1. Introduction

In this paper I highlight some fundamental character-istics of the recent financial crisis and identify ways tomake the financial system stronger. The financial sys-tem has evolved technologically, with an ever wideruse of securities and derivative contracts, and geo-graphically, with the worldwide spread of the capitalistmodel of financial intermediation. I start with adescription of the effects of these developments andshow that, fundamentally, the key weak points in thebusiness of financial intermediation have remained thesame.

The recent crisis has brought to the forefront a num-ber of pitfalls in financial markets including, promi-nently, weak systems to protect retail investors as wellas retail borrowers. This paper does not address theseissues, which are important and, in some countries, mayhave been a factor in the build-up of excessive risk posi-tions. By contrast, the paper focuses exclusively on theproblem of markets instabilities, and on reforms tomake the system less vulnerable to crises.

When addressing the question of improving systemicstability, it is useful to take very few things as given. Iuse a minimalist, functional perspective and onlyassume that a financial system is a network of agree-ments (typically contractual) and rules, and that afinancial system, in order to channel economicresources to productive uses, manages information. Myapproach is to start from what I consider the bestdescription of systemic fragility in financial intermedia-tion (a model in economics). Then, I describe the salientcharacteristic of the recent crisis, and in particularaspects of it that are the result of recent evolutions ofthe financial system. I argue that the basic mechanics of

financial crises still applies, but requires certain keyextensions.

The analysis yields implications for reform. The fun-damental approach in my analysis of reform is the com-parison of financial markets functions with financialmarkets institutions.

Section 2 illustrates the model of financial crises thatI use throughout the paper. Section 3 describes themajor lessons from the events of 2007-2008. Section 4presents the argument for proper equilibrium betweenfunctions and institutions in financial market. Section 5draws the implications for financial market reform.Section 6 contains a few concluding remarks.

2.What is a financial crisis?

‘Any sudden event which creates a great demandfor actual cash may cause, and will tend to cause,a panic in a country where cash is mucheconomised, and where debts payable on demandare large. In such a country an immense credit restson a small cash reserve, and an unexpected andlarge diminution of that reserve may easily break upand shatter very much, if not the whole, of thatcredit. Such accidental events are of the most vari-ous nature: […] , the sudden failure of a great firmwhich everybody trusted, and many similar events,have all caused a sudden demand for cash.’ Bagehot (1873), p. 122.

More than 130 years later, the nature of financial criseshas not changed, and that is because the basic featuresof financial intermediation - asymmetric informationand liquidity transformation - have not changed. In anintermediated financial system, one where institutionsraise capital resources from investors, the asymmetricinformation between investors and intermediaries cancause withdrawals of capital even in the presence ofgood investments. In the case where investments arerelatively illiquid, as in the classic banking business,where banks finance long-term loans with short-termdeposits, investors (depositors) not only worry about theway a bank uses the resources they lend to it, but theyalso worry about the possibility of not getting theirmoney back in the case where the withdrawal of

JANUJANUARARYY 20102010

CE

PR

PO

LIC

YIN

SIG

HT

No.

45

POLICY INSIGHT No. 45 abcd

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

Financial system reform proposalsfrom first principlesAlberto Giovannini1

Unifortune Asset Management

a

...the combination of asymmetricinformation and illiquidity gives rise

to the possibility of a financial crisis...

1 I thank Vitor Gaspar, Garry Schinasi and Luigi Spaventa for com-ments and suggestions.

deposits is widespread. But what justifies a business of liquidity transforma-

tion, where resources are gathered from actors whowant to gain easy access to them in case of need, toemploy them in a durable way? Consider the classicbanking business model. When banks borrow short termand lend long term they carry out a socially productivefunction (See Diamond and Dybvig 1983). They alloweasier access to their funds by savers and at the sametime they allow users of capital to take on long-term,more productive plans. Liquidity transformation is asocially productive, hence desirable, function. But thecombination of asymmetric information and illiquiditygives rise to the possibility of a financial crisis, a situa-tion whereby all depositors want their cash back - andthis can happen simply out of suspicion, with a perfect-ly healthy loan portfolio.

Because there is a possibility of equilibria that aresocially inferior (when a financial crisis occurs, liquiditytransformation breaks down, thus leading resourcesaway from higher yielding investments), these are mar-ket failures and therefore there is an economicallysound argument for a role of public institutions andregulation.

Armed with the justification that market failures arepossible, the regulatory tools to avoid financial crisesare of two kinds: permanent, ex-ante mechanisms andemergency, ad hoc, devices. The ex-ante mechanismsare banking supervision, deposit insurance and regula-tory constraints on the management of the bankingbusiness. The emergency devices include various liquid-ity provisions, typically managed by central banks.

With supervision, authorities are given the power topierce the veil of asymmetric information, which inprinciple they can do very well since they are not con-flicted entities. By assessing banks' investment portfo-lios they can determine whether such portfolios are toorisky, thus making it more likely that depositors with-draw their funds. In some cases, authorities have thepower or the persuasion to force banks to close out riskypositions. Deposit insurance provides safety to individ-ual depositors, and erases the risk that any one deposi-tor becomes a victim of a justified - or unjustified - runon a bank (typically, only small deposits are covered byinsurance). Regulations, of which the so-called Baslecapital requirements are the prime example, provide risklimits to banks, expressed as ratios of capital and liquidresources to assets.

Of course, these time-honoured devices to preventfinancial crises have all been in place during the recentcrisis. But, in the current crisis, as well as in pastepisodes, they were clearly insufficient. In the next sec-tion I explore potential causes of this failure.

3. Contemporary Financial Crises

The 2007-2008 financial crisis, similar to previousepisodes, more vividly highlighted the profoundchanges in the financial system, and drew the attentionof public opinion and authorities on the mechanismsand weaknesses of the financial system. I will argue thatthe events have shown three important facts:

• That a securities based financial system has crisesthat are, fundamentally, identical to those of thetraditional bank-based system, with some fea-tures that possibly exacerbate their disruptiveimpact;

• That the counterparty risk problem is key in asecurities based financial system;

• That the authorities had developed an unsus-tainable information deficit relative to the mar-kets they were supposed to supervise and safe-guard.

The financial system has changed dramatically in thepast 20 years. The last decade of the second millenni-um was the decade of fastest development of commu-nication and computation technologies, both of whichare used intensely to make securities and derivativesmarkets work. This phenomenon, coupled with a gener-alised embrace of market capitalism by essentially alllarge economies in the world, fully explains the evolu-tion of the financial system up to 2007.

The most notable feature of the world financial sys-tem on the eve of the recent crisis is the tremendousdevelopment of securities and derivatives.2 In 2004 theDTCC settled 1015 dollars worth of transaction ($1quadrillion). Two years later the number was already50% higher! What stands behind such huge volumes?

A securities-based financial system has some veryattractive attributes. The non-proximity between issuersand investors is one. It broadens the potential marketfor any issuer enormously. An additional attractiveattribute of securities is that they, typically, have a sec-ondary market, where their stock changes hands amonginvestors. If you are not compelled to keep a security inyour investment portfolio until maturity (or forever) youmay be more willing to buy it if you believe it is worth-while - the result is a wider potential set of investors.This is the value of liquidity in securities markets; justlike a bank aggregates diverse depositors with diversifi-able short-term liquidity needs to commit long-termresources to higher-yielding and productive invest-ments, the securities market aggregates diverse investorswith diversifiable short-term liquidity needs to providelong-term resources (debt and equity capital in the formof bonds and stocks) for productive purposes. Therefore,liquid secondary markets represent a fundamentallyproductive resource in an economy, in that they allowgreater access to financial resources for productive use.

For securities markets to function properly, potentialbuyers and sellers need to obtain accurate informationabout the prices of the securities they are interested in,and actual buyers and sellers need to find willing coun-terparties and need to complete their transactions

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

CE

PR

PO

LIC

YIN

SIG

HT

No.

45JANUARY 2010 2

The most notable feature of the worldfinancial system on the eve of the

recent crisis is the tremendous development of securities and

derivatives.

2 This is not to say that securities are a development of the last 20years. Just that securities trading has mushroomed in recent years.

smoothly and with minimum risk. Therefore a securities-based financial system necessitates intermediaries,whose functions are to service issuers on one side andinvestors on the other, but also to support the price-set-ting process as well as the settlement of transactions.

The second notable development of the financial sys-tem in the past 20 years has been the spreading ofderivatives. Derivatives are a natural by-product of theboom in securities which has made available a plethoraof prices in an equal number of markets, each repre-senting different risks or combinations of risks. Theseprices provide references for derivative contracts, whichthemselves allow buyers and sellers to gain exposure to,or eliminate exposure from, those different risks. Thedevelopment of derivative contracts has multipliedhedging and risk-taking opportunities to all actors.

A key implication of the development of the securitiesand derivatives businesses is the proliferation of trans-actions and counterparties. The "production" of deriva-tive contracts requires high-frequency trading (in theo-ry it requires a trade every instant): it is easy to conceivethe piling up of bilateral credits and debits in a marketwith widespread use of derivatives and the attendantintense and continuous trading activity. In general,high-frequency trading allows more efficient hedging,and therefore, for any given risk level, it minimises cap-ital usage, with the effect of making capital more abun-dant.

The important lesson from the recent crisis to outsideobservers, public opinion, and political leaders, is thecentral role played by securities in the present-day glob-al financial system. In particular, the experience of2007-08 (as well as earlier cases) has dramaticallyshown that the standard financial crisis described insection 2 occurs with the same characteristics in securi-ties markets, with distinct properties.

In general, as the market for securities stops working,it stops providing the liquidity services described above.People become unwilling to trade and subsequentlyprices no longer convey information about the value ofthe securities. This happens either when market partic-ipants themselves become unable to value the securitiestraded, or when participants lose confidence in othermarket participants' capability to settle their obligationsin the market. At the peak of the financial crisis in 2008a long list of markets almost ground to a halt, includ-ing the markets for mortgage bonds, the markets forcorporate bonds and the interbank deposit market (inthis case, however, loss of confidence towards counter-parties coincides with loss of confidence on the under-lying asset: deposits are not securities). This producedmassive damage in other securities markets, in particu-lar for the stock markets.

How is does the standard "bank" crisis manifest itselfin securities markets? Practically all financial intermedi-aries, banks and non-banks alike, hold securities in theirbalance sheet. A number of them hold securities thatare more vulnerable to the problems described in theprevious paragraph - securities which are normallycalled less-liquid because they have markets with fewerwilling counterparties than, say, US Treasury securities.A liquidity mismatch between securities in the assets

versus the liabilities side of the balance sheet has thesame business justification of the traditional bankingbusiness: providing liquidity in the loans and depositmarkets is a risky activity, since liquidity demand fluc-tuates randomly, and as such it carries a risk premium,or return. Providing liquidity in securities markets bybuying relatively illiquid securities and selling more liq-uid securities is the same risky activity and also carries areturn. As I pointed out above, it is also a socially desir-able activity because, by making securities markets workbetter, it facilitates the fundamental function offinance, channelling economic resources to worthwhileproductive activities.

Just as in banking crises, and as in the 1873 descrip-tion of Bagehot, a security crisis is associated with anincrease in demand for liquidity, or more liquid securi-ties. This puts strain on the balance sheets of thoseintermediaries who provide liquidity in securities mar-kets: their assets fall in value, their liabilities increase invalue. To restore their own financial equilibrium, thoseintermediaries have to sell their assets, in a situationwhere buyers are relatively fewer. This is evidently anunstable situation, well described by, for example,Brunnermeier and Pedersen (2008) and Geanakoplos(2009).

A distinct property of liquidity crises in securities mar-kets is the way they spread, the so-called contagionphenomenon. Those actors that find themselves in abind, unable to sell securities that do not have very liq-uid markets, because their attempts would produce toosteep falls in their value, are forced to sell securities thatare more liquid. The result is that the fall in value inilliquid securities spreads to the more liquid securities:this is why, for example, stock markets experienced col-lapsing quotations across the board during the worstweeks of the crisis. (The reasoning in the previous para-graphs applies to derivative markets as well.)

The most telling example of a securities-driven crisisin 2008 is that of US money market mutual funds. On15 September, Lehman Brothers filed for bankruptcy.The following day Reserve Primary Fund, a US moneymarket mutual fund, "broke the buck" because it heldLehman debt. This triggered a run on money marketmutual funds: investors feared for the quality of theirassets. Three days later, to avoid contagion, the USTreasury announced a guarantee program for the USmoney market mutual funds while the Federal Reserveannounced it would extend non-recourse loans tobanks to finance purchases of asset-backed commercialpaper from money market mutual funds. Money marketmutual funds were major holders of asset-backed com-mercial paper, highly rated and short term; however, themarket for asset-backed commercial paper frozebecause much of it had been issued to finance invest-ment vehicles into mortgage debt, including subprimemortgages. The last public initiative was the MoneyMarket Investor Funding Facility created by the FederalReserve on October 21. (See Acharya and Richardson2009)

Money market mutual funds were viewed as thetamest financial entities offering maximum liquidity totheir unit holders and were meant to be close substi-C

EP

RP

OL

ICY

INS

IGH

TN

o.45

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

JANUARY 2010 3

tutes for cash. Yet, in 2008 they experienced a run, justlike banks, and needed to be bailed out, just like banks.

The second issue that became clear in 2007-08 wasthe importance of counterparty risk.3 The discussionabove has explained why functioning securities andderivative markets produce very large numbers of trans-actions, the DTCC volume of business confirms thatobservation. Each security or derivative transactionexposes the parties to counterparty risk: the default ofthe counterparty in a trade creates an imbalance of theother counterparty's financial position, which needs tobe hedged.

To illustrate the size of counterparty risk in a typicallarge financial intermediary, consider Morgan Stanley'sbalance sheet, reported in Table 1. Most assets and lia-bilities are in the form of securities, or securities lend-ing or borrowing. In addition, the off-balance sheetgross total of derivative contracts at the same date was$3,728,464 million, i.e. almost $4 trillion.

It is evident, despite the alarmist statements of manyobservers, that the stock of derivative contracts is, large-ly, inert from a financial perspective; it is assets and lia-bilities that almost perfectly match each other (it couldnot be otherwise with a size of almost $4 trillion!). Butthose very large debits and credits become open posi-tions if Morgan Stanley were to become insolvent. Theywould become extra risk thrown into the financial sys-tem. In other words, a very large financial intermediarylike Morgan Stanley is a counterparty to a myriad of dif-ferent actors in the market, and, if it were to becomeinsolvent, the total risk in the financial system wouldshoot up by the size of un-hedged positions of all ofMorgan Stanley's counterparties. It would be like theinsolvency of a clearing house.

This is exactly what happened when Lehman Brothersdeclared bankruptcy; the balance sheet of Lehman, ofthe same order of magnitude as that of Stanley, wasthrown open, multiplying the risk in the system andprovoking a near collapse of world finance.

The final issue that the crisis laid bare is the informa-tion deficit of authorities, including financial supervi-sors, on the actual risks present in financial markets.And this is despite the logic of financial supervision

described in section 2 which would require supervisorsto have full knowledge of risks present in the system inorder to be able to anticipate problems. As an illustra-tion, I report excerpts of a speech by the President andChief Executive Officer of the Federal Reserve Bank ofNew York at the end of 2006:

‘The resiliency we have observed over the pastdecade or so is not just good luck. It is the conse-quence of efforts by regulatory, supervisory andprivate financial institutions to address previoussources of systemic instability. Risk managementhas improved significantly, and the major firmshave made substantial progress toward moresophisticated measurement and control of concen-tration to specific risk factors. What seems to havebeen most critical in preventing financial marketturmoil from translating into a significant reductionin credit provision by banks and other financialinstitutions were the steps taken by regulatoryauthorities and financial institutions alike tostrengthen capital in the core of the financial sys-tem, and to measure and manage risk. These effortshave most notably manifested themselves inincreased levels of risk adjusted capital in the coreof the system relative to what prevailed in the early1990s. In the United States, for example, tier-onerisk-based capital ratios have stabilised near 8.5percent, considerably higher than the estimatedlevels around 6.5 percent for the early 1990s. Thisis based on a relatively crude measure of risk, butthe direction of the improvement is right and themagnitude of the change is significant. Relative tothe conditions that prevailed in the early 1990s, thehigher levels of capital in the core now provide alarger buffer against shocks and enhance the abili-ty of the banking industry to act as a critical sta-biliser in times of stress by providing liquidity to thecorporate sector. When financial markets dry up,firms turn to banks and their unused loan commit-ments and lines of credit. Banks are in a position tofund this liquidity because transaction depositstend to flow into the banking sector. In times ofcrisis, it appears that U.S. investors now run tobanks, not away from them.’ Geithner (2007).

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

CE

PR

PO

LIC

YIN

SIG

HT

No.

45JANUARY 2010 4

Table 1 Morgan Stanley’s balance sheet

ASSETS US$ million LIABILITIES US$ million

Cash and cash equivalents 16,636 Commercial paper and other short-term borrowings 33,829Cash and securities deposited with clearing organisations 35,739 Deposits 37,313Financial instruments owned 407,754 Financial instruments sold, not yet purchased 157,807Securities received as collateral 82,684 Obligation to return securities received as collateral 82,684Collateralised agreements 469,131 Collateralised financings 502,105Receivables 138,262 Payables 146,473Office facilities and equipment 4,313 Other liabilities 24,063Goodwill 3,131 Long-term borrowings 159,833Intangibles and other assets 24,411 Equity 37,954

1,182,061 1,182,061

Source: Securities and Exchange Commission Form 10-Q, February 28, 2007

3 Counterparty risk has attracted policy-makers' attention since thesetup of the Counterparty Risk Management Policy Group, chairedby Gerald Corrigan, in 1999. Since then the Group, in three incar-nations, has produced an equal number of reports, the last in2009.

CE

PR

PO

LIC

YIN

SIG

HT

No.

45JANUARY 2010 5

The speech provides aggregate data on regulatory cap-ital requirements for US banks. It concludes with aguardedly optimistic note, indicating that increasingregulatory capital levels at banks, "the core of the sys-tem", implies decreasing systemic risks. With the bene-fit of hindsight, systemic risk was probably at an all-time high at the time of this speech. Just a few monthslater, US banks went collectively nearly bankrupt due tohaving taken on risks that were large multiples of theircapital base, often by arbitraging regulations. In April2009 the IMF (2009) estimated $2.7 trillion of loansand securities losses by US banks, and concluded thatthe new equity capital needed by US banks to bringleverage down to 17 times (17 times!) would be about$500 billion. Without government intervention, allmajor US banks may well have gone bankrupt. Theinformation-deficit problem was in no way limited tothe US. In Europe - in particular in the UK - authori-ties were all surprised by the events that unfolded in2007 and 2008.

Thus far, I have shown that major financial marketsdevelopments have changed the way financial crisesunfold, though the fundamental mechanisms at workhave not changed. I have also shown that a securitiesbased financial system has changed the global risk map,and that authorities in charge of supervision seemedunable to evaluate how much systemic risk there was infinancial markets on the eve of the crisis. Why were thestandard tools to limit systemic risk insufficient? In thenext section I argue that the boom in the securitiesbusiness has changed the structure of financial interme-diaries and has generated imbalances that may be at theroot of the instabilities we have been experiencing. Thechanges in the financial industry will require a changein the tools and functions of financial regulators.

4.The boundary problem

Organisations in financial markets are entities definedby a set of governance rules, contractual arrangementswith counterparties and, typically, a regulatory frame-work - the finance industry is one of the most regulat-ed. Thus, financial markets institutions define individualorganisations.

In this section I follow the method suggested byMerton and Bodie (2005) and treat functions (i.e. busi-nesses) as the conceptual anchors. Merton and Bodiealso offer a theory on how institutions endogenouslyevolve. By their very nature, institutions evolve moreslowly than functions/businesses. The evolution of insti-tutions requires a complex coordination exercise, withdifferent roles played by authorities and private market

participants. Functions, i.e. business opportunities,evolve because of technological progress, because ofinnovations by the industry, and in response to changesin needs of final users.

The progressive spreading of securities and derivativeshas led to a multiplication of business opportunities,and a transformation of financial intermediaries. Thesecurities business has moved from specialised broker-dealers and investment managers to the whole of thefinancial system, including all kinds of banking organi-sations as well as insurance companies. But the evolu-tion of businesses (or, to use a term more familiar toregulators and academics, functions) has not been insynch with the evolution of institutions. To be moreprecise, certain private-market contracts have necessar-ily developed with the booming securities- and deriva-tives-based system. Consider for example standardisedcontracts, tri-party agreements, netting provisions, dailysymmetric mark to market, and so on. Other aspects ofinstitutions, and in particular regulatory frameworks,appear to have dismissed the notion that there shouldbe a satisfactory correspondence between functions(businesses) and the organisations that run them.

4

Without doubt, such lack of correspondence is large-ly due to a very basic fact; institutions are static; busi-nesses (functions) are dynamic.

As much as this logic is hardly disputable, institutionshave two roles to play in the financial system:

• Institutions provide constraints in business con-duct; such constraints should be efficient, that is,appropriate for the nature and the riskiness ofthe businesses they carry out, as for example byeliminating incentives to excessive risk taking orother distortions like abuse of dominant marketposition,

• Since finance is pervaded with asymmetric infor-mation, knowledge that a given entity is a mem-ber of a given organisational class should helpestimate its riskiness; institutions should be reli-able signalling devices.

I define the boundary problem as follows: "Do institu-tions provide efficient boundaries for the differentfunctions in the financial market?" Efficient boundariesare those that perform the functions 1) and 2) above;they provide the appropriate constraints and incentivesto those that run the businesses and they are reliablesignals to all market participants about the riskiness ofeach individual institution. To answer this question Ineed to first identify the basic functions performed inthe financial system. Then I will review the way differ-ent financial organisations perform these functions anddiscuss whether the boundaries are efficient.

The boom in the securities businesshas changed the structure of financial

intermediaries and has generatedimbalances that may be at the root of

the instabilities we have been experiencing.

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

4 The abandonment of financial institutions as the starting point ofan efficient regulatory framework is illustrated, for example, by theFinancial Services Authority's (FSA) emphasis on functional regu-lation. Such emphasis de-facto dismisses the notion that a satis-factory correspondence between institutions (by which I meanentities with uniform regulation) and functions or businessesshould be an appropriate objective for a financial regulatoryframework.

4.1 Functions of a financial system

The primary functions of a financial system are to pro-vide clearing and settlement of payments, to provideresource pooling, to transfer economic resourcesthrough time and space, to manage risk, to provideprice information and to deal with incentive problems(see Merton and Bodie 1995). These primary functionsare bundled in a number of different businesses. Ichoose to order the different financial businesses inthree classes that are very well known to market partic-ipants, and correspond to three broad functions. Thesefunctions are depicted in Figure 1.

In the Figure, I portray exclusively functions per-formed by private market participants. The picturewould be completed by a number of public functions,such as supervision, public liquidity provision, etc.,which I cover below.

The business of client services is one that is definedby a large overlap between market counterparties andclients, who are the core of the franchise value of thebusiness. A client servicer lends money or securities toclients, it produces contingent payoffs (OTC derivativecontracts) that clients need for the purpose of manag-ing their own risk, it arranges capital market transac-tions for clients (debt or equity issues), and performs anumber of ancillary services, such as research and advi-sory. The overriding objective of client services is tomaximise profits through the enlargement of the clientfranchise, more clients and more transactions with eachclient. Client servicers gain their profitability from expo-sure to, and efficient management of, two types offinancial risk: credit (or counterparty) risk and liquidityrisk. That client business is a means of achieving expo-sure to counterparty risk is self-evident. That exposureto liquidity risk is also a key characteristic of the clientbusiness can be explained intuitively; typical services toclients include liquidity provision, in all of its forms like,e.g., long-term loans and broker-dealer services. In gen-eral, clients seek from their intermediary easy (andtherefore liquid) access to securities markets; client ser-vicers are the principal suppliers of liquidity services(buyers of liquidity risk) in securities markets.

Capital management is the second important func-tion (business) in the securities market. It is, essentially,

the business of investing (from now on, I use capitalmanagement and investment management inter-changeably). Capital management consists in creatingand managing exposure to all kinds of financial risksaccessible in securities markets, with a view of produc-ing attractive returns. The balance sheet of a capitalmanager may have securities both in the asset and inthe liabilities side (exposure to certain kinds of riskrequires going short certain securities). A capital man-ager may use either debt or equity, and all the variationsthereof. The fundamental distinction between capitalmanagement and client servicing is that, in the former,market counterparties are not clients. The clients of acapital manager are the shareholders. A capital manag-er's objective is to maximise shareholders' returns with-out regard to market counterparties.

Thus, client business is all about counterpartieswhereas an investment manager does not care aboutcounterparties. Indeed, much of the profitability of acapital manager may come at a loss to counterparties;a capital manager cares about counterparty risk, ofcourse, but not about the satisfaction of his/her coun-terparties in the trades he/she does with them. Supposethat, after appropriate research, an investment managerdecides that a given security (say, a stock) is underval-ued. The right course of action is to buy the security inthe most inconspicuous way, hoping that the buyingactivity does not attract the attention of other marketparticipant. By contrast, in the client business, ifresearch suggests the attractiveness of a certain securi-ty, this research will be immediately shown to clients, inorder to give them the chance to take advantage of agood opportunity.

Another inconsistency between client business andcapital management relates to the so-called problem of"front running". Client business often carries with itvaluable information about supply/demand imbalancesin the market (consider the market impact of largetransactions); if capital management could access suchinformation it would use it to its own advantage. Bynecessity the actions of a capital manager, taken as aresult of knowledge of client transactions, will negative-ly affect the cost of such transactions to clients.

The last business in the classification is that of infra-

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

CE

PR

PO

LIC

YIN

SIG

HT

No.

45JANUARY 2010 6

Issuers

Final Investors

Capital Managers

Client Servicers Infrastructures

Figure 1 Functions in securities markets

structure. Infrastructures facilitate the functioning ofsecurities markets. The salient characteristic of infra-structures is that they are information-processing plat-forms, with no human intervention; hence, they approx-imate natural monopolies. Take, for example, the caseof post-trading; the business of clearing and settlementof securities transactions. In today's securities marketsthe technology to provide such services is entirely auto-mated. It is, for all intents and purposes, a zero-margin-al-cost business. As such, it achieves maximum costefficiency when it reaches maximum scale, that is, thesize of the entire market.

4.2 Institutions

Where are the functions described above performed? Iillustrate some salient cases in Figure 2. The figurereports some organisations and their businesses in secu-rities markets. As mentioned above, an organisation isdefined by a set of institutional arrangements; laws,regulations governance rules, private contracts and mar-ket conventions. The aim of the figure is to highlightthe most interesting cases, not to provide an exhaustivelist. In the Figure I list, for each class of organisationsthe different businesses (functions) performed. Theshaded organisations are not discussed, but are in thepicture to indicate the many varieties of organisationsactive in financial markets.

I now turn to the boundary question, that is, whetherinstitutional boundaries are appropriate for the busi-nesses run by the different organisations.

I start with fund managers. As a whole, the commu-nity of fund managers seems to be specialising in onlyone business and, perhaps because of the specialisationin just one function, the private contractual frameworkto which they are subject is already rather robust.However, there are some boundaries that need to be

considered. For example, hedge funds, one sub-class ofmoney managers, have been the fastest growing seg-ment of the investment management industry. Hedgefunds are also universally regarded as major actors infinancial markets. Most hedge funds are corporationsbased in offshore centres. This stands in contrast to therest of the investment management industry, whichoperates through onshore organisations, with specificregulatory status. Thus, the size and significance of thehedge fund industry suggests the presence of a sort ofinstitutional imbalance; the fact that most hedge fundsare to a large extent outside the realm of standardadvanced financial regulatory frameworks is clearly anembarrassment for the latter (the advanced financialregulatory frameworks) and, in my opinion, also anembarrassment for hedge fund themselves.5

In addition, while the crisis has exposed potentiallydevastating liquidity risks even in the money marketmutual funds, i.e. supposedly the least risky of all man-aged funds, it is self evident that a significant numberof hedge funds are involved in liquidity transformationin the securities space and therefore are, just like banks,vulnerable to liquidity crises.6 However, in the currentinstitutional setup, authorities have little or no visibilityon the extent to which offshore hedge funds areinvolved in liquidity transformation. Authorities, in theirrole as systemic risk managers, would certainly behelped by having full access to information on hedgefunds' risks and trades.

Consider now the financial market infrastructure

CE

PR

PO

LIC

YIN

SIG

HT

No.

45

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

JANUARY 2010 7

Issuers

Final Investors

Banks:Capital Management

Client BusinessInfrastructure Business

Insurance Co�s:Client Business

Capital Management GSE�s:Client Business

Capital Management

Rating Agencies:Client Business

Infrastructure Business

CSDs, CCPs, Exchanges:

Infrastructure BusinessClient Business

(Mutual, Pension Hedge) Funds & Managers:

Capital management

Figure 2 Organisations in securities markets

5 They are routinely defined in the press as secretive schemes choos-ing offshore status to escape regulation and controls.

6 Hedge funds have dealt with sudden liquidity demands using toolsat their disposal, restrictions of equity reimbursements in the formsof gates and side pockets, which are part of their contractualagreement with investors.

providers. Central securities depositories (CSD), centralcounterparties (CCP) and exchanges CSDs are the enti-ties that hold the dematerialised securities, and performthe function of assuring, at every point in time, that thestock of securities issued equals the stock of securitiesheld. This implies that every transaction involving asecurity in a CSD is duly recorded. CSDs provide, there-fore, settlement services in securities transactions. CCPs,instead, perform in a centralised fashion the clearingbusiness - they become the counterparty of every trans-action (e.g. instead of A selling a security to B, the CCPbuys the security from A and sells it to B) - i.e. the busi-ness of determining credits and debits in all securitiestransactions. This way, CCPs facilitate risk reduction bypooling counterparty risk. Centralised trading, clearingand settlement are typical infrastructural businesses;they are high volume, near-zero-marginal cost. But theyare also closely linked to client businesses through lend-ing and asset servicing. It is evident that bundling infra-structural services with banking services is an attractivebusiness opportunity, and this is what has been done byseveral providers of infrastructural services (seeGiovannini, forthcoming). While originally infrastructur-al service providers were either government-owned enti-ties or mutual companies, throughout the 1990's manyhave progressively transformed themselves into commonstock companies. Thus, companies that offer mainlyinfrastructural services are moving towards client serv-icing and businesses are often dubbed as banks becausethey offer lending.7 These critics also suggest that suchinitiatives are being "subsidised" by the near-monopoly,standard infrastructural business. Should there be aninstitutional boundary (governance, regulatory, privatecontractual) between the natural monopoly functionsand the client service functions? This has been dis-cussed at length in Europe and has led to proposals tore-introduce user-ownership of infrastructure business,to avoid vertical integration of trading and post-tradingand to setup regulations restricting natural monopoliesin market infrastructures from exploiting their domi-nant position (see Giovannini forthcoming).

I now turn to banks, where the boundary problem ismost critical. In recent years the institutional setup ofbanks has converged. For example, the vast majority ofbanks are now common stock companies (and this con-vergence has accelerated during the financialcrisis).Meanwhile, the partnership structure, which was adistinctive characteristic of investment banks, has disap-peared from large institutions. In addition, investmentbanks have all but disappeared; they were born in theUS in the 1930s, when Glass-Steagall legislation sepa-rated commercial and investment banking, but theywere de-facto reabsorbed into commercial banks withthe Gramm-Leach-Bliley Act, which ended the separa-tion of securities business and lending and deposit busi-

ness. Finally, in the immediate aftermath of the recentcrisis, some of the most prominent investment bankshave transformed themselves into commercial banks.The basic regulatory framework of banks - capitalrequirements - allow for leverage equal to 12.5 to 1;banks can hold (risk-weighted) assets 12.5 times theircapital. In addition, banks have access to liquidity facil-ities provided by central banks. They are also subject tosupervision.

The significant trend in banks business has been theincrease in securities and derivatives dealing, relative totraditional lending and deposit taking activities. As Iargued above, this simply reflects technical progress infinancial intermediation. However, banks have alsomixed their client business and investment (or capitalmanagement) business.

Indeed, recent years can be characterised as a periodduring which investment business has slowly taken overclient business in many banking organisations. Startingfrom the very large losses suffered by banks in therecent financial crisis, it is straightforward, almost tau-tological, to conclude that the size of the losses can bejustified only by very large risk positions which couldonly be undertaken in a (rather reckless) investmentbusiness, and could not possibly arise from client busi-ness. While there has been a dramatic fall in liquidity,and liquidity risk, this is inherent in client business andI would expect those losses to account for a limitedfraction of total losses. This view is shared by manyobservers. Indeed, it has been claimed that the financialcrisis was, essentially, caused by excessive risk taking ofbanks involved in investment or capital managementbusiness:

"We believe that, although the originate-to-distrib-ute model of securitisation and the rating agencieswere clearly important factors, the financial crisisoccurred because financial institutions did not fol-low the business model of securitisation. Ratherthan acting as intermediaries by transferring riskfrom mortgage lenders to capital market investors,these institutions themselves took on this invest-ment role. But unlike a typical pension fund, fixedincome mutual fund, or sovereign wealth fund,financial firms are highly levered institutions". Richardson (2009)

The mixture of capital management and client businesshas three implications:

• It exposes banks' balance sheets to an array ofrisks other than liquidity and counterparty risks,which characterise client business;

• It uses banks' leverage potential in a way thatmay be inappropriate for the kinds of risks in acapital management business-in other words,12.5 to 1 leverage is to be considered very high,and rather unusual, in a straightforward capitalmanagement business, like for example that of ahedge fund. Such a high leverage ratio may bemore appropriate for a client business, where thegreatest majority of financial risks are hedgedout;

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

CE

PR

PO

LIC

YIN

SIG

HT

No.

45JANUARY 2010 8

7 Indeed, the debate on financial market infrastructure reform in theEU witnessed a debate between an impromptu association called"Fair and Clear" and Euroclear, the banking group with access toall of the main EU CSDs. The "Fair and Clear" group were banksconcerned with the fact that Euroclear, in their view, was using theadvantage it had as a manager of the "notary function" to bun-dle banking services together with settlement services.

• It exposes banks to those conflicts of interestdescribed above.

One of the most evident implications of the manage-ment of investment businesses within banks is thatbanks and investment managers operate the same busi-ness with very different constraints. I illustrate as anexample some of the different constraints on the invest-ment business of banks and hedge funds. To start, takeFigure 3.

The figure reports the size of investments in a certainsecurity with 100% risk weighting (I use MBS, but itcould be any other security with that feature) that canbe implemented by a bank and by a hedge fund. Itshould be stressed that the constraints in figure 3 arenot of the same kind; in the case of a bank I report theregulatory capital requirement, in the case of a hedgefund, I report the typical financing constraint of ahedge fund (this is pre-crisis; right now there is no debtfinancing available for mortgage backed securities). TheFigure shows that under bank regulation a capital of100 in a bank can be deployed to invest in 1250 ofMortgage Backed Securities. The same capital in ahedge fund can finance at most 200 of MBS. Thusbanks can, in principle, take on far more risk than stan-dard investment management businesses in the samecapital management activities.

An additional asymmetry between banks and fundmanagers that manage the same investment businessregards accounting rules. Banks have exercised discre-tion in the accounting treatment of assets that weremost affected by the crisis, as Huizinga and Laeven(2009), and Ross Sorkin (2009, pp. 102-104) haveshown. Indeed, Huizinga and Laeven claim:

"Overall, accounting discretion enable the banks tosoften the impact of the crisis on the book valua-tion of assets during the present financial crisis. […]In the present crisis, the financial statements ofbanks appear to overstate the book value of assetsto the point of becoming misleading guides toinvestors and regulators alike."

The discretion available to fund managers is much morelimited; their balance sheet is generally marked-to-mar-

ket by external entities (administrators, custodians).Offshore hedge funds, though they are not subject tospecific regulations, have market-imposed valuationpractices which are typically the standard ones (that is,mark-to-market valuation). Due to a lack of a uniformregulatory framework, however, exceptions are to beobserved here as well.

Finally, the investment business is not highlighted inbanks' financial reports even when, as argued above,such businesses explain most of the volatility of bankprofit and loss.. Indeed, bankers do not even admit inpublic to perform hedge-fund-like activities, that is dis-tinct from, but along with, their client business andwithin their balance sheet, as illustrated by this descrip-tion of the business of Goldman Sachs offered by itsChairman and CEO, Lloyd Blankfein, at a congressionalhearing in the US (Blankfein, 2010):

"Our activities are divided into three general areas:Our investment banking business provides strategiccorporate services, matching the resources of thefirm to specific client needs. This frequently meanscombining advisory, financing and co-investmentcapabilities. We help clients access equity and debtcapital markets in order to grow their businesses,restructure their balance sheets to improve or tosolidify their financial strength and to manage theirassets and liabilities. We also assess and facilitatestrategic options for M&A, divestitures and corpo-rate defense activities. Through our merchantbanking activities, we create and manage invest-ment funds consisting of both our own and ourclients' money in order to invest in growing busi-nesses. Our market making or securities sales andtrading business facilitates customer transactionsfor corporations, financial institutions, govern-ments and individuals through market making andtrading of fixed income, equities, currencies, com-modities and derivatives products. As a marketmaker, we provide the necessary liquidity to helpensure that buyers and sellers can complete theirtransactions and markets can function efficiently.In dislocated markets, we are often required tocommit capital to hold client positions over alonger term while a transaction is completed. Ourasset management and securities services business-es help public and private pension funds, corpora-tions, non-profit organisations and high net-worthindividuals plan, manage and invest their financialassets for the long-term. We also provide theseentities as well as mutual funds and hedge fundswith prime brokerage, securities lending andfinancing services."

In this description the investment business is mentionedas a separate business in merchant banking transactionsand in asset management. However, there is no indica-tion that sales and trading profits originate both client-driven transactions and principal transactions, except forthe fact that sometimes capital is committed to holdpositions, but just for clients, for the purpose of com-pleting a transaction (this is the typical case of securi-C

EP

RP

OL

ICY

INS

IGH

TN

o.45

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

JANUARY 2010 9

Bank Hedge Fund

100 Capital 100 Capital

1250 MBS 1150 Debt 200 MBS 100 Debt

Figure 3 Balance sheets

...banks and investment managersoperate the same business with very

different constraints.

ties underwriting in capital markets).The boundary problem of banks seems to be observed

elsewhere as well, for example in insurance companies,as Ben Bernanke has vehemently remarked in the after-math of the AIG crisis (Bloomberg, March 3):

Federal Reserve Chairman Ben S. Bernanke saidAmerican International Group Inc. operated like ahedge fund and having to rescue the insurer madehim "more angry" than any other episode duringthe financial crisis: "If there is a single episode inthis entire 18 months that has made me moreangry, I can't think of one other than AIG,"Bernanke told lawmakers today. "AIG exploited ahuge gap in the regulatory system, there was nooversight of the financial- products division, thiswas a hedge fund basically that was attached to alarge and stable insurance company." Thus insteadof expressing concern, as some observers have,about "shadow banks", it may seem more appropri-ate to worry about "shadow hedge funds", both inbanks and in insurance companies.

I do not cover insurance companies in this discussionsince my ignorance of them of is even greater than thatof banks. I conjecture, however, that the general issuesrelating to the proper boundaries between the clientand investment business are likely to be the same (wit-ness the AIG case).

To summarise the discussion, this section has illustrat-ed the effects of a natural phenomenon, namely thedivergences between functions and institutions in thefinancial system, caused by the differences between thedynamics of businesses and business opportunities andthe dynamics of rules and regulations that constrainfinancial businesses. I have illustrated cases where cur-rent institutional boundaries may be inappropriate forthe financial businesses subject to them, given the waythey are currently run. In particular, one important casehas been the growth of risky investment businesseswithin banking organisations, whose regulatory con-straints, relating to leverage and accounting, appear tobe designed for client services, and not for an invest-ment business. Indeed, I have also shown that there arenoticeable differences in the constraints to investmentbusiness as run by banks and that run by investmentfunds and/or hedge funds.

5. Financial Reform

In section 2 I have argued that a financial crisis is a mar-ket failure characterised by a generalised and wealth-destructive spike in liquidity demand. Regulatory reme-dies against financial crises include financial supervi-sion, capital requirements, deposit insurance and liquid-ity facilities provided by central banks. These tools, withthe possible exception of liquidity facilities, have failedin the most recent financial crises. In section 3 and 4 Ihave attempted to identify reasons why the standardtools have failed. They include:

• The spreading of liquidity and counterparty riskthroughout the financial system, caused by the

huge growth of securities issuance and securitiesand derivatives trading - thus liquidity crises hitwhere none of the remedies mentioned abovewas present or effective;

• A progressive inadequacy of existing institutionsin providing efficient constraints to financialmarket organisations in the current securities-based financial system - which led to excessiverisk taking (see Rajan (2005) and Lowenstein(2004);

• A deep information deficit of authorities, whohave proved unable to effectively identify the keyweaknesses and risk concentrations in the finan-cial system and take preventive action, as wouldbe expected of a systemic risk manager;

A strategy for financial market reforms flows naturallyfrom the analysis I have provided. While banks are thefirst candidates for regulatory reform, I believe that theeffort to redesign financial regulation should be effi-ciently guided by the principles I laid out above in sec-tion 4. The overall objective of regulatory reform wouldbe to re-establish a situation whereby the regulatoryframework of different actors is appropriate for thebusiness they run.

It is immediately apparent, following the method Ilaid out, that proposals to re-instate Glass Steagall donot seem to be appropriate solutions. Glass-Steagallregulations separated banking businesses from securi-ties businesses. Yet, in my analysis I have argued thatthe financial system is largely based on the securitiesbusiness. It is unthinkable to provide financial servicesto a client separating securities services from cash busi-nesses, since in the financial market cash and securitiestransactions are very closely linked together. Yet, itmakes a lot of sense to separate out client businessfrom investment business, for all the reasons that I havejust recalled. The separation of client business from theinvestment business would confine client servicers -andI believe the name "banks" is appropriate here - to man-age credit risk and liquidity risk, just like the old banksused to do, but this time also in securities space.8

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

CE

PR

PO

LIC

YIN

SIG

HT

No.

45JANUARY 2010 10

8 The typical "practical" exception to a separation of client businessfrom investment business is that to better serve clients banks oftenhave to take securities risks other than credit or liquidity. Thisobjection is true at face value, but not empirically significant.Securities services surely are not only brokerage services, theyinclude dealers' services. These require holding securities invento-ries for some time. However, such holdings have to be limited intime: if they were not they would be proof of the inability of thebank to intermediate client needs. The typical tool used in banksto separate client business from investment business is a very tightrisk budget to the former, and looser risk budget to the latter.

The securities-based financial systemis characterised by an exponential

growth of total counterparty risk, andany mechanism designed to reabsorbcounterparty risk is a move in the

right direction

Continuing with the argument, I find proposals toreturn to narrow banking or so-called utility banksinappropriate, because I still trust, in principle that asystem combining liquidity transformation with institu-tions designed to correct market failures will work.

Along the same lines, it would be very desirable tohave a uniform regulatory framework for capital man-agers bringing hedge funds onshore. Hedge funds aremajor actors in securities and derivatives markets, yetthey are outside mainstream regulatory frameworks andlargely invisible to supervisory authorities. Such reformshould focus on the exposure of different funds to liq-uidity risk in the securities space (again, the experienceof money market mutual funds in the US has been aneye-opener).

Proper boundaries for infrastructure service providerswill certainly complete the buildup of a more robustfinancial system. The standard for evaluating marketinfrastructure is risk management, but this should beaccompanied by governance structures that aredesigned to avoid taking inappropriate risks, either tobeat competition or to increase profitability by access-ing non-infrastructure businesses.

Recent initiatives, in the US and in the EU, to incen-tivise or force the usage of central counterparties forstandardised derivative contracts, are a very welcomestep in the direction of counterparty risk reduction andcounterparty risk control from a system-wide perspec-tive. As I have pointed out above, the securities-basedfinancial system is characterised by an exponentialgrowth of total counterparty risk, and any mechanismdesigned to reabsorb counterparty risk is a move in theright direction; in the same category should be includ-ed rules aimed at regulating orderly resolution of large,systemic financial intermediaries. These initiatives, how-ever, should be accompanied by a careful reassessmentof institutional boundaries of infrastructure providers,along the lines I mentioned above.

After the appropriate institutional boundaries forclient servicers, market infrastructures and capital man-agers have been re-established; there will be a need toredesign the role of authorities. I stick with the basicmodel where one of the main functions of authorities isto perform the role of the systemic risk manager. Thisstarts with gathering and processing information on riskpositions from all financial market participants, espe-cially information relevant to assess liquidity risk. Thismeans giving authorities full access to all activities of allrelevant actors. While this may appear a gigantic task, itis a requirement if we want to have any hope to build afinancial system less prone to systemic crises. To theextent that the information collection preserves thelawful business interests of individual market partici-pants, and there are several ways to maintain confiden-tiality, starting from the obvious requirement thatauthorities do not publish data on specific entities butonly aggregate data,9 the implementation of a global

data gathering project is certainly within reach. Itrequires clear vision by policy-makers who need tounderstand the importance of endowing financial mar-ket authorities with the informational advantage theyneed to fulfil their role in managing systemic risk.

The proper design of the management of systemic riskpresents a vast problem in need of urgent and carefulattention. I note here that markets' awareness ofauthorities' ignorance is a powerful volatility multiplier;anybody who has no view on the systemic risk implica-tions of any major market shock, and knows thatauthorities do not have a view either, has to react to theshock in the most conservative way, by decreasing riskas much as possible. The events of August 2007 are inmy view a good illustration of the fact that markets per-ceived the authorities' lack of information on theunfolding market events, therefore reacting excessively.As soon as the Fed lowered the discount rate on August17, there was a swell of relief that only underlined theconfusion and concern of the previous weeks.Authorities with sufficient, valuable information onfinancial markets systemic risks would be in a positionto publish such information in an aggregate fashion,and offer useful guidance to individual market partici-pants, who typically have limited, incomplete views onaggregate market developments. The various financialmarkets stability reports, published by authoritiesaround the world, could be crucial tools for all marketparticipants if they contained data that are not availablethrough Bloomberg or other services, but are producedby a properly designed data gathering system forauthorities. Instead, until very recently they have beenjust another piece of research, often at par or worsethan that produced by the financial or academic com-munities. In addition to provide the valuable externali-ty of informing the market as a whole about aggregatesystemic risk, authorities with appropriate informationwould be in a position to adopt ad-hoc, ex-post meas-ures able to mitigate systemic risk with much greatereffectiveness.

Even if the strategic directions for financial reformthat I have described in this section were to be agreedupon, it is important to address a few questions relatedto implementation. The standard criticism to marketreform and new regulations, especially in the case offinancial markets, revolves around two themes. The first,as Williamson (2000) following Coase (1964), Demesz(1969) and Dixit (1996) highlight, is that governmentscannot be expected to deliver flawless rules and super-vision. That information to the systemic risk manager iscertainly not sufficient to assure that systemic risks willbe identified and dealt with effectively. There is noquestion that this criticism is correct. However, it doesC

EP

RP

OL

ICY

INS

IGH

TN

o.45

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

JANUARY 2010 11

9 Such information collection could be implemented in an incentive-compatible way, for example by establishing that entities cannotbe prosecuted by authorities on the basis of data releases, but onlyon the basis of specific inspections or external complaints.

A regulatory system based on a number of tools, including reinforceddisclosure to authorities, is certainly

a more robust system than the current one.

not invalidate my argument; a regulatory system basedon a number of tools, including reinforced disclosure toauthorities, is certainly a more robust system than thecurrent one. This criticism is particularly relevant forfocusing on the costs and expected benefits of reforminitiatives, which always need detailed and carefulassessment - though the cost of failure has proven to beastronomical.

The second criticism, often associated with the workof Merton and Bodie (2005), is that financial regula-tions can be bypassed by financial innovation. Indeed,these authors argue that financial organisations inno-vate their way away from the constraints imposed onthem by regulations when their business evolves indirections inconsistent with existing rules. While thephenomenon described by Merton and Bodie is certain-ly present and needs to be taken into account, I do notthink (and I believe that I am also interpreting theseauthors correctly) that this reasoning implies total inef-fectiveness of regulatory constraints on financial activi-ties. The characteristic speed, flexibility, and mobility offinancial organisation certainly needs to be taken intoaccount when considering regulatory design, but doesnot justify inactivity on the regulatory front. Onceagain, this criticism brings the attention on anotheraspect of implementation, essential in financial reform,which is the requirement of uniform initiatives by allrelevant national authorities. Indeed, Initiatives on reg-ulatory reform are being discussed in international fora,such as the Group of 20, the Financial Stability Board,the Bank for International Settlements. The US andEuropean authorities, including the EuropeanCommission, have repeatedly highlighted the need tocoordination and their efforts to achieve it.

6. Concluding remarks

A financial crisis is a market failure which destroys afundamental function of financial intermediation, thefunction of liquidity transformation, with substantialsocial costs. The recognition of the mechanics of finan-cial crises has led to a number of regulatory institutionsthat were designed to minimise the probability that liq-uidity transformation breaks down. Such institutions,however, were designed in a world that was differentfrom today. Financial markets are now characterised bya wide diffusion of securities, derivatives and relatedbusinesses. These developments have had two effects;they have multiplied exponentially transactions andcounterparty risk, and they have led to a progressivedivergence between functions and institutions. In otherwords, individual actors in the financial marketplacehave run businesses within institutions that were notconceived for those businesses. These institutions areinefficient in the current financial system. The end

result is that the mechanics of financial crises havechanged.

The intended contribution of this paper is to offer ananalytical framework to organise a reform strategy. Itthus differs from many reform templates that consist oflists of (generally eminently sensible) reform initiatives.The strategy for financial reform that I propose startsfrom the premise that institutions are very important infinancial markets, and have to be appropriate for thebusinesses they run. I thus propose ways to improveexisting institutions, in the interest of more efficientmarkets. A major aspect of my proposal is a very signif-icant increase in the supervisory responsibilities offinancial authorities, necessary for them to regain effec-tiveness as managers of systemic risk.

References

Acharya, Viral V. and Matthew Richardson eds. (2009),Restoring Financial Stability, New York: John Wileyand Sons

Bagehot, W. (1873), Lombard Street, New York:Scribner, reprinted by John Wiley and Sons

Blankfein, Looyd (2010), "Testimony to the FinancialCrisis Inquiry Commission," January 13.

Brunnermeier, Markus K. and Lasse H. Pedersen (2008),"Market Liquidity and Funding Liquidity", Review ofFinancial Studies, 22(6), 2201-2238.

Coase, Ronald (1964), "The Regulated Industries:Discussion," American Economic Review, 54:3, pp.194-97.

Demestz, Harold (1969), "Information and Efficiency:Another Viewpoint," Journal of Law and Economics,12:1, pp. 1-22.

Diamond Douglas W. and Philip Dybvig (1983, "Bankruns, deposit insurance, and liquidity" Journal ofPolitical Economy 91 (3): 401-19.

Dixit, Avinash (1996), The Making of Economic Policy:A Transaction Cost Politics Perspective, CambridgeUniversity Press.

Geanakoplos, John (2009), "The Leverage Cycle,"Cowles Foundation Discussion Paper 1715, July.

Geithner, Timothy F. (2006), "Hedge Funds andDerivatives and Their Implications for the FinancialSystem," Remarks at the Distinguished Lecture 2006,sponsored by the Hong Kong Monetary Authority andHong Kong Association of Banks, September 15.

Giovannini, Alberto (forthcoming), "Why the EuropeanSecurities Market Is Not Fully Integrated," in AlbertoAlesina and Francesco Giavazzi (editors), Europe andthe Dollar, Chicago, University of Chicago Pres andNBER.

Huizinga, Harry and Luc Laeven (2009), "AccountingDiscretion of Banks During a Financial Crisis", IMFWorking Paper 09/207, October.

International Monetary Fund (2009), Global FinancialStability Report, Responding to the Financial Crisisand Measuring Systemic Risk. Washington, D.C., April.

Lowenstein, Roger (2004), Origins of the Crash, NewYork, The Penguin Press.

Merton, Robert C. and Zvi Bodie (1995), "A ConceptualFramework for Analyzing the Financial Environment,"

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

CE

PR

PO

LIC

YIN

SIG

HT

No.

45JANUARY 2010 12

Individual actors in the financialmarketplace have run businesseswithin institutions that were not conceived for those businesses.

in Dwight Crane et al, (eds) The Global FinancialSystem, Cambridge, MA, Harvard University Press.

Merton, Robert C. and Zvi Bodie (2005), "Design ofFinancial Systems, Towards a Synthesis of Functionand Structure," Journal of Investment Management,vol. 3, No. 1, pp. 1-23.

Rajan, Raghuram (2005), Has Financial DevelopmentMade the World Riskier?", manuscript, September.

Richardson, Matthew (2009), "Causes of the FinancialCrisis of 2007-2009", in Acharya and Richardson(2009).

Ross Sorkin, Andrew (2009), Too Big to Fail, New Yorkand London, Penguin.

CE

PR

PO

LIC

YIN

SIG

HT

No.

45

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

JANUARY 2010 13

Alberto Giovannini is Chief Executive Officer and a founding shareholder of Unifortune SGR SpA, an asset man-agement company based in Milano, Italy. He has also been Principal Policy Advisor of the Commission’s Clearing andSettlement Advisory and Monitoring Group (CESAME).

Mr.Giovannini started his professional career as an academic.He was the Jerome A.Chazen Professor of InternationalBusiness at Columbia University, where he taught and conducted research from 1983 to 1995. He has held a num-ber of positions in the public and private sector, including Co-Chairman of the Council of Experts at the Ministry ofthe Treasury in Rome, Italy (in charge of the Republic of Italy’s international debt program), Senior Strategist at Long-Term Capital Management and Deputy General Manager of Banca di Roma.

Mr. Giovannini graduated from the University of Bologna, and holds a PhD in Economics from the MassachusettsInstitute of Technology.

The Centre for Economic Policy Research, founded in 1983, is a network of over 700 researchers based main-ly in universities throughout Europe, who collaborate through the Centre in research and its dissemination. TheCentre’s goal is to promote research excellence and policy relevance in European economics. CEPR ResearchFellows and Affiliates are based in over 237 different institutions in 28 countries. Because it draws on such a largenetwork of researchers, CEPR is able to produce a wide range of research which not only addresses key policyissues, but also reflects a broad spectrum of individual viewpoints and perspectives. CEPR has made key contribu-tions to a wide range of European and global policy issues for over two decades. CEPR research may include viewson policy, but the Executive Committee of the Centre does not give prior review to its publications, and the Centretakes no institutional policy positions.The opinions expressed in this paper are those of the author and not neces-sarily those of the Centre for Economic Policy Research.