The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
New Financial Order Recommendations by the Issing Committee Preparing G-20 - Washington, November 15, 2008 Otmar Issing (Chairman) JorgAsmussen Jan PieterKrahnen Klaus Regling Jens Weidmann William White New FinancialArchitecture (ShortVersion) 1. Mandate and Overview (OI) 2. Improving the Framework 2.1. Incentives (JPK) 2.2. Transparency (JPK) 2.3. Regulation and Supervision (KR) 2.4. International Institutions (KR) 3. Concluding remarks Appendix (Fulltext) A 1. Overview A 2. Improving the Framework A 2.1. Incentives A 2.2. Transparency A 2.3. Regulation and Supervision A 2.4. International Institutions A 3. Concluding remarks 1. Purpose of the paper - causes of the crisis In linewith the mandate ofthe committee thispaper does not discuss problems ofmanaging the presentcrisis.Itratherdealswith elements ofa future financialarchitecturewhich should help to avoid arepetition ofpast developments. The recommendations include: Higher transparency, proper incentive structures,betterregulation, efficientsupervision, and improved internationalcooperation among authorities. The globalcrisisin financialmarkets isthe consequence ofadynamic interaction between factorsin the macro and the micro sideof the economy. The combination ofmassive liquidityand low interest rateshas caused sharp increases in assetprices,especiallyin the housing sector ofmany countries. In an environment ofinadequate regulation and important gaps in supervisory oversight,inappropriate J incentive structures have promoted short-termism and encouraged theproduction ofcomplex f financialinstruments, supported by high degrees ofleverage. Overall, the situationin booming financialmarkets became more and more unsustainable. Itneeded only atriggertocollapse. This triggerwas deliveredwhen house pricesin theUS startedto fallwith the expectation thatthiswould continue foran extended period oftime. 2. Recommendations 2.1. Incentives This section discusses three major cases ofincentive misalignment that have gready contributed, on themicro level,to the current financialcrisis.These cases referto structured finance, ratingagencies, and management compensation. They all contributed to an understatement of true default risk, generating mispricing of credit instruments. We deal with the structuring of securitization transactions first, since itis here where the current crisis started. We then turn to the role of rating agencies, and lasdy to management compensation. To the extent thatour recommendations relateto betterinformation forinvestors,itwillbe covered inthe following subsection on transparency a) Rating agencies Rating agencies are the major information provider forbond investors in capitalmarkets around the world. By inappropriately applying their long-established bond rating standards to the valuation of structured finance products (tranches), rating agencies have compromised their credibility. In order to re-establish trust in structured finance ratings, a number of new rules should be enforced. For this purpose, agencies should be subjected to regulatory oversight, albeitin alimitedway. Rating performance (i.e. the long-term statistics relating initial ratings to subsequent defaults) should be monitored by the supervisors, applying high standards of transparency and statistics. Rating performance will be published regularly (e.g., once a year). To minimize rating shopping, unsolicited ratings are encouraged (e.g. by mandatory ratingdisclosure). Authorities should continue to review their use of structured finance ratings in the regulatory and supervisory framework und reduce the use to the extent possible in order to limitthe pressure on agencies, e.g.inconsumer protection regulation. Furthermore, rating fees may be linked to rating performance (either by linking it to default statistics, or by linking it to individual tranche performance; in the latter case, agencies essentially buy into the tranches they rate). An annual report on ratingpractice 2 and ratingcompetition by a central oversight body may help to monitor market integrity and quality. b) Structured finance The pooling and tranching of loan portfolios requires special provisions to ensure that the underlying relationships between borrowers and lenders remain intact. Securitizations issued during the past few years typically did not meet these requirements. In order to realign incentives, the whereabouts of the tranches that carry most of the risk, the firstloss pieces, has to be public knowledge. This will ensure that the market can once again price individual securitiesproperly (see the corresponding paragraph under “Transparency”). c) Compensation Management incentives have increasingly focused on short-term performance, disregarding longer-term risks. For instance, various instruments were used to frontend-load financial transactions, and by carving out income via short-term bonuses to management and employees. Since the compensation system influences management behaviour, a stronger reliance on long term performance of investment strategies is required in order to achieve an appropriate incentive alignment. New compensation models may be needed that allow striking a proper balance between long and short term orientation, involving bonus and malus components. Such schemes could resemble pension annuities rather than all-cash-out payments. Since compensation schemes should not be dictated by government decree, disclosure toinvestors will be areasonable regulatory strategy (see the corresponding paragraph under “Transparency”).Of course, regulators may contribute to identifying sound practice principles for compensation schemes. 2.2. Transparency The objective of an increase in transparency is to overcome opaqueness which has prevented investors to assess the risks of financial instruments correctly. The opaqueness, in turn, is to a large extent the consequence ofmisaligned incentives in many areas of the financialvalue chain. Thus, by specifying rules of disclosure, the market will be once again enabled to pricedifferentiatebetween financialinstruments, takingincentive alignment differentialsinto account. a) Structured finance: We propose tomake itmandatory thatin securitization transactions the economic firstloss position isdescribed and itsallocation in the market istreated as a bond covenant, and isdisclosed to the market, thereby avoiding opaqueness ofbank portfoliorisk. Disclosure isneeded for the market to adequately price risks and thus induce banks to retain firstloss pieces, and to apply sound banking practices. Since the size of the firstloss piece alone does not determine the amount of risk retained, the stipulation of any mandatory retention rate has to be discussed carefully. Rating agencies and other information providers (e.g. auditors) are supposed to report regularly on compliance. The revelation of first loss pieces allocation among market participants may even be used to ease tensions on today’s interbank markets. b) Compensation: We propose disclosing, on a regular basis, the incentive components in management compensation schemes, and to encourage rating agencies (and other information providers, like auditors) to report a firm-level metric that captures incentive alignment. Mandatory rules on management compensation, such as salary caps or bonus limits,are not advocated, as they areexpected to backfire. 3 c) Risk map: An internationally coordinated effortin creating the institutionalbackground for drawing, on aregularbasis,aglobal financialriskmap isrecommended. The map involves all major international financial institutions, e.g. banks, insurance companies, hedge funds, and allmajor financial products, e.g. loans, credit insurance and CDS, and ABS. The details of such a riskreporting system stillneed to be worked out, possibly by a task force consisting ofmarket participants,regulators and academics. d) Credit register: Either as a by-product of the risk map, or as a stand-alone initiative, we propose the creation of aglobal creditregister. Such aregistercompiles major interbank and customer-specific exposure data to allow individual market participants, typically those institutions that also deliver the raw data, to learn about major exposures of their counterparties. While the value of such information is appreciated almost universally on a national level, there is nothing commensurate on an international level. However, even the bestregisterwillnot be able toprovide an account oftotalbank exposures inrealtime. e) Accounting standards are to be reviewed. 2.3. Regulation and Supervision The existingregulatory and supervisory system has importantweaknesses and gaps (see 1.)and should be reformed to avoid the build-up ofexcessive leveragein the future and toreduce the system’spro-cyclicality. a) All gaps in the supervisory system should be closed: Financialinstitutions should be supervised on a fullyconsolidated basis thus inhibitingthe scope forregulatory arbitrage. Investment funds operatinginsidebanks, insurance companies or non-financial corporates should be supervised comprehensively. All systemicallyimportant financialinstitutions should be subjected to appropriate supervision and regulation The “indirect” supervision ofhedge funds viatheirregulated creditors needs to be strengthened by consolidating the information available from differentlenders. The activitiesofcreditratingagencies should be monitored, including through the implementation ofthe IOSCO srevised code ofconduct (see2.1.). Market participants should urgentlyimplement centralcounterparty clearing for OTC credit derivatives. Central counterparty clearingwould help to monitor market developments and reduce settlementrisks. Compliance ofoffshore-centers with relevantregulations could be improved through an “indirect” approach ifallmajor financialcenters cooperate. b) Capital requirements should be tightened after the end of the current crisis: Introduction ofan additionaloverallleverage ratioinaddition to the risk-weightedBaselratio. Additional capitalrequirements forSIVs, conduits and off-balance sheetactivitiesunless they are fullyconsolidated. 4 Additional capitalrequirements forlendingtohedge funds and lending to non-cooperative off shore centers. Allowance forliquidityrisks. In addition, capitalrequirements, provisioningrules and accounting rules should be reviewed to make the system lesspro-cyclical. c) Cooperation among the differentauthorities and bodies,which are directlyorindirectly involved inthe supervision offinancialinstitutions, should be improved: Improved information flow between nationalsupervisors to obtain a complete picture of the situationofcross-border banks, bank holdingcompanies, hedge funds, and otherinvestment funds. Creation of“global colleges of supervisors” forcross-border banks. Greater consistencyin the regulation ofdifferent financialentities (banking,insurance, securities). Enhanced interaction between institutionswith experience inassessingmacro-prudential risks and those in charge ofassessingmicro-prudentialrisks. 2.4. International Institutions Internationalinstitutions should play a strongerroleincrisisprevention. Internationalinstitutions can help to build amore robust macro-economic policy framework, akey condition forpreventing futurecrisis. The International Monetary Fund (IMF)’s roleasan international “watchdog” should be strengthened. This relates to macro-economic issues,likeglobalimbalances and exchange rates,and to financialmarkets but the concept of“Bretton Woods II”ismisleading. In particular,the IMF should re-directitswork towards adeeper and more comprehensive analysis of financialmarket developments in the followingway: A stronger focus on spillovers between financialmarkets and the realeconomy. An assessment ofmacro-prudential risks. FinancialSectorAssessment Programs (FSAPs) should become mandatory forallIMF members states. Analysis ofcomprehensive data on global financialmarket developments (asapre-requisite fora globalriskmap, see 2.1.)in close cooperation with theBIS. Itisimportant to note the limits ofwhat can be reasonably expected from the IMF: There isan unavoidable trade-offbetween the neutralityof thework of the IMF staff,based on rules,and the politicalinterestsifitsshareholders. The IMF does not have the expertise to actas standard setterfor financialmarkets. The IMF does not appear to be suitableto become aglobal supervisor forthe largest international financialinstitutions. 5 The Bank for International Settlements (BIS), includingitscommittees, isthe standard setterin financialmarkets. Itwillbe the key body to adopt most ofthe measures proposed above (under 2.3.).However, the BIS suffers from alackoflegitimacyand should broaden itsmembership (particularlyinitscommittees). The Financial Stability Forum (FSF) isthe key forum to coordinate and advocate regulatory and supervisoryreforms. But italso suffers from alackoflegitimacyasitisG7-dominated. The FSF should continue to be anchored attheBIS but the linksbetween the FSF and the IMF should be strengthened. A merger ofthe FSF with the IMF appears neither feasiblenor efficient. The membership ofthe FSF should evolve inlinewith the'future enlargement ofthe G7/8. The European dimension. The current financialcrisishas also highlighted theweaknesses in the EU’ssupervisory framework, which remains fragmented along nationallines.A reform of the European supervisory framework isessentialtoprotect the SingleMarket and the Euro and needs to go beyond what isfeasibleglobally. European issueswillnot be discussed by the G-20 but the EU should take the leadinimplementing theproposals presented inthisnote. 3. Concluding remarks A number ofour recommendations aremore orlessstraightforward and inlinewith those by the FSF, otherinstitutions,or individualauthors. On some issuesmore information and researchis needed. This isone reason towarn againstpremature conclusions and even more againsthasty decisions.The elements ofanew and better financialarchitecturemust interactinapositiveway and support each other. This cannot be achieved by quick and unavoidably short-sighted actions. Itis alsoimportant thatthe crisismanagement —increasinglywith the duration and dimension of measures- takes into account theimpact on the futuredevelopment of financialmarkets. This refers not only to exitstrategies for stateinterventions inbanks etc.,but also to the creation fora stable, non-inflationaryenvironment. Current proposals for strengthening the financial architecture are basically focused on providing betterrules for private actors. This isvery important, but not enough. An analysis of the underlying causes of the crisis show that it would not have gained such significant and cosdy dimensions without a macroeconomic environment of massive liquidity and persistendy low interestrates. As a consequence, sounder macro policies and more prudent rules for individual actors have to go hand inhand. 6 Appendix A 1. Causes of the Crisis Since the end ofthe 1990s and thereafter,interestrateshave been low and liquidityhigh around theworld. This was due inno smallpart tocurrency pegginginkey regions ofthe world economy. The increasingintegration ofChina and other emerging markets into theworld economy resultedinglobalpressure on commodity prices.As itpicked up speed, globalizationlimited the scope ofbusinesses toincrease prices and restrainedwage growth inindustrializednations. The huge riseinliquiditytherefore did not make itselffelton commodity prices forquite a while. Nor did the strategyof“inflationtargeting”,whilstignoringmoney supply and credit volumes, encourage centralbanks toraiseinterestrates. However, the high levelofliquiditydid not remain without consequences. Itled torisingasset prices around theworld. The combination ofhigh liquidityand low interestratesproduced a strongincentive to choose everriskierinvestments. As aresultthe premium over secure investments became ever smaller. This trendwas enhanced by numerous financialinnovations. The securitizationand packaging offinancialproducts — without the originatorretainingany risk— compounded this trend. Complex products were bought world-wide, and inparticularin Europe, in Germany not leastby institutionswith no viable business model oftheirown. The certificatesissued by the ratingagencies underpinned thisbehaviour. Transparency had largelybeen lost. Ithad barelyever existed atgloballevel,and atsome stage theplayers themselves retained only arudimentary overview oftheirown transactions.A lack ofregulationwas revealed, especiallyintheUSA, where politicallyleveraged institutions (Freddie Mac and Fannie Mae) amplified therealestateboom yet further. In numerous countries the banks circumvented capitalregulations by establishingnew entities for their activities.The failureof the oversightregime cannot be excused by the factthatthese activities were entirelylegal. This development occurred againstabackdrop of financialincentives which rewarded short term success — inparticularthe saleofnew financialproducts — with no reference to thelong term. Risk management shortcomings within financialinstitutions have now also been revealed. While maintainingliquidityand protectingdepositors are centralmantras ofbanking regulation, the problem of "market liquidity”has been largelyneglected. Itwas blithely assumed thatitwould always be possible torefinance short-term debt via the markets atany time. These factors combined to produce a situationinwhich only a slightprod was needed to shake the entire (inverse) pyramid, and indeed to bringittoppling down. This prod was delivered by the subprime crisisinthe USA, theimpact ofwhich firstshook the German bankingworld for thereasons familiartous alland continued to spread. In me aftermath, these ripplesproduced by therealestatecrisisspread to the countries thatwere more orless predestined foritbecause theirhousing markets had been overheated foryears. Itwould however be too simplisticand indeed misleading to blame the crisisentirelyon the subprime troubles and the realestatemarkets. As important as this factoris— and asimportant as an end to the housing-market crisisinthe US isforworldwide recovery — the subprime disasterwas only the triggerthatsparked offaworldwide financialcrisis.Risks were re evaluated everywhere — the spreads, riskpremiums rose abruptly, and the high volatilityofthe 7 evaluations contributed to furtherinsecurity.Access to short-term refinancingon the markets was abruptly cut off.This caused an undreamt ofliquiditycrisis.With the collapse oftheir assets,the banks’capitalratiowas put under immense pressure, and the firstinstitutions had to be taken over or ’’rescued”.The sense ofinsecurityreached apeak with the collapse of Lehman Brothers. Since thattime, the authoritiesinmany countries have taken ever more aggressive steps to stabilizethe financialsituation.Liquidityinfusions have been increased,government guarantees have been offered forvarious forms ofprivate debt, and public sector funds have been offered to support the recapitalizationofbanks. While we are stillavery longway from restoringtrustin the markets and between stakeholders, these efforts do seem to have improved the financialsituation somewhat. There have been some signs ofagrowing willingness forbanks to lend to each other forperiods longer than justa few days. Not surprisingly,however, these unusual initiativeshave alsoraised concerns about “moral hazard” looking forward. Nor can itbe judged thattheworst ofthe crisisisbehind us.While the financialsituation might temporarily appear better,the deteriorationin the realside of the globaleconomy is becoming increasinglyevident. Household consumption has weakened sharply,particularlyin countries where savingrates fellmarkedly under theinfluence ofeasy creditconditions. This tendency to retrenchment seems likelyto be exacerbated both by fallingassetprices and tighteningcreditconditions. Investment spending might also fallagainst such abackground. Evidendy, as the economy slows, both household and corporate bankruptcies willrise.This implies furtherlosses to financialinstitutions and potentially stilltightercreditconditions,with furtherimpacts on both the realeconomy and assetprices. In sum, the jointprocess of deleveraginginboth the realand financialsectors might stillhave a significantway to run. The seriousness of the current situationevidendy raises firstthe challenge of furtherpolicy actions to help manage the crisis.At the same time, these circumstances alsopresentus with a politicalwindow ofopportunity forreforms designed to help minimize the severityoffuture crises.This opportunity should not be missed, even ifarepetition ofrecent events might be thought unlikely for some time to come. This fact,thatwe do have time means thatsuggested reforms can be wellresearched and carefullythought outinadvance. Impetuous action, particularlyifmotivated by the desire topunish those held to blame forour current troubles, could easilyprove counterproductive. Well considered and crediblereforms, valuable in themselves, could alsoplay a significantrolein offsettingthe moral hazard associatedwith the measures currendy needed forcrisismanagement. In thisfundamental sense, crisis management and crisisprevention should not be seen as substitutes but complements. A 2. Improving the Framework A 2.1 Incentives In this section we discuss severalincentive misalignments that are typical for the industry. They allwork in the same direction, namely they underestimate the true probability of default on the debt instruments being offered for sale. We startwith the structuring of transactions, then turn to the roleofratingagencies, and lasdy tomanagement compensation. The 2007 credit crisis evolved initially around the subprime mortgage market in the US, which entailed lending programs, often government sponsored, targeting low wealth individuals. Since personal income of debtors tended to be low as well, lending in these markets was typically based on asset value alone. Mortgage Backed Securities (MBS) offered an easy ways to access capital market funding, despite the low incomes of those taking out mortgages. Government 8 sponsored programs (Fannie Mae and Freddie Mac) played an important role in market development, both in terms ofincreasingmarket volume, and in terms ofpushing debtor quality tolower levels. Using mortgage lending to enhance home ownership among lower income strata was an explicit objective of US Congress, and government sponsored programs were eager to comply. Securitization of mortgage loan portfolios, and more generally of all types of consumer and commercial lending, is in general a useful way to mobilize funding for retail and commercial credit, and as such a viable alternative to bank deposits. Furthermore, properly designed securitization programs can address successfully the destruction of lender incentives that might be expected to go along with outright loan sales. Careful borrower selection, continuous loan monitoring, timely restructuring and intervention in default situations — all these elements of prudent and efficient banking can be upheld in securitization transactions. However, incentive alignment requires securitization transactions to obey specific technical rules of structuring, which in essence put a cap on the amount of risk that can be shed on the market. This rule, however, was not respected in securitizationmarkets inrecentyears. This led to the over-pricing of low qualityMBS atissue. Similar qualityproblems have begun to surface in other markets as well (e.g. credit card and car loans, and cov-lite corporate loans) with the fullimpact likely still yetto be seen. Why did investors not respond rationally by increasing the required risk premium, thereby demanding an adequate compensation for the increase in risk? One reason is,in addition to the increasing appetite for risk prevailing across virtually all markets at the time, was the lack of information on the true risk properties of the underlying loan portfolios, in particular on the whereabouts of the first loss position. One has to recall that up until the early years of the century, perhaps until 2004, the standard securitization practice entailed retaining the firstloss position. When, where, and to what extent retention standards in the industry were changing, is stilllargelyunknown. But itisconsensus opinion that these standards have been weakened over time, and thatinvestors were ignorantof these changes foralong time. When investors began to focus on the possibility of such changes, probably inmid-2007, itbecame evident that the prices of structured instruments could no longer be based on originally assigned ratings. Combined with the more general retreat from imprudent risk taking, this contributed to the virtual shut down ofmarkets for structuredproducts. This raises the issue of the role of rating agencies (RAs) in these markets. In short, RAs have played the role of trusted gatekeepers. ABS market development would not have been possible without RAs providing quality assurance about inherentiy complex financial products to unsophisticated investors. Most observers believed that RAs were technically well equipped to perform this task, and in particular that the RAs reputational capitalwould serve as a strong line ofdefence againstany compromising ofratingstandards. We have learned otherwise. First, the self-disciplining role of reputation cannot always be relied upon, even under normal market conditions. This isparticularly true for high quality bonds and notes, mainly because the implied default probabilities are so small. Itwould need very long periods to verify statistically that rating standards have been compromised, and it therefore remains unclear how agencies thatcheated would be punished by themarket. Second, ratings could be gamed, i.e.clientsmay have been able to outsmart the rating standards. Consider ratings assigned to structured products, like MBA tranches. They are based on estimating the loss distribution of the underlying loan portfolio. These estimations rely on models that are not fully transparent to the industry. However, RAs provided “customer end” tools to their clients which allowed banks to run pre-tests of their new securitization portfolios before submitting to the RA. As aresult,loan portfolios could be designed in away thatjustmet 9 the criteriaincluded in the relevantmodel, but may have additional riskpertaining to criterianot included in the model. As an example, consider information on whether firstloss tranches are retained or not. Although we know that the saleof the so-called firstloss piece lowers expected portfolio returns, thisisnot an explicitpart of the ratingmodel currentlyinuse, and istherefore not reflected in the assigned rating. Hence the issuer can raise its profits by selling first loss pieces,without disclosingitto the investors. Note that the gaming argument refers to Structured Finance (SF) products only, not to corporate bonds in general. The reason is that in SF, the tailoring of portfolio composition is feasible (easy, low cost), while it is infeasible (difficult, expensive) ifthe underlying assetpool isan entirecorporation with itsfixed assets. Third, ratingshopping by issuers also contributed to agradual erosion ofratingstandards among structured finance products. The negative effect follows from the right of issuers to suppress ratingswhich itdeems to be unwelcome (unsolicited), thereby exertingpressure on the agencies. Incentive misalignment in the industryisnot confined to inappropriate securitization techniques and incentive conflicts at rating agencies. It also entails compensation systems, both for top management and for employees and line management, particularly in trading and sales. The problem isan incentive structurewhich induces a focus on short-term profitability,and aneglect of longer-term risks. Examples are bonus payments due at, or shordy after, selling a mortgage loan, or fees forissuing structured finance securities.Alternative compensation systems thatlead to a fullappreciation ofriskand returnwillstretch the payoffatleastpartiallyover the lifeof the underlying business, thereby making it performance-dependent. The general idea is to add a malus component to the compensation scheme. Here isalistofconcrete measures suggested by the above arguments. 1. Firstlosspieces The issuer should be obliged (by law) to inform the market of any change in the size or composition of his stake in the firstloss piece. This should encourage him to make a binding commitment to retain the firstloss piece. On the other hand, ifthe issuer wants to sellthe first loss poition, then the market can respond immediately, thereby allowing price differentiation of issueswith and without recourse. 2. Ratings. Regulatory oversight addresses incentives problems in four areas. First, rating performance (i.e. the long-term statisticsrelatinginitialratings to subsequent defaults) should be monitored by the regulators, applying high statistical standards. Rating performance relative to outcomes should be published regularly (e.g., once a year). Second, to minimize rating shopping, unsolicited ratings should be encouraged (e.g. by mandatory rating disclosure). Third, the use of structured finance ratingsin public regulation has to be reconsidered (and dropped ifnecessary) in order to limit the pressure on agencies, e.g. in Basel II regulation, or in consumer protection regulation. Fourth, agencies should be encouraged to adjust theirrating methodology to innovations in the financial industry, e.g. to flag structured finance ratings, to reveal incentive alignment and first loss piece retention as part of the ratinginformation. Furthermore, rating fees should be linked to rating performance (either by linking it to default statistics, or by linking it to individual tranche performance. In the lattercase, agencies might be obliged to buy into the tranches they rate (ineffect to put theirmoney where theirmouth is). An annual report on ratingpractice and ratingcompetition by a centraloversightbody might help both to monitor market qualityand to draw attention to outstanding analyticaluncertainties ofwhich investors might be unaware. 3. Compensation 10 Since the compensation system influences management behaviour, a stronger reliance on long term performance of investment strategies is required in order to achieve incentive alignment. New compensation models maybe needed thatallow strikingaproper balance between long and short term orientation, e.g. bonus and malus systems, comparable to pension annuities. We propose that the incentive components in management compensation be disclosed on a regular basis. As well rating agencies (and other information providers, like auditors) should be encouraged to report a firm-level metric that captures incentive alignment in management compensation. 2.2 T ransparency The issue of transparency is closely related to the idea of incentive alignment. Through the disclosure of appropriate information market participants are assisted to make correctinferences on, forexample, projectriskand return, and market pricingand liquidity. Increased transparency is therefore a key requirement for improving market functionality. This being said, one has to keep in mind that there are certain institutions, like rating agencies, where some degree of secrecy is required for them to perform their task properly. This is why transparency of rating processes, as distinct from ratingperformance, isnot recommended, in order to avoid a gaming ofthe system. Turning to the inter-bank liquidity crisis that started in the fallof 2007, we see opaqueness of true risk exposures after several years of deteriorating credit standards as the major reason for the near-universal lack of trust among banks. It seems that on the level of bank assets, even experts in banks and rating agencies were no longer able to properly assess the characteristics (e.g. the moments) of the underlying credit risk. One reason for the unprecedented degree of opaqueness was once again the rising importance of structured finance. These financial instruments had loss rate distributions that could not be properly assessed without additional information which was not in fact readily available in the market. In particular, one important missing piece of information related to how the incentive alignment between originator and ultimate borrower had changed dramatically aslending standards worsened after2005. This isan example of a market where an increased degree of disclosure seems likely to improve market qualitygready. Turning to the market level, the devastating impact that a relatively small segment of the credit market, the subprime housing market, had on financial stability in the US and in Europe is notable. We see one reason for the crisis escalation in a high degree of interconnectedness between financial institutions that had escaped financial stability oversight. While we are doubtful that there will ever be powerful early warning systems in financial markets, an improved world-wide risk oversight might help to guide central bank and bank supervision activities in the future. We therefore envisage the development of a financial risk map, which would be regularly updated, summarizing both the exposures of major international financial institutions and anomalous market developments that might presage future disruptions. Such a risk map would also compile information on the shadow banking system, i.e. on SIVs, hedge funds, CDS in order to note early on any tectonic changes in the allocation of risk among financialinstitutions— something thatwas missing thistime around Ithas to be stressed that the methodology for constructing such a financial riskmap stillneeds to be developed. Furthermore, for regularly updating the information in the risk map, an institution with adequately qualified staff has to be assigned, possibly the IMF or the BIS. A closely related task, which can be embedded in the concept of a financial risk map, is the creation ofan international creditregistrar. Such institutions existin allmajor countries, but they do not yet aggregate information on cross-border transactions. We now know that cross-border financial links are abundant, and need to be considered when evaluating the solvency of any financialinstitution. The creditregistrarin Germany iscalled ccEwden%%entrale”, and itiscurrently managed by the Bundesbank. Although, of course, no real time register can be expected, a regularly matrix of liabilities between banks and major corporates may help to maintain transparencyininternational corporate financing,includinginterbank lending. Here isalistofconcrete measures suggested by the above arguments. 1. We propose to make it mandatory that in securitization transactions the economic first loss position is described and itspotential allocation in the market is treated as a bond covenant. It would therefore be disclosed to the market. Retention by the originatorneed not and should not be mandatory, only disclosure is needed for the market to sort it out, thereby avoiding opaqueness of bank portfolio risk. Rating agencies and other information providers (e.g. auditors) are supposed to report regularly on compliance. Note that the allocation of firstloss pieces among market participants has remained a secretto date. 2. Furthermore, and reiterating section A 2.1, we propose to disclose on a regular basis the incentive components in management compensation, and to encourage rating agencies (and other information providers, like auditors) to report a firm-level metric that captures incentive alignment in management compensation. Once again, mandatory rules on management compensation, such as salary caps or bonus limits, are expected to backfire. Rather than micromanaging compensation systems, which is likely to produce distortions in the economy, an improved approach to financial regulation needs to overcome opaqueness on corporate incentive structures, allowing the market toprice-differentiateaccording toincentive alignment. 3. We suggest an internationally coordinated effort to draw up a global financial risk map that could be updated on a regular basis. The map would attempt to identify building stresses at all major cross-border financial institutions and in all major markets (including derivatives and markets for credit risk transfer). The details of such a risk reporting system would take time to work out,possibly by a task force consistingofmarket participants,regulators and academics. 4. Either as a by-product of the risk map, or as a stand-alone initiative, the creation of a global credit register might be considered. Such a register would compile major interbank and customer-specific exposure data to allow individual market participants, typically those institutions that also deliver the raw data, to learn about major exposures of theircounterparties. While the value of such information isappreciated almost universally on a national level,there is nothing commensurate on an international level. Evidendy, even the best register will not be able to provide a survey about bank exposures in real time. However, such a registerwould still capture the longer term trends that history shows have often posted the biggest threat to financialstability. A 2.3 Regulation and Supervision The crisishas shown thatthe existingregulatoryand supervisory system has important weaknesses and gaps (see 1.).In particular, supervisors should thinkmore systematicallyand understand the necessityto moderate the inherentpro-cyclicalityoffinancialmarkets. The objectivemust be to avoid the build-up ofexcessive leverage and toreduce the system’spro-cyclicality. Consequendy, • Gaps in the supervisory system should be closed, 12 • Capitalrequirements should be tightened and • The cooperation ofdifferentauthoritiesinvolved in supervision should be improved. a) Gaps inthe supervisory system have been identifiedin the followingareas: off-balance sheetactivitiesofbanks investments funds (includinghedge funds and funds thatoperate insidebroader financial institutions,e.g.insurance companies) creditratingagencies derivatives offshore centers The objectivemust be to have a supervisory systemwithout any gaps (see2.1.).However, the necessarymeasures willdifferdepending on the nature ofthe activities.The followingmeasures should be considered: Regulators should supervise financialinstitutionson a fullyconsolidated basis (as foreseen in Basel IIto alargeextent) to avoid the emergence ofunsupervised riskexposure outside the balance sheets. This should help toreduce regulatoryarbitrage. Investment funds operatinginsidebanks, insurance companies or non-financial corporates should be supervised comprehensively. The “indirect” supervision ofhedge funds viatheirregulated creditors needs to be strengthened by consolidatingtheinformation available from differentlenders. The activitiesofcreditratingagencies should be monitored, including through the implementation ofacode ofconduct (see2.1.). The creation of acentralised tradingplatform forderivativeswould help to monitor market developments and reduce settlementrisks. Compliance ofoffshore-centers with relevantregulations could be improved through an “indirect” approach ifallmajor financialcenterswere to cooperate (comparison with FATF) in theiroversight.. b) Capital requirements under Basel 1have notbeen sufficientto avoid the build-up ofexcessive leverage, and this does not seem likelytoimprove (indeeditmay worsen) under Basel 2..A tighteningalong the followinglines should be considered afterthe end ofthe currentcrisis: Introduction ofan additionaloverallleverageratioof,perhaps, 20 (as suggested by the Swiss authorities) in addition to the proposed risk-weighted ratiosunder Basel 2. Additional capitalrequirements forSIVs, conduits and off-balance sheet activitiesunless they are fullyconsolidated. Additional capitalrequirements forlending tohedge funds and lending to non-cooperative off shore centers. Allowance forliquidityrisks In addition, capitalrequirements and provisioningrules should be reviewed to make the system less pro-cyclical.This could be done by activatingone ofthe following steps during periods ofhigh economic growth (likein2004-07) or high creditgrowth: o Reduction ofthe overallleverage ratio. o Grossing up (under pillar2) ofthe capitalrequirements given by pillar1. o Increase inpremiums fordepositinsurance. The activationof such stepswould likelybenefit from being “automatic”,according to a pre-definedrule.The effectofaccounting rules (fair-valueaccounting) on capitalrequirements should alsobe reviewed to establishwhether compensating supervisoryrequirements might be needed. c) Cooperation among the differentauthorities and bodies,which are directlyorindirectly involved in the supervision of financialinstitutions, should be improved. The following possibilities should be considered: The information flow between national supervisors must be improved. This isessentialto obtain acomplete picture of the situationofcross-border banks, bank holding companies, hedge funds, and otherinvestment funds. The creation of “global colleges ofsupervisors” forcross-border banks (along the linesofthe colleges now created inthe EU); thiswould facilitatethe necessaryinformation flow atalltimes but also the prospects ofquick regulatoryaction (includingthe use ofpublic funds, ifnecessary) during difficulttimes. Greater consistencyin the regulation ofdifferent financialentities (banking, insurance, securities)when conducting similarbusinesses.. Enhanced interaction between institutionswith experience inassessingmacro-prudential risks (centralbanks, International Financial Institutions) and bodies in charge ofassessingmicro prudentialrisks (nationalregulators).An effectiveearlywarning system would need to relyon the expertise ofboth groups ofinstitutions. A 2.4 International Institutions As financialmarkets become more globalinnature,internationalinstitutions should play a stronger roleincrisisprevention. However, we believe thatthe term “Bretton Woods II” could be easily misunderstood in thiscontext. The current crisiswas not caused by the prevailing“floating” exchange rate system. Rather, underlying trends toprocyclicalitywere enhanced by the effortsof countrieswith large trade surpluses to resistappreciation oftheirexchange rates.Accordingly, stronger surveillance over macro-economic policies,exchange ratepolicies and globalimbalances are desirable. A strongerrole forinternationalinstitutionsisimportant because alarge number ofindividual countries cannot coordinate complex financialmarket issuesinan inter-governmentalway. However, international standard-setting foraand informal bodies should broaden theirmembership to strengthen theirlegitimacy. One ofthe key questions inthiscontextishow tomake the Financial StabilityForum (FSF) more legitimateand more effective. Crisisprevention in the futurewillalsorequire amore robustmacro-economic policy framework. For many years,monetary and fiscalpolicieswere over-expansionary inmany countries and thus contributed to bubbles and the current crisis.The IMF, OECD and the European Commission should improve theirassessment of fiscalpolicies. Currentinstruments to calculate cyclically adjusted budget balances areinsufficient, forexample. Centralbanks that focused too narrowly on inflationtargetingand neglected the rapid growth of ofmonetary and creditaggregates may have 14 missed earlysigns ofthe emergence ofbubbles in financialassets and housing. Policylooseningin downturns has often been fasterthan policytighteninginupswings. The International Monetary Fund (IMF) istheleadinginternationalinstitutionwith aglobal membership thatsupervises theinternational financialsystem. The IMF’sroleas an international “watchdog” should be strengthened. This relatestomacro-economic issues,likeglobalimbalances and exchange rates,and to financialmarkets. To thisend, the IMF should re-directitswork towards adeeper and more comprehensive analysis of financialmarket developments in the followingway: A stronger focus on spillovers between financialmarkets and the realeconomy initscountry reports and the semi-annualWorld Economic Outlook. An assessment ofmacro-prudential risksinitscountry reports and the semi-annual Global FinancialStabilityReports. FinancialSectorAssessment Programs (FSAPs) should become mandatory forallIMF members statesand be conducted on aregularbasis. Comprehensive data on global financialmarket developments (asapre-requisite foraglobalrisk map, see 2.1.) should be collected by theBIS and analysed by the IMF. Closer cooperation between IMF and BIS should be encouraged in thisrespect. With adeeper and more comprehensive knowledge of financialmarket developments and risks (in addition to the Fund’straditionalstrength on macro-economics), the IMF could become abetter “early-warner”. However, there arelimits towhat can be reasonably expected from the IMF: The more competences and politicalpower the IMF obtains, the greater theinterestinpolitical control over itswork willbe. There isan unavoidable trade-offbetween the neutralityofthe work ofthe IMF staff,based on rules,and thepoliticalinterests ofitsshareholders. The IMF does not have the expertise to actas standard setterfor financialmarkets. This taskhas been carried out successfullyby committees ofnationalexperts meeting atthe BIS (seebelow) and elsewhere. Double-work should be avoided. The IMF does not appear to be suitable tobecome aglobal supervisor for the largest internationalbanks since thatwould interferewith national sovereignty,including the rightof Parliaments to approve budgetary expenditures incase ofsolvency problems (more suitable solution: colleges of supervisors, see above). Finally,a stronger role for the IMF requires thatthevoting structure (quotas) ofitsmembers iskept under permanent review. The growing weight ofemerging economies willrequire furtherreductions intheweight ofadvanced economies in the IMF. The Bank for International Settlements (BIS) has aclearmandate forglobal financialstability, has more expertiseinregulatory and supervisoryissues than any otherinternationalinstitution,was more successful than others inwarning earlyabout the currentcrisisand collects ahuge amount of financialmarket data.These strengths oftheBIS should be preserved and used effectively. Most ofthe measures proposed above (under 2.3.)would have to be discussed and agreed in the framework of theBIS or one ofitscommittees. However, the BIS suffers from alackoflegitimacybecause itscore decision-making bodies and committees are dominated by G-10 countries. For example,Japan isthe onlyAsian country in the key “Basel Committee on Banking Supervision”;China and Mexico are the only representatives from emerging markets on theBIS Board thatcomprises 20 members. The Financial Stability Forum (FSF), createdin 1999 tobring together Finance Ministries, CentralBanks and Supervisors from G7 countries,has become the key body to coordinate and 15 advocate regulatory and supervisoryreforms. Ithas worked well for the G7 but also suffers from a lackoflegitimacygloballyas only avery smallnumber ofnon-G7 financialmarket centers attend as observers. No large emerging economy isrepresentedinthe FSF. A key question willbe how to use the expertise ofthe FSF effectivelyatthe globallevel.We believe thatthe FSF should continue to be anchored attheBIS (where currently the Secretariatofthe FSF islocated) but thatthe links between the FSF and the IMF should be strengthened. A merger ofthe FSF with the IMF appears neither feasiblenor efficient. However, ways to strengthen the legitimacyofthe FSF, without loosingitseffectiveness, need to be considered. The membership of the FSF could evolve inlinewith the futurecomposition ofthe G7/8. The G7 has played the principal steeringrole fortheglobal financialsystem forthree decades. This is changing and the meeting ofthe G20 inWashington on 15 November 2008 isa clearindication of that. Ifthe G7/8 were enlarged by the largestemerging economies, such as China, India and Brasil, this should have automatic consequences atthe “working level”,such as the FSF, but also the Basel Committee on Banking Regulation. The European dimension. The current financialcrisishas alsohighlighted theweaknesses in the EU’ssupervisory framework, which remains fragmented alongnationallinesdespite the substantial progress achieved in financialmarket integrationover the lastdecade and theincreased importance ofcross-border entities.A reform of the European supervisory framework isessentialto create a trulyintegrated financialmarket in the EU and toprotect the Single Market and the Euro. Although these issues are unlikely to be discussedby the G-20 on 15 November, the European experience can be usefulinrevealing the problems and pitfallsincreatingamore efficient,lesscrisisprone regulatory system. For example, while itappears essential,over time, to establish aSingle Regulator for thelargestcross-border banks in the EU, thisseems to be impossible atthe global levelbecause ofpotentialbudgetary implications and unresolved accountabilityissues.The EU has a 50-yearhistoryoftransferringsovereignty to theEuropean levelifnecessary forthe good functioningofthe SingleMarket. This willmake itmuch easier (not easy) for the EU tomake progress inallthe areas mentioned above (2.1 to 2.3)related to transparency,incentives,regulation and supervision. The EU should therefore take theleadinimplementing these proposals. A 3. Concluding remarks A number ofour recommendations aremore orlessstraightforward and inlinewith those by the FSF, otherinstitutions,or individualauthors. On some issues more information and research is needed. This isone reason towarn againstpremature conclusions and even more againsthasty decisions.The elements ofanew and better financialarchitecturemust interactin apositiveway and support each other. This cannot be achieved by quick and unavoidably short-sighted actions. Itis alsoimportant thatthe crisismanagement —increasinglywith the duration and dimension of measures- takes into account theimpact on the futuredevelopment of financialmarkets. This refers not only to exitstrategies for stateinterventions inbanks etc.,but also to the creation fora stable, non-inflationary environment. Proposals for strengthening the financial architecture are basically focused on providing better rules for private actors. This isvery important, but not enough. As the analysis of the underlying causes shows, the crisiswould not have gained such adimension without amacroeconomic environment of massive liquidity and persistent low interest rates. As a consequence sound macro policies and prudent rules forindividualactors have togo hand inhand. 16 17
@FedFRASER