On the Record

Enhancing Financial Stability: The Role of Transparency

Donald Kohn

On September 6, 2011, Donald Kohn addressed the Financial Markets Group at the London School of Economics in his first public speech as an external member of the Bank of England’s Interim Financial Policy Committee.

I’m pleased to be here at the London School of Economics to present my first public speech in my role as
external member of the Interim Financial Policy Committee. I can’t think of a better venue; the Financial
Markets Group at the LSE has been in the forefront of thinking on financial stability, and I am greatly
indebted to Charles Goodhart for all that he has patiently taught me about central banking and financial
markets over the years at innumerable conferences and in many private conversations.

I am honoured to have the opportunity to serve on the FPC. In the wake of the financial crisis, I can think of
no more important task than to bring a broad economic and financial perspective to the supervision and
regulation of the financial system. An element of that perspective has always been applied to the oversight
of the financial system, but at least in advanced economies it hadn’t been applied systematically and
forcefully, bringing to bear a wide range of expertise and analysis. It is important that we succeed, and I am
encouraged by our initial meetings, which were marked by probing questions, varied contributions from all
participants, and very useful and fruitful dialogue between the microprudential supervisors and those of us
coming to the task from the more macro side. Although, as I will point out, transparency was an important
theme of these meetings, the views you are about to hear are my own, and not necessarily those of my

However well we do our task, we need to recognize that the authorities can’t do it all. The stability of the
financial system ultimately rests on the decisions of its private sector participants. To be sure, government
regulations can and should help shape those decisions and provide a structure within which private
participants are more likely to help to stabilize the system. But, in the end, it is those private decisions that
will determine the allocation of capital, economic growth, and economic and financial stability.

Those decisions cannot contribute to financial stability if they are not well informed, and the subject of my
talk tonight is the need for better transparency among financial institutions, markets, and instruments. The
available information needs to be comprehensive and accurate, it needs to focus on the important risks and
characteristics, it needs to enable comparisons across institutions and time to facilitate analysis, and it needs
to be understandable enough that market participants can assess and act on its implications. We’ve made
considerable progress over the past few years, but more can be done. I’ll begin with some thoughts on the
overall role of transparency in promoting financial stability. Next, I will make some suggestions for
enhancements to transparency in various aspects of the UK financial system. Transparency has positive
externalities, but it is not free, and we will need to consider the costs and benefits as we move forward.
Finally, I will outline how I believe transparency by the FPC itself can contribute to financial stability as well
as to the accountability of the committee.

Transparency in perspective
Clearly, in the build up to the crisis private sector decisions did not result in financial stability. A variety of
risks were poorly understood, poorly managed, and badly priced. And these decisions, inadequately
overseen by regulation and supervision, encouraged a housing bubble backed by excessive debt in the
United States, and over borrowing and spending in a number of European countries, as well as an explosion
in the size, leverage, and maturity mismatch of the financial system itself. Incentives were skewed – often
toward short-run profits; some investors did not see their money at risk because it was placed with large
institutions that were thought to be too important to be allowed to fail; and, in my view, above all, people,
including the supervisory authorities, became way too complacent about risks – especially tail risks – after
years of good growth, mild recessions, and low and stable inflation.

But a lack of information also contributed to the mispricing of risk. Institutions, markets, instruments, and the
interactions in the financial system became more opaque in the years leading up to the crisis. Markets
became characterized by long complex chains of intermediation and risk transfer; the risk characteristics of
complex structured instruments were poorly understood; financial institutions were difficult to evaluate and
often were exposed to other institutions in obscure ways that became evident only in the crisis. Increased
complexity requires increased information for risks to be managed well; available information did not keep up
with the substantial jump in complexity. Market participants came to rely on third parties – credit rating
agencies and insurers like AIG and the monoclines – in part because they couldn’t themselves evaluate the

And the lack of information exacerbated the downturn. Contagion fed on uncertainty about the financial
health of counterparties, spreading and intensifying runs and withdrawal from market making that led to
falling prices of assets due to fire sales and premiums in illiquid markets. Impaired and fearful lenders cut
back sharply on credit available to many households and businesses, which in turn reduced their spending.

Promoting greater transparency was an important theme of the initial discussions of the FPC.1 Our first two
policy recommendations focused on public release of information about sovereign and banking sector
exposures. But Recommendation 1 also states that the FSA will work with the FPC “to consider further
extensions of disclosure in the future”. The Record of our meeting notes that “a number of members”
argued for extending disclosure to other sectors and finding other dimensions in which information to
investors could be improved – and I count myself in that group of members. In addition, in its discussion of
complex instruments and interconnections, the FPC was concerned not only about the information available
to regulators but also the information that investors have – it is crucial that they understand the risks
associated with the instruments they are buying.

Greater transparency is not by itself a sufficient condition for improved market discipline. Market participants
need to know they have funds at risk to give them incentives to monitor and price risk appropriately. In that
regard, the efforts under way to roll back the perception of ‘Too Important to Fail’ with, among other things,
new resolution authorities and the living wills that will allow those authorities to act effectively, are critical. A
market-based system can function effectively only when institutions can – and do – fail. Owners,
management, and creditors must perceive their stakes are potentially at risk if their actions are to promote
resilience and stability. The important point is that the institutions must be able to fail in a way that does not
imperil the system and many innocent bystanders.

And these stakeholders need to be able to act on their perceptions in ways that discipline the institutions and
affect asset prices. Corporate governance structures should allow all shareholders an effective voice in
oversight of the enterprise. Market participants should be able to make their views felt through purchases
and sales – including short sales – of instruments in as liquid markets as possible.

Market discipline by itself cannot be sufficient to protect financial stability, however. As has been so
graphically and tragically demonstrated in the past few years, financial market instability is replete with
externalities for the broader economy. And because of those externalities, governments have put safety
nets under the financial system, including deposit insurance and central bank liquidity provision, that
engender moral hazard. Moreover, strengthening financial systems often runs into collective action
problems, where policymakers acting to coordinate behavior, for example in clearing and settlement, can
produce better outcomes for financial stability than each bank acting in its own self interest. All these
market failures mean that strong micro- and macro-prudential oversight is critical to building more resilient
financial systems.

Although better market discipline may not be sufficient for financial stability, it is essential, and better
transparency is a necessary condition for better market discipline. The authorities have a critical role to play
in promoting adequate transparency. Those same market failures that call for an important role in the
regulation of financial institutions – moral hazard, systemic externalities, collective action problems – also
imply that institutions and investors do not have the incentives to release as much or as useful information as
is needed to judge fully and accurately the risks to the institutions or to the broader financial system. Many of
the characteristics of “useful” information are things the authorities can help banks and other market
participant coordinate on, for example enforcing consistency across institutions, jurisdictions, and time. In
this regard, the macroprudential perspective has much to contribute to determining what information is most
likely to allow the market decisions of private parties to enhance overall financial stability.

A substantial amount of work is underway in the UK and internationally to provide more and better
information to market participants so they have the opportunity to understand the risks they are taking. It is
not my objective to review those work streams in detail. Rather I want to highlight some broad principles and
areas I believe would be especially useful to push further on with respect to transparency of institutions,
instruments and markets. Clearly these are linked. Market discipline on institutions requires an
understanding of the risk characteristics of the instruments they hold and how institutions and instruments
are interconnected in markets. And, to the extent intermediation bypasses traditional financial institutions,
financial stability requires the ultimate investors to examine and understand the characteristics of the
instruments they purchase and the markets they are traded in.


For market discipline to have a chance of enhancing financial system stability, counterparties and investors
need to be able to make a reasonably accurate assessment of the financial health of an institution and how
that health would be affected as economic and financial conditions change. I will be concentrating on banks
in this discussion because they are the key and dominant players in the British financial system. An accurate
gauge of financial health of these institutions – their source of profits, their asset quality, the structure of their
funding – is a particularly challenging objective as they, the instruments they hold, and the interconnections
among them become more complex. Prompted by the crisis, considerable progress continues to be made
on institutional transparency, in part under the leadership of the FSA here in the UK, the FSB globally on risk
disclosure practices,2 and pillar 3 disclosures within the Basel capital requirement regime.3 For example, as
a consequence of these efforts, counterparties and investors have more information about Special Purpose
Entities, complex securitizations and structured credits and how they are valued. Banks themselves have
greatly increased the amount of information in their annual reports, and in some cases in their interim reports
as well.

But transparency is still a work in progress. I will highlight some principles for implementing the overall
objective of enough transparency to give counterparties an accurate assessment of risks and suggest the
areas for additional disclosure I’d like to see being given serious consideration in the UK.

Information should be timely, up to date, and sufficiently comprehensive. Investors can’t gauge
creditworthiness based on data that are out of date. The current practice of UK banks of releasing detailed
reports semi-annually with less complete interim updates in the other quarters may not allow the public to
update its assessment of the financial strength of these institutions frequently enough. In a world in which
positions can be adjusted within minutes and soundness affected by daily economic and financial
developments, even quarterly updates are not enough really to keep on top of a reporting institution. Market
participants will naturally adjust their assessments as conditions change based on what they know about the
institution. But at least detailed quarterly reporting would allow more frequent benchmarking of those


Comprehensive data should include information about credit exposures to other financial institutions, and
concentration, market and liquidity risks, as the FSB has highlighted.4 As we now know, these are critical
aspects of the risk profile of a bank and for all banks together they are critical to the assessment of systemic
risk. Contagion often ran through interbank relationships; banks had concentrated lending in narrow sectors;
the originate-to-distribute model depended on liquid markets for potentially risky assets; and maturity
mismatches exposed banks to liquidity and rollover risk.

We also found in the crisis that transactions between affiliates can create risk for core institutions, and
problems in one affiliate can undermine confidence in other parts of the group. So reporting should also
include enough information on group structures and the relationships and transactions among the affiliates in
the group to allow people to judge mutual dependencies and vulnerabilities.

Information should fairly represent the condition of the firm. We will not realize the full value of reporting for
preserving financial stability if the data, however frequently reported, cannot be readily utilized to assess the
condition of the firm. In that regard, we should be looking at several ways to improve the usefulness of UK
bank reporting.

First, I believe serious consideration should be given to requiring reporting of quarterly averages and
intra-period highs and lows for a number of balance sheet categories. For many balance sheet items,
single-day data may not be indicative of the true position of the firm – either because the circumstances of
that particular day caused the bank to adjust its risk positions in an atypical fashion, or because it
deliberately engaged in window dressing to achieve a certain configuration for reporting purposes. This is a
long-standing issue with the reporting of British banks that has frustrated those trying to understand the true
circumstances of the banks. Hartley Withers in his foreword to the 1910 edition of Lombard Street remarks
that “Banking reformers press continually for more frequent and clearer statements of their positions by the
country banks, and for the adoption of the average system in all bank statements, so that there may be no
possibility of specially arranged displays”.5 Quarterly averages themselves may not tell us a lot about how
positions are evolving over the quarter, but they do give a more complete picture than the single day.

Second, investors need to be able to evaluate how the banks reach judgments on the fair value of their
mark-to-market assets. When liquid markets are not available in which to price assets, elements of model
and judgment naturally come into play. We know considering more about how these are used than we did a
few years ago, but we ought to ask whether the public as yet has sufficient information to judge the validity of
these marks and to compare them to the techniques used by other institutions on similar assets.

Third, the banking book presents very considerable challenges in assessing the health of a financial
intermediary. Questions about bank loans continue to colour market commentary on the bank resilience,
several years after the onset of the problems. Restoring confidence in the banks may well entail publishing
more and better information about the banking book. To inform itself better about threats to financial
stability, in one of its recommendations, the FPC asked the FSA to obtain more comprehensive information
from the banks about forbearance and its link to provisioning. Forbearance can be a stabilizing strategy for
an institution and for the system as a whole. But loans that have been forborne are much more likely to
suffer losses eventually. We wanted to know the prevalence of forbearance and whether the heightened
likelihood of losses had been adequately reserved against; if they had the firm’s capital would fairly represent
its ability to meet unexpected developments. It would seem that market participants would benefit from
similar information at some level of aggregation.

As this suggests, ensuring that banks provision appropriately for loan losses is crucial to assessing their
underlying ability to generate earnings on loans and absorb losses in the future. Information on provisioning
will help most if it is related in a systematic and credible way to expected losses – not just those already
incurred. Accounting standard setters are working toward agreement on an expected loss impairment
approach – it is important that they be successful.

Investors should be able to update their assessment of the condition of a banking firm between periodic
reports as circumstances change.
The positions and risk profile of banks are constantly evolving; they are
not frozen in amber between periodic reports. And lack of knowledge about that evolution can create
uncertainty and adversely affect financial stability in a stressful situation. Firms can help relieve uncertainty.
One way is by publishing enough information in each report to enable market participants to apply their own
informed judgment about the effect of changing circumstances – actual or projected – on the firm. These
sorts of judgments are especially difficult for products whose valuations can be based on complex models.
Publication of more details of the firm’s sensitivity analysis of these credits would be useful.

Investors should be able to compare across firms and across time, to the extent possible. Each firm is
different and some of its reporting will be idiosyncratic to its own characteristics. But considerable insight
can be gained from comparisons of similar data across firms. The required conformity of reporting won’t
happen without the intervention of the regulatory authorities or other oversight bodies.

The FSA has agreed a Code with the BBA on enhanced transparency for UK banks. But it has also noted
that they would keep the impact of the BBA code under review and assess the need for further policy
initiatives to strengthen disclosure. As our recommendations suggest, the FPC will be monitoring this
process closely to assure itself that the required information to allow market participants to contribute to
financial stability is reaching the public.

Disclosure by regulators of the information they collect can materially aid transparency and market discipline.
Such disclosure can occur without additional cost to the reporting institutions, can be focused on key risks
and allows comparability by setting the parameters of the disclosures. It also allows for periodic
adjustments to reflect shifting risks. Coordination among regulators in different markets would be helpful in
allowing a broader view of changing risks in the many circumstances in which transactions and instruments
involve more than one market. The PRA has committed to publishing some of the data it collects6, and in my
view this is an area that can be exploited for significant useful increases in effective and focused
transparency at low cost.

Complex and poorly understood instruments were at the heart of the crisis arising from the mortgage market
in the United States. The complexity and the interconnections – the long chains of claims embodied in the
securitization, resecuritization, and derivatives based on securitizations made it almost impossible for people
to understand and price the risks they were taking. The complexity of the securities and their risk
characteristics meant that the models used to price them were exceedingly difficult to understand. Moreover,
detailed granular information about the underlying mortgages and securities was often not readily available.

The chains of borrowing and lending made it difficult to trace interdependencies among counterparties and
the amount and character of the collateral securing the obligations. This was a problem not only with
respect to the underlying real properties, but also the various securities that were financed along the chain
through, for example, repurchase agreements and securities lending. Under these circumstances it’s not
surprising that risk was not well priced.

In response to these difficulties a lot more information is now available about these instruments and
structures are now simpler.7 But the FPC is concerned about backsliding as confidence returns – about the
re-emergence of complex instruments with chains of counterparty exposures that are not transparent or well
understood. In June, we pointed to synthetic ETFs as an example in this regard, with counterparty risk hard
to evaluate because it often involved swaps of underlying collateral that was illiquid and difficult to value. But
we will want to monitor other instruments and circumstances where complexity complicates valuation by
market participants.

Transparency about these structures – full information about them readily available to all market participants
– is required to protect financial stability. Investors in the instruments and counterparties of those involved in
the chains need to have the opportunity to evaluate their risk.

For some instruments, even if disclosure had kept up it would have been futile – instruments were so
complex that the required information to appropriately monitor risks was overwhelmingly large.8 Indeed
excessive complexity and information overload may be limiting factors on the effectiveness of disclosures.
This possibility perhaps points to a need for avoiding the re-emergence of such complexity and for
encouraging appropriate design of disclosures for unavoidably complex instruments, for example a good
summary of properties of the instruments and clarity about the assumptions in valuation models.

Complexity in the instruments and the models used to construct and price them also bred reliance on the
opinions of credit rating agencies about their risk characteristics. That meant that market discipline
depended on the opinions of a relative few agents, who were being compensated by the issuers. Meaningful
transparency about these ratings, how they are arrived at, the characteristics of the securities and their risk
profiles is an essential element for market discipline to help stabilize the financial system. The CRAs are
now publishing considerably more information about their techniques and the data and models used to arrive
at ratings of structured credit. This information will facilitate the analysis of market participants, increasing
the variety of views that are brought to bear on pricing securities, and increasing the scrutiny of CRAs.

However, both the CRAs and investors have resisted separate and distinctive ratings nomenclatures for
different kinds of securities. As we saw in the crisis, structured credits respond very differently to economic
developments affecting their underlying securities than do plain vanilla bonds, yet they carry the same
ratings designations. Innumerable anecdotes suggest that many investors did not understand this. An AAA
corporate bond is not an AAA structured credit is not an AAA sovereign debt. Market discipline will be
strengthened when different designations force investors to look more carefully at what they are buying.

The increased use of central counterparties for clearing derivative trades will reduce the complexity of many
of the connections and interdependencies in this huge and growing sector of the financial markets. CCPs
imply more transparent, less complex networks, though risk is concentrated in the CCP, and financial
stability requires that it be managed very carefully. At the same time, information about centrally cleared and
other derivatives transactions will be gathered in trade repositories.

These data warehouses present an opportunity to improve transparency in a number of dimensions. More
complete reporting of price information should help in price discovery and enhance competition. Release of
information about aggregate trade volumes should help policymakers and market participants assess how
participation in markets for risk is changing over time.

But the fragmentation of markets, CCPs, and data warehouses threaten to limit the value of the data and the
extent of network simplification. Fragmented markets themselves create complex interdependencies as
institutions and final users operate in more than one venue. And the data available from each market may
not be consistent or in a common format that permits market participants or regulators to see the
relationships readily or aggregate positions by counterparty and across markets. Continued industry effort to
develop common data classification systems is needed to enable effective utilization of this new information.
In addition, realizing the full value of the repositories may require filling in some data gaps. For example, the
current data sets reported to existing data warehouses say nothing about whether two parties to a trade
have agreed to take collateral from each other. More generally consideration might be given to expanding
the reach of data warehouses to such transactions as securities lending and collateral swaps, where risk is
often difficult to track. Some change to market trade reporting practices, new or at least enhanced
arrangements may be necessary to fill those gaps.

Balancing costs and benefits
Collecting and publishing information is not a costless undertaking, and requirements and expectations must
be based on considerations of the balance of benefits and costs.

Transparency operates not only by giving information to investors and other counterparties, but it also tends
to focus management attention on the information being released. A benefit of increased reporting is that it
can force senior management to improve its understanding of the risks the institution is taking. But it’s
important that transparency be centered on the key issues; poorly structured or outmoded requirements can
distract management attention from more important matters.

There are important limits on the benefits side of the equation. In transparency, as in many other aspects of
life, more is not necessarily better. Great amounts of information can make it difficult for investors to pick
out the important bits that have a significant bearing on risk.

Ideally, requirements for transparency would be based on an analytic framework that facilitated identification
of the most useful pieces of information. Right now, our framework is importantly based on the crisis – what
that experience taught us about what information was missing as the imbalances were building and then as
they imploded. That’s a useful and necessary exercise. But the situation will change; markets will learn the
lessons of the last episode; risks will come from different directions over time as a consequence of regulatory
arbitrage, innovation, and a shifting macroeconomic landscape. Let me suggest to this audience that
theoretical and applied research on optimal transparency and how transparency should evolve as
circumstances change could contribute importantly to financial stability in the UK and around the world.

Adapting transparency to shifting risks will be a substantial challenge. Identifying those risks is the major
task for the FPC and we should be prepared to work with the supervisors and the financial sector to translate
our concerns into evolving requirements for public disclosure. As I’ve already noted, one advantage of the
use of information gathered by the regulators for informing the public is that it should be faster and easier to
adapt the collection and release of data to changing circumstances.

The experience in the United States was that the migration of intermediation to outside the banking sector
gave rise to critical nonbank points of vulnerability in the financial system. It’s difficult to predict how
intermediation will evolve in the UK after the crisis and the regulatory response, but we need to keep a
careful eye on potential risks outside of banks. Finding the right level and mix of information to be released
beyond the closely regulated banking sector will be a particularly difficult challenge. We should be able to
expand the perimeter of data collection and expectations for transparency, and to do so even more readily
than the perimeter of regulation itself. Indeed strengthening market discipline may be a substitute in some
circumstances for expanding the reach of regulation – at least it probably should be tried first.

We do need to balance the benefits of gathering and releasing information against the costs incurred by
institutions and market structures. One such potential cost would be the loss of legitimate competitive
advantage from revealing proprietary business strategies. The public should have access to enough
information to evaluate risks and prospects for a firm without compromising its competitive position or
reducing growth-enhancing innovation or expansion of market share. This may not be an easy balance to
strike, and in doing so we do need to keep in mind the costs for financial stability of inadequate market

Institutions often cite the cost of gathering and reporting information when they resist new requirements.
However, in many respects the information a bank or other financial firm is expected to release to the public
should be a subset of the information it uses for its own risk management – assuming its risk management is
up to best practices standards – or that it is required to forward to the regulators. That’s not to argue that the
burden is non-existent. Transparency requirements may not align perfectly with the institution’s own risk
management data; for example the authorities may be gathering information by legal entity and the firm may
be operating by business or functional lines across legal entities. And anything released to the public will
probably require more verification than data used internally. But, regulators and other authorities should
strive to make their data requirements lineup as closely as possible with best-practices risk management to
keep so the added burden minimal.

Both regulators and banks express concern that publishing some data in the midst of a crisis could make the
situation more serious by suggesting weaknesses in reporting institutions. This is a legitimate concern; for
example inferences from cent