Risk management for central banks and other public investors

List of figures page xJose´-Manuel Gonza´lez-Pa´ramo Ulrich Bindseil, Fernando Gonza´lez and Evangelos Tabakis 1 Central banks and other public institutions as financial investors 3 Ulri

and Other Public Investors

Domestic and foreign financial assets of all central banks and public wealth funds worldwide are estimated to have reached more than USD 12 trillion in 2007 How

do these institutions manage such unprecedented growth in their financial assets and how have they responded to the ‘revolution’ of risk management techniques during the last fifteen years? This book surveys the fundamental issues and techniques associated with risk management and shows how central banks and other public investors can create better risk management frameworks Each chapter looks at a specific area of risk management, first presenting general problems and then showing how these materialize in the special case of public institutions Written by a team of risk management experts from the European Central Bank, this much-needed survey is an ideal resource for those concerned with the increasingly important task of managing risk in central banks and other public institutions.

Ulrich Bindseil is Head of the Risk Management Division at the European Central Bank.

Fernando Gonza´lez is Principal Economist at the European Central Bank.

Evangelos Tabakis is Deputy Head of the Risk Management Division at the European Central Bank.

Risk Management for

Central Banks and

Other Public Investors

Edited by

Ulrich Bindseil,Fernando Gonza´lez and

Evangelos Tabakis

Cambridge University Press

The Edinburgh Building, Cambridge CB2 8RU, UK

First published in print format

ISBN-13 978-0-521-51856-7

ISBN-13 978-0-511-47916-8

© Cambridge University Press 2009

2009

Information on this title: www.cambridge.org/9780521518567

This publication is in copyright Subject to statutory exception and to the

provision of relevant collective licensing agreements, no reproduction of any partmay take place without the written permission of Cambridge University Press

Cambridge University Press has no responsibility for the persistence or accuracy

of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain,

accurate or appropriate

Published in the United States of America by Cambridge University Press, New Yorkwww.cambridge.org

eBook (EBL)hardback

List of figures page x

Jose´-Manuel Gonza´lez-Pa´ramo

Ulrich Bindseil, Fernando Gonza´lez and Evangelos Tabakis

1 Central banks and other public institutions as financial investors 3

Ulrich Bindseil

3 How policy tasks have made central banks large-scale investors 10

4 Optimal degree of diversification of public institutions’

5 How actively should public institutions manage their

2 Strategic asset allocation for fixed-income investors 49

Matti Koivu, Fernando Monar Lora, and Ken Nyholm

v

4 Forward-looking modelling of the stochastic factors 75

3 Credit risk modelling for public institutions’ investment

Han van der Hoorn

2 Credit risk in central bank and other public investors’ portfolios 118

3 The ECB’s approach towards credit risk modelling: issues

4 Risk control, compliance monitoring and reporting 157

Andres Manzanares and Henrik Schwartzlose

2 Overview of the distribution of portfolio management tasks

Herve´ Bourquin and Roman Marton

Roman Marton and Herve´ Bourquin

Part II: Policy operations 269

7 Risk management and market impact of central bank

Ulrich Bindseil and Francesco Papadia

3 A cost–benefit analysis of a central bank collateral

8 Risk mitigation measures and credit risk assessment in central

Fernando Gonza´lez and Phillipe Molitor

9 Collateral and risk mitigation frameworks of central bank

policy operations – a comparison across central banks 340

Evangelos Tabakis and Benedict Weller

10 Risk measurement for a repo portfolio – an application to the

Eurosystem’s collateralized lending operations 359

Elke Heinle and Matti Koivu

7 Residual risk estimation for the Eurosystem’s credit operations 387

4 Financial stability role of central bank operational framework 416

5 The inertia principle of central bank risk management

Part III: Organizational issues and operational risk 441

12 Organizational issues in the risk management function

Evangelos Tabakis

4 Six principles in the organization of risk management

2.1 Evolution of Strategic Asset Allocation page53

2.3 Adapted efficient frontier and VaR constraint 65

2.6 The overall policy structure of the investment process 73

2.9 Normal macroeconomic evolution: (a) GDP YoY % Growth;

2.10 Projected average evolution of the US Government yield curve

2.11 Projected distribution of yields in a normal example:

2.12 Distribution of returns in a normal example: (a) US Gov 0–1Y;

2.13 Inflationary macroeconomic evolution: (a) GDP YoY %

2.14 Projected average evolution of the US Government yield

2.15 Projected distribution of yields in a non-normal example:

2.16 Distribution of returns in a non-normal example:

3.1 Asset value and migration (probabilities not according to scale) 130

3.2 Impact of asset correlation on portfolio risk (hypothetical

portfolio with 100 issuers rated AAA–A, confidence level 99.95%) 142

3.3 Comparison of portfolios by rating and by industry 144

3.5 Comparison of simulation results for Portfolios I and II 152

x

3.6 Lorenz curves for Portfolios I and II 153

7.1 Marginal costs and benefits for banks of posting collateral

7.2 One-week moving average spread between non-EEA and EEA

7.3 Spread between the three-month EURIBOR and three-month

EUREPO rates since the introduction of the EUREPO in

7.4 Evolution of MRO weighted average, 1 Week repo, and

1 Week unsecured interbank rates in 2007 298

7.5 Evolution of LTRO weighted average, 3M repo, and 3M unsecured

8.1 Risks involved in central bank repurchase transactions 305

8.2 Basic determinants of haircut calculations 319

8.4 Relationship between position size and liquidation value 324

8.6 Value-at-Risk due to credit risk for a single exposure 334

10.1 Important types of concentrations in the Eurosystem

10.4 Herfindahl–Hirschmann Indices (HHI) of individual

counterparties with respect to their collateral submitted 375

10.5 Variance reduction factors, for varying values of ^Ł and

10.6 The effect on Expected Shortfall of changed liquidation

time assumptions Source: ECB’s own calculations 390

10.7 The effect on Expected Shortfall of changed credit quality

assumptions Source: ECB’s own calculations 391

10.8 The effect on Expected Shortfall of changed assumptions

11.1 Liquidity shocks and associated marginal costs to a specific bank 424

13.2 Drivers of the risk impact-grading scale of the ECB 474

13.3 Operational risk tolerance: illustrative principles 475

1.1 Foreign reserves (and domestic financial asset of G3 central

1.2 Different reasons for holding foreign exchange reserves –

importance attributed by reserve managers according to

1.3 Risk quantification and economic capital, in billions of EUR,

1.4 Modified duration of fixed-income market portfolios 19

1.5 Asset classes used by central banks in their foreign

1.6 Asset classes currently allowed or planned to be allowed

1.7 Derivatives currently allowed or planned to be allowed

1.8 Trading styles of central bank reserves managers according

2.1 Example of the eligible investment universe for a USD portfolio 100

2.4 Intercepts of the Nelson–Siegel state equation 102

2.5 Autoregressive coefficients of the Nelson–Siegel state equation 103

2.6 Returns in a normal example: average and standard deviation 108

2.7 Optimal portfolio composition in a normal example 110

2.8 Summary information for the optimal portfolio in a normal example 110

2.9 Returns in a non-normal example: average and standard deviation 114

2.10 Optimal portfolio composition in a non-normal example 116

2.11 Summary information for the optimal portfolio in

3.1 Migration probabilities and standard normal boundaries for bond

xii

3.2 Risk-weighting of Standardized Approach under Basel II 135

3.3 Original and augmented migration probabilities for bond

3.4 Common migration matrix (one-year migration probabilities) 146

3.7 Decomposition of simulation results into default and migration 148

3.8 Simulation results for Portfolio II, including decomposition 151

4.1 Rating scales, numerical equivalents of ratings and correction

7.1 Shares of different types of collateral received by 113 institutions

responding to the 2006 ISDA margin survey 279

7.2 Comparison of the key recommendations of ISDA Guideline

for Collateral Practitioners with the Eurosystem collateralization

7.5 Social welfare under different sets of eligible collateral and

7.6 Information on the set of bonds used for the analysis 292

7.7 Spreads containing information on the GC and Eurosystem

collateral eligibility premia – before and during the 2007 turmoil 299

8.1 Summary of ECAF by credit assessment source in the context

8.3 Eurosystem liquidity categories for marketable assets 332

8.4 Eurosystem levels of valuation haircuts applied to eligible

marketable assets in relation to fixed coupon and zero-coupon

8.5 The distribution of bond values of an A rated bond 335

8.6 ‘Through-the-cycle’ credit migration matrix 336

8.8 99 per cent credit risk haircut for a five-year fixed coupon bond 337

9.1 Differentiation of collateral policy depending on type of operation 343

9.2 Comparison of sizes of credit operations (averages for 2006,

9.3 Comparison of eligibility criteria 350

9.4 Comparison of haircuts applied to government bonds 355

9.5 Comparison of haircuts of assets with a residual maturity

10.1 Default probabilities for different rating grades 363

10.2 Liquidation time assumptions used for the different asset classes 364

10.3 Comparison of various variance reduction techniques

10.4 Comparison of various variance reduction techniques

10.5 Breakdown of residual risks in the base case scenario 389

10.6 Composition of submitted collateral over time and composition

11.1 FCM typology and illustration from August–December 2007 400

2.1 The VAR macro model page 78

2.2 Transformation of yields and relative slope 83

3.1 Credit spreads and the limitations of diversification 121

4.2 Calculation of rate reasonability tolerance bands at the ECB 184

4.4 The systems used by the ECB Risk Management Division (RMA) 201

8.1 Historical background in the creation of in-house credit

assessment systems in four Eurosystem central banks 310

8.2 In-house credit assessments by the Bank of Japan 311

8.3 The Qualified Loan Review programme of the Federal Reserve 312

9.1 Survey of credit and market risk mitigation in a collateral

xv

The reader familiar with central bank parlance will have certainly noticedthat our vocabulary is full of references to risks It seems that no speech ofours can avoid raising awareness of risks to price stability or evade thesubject of risks to the smooth functioning of the financial system Indeed,one way to describe our core responsibility is to say that the central bankacts as a risk manager for the economy using monetary policy to hedgeagainst inflationary risks However, we tend to be less willing to shareinformation on the ways we manage financial risks in our own institutions.

It is thus not surprising that a book that sheds light on risk management inthe world of central banks and other public investors in a systematic andcomprehensive way has not been published so far And I am very happy thatthe initiative to prepare such a book has been taken by staff of the EuropeanCentral Bank

Central banks’ own appetite for financial risks is not always easy tounderstand Our institutions have historically been conservative investors,placing their foreign reserves mostly in government securities and takingvery little, if any, credit risk Progressively, the accumulation of reserves insome countries, either as a result of their abundant natural resources or offoreign exchange policies, has led their central banks to expand theirinvestment universe and, with it, the financial risks they face More recently,the landscape of public investors has been enriched by sovereign wealthfunds, state-backed investors from emerging economies that made theirpresence more than noticeable in international capital markets and haveoccasionally created controversy with their investment strategies

While managing investment portfolios is one area where risk ment expertise is needed, central banks have other core concerns They are

manage-in charge of monetary policy manage-in their jurisdiction They are also expected tointervene when the stability of the financial system is at stake In order tosteer the system out of a crisis, they are prepared, if needed, to take thosexvii

risks which other market participants rush to shed They are prepared toprovide additional liquidity to the system as a whole or lend to specificbanks on special conditions Such behaviour, which may seem to put riskmanagement considerations on hold, at least temporarily, further compli-cates the effort of an outsider to understand the role of risk management inthe central bank.

Being responsible for risk management in a public institution, like acentral bank, does not simply rely on technical risk management expertise.Although the requirement for a high degree of fluency in quantitativetechniques is not less important than in private financial institutions, itmust be combined with a deep understanding of the role of the publicinstitution and its core functions In our institutions, financial decisions arenot taken based only on risk and return considerations but also take intoaccount broader social welfare aspects

Central bank risk managers provide decision makers with assessments offinancial risks in the whole range of central banks’ operations, whether theseare linked to policy objectives or are related to the management ofinvestment portfolios They should be able to deliver such assessments notonly under normal market conditions but, even more so, under conditions

of market stress Decision makers also seek their advice to understand anddraw the right conclusions from the use of the latest instruments of risktransfer in the markets and the implementation of risk managementstrategies by financial institutions in our jurisdictions

The European Central Bank placed, from the very beginning, particularattention to risk management As a new member of the central bankcommunity, it had the ambition of fulfilling the highest governance standards

in organizing its risk management function within the institution andapplying state-of-the-art tools No less than that would be expected from anew central bank that would determine monetary policy and overseefinancial stability for an ever-increasing number of European citizens, playingthe lead role in a system of cooperating central banks

Central banks and other public investors have been entrusted with themanagement of public funds and are expected to do so in a transparent waythat is well understood by the public This book systematically explains howcentral banks have addressed financial risks in their operations It discussesissues of principle but also provides concrete practical information Itexplains how risk management techniques, developed in the private sector,apply to central banks and where idiosyncrasies of our institutions meritspecial approaches The blend of analysis and information provided in the

next pages makes me confident that this book will find an eager readershipamong both risk managers and central bankers.

Jose´ Manuel Gonza´lez-Pa´ramoMember of the Executive Board ofthe European Central Bank

Ulrich Bindseil, Fernando Gonza´lez and Evangelos Tabakis

Domestic and foreign financial assets of central banks and public wealth fundsworldwide are estimated to have reached in 2007 more than USD 12 trillion,which is more than 15 per cent of world GDP, and more than 10 per cent of theglobal market capitalization of equity and fixed-income securities markets.Reflecting unprecedented growth of their financial assets, and the revolution ofrisk management techniques and best practices during the last fifteen years, theinvestment and risk management policies and procedures of central banks andother public investors have undergone a profound transformation The pur-pose of this book is to provide a comprehensive and structured overview ofissues and techniques in the area of public institutions’ risk management Oneach of the main areas of risk management, the book aims at first presentingthe general problems as they also would occur in private financial institutions,then to discuss how these materialize in the special case of public institutions,and finally to illustrate this general discussion by describing the EuropeanCentral Bank’s (ECB) specific approach Due consideration is given to thespecificities of public institutions in general and central banks in particular Onthe one side, their public character relates to certain policy tasks, which will alsoimpact on their investment policies, in particular with regard to assets whichare directly considered policy assets (e.g monetary policy assets, foreignreserves to stand ready for intervention purposes) Secondly, the publiccharacter of these institutions has certain implications regardless of policytasks, such as particular duties of transparency and accountability, lessflexibility in terms of human resource policies and contracting, being outsidethe radar of regulators, etc These characteristics will also influence optimalinvestment policies and risk management techniques of public institution.The book targets portfolio managers, risk managers, monetary policyimplementation experts of central banks and public wealth funds, and staff insupranational financial institutions working on similar issues Moreover, stafffrom the financial industry who provide services to central banks would alsohave an interest in this book Similarly, treasury and liquidity managers ofbanks will find the risk management perspective of central banks’ liquidityxx

providing operations useful in understanding central bank policies Around ahalf of the chapters also provide methodological discussions which are notreally specific to central banks or other public investors, but which are equallyrelevant for any other institutional investors Finally, students in both financeand central banking will find the book important as bridging theory andpractice and as providing insights in a key area of central banking other thanmonetary policy on which very little has traditionally been made public.The authors of this book all work or worked in the ECB’s Risk Mana-gement Division (except two, who work in the ECB’s Directorate GeneralMarket Operations), and the topics covered reflect the area of expertise ofthe respective authors Thus, the book obviously reflects the experience ofthe ECB and the specific challenges it has had to address Nevertheless, thebook aims at working out the generic specificities and issues relating to allpublic institutions’ risk management functions.

There are two types of books with which the present one can be pared First, there are a number of books on central bank investment policiesand risk management, like Bernadell et al (2004), Danmarks Nationalbank(2004), Pringle and Carver (2007, but also previous editions), Johnson-Calari and Rietveld (2007) or Bakker and van Herpt (2007) These bookshowever do not aim at being comprehensive and conceptually structured,nor do they go really into depth In contrast, the present book is intended

com-to be a comprehensive reference book, structured along the main areas ofcentral bank investment and risk management, reviewing systematically theexisting literature, going into depth, and using state-of-the art methods.Second, there are at least two recent books by teams from the institutionalinvestor/asset allocation area of major investment banks, namely Litterman(2003) and Dynkin et al (2006) These books are similar in authorship asthey are produced by a team of experts from one institution and covertopics in the broader area of financial management, including risk mana-gement However the two books have a different perspective, namely that ofinvestment management, and do not cover the risk control and risk miti-gation aspects of risk management

Structure of the book: Investment vs policy operations;

different risk types

The book is structured into three main parts: the first deals with the riskmanagement for investment operations of public institutions Investment

operations are defined broadly as financial operations of public institutionswhich are not or only limitedly constrained by the policy mandates of thepublic institution Still, the public character of the institution shouldinfluence its investment and risk management policies, relative to a non-public institutional investor The second part deals with policy operations

of central banks, whereby the focus is on collateralized lending operations,

as such monetary policy operations are standard today for central banks

to control short-term interest rates Most issues arising in this context are,however, also relevant for collateralized lending programmes that a finan-cial institution would establish, and techniques discussed are thereforerelevant for the financial industry Finally, a short third part deals withorganizational issues and operational risk management in public financialinstitutions

While the segregation of risk management approaches into those relating

to investment and those relating to policy operations may seem forward for central bankers, its compatibility with the idea of integratedfinancial risk management may be questioned Why wouldn’t all risks bemapped eventually into one risk framework? It appears a standard problem

straight-of any bank that risks from different business lines seem at a first lookdifficult to aggregate, but that these problems need to be overcome becausesegregated risk management is inferior In contradiction to this, in manycentral banks, the organizational units for risk management are segregated:one would be responsible for investment operations, and the other forpolicy operations In the case of the ECB, both risk management functionsare assigned to one division, not to aggregate risk across the two ‘businesslines’, but for achieving intellectual economies of scale and scope Aprobably valid explanation in the case of the ECB for not integrating the twobusiness lines in terms of risk management is that monetary policy oper-ations are in the books of the national central banks (NCBs) of the Euro-system, and not in the books of the ECB Therefore, also, losses would arisewith NCBs The responsibility of the ECB’s risk management for definingthe risk framework for policy operations is based on the fact that lossesrelating to monetary policy operations are shared across NCBs In contrast,the ECB’s investment operations are genuinely in the books of the ECB, anddirectly affect its P&L Therefore, integrating the two types of operationswould mean ignoring that the associated P&Ls are not for the same insti-tutions, and thus should be part of different risk budgets, etc While theECB has thus a valid excuse for keeping the two issues separated, whichaffects the structure of the current book, other central banks should

probably not follow this avenue, as all types of operations end up affectingtheir P&L.

The structure of this book from the risk type perspective may appear lessclear than for a typical risk management textbook While Chapter 3 ispurely on the credit risk side, Chapters 2, 5 and 6 are about market riskmanagement Chapters 7 10 are mainly on the credit risk side; however,potential losses in reverse repo operations are also driven by liquidity andmarket risk when it comes to liquidating collateral in the case of a coun-terparty default Chapter4addresses risk control tasks aiming at both creditand market risk Operational risk management as discussed in Chapter13is

a rather different animal, but as operational risk contributes in Basel II

a third component to capital requirements, it is thought that a book onpublic institutions’ risk management would be incomplete if not also dis-cussing, at least in one chapter, issues relating to operational risk in publicinstitutions In the ECB, the more limited interaction between operationaland financial risk management is reflected by having separate entities beingresponsible for each

Part I: Investment operations

Part Iof the book, on investment operations, begins with a chapter (Centralbanks and other public institutions as financial investors) discussing the

‘nature’ of central banks and other public institutions as investors Thechapter aims at providing tentative answers to questions like: What are thespecial characteristics of such investors implied by their policy mandates?What are the basic risk–return properties of their balance sheets? Whatcapital do they need and what are their long-run financial perspectives? Inwhich sense should they be ‘active’ investors and how diversified shouldthey be? Are they unique in terms of aversion against reputation risk? Thechapter suggests that while on one side, many financial industry riskmanagement techniques (like VaR, limit setting, reporting, performanceattribution) are directly applicable to public institutions, the foundations

of integrated risk management (e.g risk budgeting, economic capital culation, desired credit rating) are very special for public institutions, and infact are more difficult to derive than in the case of a private financialinstitution

cal-Chapter 2 (Strategic asset allocation for central banks and publicinvestors) contains a general introduction to strategic asset allocation and a

review of the key issues relating to it It also provides a review of centralbank practice in this area (also on the basis of available surveys), and adetailed technical presentation of the ECB’s approach to strategic assetallocation The importance of strategic asset allocation in public institutionscan hardly be overestimated, since it typically drives more than 90 per cent

of the risks and returns of public institution’s investments This also reflectsthe need for transparency of public investments, which can be fulfilled inprinciple by a strategic asset allocation approach, but less by ‘active man-agement’ investment strategies

Chapter 3 discusses Credit risk modelling for public institutions’investment portfolios Portfolio credit risk modelling in general hasemerged in practice only over the least ten years, and in public institutionsonly very recently Its relevance for central banks, for example, is on the onehand obvious in view of the size of the portfolios in questions, and theirincreasing share of non-government bonds On the other hand, publicinvestors tend to hold credit portfolios of very high average credit quality,still concentrated in a limited number of issuers, which poses specificchallenges for estimating sensible credit risk measures

Chapter4on Risk control, compliance monitoring and reporting turns

to the core regular risk control tasks that any institutional financial investorshould undertake There is typically little systematic literature on thesetopics which are so relevant and also often challenging in practice

Chapter5on Performance measurement again deals in more depth withone core risk control subject of interest to all institutional investors While

in principle being a very practical issue, it often raises numerous technicalimplementation issues Chapter 6, on Performance attribution comple-ments Chapter5 While performance attribution is a topic which can fill abook in its own right, this chapter includes a discussion of the most fun-damental principles and considerations when applying performance attri-bution in the typical central bank setting In addition, the fixed-incomeattribution framework currently applied by the European Central Bank isintroduced

Part II: Policy operations

Chapters7to 11 cover central bank policy operations conducted as reverserepo operations Chapter 7 on Risk management and market impact ofcentral bank credit reviews the role and effects of the collateral framework

which central banks, for example, use in conducting temporary monetarypolicy operations First, the chapter explains the design of such a frameworkfrom the perspective of risk mitigation It is argued that by means ofappropriate risk mitigation measures, the residual risk on any potentiallyeligible asset can be equalized and brought down to the level consistent withthe risk tolerance of the central bank Once this result has been achieved,eligibility decisions should be based on an economic cost–benefit analysis.The chapter looks at the effects of the collateral framework on financialmarkets, and in particular on spreads between eligible and ineligible assets.Chapter 8 goes in more depth with regard to methodological issues ofrisk mitigation measures and credit risk assessments in central bankpolicy operations It motivates in more detail the different risk mitigationmeasures, and how they are applied in the Eurosystem In particular,valuation issues and haircut setting are explained To ensure that acceptedcollateral fulfils sufficient credit quality standards, central banks tend to rely

on external or internal credit quality assessments While many central bankstoday rely exclusively on ratings by rating agencies, others still rely oninternal credit quality assessment systems

Chapter 9 provides a comparison of risk mitigation measures andcredit risk assessment in central bank policy operations across in par-ticular three major central banks, namely the Federal Reserve, the Bank ofJapan and the Eurosystem

Chapter10(Risk measurement for a repo portfolio) presents a the art approach to estimating tail risk measures for a portfolio of collate-ralized lending operations, as it is relevant for any investor with a large repoportfolio, and as it has been implemented for the first time by a central bank

state-of-in 2006 by the ECB

Chapter 11 turns to central bank financial crisis management from

a risk management perspective Financial crisis management is a keycentral bank policy task and unsurprisingly financial transactions in such anenvironment will imply particular risk taking, which needs to be well jus-tified and well controlled The second half of 2007 provided multipleillustrations for this chapter

Part III: Organizational issues and operational risk

Part three of the book consists of Chapters 12 and 13 Chapter 12 is onOrganizational issues in the risk management function of central banks,

and covers organizational issues of relevance for any institutional investor,such as segregation of duties; Chinese walls; policy vs investment opera-tions, optimal boundaries of responsibilities vis-a`-vis other business areasetc The final Chapter 13 treats Operational risk management in centralbanks and presents in some detail the ECB’s approach to this.

Part I

Investment operations

institutions as financial investors

of interest How far would such a theory deviate from a normative theory ofinvestment for typical private large-scale institutional investors, such aspension funds, endowment funds, insurance companies, or mutual funds?Can we rationalize with such a theory what we observe today as centralbank investment behaviour? Or would we end concluding like Summers(2007), who compares central bank investment performance with thetypical investment performance of pension and endowment funds, thatcentral banks waste considerable public money with an overly restrictiveinvestment approach?

In practice, central bank risk management is extensively using, as itshould, risk management methodologies and tools developed and applied

by the private financial industry Those tools will be described in moredetail in the following chapters of the book While public institutions are inthis respect not fundamentally different from other institutional investors,important specificities remain, due to public institutions’ policy mandate,organizational structure or financial asset types held This is what justifiesdiscussing all these tasks in detail in this book on central bank and otherpublic institutions’ risk management, instead of simply referring to generalrisk management literature The present chapter focuses more on the mainidiosyncratic features of public institutions in the area of investment and3

risk management, which do not relate so much to the set of risk ment tools to be applied, but more on how to integrate them into oneconsistent framework reflecting the overall constraints and preferences of,for example, central banks, and how to correspondingly set the basic keyparameters of the public institution’s risk management and investmentframeworks.

manage-The rest of this chapter is organized as follows: Section2reviews in moredetail the specificities of public investors in general, which are likely to berelevant for their optimal risk management and investment policies Section3

turns to the specific case of central banks, being by far the largest type ofpublic investors It explains how the different central bank policy tasks onthe one side have made such large investors out of central banks, and on theother side may constrain the central bank in its investment decisions.Sections 4 and 5 look each at one specific key question faced by publicinvestors: first, how much should public investors diversify their assets, andsecond, how actively should they manage them Sections 6 and 7 aredevoted again more specifically to central banks, namely by looking moreclosely at what non-alienable risk factors are present in central bank balancesheets, and at the role of central bank capital, respectively Section 6, asSection3, reviews one by one the key central bank policy tasks, but in thiscase to analyse their role as major non-alienable risk factors for integratedcentral bank risk management Also on the basis of Sections6and7, Section8

turns to integrated financial risk management of public institutions, which is

as much the holy grail of risk management for them as it is for privatefinancial institutions Section9 draws conclusions

2 Public institutions’ specificities as investors

Public institutions are specific as financial investors as they operate underunique policy mandates and are subject to constraints which do not exist forprivate institutional investors These specificities will have implication foroptimal investment behaviour The following specificities 1) to 5) arerelevant for all public investors, while 6) to 10) only affect central banks.1) Public institutions may appear to be, relative to some privateinstitutional investors (like an insurance, or an endowment fund), subject

to some specific constraints: (i) Less organizational flexibility, includingmore complex and therefore more costly decision-making procedures Thismay argue against ‘decision-intensive’ investment styles; (ii) Decision

makers less specialized on investment For instance central bank boardmembers are often macroeconomists or lawyers, and come more rarelyfrom the investment or risk management side; (iii) Higher accountabilityand transparency requirements, possibly arguing against investmentapproaches that are by nature less transparent, such as active portfoliomanagement; (iv) Less leeway in the selection and compensation of port-folio managers due to rules governing the employment of public servants.This may argue against giving leeway to public investors’ portfolio man-agers, as compared to less constrained institutional investors There arecertainly good reasons for these organizational specificities of public insti-tutions They could in general imply, everything else being equal, a certaincompetitive disadvantage of central banks in active portfolio management

or in diversification into less standard asset classes, relative to privateplayers

2) Public institutions being part of the consolidated state sector Itcould be argued that when investing into domestic financial assets, publicinstitutions should have a preference for Government securities as they arepart of the state sector, and as the state sector should not lengthenunnecessarily its consolidated balance sheet (i.e the consolidated statebalance sheet should be ‘lean’) A lean state sector may be defended on thebasis of the general argument that the state should concentrate on its corebusiness, and avoid anything else, since it is likely to be uncompetitiverelative to private players (which are ‘fitter’ as they survive free marketcompetition) The Fed may be viewed as a central bank following the ‘leanconsolidated state sector’ approach most closely; as more than 90 per cent ofits assets are domestic Government bonds held outright (see Federal ReserveBank of New York 2007, 11) Thus, if one consolidates the US federalGovernment and the Federal Reserve System, a large part of the Fed balancesheet can be netted off

3) Public institutions have a very special owner: the Government, andtherefore, indirectly, the people (or ‘the taxpayer’) When discussing how

a specific institutional investor should invest, it is natural to first look atwho ‘owns’ the institutional investor or, more generally, who owns thereturns on the assets that are managed One tends to describe (or to explain)the preferences of investors with (i) an investment horizon, (ii) relativerisk–return preferences, expressed in some functional form, (iii) possiblysome non-alienable assets or liabilities (for individuals, this would forinstance be human capital), which exhibit specific correlations with finan-cial assets, and thereby determine the optimal asset allocation If one would

view the central bank in its role as investor as a pure agent of the ernment or of the people, one needs to look in more detail to these threecharacteristics of its owner The opposite approach is to view a publicinstitution as a subject on its own, and to see payments to its owners (towhich it is obliged through its statutes) as ‘lost’ money from its perspective.Under this approach, the three dimensions (i)–(iii) of preferences aboveneed to be derived taking directly the perspective of the public institution.4) Public institutions do not have the task to maximize their income.Instead, for instance the ECB has, beyond its primary task to conductmonetary policy, the aim to contribute to an efficient allocation of resources,i.e it should have social welfare in mind According to article 2 of the ESCB/ECB Statute: ‘The ESCB shall act in accordance with the principle of anopen market economy with free competition, favouring an efficient allocation

Gov-of resources .’ The question thus arises in how far certain investmentapproaches, such as e.g active portfolio management, are socially efficient

As Hirshleifer (1971) had demonstrated, there is no general insurance thatprivate and social returns are equal in the case of information producingactivities Especially in the case of what he calls ‘foreknowledge’, it seemslikely that private returns of information producing activities tend to exceedsocial returns, such that at the margin, investment into such informationwould tend to be detrimental to social welfare (i.e to an efficient allocation

of resources) In his words:

The key factor .is the distributive significance of foreknowledge When private information fails to lead to improved productive alignments (as must necessarily be the case in a world of pure exchange .), it is evident that the individual’s source of gain can only be at the expense of his fellows But even where information is disseminated and does lead to improved productive commitments, the distributive transfer gain will surely be far greater than the relatively minor productive gain the individual might reap from the redirection of his own real investment commit- ments (Hirshleifer 1971 , 567)

One could thus argue that it is questionable that an institution, whichaccording to its statute should care about social welfare, engages into activeportfolio management On the other side, it could be felt that this argumentapplies to a lesser extent to foreign reserves, since a central bank shouldprobably always care more about the welfare of its own country than aboutthe one of others, such that egoistic profit maximization in the case offoreign reserves would be legitimate Also beyond the issue of active mana-gement, the question is to be raised whether what is rational from the

perspective of a private, selfish investor would be economically (or

‘socially’) efficient if applied by the central bank Unless one has concreteindications of the contrary, public institutions should probably assume thatthis is the case, i.e that by adopting state-of-the-art investment and riskmanagement techniques from the financial industry; they also contribute tothe social efficiency of their investments

5) Public institutions and reputation risk Reputation risk may bedefined as risks arising from negative public opinion for the P&L of aninstitution or more generally for the ability to conduct relevant tasks Thisrisk may be related to the risks of litigation and loss of independence It isalso called sometimes ‘headline’ risk as events damaging the reputation of apublic institution are typically taken up by the media Reputation risk isoften linked to financial losses (i.e in case of losses due to the failure of acounterparty), but not necessarily For instance, it may be deemed a

‘scandal’ in itself that a central bank invests into some issuer, be it public orprivate, which is judged not to adhere to ethical standards Or it could beconsidered that the central bank should not invest into some ‘speculative’derivatives, although these derivatives are in fact used for hedging, what thepress, the government or the public however may not understand Allinvestors may be subject to reputation risk, but clearly to a varying degree.Central banks’ rather-developed sensitivity for reputation risk may stemfrom the following three factors:

(i) Their need for credibility for achieving their policy tasks, such asmaintaining price stability Credibility is not supported by beingperceived as unethical or amateurish

(ii) Central banks tend to ‘preach’ to the rest of the world what is right andwrong For instance, they often criticize the spending behaviour and lack

of reform policies of Governments Or, as banking supervisors, theyimpose high governance standards on banks, and search for weaknesses

of banks to intervene against them Again, such roles do not appearcompatible with own weaknesses, which again is a credibility issue.(iii) Central banks worry about preserving their independence Independ-ence is a privileged status, and it is obviously endangered if the centralbank shows weaknesses which could help the adversaries of centralbank independence (and those which were criticized or lectured by it)

to argue that ‘these guys need to be controlled more closely bydemocratically elected bodies’

A classical example for central bank headline risk is the attention the smallexposure of Banca d’Italia to LTCM got in 1998, including a need for the

Governor to justify the Bank in the Italian Parliament Reputation risk maydepend first of all on whether a task is implied by the statutes of a publicinvestor If for instance holding foreign reserves is a duty of a central bank,then associated financial risks should imply little reputation risk The moreremote an activity is to the core tasks assigned to the public investor, thehigher the danger of getting questions like: ‘How could you lose publicmoney in this activity and why did you at all undertake it as you have notbeen asked to do so?’ If taking market or credit risk for the sake of increasingincome is not an explicit mandate of a public institution, then market orcredit risk will have a natural correlation to reputation risk.

Reputation risk is obviously closely linked to transparency, and maybetransparency is the best way to reduce reputation risk What has been madepublic and explained truthfully to the public can less be reproached to thecentral bank in case of non-favourable outcomes – in particular if nocriticism was voiced ex ante Central banks have gone a long way in terms oftransparency over the last decades, not only in terms of monetary policy(e.g transparency on their methodology and decision making), but also

in the area of central bank investments For instance the ECB has published

in April 2006 an article in its Monthly bulletin revealing a series of keyparameters of its investment approach (ECB 2006a, 75–86) Principles ofcentral bank transparency in foreign reserves management are discussed insection2of IMF (2004)

6) Central banks are normally equipped with large implicit economiccapital through their franchise to issue banknotes This could be seen toimply that they can take considerable risks in their investments, and harvestthe associated higher expected returns At least for a majority of centralbanks, the implicit capital is indeed considerable, which is discussed in moredetail in Section7 Still, for some other central banks, financial buffers may

be less extensive For instance, central banks which are asked to purchasesubstantial amounts of foreign reserves to avoid revaluation of their currencymay be in a potentially loss-making situation, in particular if, in addition:(i) the demand for banknotes in the country is relatively limited; (ii) domesticinterest rates are higher than foreign rates; (iii) their own currency is underrevaluation pressure, which would imply accounting losses

7) Central bank independence (relevant mainly for domestic financialassets) The need for central bank independence may be viewed to berelevant in this context as implying that the central bank should stay outfrom investing into securities or other assets issued by its own countries’Government In particular World War I taught a lesson in this respect to

e.g the US, the UK, and more than to anyone else, to Germany UnderGovernment pressure, the central banks purchased during the war massiveamounts of Government paper and kept interest rates artificially low It hasbeen an established doctrine for a long time that the excessive purchase ofGovernment paper by the central bank is a sign of, or leads to, a lack ofcentral bank independence For instance article 21.1 of the ECB/ESCBStatutes reflects this doctrine by prohibiting the direct purchase of publicdebt instruments by the ECB or by NCBs.

8) Central banks have insider information on the evolution of term rates, at least in their own currency, and thus on the yield curve ingeneral One may argue that insider information should not be used forethical or for other reasons, and that therefore certain types of investmentpositions (in particular yield curve and duration positions in domesticfixed-income assets) should not be taken by central bank portfolio mana-gers As a possible alternative, ‘Chinese walls’ or other devices can beestablished around active managers of domestic portfolios in the centralbank For foreign exchange assets, the argument holds to a lesser extent.9) Central banks may have special reasons to develop market intelligence,since they need to implement monetary policy in an efficient way, and need

short-to stand ready short-to operate as lender of last resort Especially the latter requires

an in-depth knowledge of financial markets and of all financial instruments.While some forms of market intelligence may be developed in the context ofbasic risk-free debt instruments, a more advanced and broader understanding

of financial markets may depend on diversifying into more exotic asset classes(e.g MBSs, ABSs, CDOs, equity, hedge funds) or on using derivatives (likefutures, swaps, options, or CDSs) Also active portfolio management may beperceived as a way to understand best the logic of the marketplace, as it might

be argued that only with active management do portfolio managers havestrong incentives to understand all details of financial markets For instance theReserve Bank of New Zealand has stated this doctrine, motivating activeportfolio management openly (taken from the IMF 2005, statement 773 – seealso the statement by the Bank of Israel, IMF 2005, statement 663):

773 The Bank actively manages foreign reserves It does so because it believes that active management: generates positive returns (in excess of compensation for risk and of active management overheads) and so reduce the costs of holding reserves; and encourages the dealers to actively participate in a wider range of instruments and markets than would otherwise be the case and so improves the Bank’s market intelligence and contacts, knowledge of market practices, and foreign exchange intervention and risk management skills The skills and experience gained

from reserves management have been of value to the Bank in the context of its other roles too For instance, foreign reserves dealers were able to provide valuable input when the Bank, in the context of its financial system oversight responsibilities, was managing the sale of a derivatives portfolio of a failed financial institution It is not possible to be precise about how much added-value is obtained from active management but, in time of crises, extensive market knowledge, contacts and experience become invaluable.

10) At least some central banks tend to be amongst the exceptionallybig investors The most striking examples are the Asian central banks and inparticular China and Japan with reserves, mostly in USD, at or beyond

1 trillion USD The status as big investor has two important consequences.First, such central banks should probably go further than others in diver-sifying their investment portfolio In the CAPM (Capital Asset PricingModel), all investors should hold a widely diversified market portfolio, but

in reality, transactions and information costs of many kinds are makingsuch full diversification inefficient Participation in a diversified fund canreduce these costs, but will not eliminate them The easiest way to modelthese costs preventing full diversification is to assume fixed set-up costs perasset type, which may be viewed as the costs for the front, back and middleoffice to understand the asset type sufficiently and to prepare for the inte-gration and handling of associated transactions These fixed set-up costs will

be lower for some and higher for other asset types Under such assumptions,

it is clear why smaller investors will end up being less diversified Set-up costscan be economized to some extent through outsourcing or through pur-chasing investment vehicles like funds Also, some important forms of diver-sification, like e.g into an equity indices, may require relatively low set-upcosts, and hesitations of central banks (large or small) with their regard may

be due to other reasons Second, large central banks with a substantial weight

in some markets (e.g US Treasuries) may influence relative prices in thesemarkets, in particular when doing large transactions This may potentiallyworsen their returns, and implies the need to smooth transactions over time,and, again, to diversify Also it increases liquidity risks, i.e the risks that thequick liquidation of relevant positions is only possible at a discount

3 How policy tasks have made central banks large-scale investors

The starting point in analysing the specificities of central banks as investors

is clearly the question why central banks are at all facing ‘investment’ issues

Return-oriented investment is clearly not amongst the standard policy tasks

of central banks To be a bit more specific on the core central bank tasks andhow they have made large-scale investors out of central banks, consider thetasks specifically prescribed for the ECB, which are quite standard forcentral banks with statutes defined in the last decade or so According to theTreaty establishing the European Community (article 105.2), the basic tasks

of the ECB are: (i) the definition and implementation of monetary policyfor the euro area; (ii) the conduct of foreign exchange operations; (iii) theholding and management of the official foreign reserves of the euro areacountries (portfolio management); (iv) the promotion of the smooth oper-ation of payment systems Further tasks according to the ECB/ESCB Sta-tutes relevant in this context are: (v) banknotes – the ECB has the exclusiveright to authorize the issuance of banknotes within the euro area (article 16

of the ECB/ESCB Statute); (vi) financial stability and supervision – theEurosystem contributes to the smooth conduct of policies pursued by theauthorities in charge related to the prudential supervision of credit insti-tutions and the stability of the financial system (see Article 25 of the ECB/ESCB Statutes) Finally, article 2 (‘Objectives’) of the ECB/ESCB Statutesalso prescribes that: (vii) ‘The ESCB shall act in accordance with theprinciple of an open market economy with free competition, favouring anefficient allocation of resources The rest of this section explains how suchtasks made “investors” out of today’s central banks.’

3.1 Banknotes issuing and payment systems

As long as there is demand for banknotes, and the central bank has anissuance monopoly, banknotes will constitute an important and uniqueunremunerated liability in the balance sheets of central banks As any otherliability, banknotes need to be counterbalanced by some assets Unless thereare specific constraints on the asset composition derived from other policytasks, the existence of banknotes in itself thus creates the need for invest-ment decisions by the central bank Although academic and even centralbank visionaries have forecast the end of banknotes for a long time, thetrend growth of banknotes of the large currencies (in particular USD andEUR) has been even above the nominal growth rate of the respectiveeconomies For example, at end June 2007, euro banknotes in circulationstood at EUR 633 billion, more or less the same as USD banknotes In thepast, there were examples of central bank payment systems creating unre-munerated liabilities of the central bank of considerable size and of almost

similar magnitude as banknotes (e.g the Reichsbank in 1900, see table2.2inBindseil (2004, 52)) Today, however, payment systems tend to be so effi-cient as to create very little unremunerated central bank liabilities In so far,they have a negligible impact on central bank balance sheets.

3.2 Monetary policy implementation

In short, central bank monetary policy consists in setting short-term moneyinterest rates such that price stability is maintained over time Short-terminterest rates are controlled by steering the scarcity of deposits of bankswith the central bank, as the short-term money interest rate is essentiallyanchored in the interbank lease rate for such deposits The supply ofdeposits can be influenced by the central bank by injecting or absorbingdeposits, namely by purchasing or selling securities, or by lending orabsorbing funds through so-called ‘reverse operations’ The demand fordeposits can also be influenced by the central bank, notably by imposingreserve requirements on banks What matters in the present context is thatthe steering of short-term rates eventually consists in manipulating thescarcity of deposits ‘at the margin’, i.e the lease price of deposits is set, asany other price, by the marginal demand and supply This means that thecentral bank is constrained in its asset management decisions from theperspective of monetary policy only marginally in the sense that the assetsthat are used to steer the scarcity of deposits at the margin need to besuitable to do so, while the choices regarding the entire rest of the assets remainunaffected (for a survey of monetary policy implementation explaining theimpact on the asset side of the balance sheet, see e.g Bindseil2004, chapters2

and 3) It is noteworthy that in the more distant past, there had been theview that the entire asset composition of central banks (so not only theone at the margin) does matter for its ability to control inflation: this wasthe case under the famous real bills doctrine, according to which ‘real’ (incontrast to ‘financial’) trade bills would be good, i.e non-inflationary assets(see Bindseil 2004, 107 and the literature mentioned there) Recognizingthat monetary policy is implemented only ‘at the margin’, the central bankcan segregate a large part of its domestic financial assets from the monetarypolicy operations to consider them domestic financial investment port-folios Monetary policy portfolios typically consist of short-term reverserepo operations which the central bank conducts through tender procedures

in the market Financial risks of those tend to be very limited: credit risk ismitigated through collateralization (see Part II of the book), while market