.” In light of the current financial crisis, in which ordinary citizens still struggle to
understand the complex financial instruments at the heart of the market meltdown, this critique
has a certain counterintuitive seductiveness and could be paraphrased as follows: policymakers
should deemphasize or roll back securities disclosure regulations because disclosure cannot
adequately describe complex financial instruments and, even if it attempted to do so, investors
could not process the information in any event.
But the critique is a little too counterintuitive. The financial crisis stemmed in large
measure from information failures, and it would thus be a grave mistake to dilute securities
disclosure for failing to stop the crisis. Instead, disclosure must be radically rethought and
enhanced to remedy the following information failures revealed by the crisis:
¶
¶
¶
¶
¶
¶
The inability of investors to identify the ultimate assets that underlay asset-backed
securities and derivatives;
The destruction of information on underlying assets as those assets are
securitized;
The additional progressive information destruction as the resultant asset-backed
securities are then re-securitized or hedged with credit default swaps (which
information destruction worsens further when the products of those new
securitizations or swaps are themselves securitized or hedged);
The intermittent disclosure that comes with periodic financial reporting, which
obscures changes in a company’s balance sheet and risk profile between reporting
periods and enables accounting gamesmanship;
The failure of securities disclosure to give granular information on a company’s
investments, which can mask dangerous homogeneity among the portfolios of
various companies;
The opacity of models used to price asset-backed securities and derivatives, set
firm risk management policies, and make credit rating agency ratings, which
1
Troy A. Paredes, Blinded by the Light: Information Overload and its Consequences for Securities Regulation, 81
WASH. U. L.Q. 417 (2003).
2
Steven L. Schwarcz, Rethinking the Disclosure Paradigm in a World of Complexity, 2004 U. ILL. L.REV. 1
[hereinafter, Schwarcz, Rethinking the Disclosure Paradigm]; Steven L. Schwarcz, Disclosure’s Failure in the
Subprime Mortgage Crisis, 2008 UTAH L.REV. 1109 [hereinafter, Schwarcz, Disclosure’s Failure].
2
Erik F. Gerding
Draft of September 19, 2009
¶
opacity hides faulty assumptions, gamesmanship of models, and dangerous
homogeneity in risk management;
The limitations in using firm specific disclosure to uncover systemic problems
such as those created by either excessive homogeneity in investment portfolios,
trading strategies, risk management, and risk exposures or complex market
interactions among the portfolios and trading strategies of separate firms.
The complexity critique is useful less for the deregulatory arguments made by Paredes
and Schwarcz and more for highlighting a deep tension that disclosure requirements must
negotiate. On the one hand, securities disclosure must present financial information in sufficient
detail and nuance to enable investors to understand the complex risks and rewards embedded not
only in complex financial products but even in the plain vanilla securities of modern firms
operating in a dynamic marketplace. On the other hand, disclosure must take into account the
cognitive limitations of investors and the costs they incur in sifting through massive amounts of
information.
Computer science, software, and information systems have long grappled with similar
tensions, and insights from these fields can vastly improve securities disclosure. More
particularly, technological advances in software and information systems can play a key role in
enhancing securities disclosure and disclosure regulations. This article outlines the following
technologies and technological insights that can enhance disclosure:
¶
¶
¶
¶
¶
¶
the SEC’s XBRL initiative that requires firms to include data tags on electronic
disclosure to allow investors to download disclosure items into spreadsheets and
other analytical software;
affixing data tags to assets that underlay asset-backed securities and derivatives to
allow investors to trace which assets underlay which financial instruments;
real-time financial disclosure;
“open source” risk models;
interactive disclosure that allows users to change certain assumptions or
accounting methods to see how financial disclosure would change; and
software designs that improve the look and feel of computer programs through
drag down menus and other devices.
These innovations pose certain risks and costs, which this article analyzes.
This article proceeds as follows: Section 2 outlines the two sides of the complexity
critique articulated by Professors Paredes and Schwarcz. Section 3 offers a critique of the
complexity critique. Section 4 offers a typology of three kinds of complexity that may afflict
complex financial instruments – contractual complexity, derivative complexity, systemic
complexity. Section 5 elaborates on the information failures in the current financial crisis that
disclosure must address. Section 6 provides more details on possible technological enhancements
to disclosure that can remedy those failures. Section 7 outlines potential costs and risks involved
with these technological enhancements. Section 8 concludes by outlining the advantages of
“open technology” and advocating for experimental testing of the effectiveness of new disclosure
technologies in improving the understanding of investors.
2.
The “Complexity Critique”
3
Erik F. Gerding
Draft of September 19, 2009
Mandatory securities disclosure has endured periodic attacks by legal scholars that have
focused on whether required disclosure adds any information that investors and financial markets
cannot demand or otherwise obtain.3 In the past five years, several legal scholars have revived
this critique, but added a novel twist. First, in 2003, Professor Troy Paredes authored a paper
arguing that securities disclosure overloads investors with too much information.4 The cognitive
limitations and behavioral biases of investors and the costs of processing massive amounts of
financial disclosure, Paredes argued, meant that securities disclosure has not only become less
effective in informing the investing public, but also counterproductive.5 Investor decisionmaking can become suboptimal, Paredes contended, as investors take mental shortcuts to sift
through the massive amount of information a securities issuer is required to disclose.6 Investors
inevitably select only a portion of any issuer’s disclosure to process and on which to base their
investment decisions. Often, investors pick the wrong information.7 Paredes recognizes that the
Efficient Markets Hypothesis (EMH)8 suggests that it does not matter that certain individual
investors may be overwhelmed by information; the marketplace collectively will process
information. This is thanks largely to sophisticated investors who drive changes to the market
price of a firm’s securities upon learning new information relevant to that firm. Paredes
responds to the EMH with arguments that sophisticated investors can also face information
overload from disclosure and that behavioral finance has shown that the EMH often does not
apply.9 Paredes concludes from his findings that policymakers should consider pruning
securities regulations on disclosure to reduce information overload and improve investor
decisions.10
Professor Stephen Schwarcz added a different side to this critique in a series of articles
that argued that securities disclosure failed to capture the “complexity” of modern financial
instruments and markets. In 2004, he argued that the Enron fraud revealed a larger failure of
securities disclosure to remedy information asymmetries in securitization, derivatives, and other
structured finance transactions because of the growing complexity of those transactions.11
Schwarcz argued that markets and policymakers could not take comfort from the EMH; the
complexity of structured finance transactions would prevent a critical mass of sophisticated
investors from sufficiently understanding particular transactions and driving market prices to an
efficient level.12 After the financial crisis took root, Schwarcz reiterated this argument in a 2008
article.13 Schwarcz argues that, in light of complexity, securities law relies too heavily on
securities disclosure and that policymakers must explore alternatives, including banning
excessively complex financial products.14
3
Cite.
Paredes, supra note 1.
5
Id. at 441-43.
6
Id. at 440-42.
7
Id. at 442-43.
8
Explanation of EMH to come.
9
Id. at __.
10
Id. at 484
11
Schwarcz, Rethinking the Disclosure Paradigm, supra note 2.
12
Id. at 17-19.
13
Schwarcz, Disclosure’s Failure, supra note 2.
14
Schwarcz considers banning excessively complex structured finance transactions a suboptimal option. Schwarcz,
Rethinking the Disclosure Paradigm, supra note 2, at 20-22; Schwarcz, Disclosure’s Failure, supra note 2, at __.
4
4
Erik F. Gerding
Draft of September 19, 2009
Paredes and Schwarcz’s arguments represent two sides of what this article labels the
“complexity critique” of modern mandatory securities disclosure. In the first side of the critique,
disclosure fails to provide investors with sufficient information on the complex nature of modern
financial instruments (most notably asset-backed securities, derivatives, and other structured
finance products, and derivatives). On the other side, investors may not be able to process
additional information that may be disclosed due to cognitive limitations.
3.
Critiquing the Critique
Both sides of the complexity critique display limitations, raise additional questions, and
demand further refinement. The Paredes critique does not specify how much information is
optimal and how much is overload. Professor Paredes, in fact, admits that empirical evidence
would be necessary to determine which items of financial disclosure are helpful for investors and
which are superfluous.15 Paredes also recognizes that this in turn begs the question of how to
address the demands of different investors for different kinds of information.16
The Schwarcz critique also needs refinement. Schwarcz offers a very general definition
of “complexity.”17 This general definition has the benefit of being comprehensive, but
forecloses the development of both ways to measure complexity and testable hypotheses.
Further definitional clarity would allow empirical evidence to be gathered to evaluate
Schwarcz’s thesis – that complexity of certain financial instruments renders disclosure
ineffective. Several definitions and metrics for complexity are discussed in the following
section.
4.
Types of Complexity
The importance of the EMH and the role that sophisticated investors play in analyzing
disclosure and setting the market price argues for creating a more refined definition of
“complexity” and breaking it into several components. This article argues that asset-backed
securities, derivatives, and other structured products may exhibit at least three kinds of
complexity. First, these financial products may exhibit contractual complexity in that they may
have numerous contractual terms that define the rights of investors under the products. These
terms may include rights to payments and collateral that may vary, in the case of asset-backed
securities, for different tranches of securities of the same issuer. Contractual complexity might
be measured by the number of tranches of securities being issued or the presence or absence of
particular contractual structures.
Second, these products may exhibit derivative complexity in that asset-backed securities,
derivatives, and other structured products derive their value from underlying assets. For
Schwarcz also proposes supplementing disclosure requirements with other regulations, such as requiring originators
in securitization transactions to provide guaranties, developing private sector or government certifications of
complex financial transactions, and enhancing corporate governance rules to mitigate conflicts of interest in those
transactions. Schwarcz, Rethinking the Disclosure Paradigm, supra note 2, at 22-35; Schwarcz, Disclosure’s
Failure, supra note 2, at __.
15
Paredes, supra note 1, at 473-74.
16
Id. at 459-60.
17
Schwarcz, Rethinking the Disclosure Paradigm, supra note 2, at 2, n.3 (“I use the term ‘complexity’ in its
ordinary meaning: the state of being complicated”).
5
Erik F. Gerding
Draft of September 19, 2009
example, mortgage-backed securities pay out according to payments received on the mortgages
that underlay the securities. Often, asset-backed securities may be backed by other securities,
which might in turn be backed by a third layer of securities. This layering of securitization of
securitization has analogues in derivatives, as a counterparty to a derivative contract can hedge
various risks by entering into a second contract. Derivative complexity increases with (i) the
number of layers of assets underlying a particular financial instrument, (ii) the diversity of assets
in each layer, and (iii) the complexity of each of those underlying assets, including the three
kinds of complexity mentioned in this section.
Third, these financial products may exhibit systemic complexity. The value and risks of
any product depend not only on underlying assets (and the value and risks of those assets) and
the contractual terms governing the product, but also changes in the marketplace. Most
basically, the financial instrument may have a market price, which may depend in part on the
market prices of substitute instruments. Potential losses from changes in market price is often
labeled as market risk, but financial instruments are also subject to liquidity risk (the risk that the
number of willing buyers for a particular product will suddenly plummet causing the market to
dry up) and systemic risk (the risk of market-wide losses). Systemic complexity captures all of
these forms of risk, which are difficult to evaluate and model. Systemic complexity increases
when the value of a financial instrument or security is more heavily dependent on the price
movements of many other financial instruments (other than underlying assets).
Of these three forms of complexity, it is most difficult to believe that sophisticated
investors will have trouble understanding contractual complexity. Retail investors may have
difficulty reading through a prospectus and indenture for an asset-backed security, but
sophisticated investors are likely to have considerable experience with the contractual
terminology and structures involved in such a security.
5.
Information Failures in the Global Financial Crisis
Sophisticated investors might face greater difficulty in evaluating financial instruments
with greater derivative or systemic complexity. It does not follow however, that the antidote to
greater complexity is less disclosure. The current financial crisis was exacerbated by a series of
information failures. This section will outline a few of these failures, several of which relate
closely to derivative or systemic complexity, and the next section will discuss how disclosure
could be improved with the aid of technological advances to address these failures.
a.
Securitization, derivatives, and the destruction of information
As noted above, asset-backed securities often derive their value from several layers of
underlying assets. In other words, because asset-backed securities are themselves securitized and
used to back new asset-backed securities, any class of asset-backed securities may be several
securitizations removed from the ultimate underlying cash-producing assets, such as mortgages
or other loans. Similarly, derivatives may be used to hedge the risks of other derivatives. For
example, credit derivatives may be used to hedge the risk of default on asset-backed securities
(or other derivatives). To value asset-backed securities and derivatives investors need
information on the value and risks associated with underlying assets. This becomes complicated
6
Erik F. Gerding
Draft of September 19, 2009
when those values and risks in turn require information on assets underlying those underlying
assets. Investors may have difficulty identifying through layers of securitization and derivatives
the ultimate underlying assets.
There are few substitutes for direct information on those ultimate underlying assets.
Investors cannot necessarily rely on the parties involved in earlier securitizations and derivatives
passing on high quality information. Often these parties lack incentives to pass on accurate and
detailed information. For example, lenders who sold the mortgages and other loans they
originated to be securitized had diminished incentives to ensure the creditworthiness of
borrowers. This created the problem of low-documentation and zero money down loans. Unable
to trace back to the ultimate underlying assets, investors might also resort to shortcuts, such as
relying on the ratings assigned to underlying assets by credit rating agencies. Rating agencies,
however, have a poor track record of predicting default, which many scholars have attributed to
conflicts of interest caused by those firms being paid by the issuers of the securities they are
rating.18
Perverse incentives and the cost of passing on information on underlying assets
contributed to what economists have called the progressive destruction of information with each
layer of securitization or derivatives.19 This information destruction can lead to serious errors in
valuing risk. Economists have also noted that securitizing and re-securitizing assets like
mortgages in multiple layers exacerbates the risk of errors; small errors in measuring the risk of
underlying assets, like mortgages, are magnified at each subsequent level of securitization or
hedging. This can lead to large errors in valuing the financial instruments at the end of a chain of
securitizations and derivatives.20 One particular error merits special attention. When losses on
when assets become highly correlated,21 losses on one asset are no longer offset by continued
profits on other assets; losses then occur in waves.22 When the correlation of losses on
underlying assets exceed expectations even slightly, valuations on the asset-backed securities or
derivatives that are backed by or based on those assets can exhibit spectacular errors. 23 In short,
a failure to identify the ultimate assets that affect an asset-backed security or derivative increases
the likelihood of serious errors in valuing that security or derivative.
b.
Financial Institution Risk Models and Model Risk
18
Frank Partnoy, The Siskel and Ebert of Financial Markets?: Two Thumbs Down for the Credit Rating Agencies,
77 WASH. U. L.Q. 619 (1999); Frank Partnoy & David A. Skeel, Jr., The Promise and Perils of Credit Derivatives,
75 U. CIN. L. REV. 1019, [1029] (2007); Frank Partnoy, How and Why Credit Rating Agencies are Not Like Other
Gatekeepers, in FINANCIAL GATEKEEPERS: CAN THEY PROTECT INVESTORS 59 (Yasuuki Fuchita & Robert E. Litan
eds., 2006).
19
Willem H. Buiter, Lessons from the 2007 Financial Crisis, CENTRE FOR ECONOMIC POLICY RESEARCH POLICY
INSIGHT NO. 18, 2–3 (Dec. 2007) available at http://www.cepr.org/pubs/PolicyInsights/PolicyInsight18.pdf.
20
Joshua Coval, Jakub Jurek, & Erik Stafford, The Economics of Structured Finance, 23 J. ECON. PERSP. 3, _ (2009).
21
Losses on mortgages and securities can become increasingly correlated due to spillover effects (for example, the
default on one mortgage lowering surrounding property values and increasing the probability of default on other
mortgages) and feedback loops (for example, widespread defaults on mortgages constricting lending, which
increases interest rates, which leads to a new round of defaults). Martin Hellwig, “Systemic Risk in the Financial
Sector: an Analysis of the Subprime-Mortgage Financial Crisis,” Max Planck Inst. Res. on Collective Goods Bonn
2008/4316 (Nov. 2008) available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1309442.
22
Erik F. Gerding, Code, Crash, and Open Source: the Outsourcing of Financial Regulation to Risk Models and the
Global Financial Crisis, 84 WASH. L. REV. 127, 172 (2009).
23
Coval, Jurek, & Stafford, supra note 20, at __.
7
Erik F. Gerding
Draft of September 19, 2009
These same errors can afflict the models that various parties (investors, rating agencies,
derivative counterparties, bond insurers, sponsors of securitizations) use to price asset-backed
securities and derivatives. These pricing models exhibited spectacular failures in measuring risk
in the current financial crisis, as did the models that financial institutions relied on to price loans,
manage investment portfolios, set overall firm risk management policies, and even establish
regulatory capital.24 These models failed in part because of faulty assumptions that were largely
hidden from the public due to the opacity of the models. Opacity of risk models has other
negative consequences. It hides potential gamesmanship of models by individuals within a
financial institution. For example, traders in financial institutions have been faulted for engaging
in high risk but low-probability trading strategies to evade detection by their firm’s risk models
(a practice called “stuffing risk into the tails”).25 In addition, the fact that the details of any
financial institution’s risk models remain largely secret prevents other financial institutions from
adequately evaluating their counterparty risk to that institution.26 The marketplace cannot
necessarily rely on regulators to police risk modeling and risk management, as regulators often
lack necessary resources and expertise.27 Moreover, regulators may lack motivation; from an
international perspective, one nation’s regulators may wish to allow their home country financial
institutions to use looser models to take on more risk, earn greater profits, and gain a competitive
advantage over institutions in other countries. Opacity of risk models and regulatory auditing of
those models obscures the failures of regulators and can undermine efforts to ensure minimum
standards for international financial regulation.28
In addition, opacity of risk models hides potentially dangerous levels of homogeneity in
the risk modeling and risk management of financial institutions. If financial institutions employ
similar investment and risk management strategies, they may purchase the same assets at the
same time (causing prices to surge) and sell those assets simultaneously (causing prices to
plummet). There is some indication that this homogeneity exacerbated the current financial
crisis, as many financial institutions rushed to invest in mortgage-backed securities,
collateralized debt obligations, credit default swaps, and similar instruments.29 The surge of
demand created higher returns, which may have further increased demand (much as economists
have argued that noise trading can create its own returns and spur further noise traders to enter
the market in a feedback loop).30 Then when the crisis hit and these instruments began suffering
losses, financial institutions began selling assets at the same time.31
c.
Firm Specific Disclosure and Systemic Exposures
The problem of excessive homogeneity in risk models, risk management, and investment
strategies points to a broader informational problem: investors need to understand systemic risks
24
Gerding, supra note 22, at 164-67. Under the Basel II Accord, certain large banks are now allowed to set their
own regulatory capital requirements according to internal risk models. Id. at 154-57. The SEC extended this
privilege to several large investment banks under its now defunct Consolidated Supervised Entity program. Id. at
157.
25
Joe Nocera, Risk Mismanagement, N.Y. TIMES MAGAZINE, Jan. 2, 2009, at 46.
26
Gerding, supra note 22, at 182-83.
27
Id. at 183.
28
Id. at 183-85.
29
Id. at 184.
30
J. Bradford DeLong et al., Noise Trader Risk in Financial Markets, 94 J. POL. ECON. 703 (1990).
31
Gerding, supra note 22, at 165.
8
Erik F. Gerding
Draft of September 19, 2009
but securities disclosure is firm-specific. For example, investors may want to know if different
financial institutions have concentrated risk exposure to the same borrowers or the same types of
loans or investments. To find this information, investors would have to compare the SEC filings
of multiple institutions. Differences in the way in which firms report financial information
frustrate this comparison. Although regulations on financial statement disclosure are designed to
ensure common yardsticks of financial performance, securities and accounting rules are replete
with examples of disclosure that gives issuers discretion to pick among various methodologies
for calculating financial results [(for example, LIFO or FIFO in valuing inventory)] or disclosure
formats.
d.
The Granularity of Disclosure
In addition, investors seeking to uncover worrisome homogeneity among the investment
portfolios and exposures of different firms need fairly granular information about the specific
assets and liabilities in a firm’s balance sheet. This detailed information is also necessary for
investors to back-test the quality of a firm’s risk modeling (and whether a firm is faithfully
describing its modeling and risk management practices). Disclosure rules that require firms
simply to list borrowers that represent the largest source of credit risk is inadequate if the firm
has numerous smaller loans to multiple borrowers that present similar risk exposures (such as
exotic mortgages which would experience a jump in default rates if interest rates rose and
housing prices stagnated, to pick a fanciful example). Regulations can add requirements to
disclose specific exposures (such as exposure to mortgages or mortgage-backed securities), but
this poses two problems. First, regulators must anticipate the next investment fad that will cause
multiple securities issuers to have similar exposures. Alternatively, regulators could put the onus
on issuers to determine emerging areas of dangerous homogeneity. But this would require firms
to gather information on the specific practices and portfolios of competitors, and the necessary
data may not be publicly available.
e.
Intermittent Disclosure and Periodic Reporting
Securities disclosure can also be gamed because financial reporting is periodic and
intermittent. Periodic reporting allows firms to obscure changes in its balance sheet and risk
profile that occurred in the period between financial statements. For example, accounting
research demonstrates that firms tend to close securitization transactions immediately before a
financial quarter ends.32 This timing may have a number of causes, including firms working to
meet earnings targets (and even manipulate earnings) and employees working to earn bonuses.
But this timing may also mask risk that firms face for most of a financial quarter, but which
disappear magically right before the date of the quarterly balance sheet. For example, a firm
may originate numerous mortgages at the beginning of a financial quarter and then sell them to
be securitized right before quarter’s end. The firm is exposed to the credit risk of the mortgages
defaulting for most of the quarter, but this risk evaporates when the mortgages are sold. An
investor can trace through the financial statements (such as the income and cash flow statements)
to find that the transactions occurred. A snapshot of risk exposures as of the quarter end,
however, does not reveal the risk to the firm for most of the quarter. This becomes a problem if
32
Patricia M. Dechow & Catherine Shakespear, Do Managers Time Securitization Transactions to Obtain
Accounting Benefits? 84 ACCOUNTING REV. 99 (2009).
9
Erik F. Gerding
Draft of September 19, 2009
the firm is unable in the future to unload its mortgages at the end of quarter for any reason (e.g.,
because of the collapse of the securitization market or a liquidity freeze). In that event, risk
would suddenly materialize in the form of impaired loans on the balance sheet. Intermittent
disclosure can thus facilitate accounting gamesmanship, including the earnings management
casually mentioned above.
6.
Technological Remedies to Information Failures; Upgrading Disclosure
The information failures outlined in the previous section prevent the marketplace from
fully understanding the risks faced by both individual financial institutions and by the financial
sector as a whole. Because the financial crisis stemmed in large measure from information
failures, it would be perverse and potentially disastrous to dilute securities disclosure on account
of its failure to prevent the crisis. Empirical evidence supports the contention that financial
markets and instruments with greater opacity and less disclosure suffer from greater mispricing
and more frequent bouts with asset price bubbles and crashes.33
a.
The Deep Tension between Nuance and Accessibility
Current disclosure is clearly inadequate. Although the complexity critique was
developed to argue for diluting or de-emphasizing disclosure requirements, the critique proves
useful in highlighting areas in which disclosure needs improvement. In particular, the two sides
of the complexity critique bring to the surface a deep tension that disclosure must negotiate. On
the one hand, securities disclosure must present financial information in sufficient detail and
nuance to enable investors to understand the complex risks and rewards embedded not only in
complex financial products but even in the plain vanilla securities of modern firms operating in a
dynamic marketplace. On the other hand, disclosure must take into account the cognitive
limitations of investors and the costs they incur in sifting through massive amounts of
information. Any enhancements to disclosure must make disclosure accessible and user
friendly,34 while still allowing investors to access information they would value in sufficient
scope and detail.
b.
Technological Innovations
Computer science, software, and information systems have long grappled with similar
tensions, and insights from these fields can vastly improve securities disclosure. More
particularly, technological advances in software and information systems can play a key role in
enhancing securities disclosure and disclosure regulations. This subsection outlines several key
technologies or technological insights that can improve disclosure and remedy some of the
information failures outlined above.35
33
Cite.
The SEC has long recognized this need, which animated a number of reform initiatives including the SEC’s
initiatives in the late 1990s requiring that disclosure be written “Plain English.” Final Rule: Plain English
Disclosure, Securities Act Release No. 33-7494, 63 Fed. Reg. 6370 (Jan. 28, 1998).
35
Other scholars have recently advocated other uses of technology to streamline disclosure to reduce costs for
securities issuers. E.g., Joseph A. Grundfest & Alan L. Beller, Reinventing the Securities Disclosure Regime:
Online Questionnaires as Substitutes For Form-Based Filings, Stanford University Law and Economics Olin
34
10
Erik F. Gerding
Draft of September 19, 2009
i.
XBRL.
The SEC’s XBRL initiative,36 which is already being implemented, provides a useful
launching point for using technology to enhance and reform disclosure.37 The XBRL initiative
builds off the decades-old SEC EDGAR regulations, which require firms and individuals that
must make filings with the SEC to do so in an electronic format that would be searchable and
available on the SEC’s online EDGAR database.38 The XBRL rules now require SEC registrants
to begin embedding data “tags” in their electronic filings. These tags would identify key items
(called “elements” in the relevant regulatory language) in financial statements and allow
investors to download these separate pieces of information directly into spreadsheets or other
analytical software. Investors can then make side-by-side comparisons of financial information
– for example, loans that are more than 90 days delinquent – from different issuers.
ii.
Tagging Assets in Securitization and Derivatives.
This same XBRL technology can be expanded to affix data tags to specific mortgages,39
asset-backed securities, and derivatives to enable investors to trace which assets underlay
specific asset-backed securities and derivatives even through multiple layers of securitization and
hedging.40 This tracing would improve the ability of investors to map the web of securitizations
and hedging in a nuanced fashion. By understanding underlying assets, investors could analyze
how underlying risks – including credit and market risk – on underlying assets course through
the system all the way to the securities they purchase. The SEC should consider expanding the
XBRL initiative to require data tags on various underlying assets.
iii.
Access to underlying financial data.
Working Paper No. 361 (Aug. 4, 2008) available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1235082
(advocating replacing SEC form filings with internet-based questionnaires for companies required to file with the
SEC).
36
See Interactive Data to Improve Financial Reporting, Securities Act Release No. 33-2461 (Jan. 30, 2009) (17
C.F.R. Parts 229, 230, 232, 239, 240, and 249 (2009)). XBRL reporting was previously an entirely voluntary pilot
project. See Final Rule: XBRL Voluntary Financial Reporting on the EDGAR System, Securities Act Release No.
33-8,529, Exchange Act Release, 34–51,129, 70 Fed. Reg. 6,556 (Feb. 3, 2005).
37
Commentators in the press have already lauded the potential of the XBRL Rule to increase transparency in
financial markets and allow an “army of citizen regulators” to police risk in financial markets. Daniel Roth, Road
Map for Financial Recovery: Radical Transparency Now!, WIRED, Feb. 23, 2009, at 81.
38
The EDGAR system was first created as a pilot program in the 1980s. See Securities Act Release No. 33-6977
(Feb. 23, 1993) (explaining the EDGAR system generally and setting forth SEC rules and procedures that apply to
electronic submissions). On December 19, 1994, the SEC required all domestic issuers to make Exchange Act
filings electronically via the EDGAR system. Rulemaking for EDGAR System, Securities Act Release No. 33-7122
(Dec. 19, 1994).
39
Chris Murphy, The Fed CIO Agenda: Should Every Mortgage Have A Data Tag?, Information Wk. Global CIO
Blog available at
http://www.informationweek.com/blog/main/archives/2009/03/the_fed_cio_age.html;jsessionid=P3FCQEYVFNZV
VQE1GHPCKHWATMY32JVN.
40
SEC officials appear to be actively considering at least tagging asset-backed securities and derivatives. See
Nicholas Rummell, Tagging, Tracking of Asset-backed Securities 'Urgent': SEC – Commission Says XBRL
Formatting of the Securities Could be Done within a Year, FIN. WK. (Oct. 15, 2008) available at
http://www.financialweek.com/apps/pbcs.dll/article?AID=2008810159987.
11
Erik F. Gerding
Draft of September 19, 2009
Data tags on specific mortgages, asset-backed securities, and derivatives can also
ultimately enable investors to see specific assets and liabilities in a firm’s balance sheet.
Required disclosure of specific assets and liabilities would uncover information that is hidden by
aggregated information and address the problems noted above when financial disclosure is not
granular enough.
iv.
Real-time reporting.
Advances in information systems will enable more rapid and frequent disclosure than
current quarterly and other periodic reports. It is conceivable that certain financial disclosure
could be made practically in real-time. Some firm disclosure to exchanges and financial
regulators already occurs daily or even more frequently. 41 More frequent disclosure would
curtail the accounting gamesmanship of many issuers noted above that is facilitated by the
periodic, intermittent nature of current disclosure.
v.
Open source models.
As noted above, the models that firms use to price asset-backed securities and derivatives
and to set investment and risk management strategies are dangerously opaque. Regulators
should require that issuers, particularly financial institutions, disclose much greater detail on the
methodologies and assumptions behind these models. More radically, an open source approach
could force disclosure of the algorithms in these models. This would allow investors to analyze
the models in greater detail and spot limitations, “bugs,” and the potential for dangerous
homogeneity among the models of different firms.42
vi.
Interactive Disclosure & Calculators.
Firms will likely resist disclosing information on their risk models and financial
statements in too much detail. As an alternative to this detailed disclosure, issuers could present
interactive disclosure – either online or embedded in an electronic document – that would allow
users to change certain assumptions or accounting methods and see how financial disclosure
would change accordingly. This technology resembles the interactive calculators that many
financial firm websites now offer and that the Obama Administration has endorsed as part of its
reform of consumer financial regulations.43
vii.
Layout of disclosure.
If disclosure would be increasingly delivered in electronic and interactive formats, this
would allow regulators, issuers, and securities lawyers to rethink the overall presentation of
disclosure to make it more accessible, malleable, and “user-friendly.” Behavioral economists
41
Cite.
Gerding, supra note 22, at 190-91.
43
U.S. DEPARTMENT OF TREASURY, FINANCIAL REGULATORY REFORM: A NEW FOUNDATION 65 (2009). These
current calculators are designed to help consumers and investors calculate financial needs not to show financial
disclosure under different scenarios and assumptions.
42
12
Erik F. Gerding
Draft of September 19, 2009
and legal scholars have explored how research on the cognitive limitations, behavioral biases,
and heuristics can be used to improve the decision-making of investors and other individuals.44
One strand of this literature focuses on using regulations to improve the “menu design” for
consumer and investment choices available to individuals.45 This approach on improving
individual decisions can also be applied to improving disclosure to consumers. The “menu”
metaphor is quite apt, as software designers have gleaned valuable experience on making
software more user-friendly. Drop down “menus” and other interactive devices have been
introduced and progressively improved in software since the advent of the graphical user
interface. These devices allowed software to perform more complex tasks and be used by more
consumers at the same time.
7.
Challenges
Employing these technologies to advance a brave new world of securities disclosure will
involve costs, pose risks, and generate opposition. This section sketches some of these
challenges.
i.
Privacy.
Regulators and issuers would need to take care that either tagging individual mortgages
or consumer loans or more granular financial disclosure on underlying assets does not lead to the
personal information of individuals being made public.
ii.
Proprietary information.
As noted above, many firms would fiercely resist requirements that they produce open
source risk models or disclose granular and real time information on individual assets and
liabilities. These firms would have legitimate concerns, as this information would give
competitors access to proprietary models, which might dull the incentives of firms to invest in
better models. Moreover, disclosure on investment portfolios and strategies would allow other
investors to trade against the firm and reap a profit. Nevertheless, even if open source models
and granular and real-time financial disclosure are not fully realized, as ideals for useful
disclosure to investors they provide a valuable starting point for policy discussion.
iii.
Investor “processing speed”.
Technologically enhanced disclosure, particularly more frequent or real-time disclosure,
will challenge the ability of investors to process information carefully. On the positive side,
higher speed of disclosure promotes market efficiency, but it also raises questions, including
Would disclosure be sufficiently audited? Audit firms could not realistically audit
financial information disclosed in real-time, i.e. as transactions occur. Some information
may be important enough to merit slower disclosure to enable auditors to catch potential
accounting errors.
44
This academic movement was popularized in RICHARD H. THALER & CASS R. SUNSTEIN, NUDGE: IMPROVING
DECISIONS ABOUT HEALTH, WEALTH, AND HAPPINESS (2008).
45
E.g., Ian Ayres, Menus Matter, 73 U. CHI. L.REV. 3 (2006).
13
Erik F. Gerding
Draft of September 19, 2009
Would higher speed disclosure promote greater market volatility? Under even the weak
form of the EMH, real-time disclosure would have immediate effects on the market price
of an issuer’s securities. This could make prices more volatile and exacerbate noise
trading.
Would higher volatility be gamed by issuers, market intermediaries, and sophisticated
traders? Instantaneous disclosure, higher volatility, and higher levels of noise trading
could allow sophisticated players in the marketplace to take advantage of retail investors
by using superior access to information to trade ahead of the market, engage inside
trading, or manipulate prices.
Would higher volatility lead to greater automated trading? Faster disclosure times may
push more sophisticated institutional traders looking to gain a competitive advantage
further into automated trading. Several episodes in the last two decades in financial
markets, beginning with the 1987 stock market crash, have demonstrated that widespread
use of automated trading can lead to unexpected market crises and place the risk
management of financial institutions on autopilot.46
iv.
Controlling the design of disclosure and Technological Elite
However, if firms do not provide open source models or the raw data behind their
disclosure, then the disclosure will inevitably have numerous embedded assumptions. The
behavioral economics literature underscores that the way in which information is framed can
exert a powerful influence over individual perceptions of risk. Thus, the individuals responsible
for choosing the format of the disclosure can wield subtle power over the decision-making of
investors.
Control over the format of technologically enhanced disclosure may fall to an elite that
possesses both the technological and financial sophistication to understand the embedded choices
and assumptions in the new technologically enhanced disclosure. Control by these elite would
create agency costs, which come in the same two traditional flavors. First, those who control the
new disclosure may fail to exercise sufficient care and make errors. In other words, the
technology may have “bugs.” Second, the elite may exploit their information advantage to
manipulate the perceptions of investors.
“Open Technology” and Beta Testing Disclosure
8.
i.
A response to agency problems: “Open technology.”
The potential for agency problems with technologically enhanced disclosure argues for
applying lessons from the Open Source movement in computer software to financial disclosure.
Open source software has several meanings, but a core element of the concept is that the source
code of the software is openly disclosed. Disclosure can also be thought of as a technology with
its own embedded assumptions and underlying data. The more that these embedded assumption
and data can be exposed to the public, the more the public can police disclosure for errors
(“bugs”) and potential manipulation of disclosure formats. An Open Source ethos underlays and
unites the technological advances outlined above. An open source approach would allow other
46
Cite.
14
Erik F. Gerding
Draft of September 19, 2009
sophisticated parties in the marketplace to police issuers and regulators responsible for the
format and content of disclosure.
ii.
Beta-testing disclosure: Need for experimental testing.
As noted above, Professor Paredes acknowledge that more empirical evidence would be
needed to identify when investors are overloaded with information, which investors are
overloaded, and which pieces of information are critical for which investors. The possibility of
information overload and questions on the effectiveness of disclosure given the cognitive
limitations and behavioral biases of investors indeed point to a need for rigorous testing of
technologically enhanced disclosure. The reforms outlined above, for example the “menu”
redesigns mentioned above, should be thoroughly tested through experiments to determine their
effect on the comprehension level of various types of investors. Legal scholars have increasingly
advocated using methodologies from experimental economics to test the effectiveness of various
legal rules.47 This article continues in that vein. Research from psychology and neuroeconomics
can be used to test the effectiveness of different disclosure technologies and approaches.
Sophisticated experimental devices, such as functional MRI, can also be employed.
This article has argued that technology can remedy information failures and created
enhanced disclosure that would improve the comprehension of investors. Like most software,
the actual 2.0 version of securities disclosure should be rigorously beta tested.
47
See EXPERIMENTAL LAW AND ECONOMICS (Jennifer Arlen & Eric Talley, eds. 2008)(surveying emergent field);
Erik F. Gerding, Laws Against Bubbles: an Experimental-Asset-Market Approach to Analyzing Financial
Regulation, 2007 WIS. L. REV.977 (arguing for using experimental asset market research to test securities
regulations).
15