Supervision of Financial Conglomerates: The Case of Chile by CONSTANTINOS STEPHANOU Abstract: This paper describes the presence of financial conglomerates and assesses the extent to which the risks they introduce to the Chilean financial system are mitigated by existing oversight arrangements (and at what cost). In particular, the paper questions whether the current silo-based supervisory framework, which has served the system fairly well until now, can continue unchanged given growing inter-linkages in the financial system. A high-level short- and medium-term supervisory reform agenda is proposed, which addresses identified vulnerabilities relating to financial conglomerates and continues the migration from a rules-oriented to a risk-based supervisory approach that has gradually been taking place in Chile in recent years. JEL Classification Code: Keywords: consolidated supervision, risk-based supervision, financial groups, financial conglomerates, risk management, Chile World Bank Policy Research Working Paper 3553, March 2005 The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Policy Research Working Papers are available online at http://econ.worldbank.org. Constantinos A. Stephanou (cstephanou@worldbank.org) is a Financial Economist in the World Bank's Latin America and the Caribbean Region. The author would like to thank the Chilean authorities for their kind cooperation, and express his gratitude to Brian Quinn, Augusto de la Torre, and Alain Ize for helpful comments and suggestions. 1 Table of Contents 1. INTRODUCTION ...........................................................................................3 2. OVERVIEW OF FINANCIAL CONGLOMERATES IN CHILE.......................4 A. Definitions.................................................................................................................. 4 B. Presence in Chilean Financial System ....................................................................... 6 C. Typical Structure........................................................................................................ 8 D. Current Supervisory Arrangements ......................................................................... 11 3. ASSESSMENT OF VULNERABILITIES AND COSTS OF REGULATION.16 A. Typology of Vulnerabilities Introduced by Financial Conglomerates..................... 16 B. Risk Aggregation, Complexity and Transparency................................................... 17 C. Inconsistent Regulatory Treatment and Insufficient Coordination.......................... 21 D. Incomplete Coverage ............................................................................................... 25 E. Contagion ................................................................................................................. 26 F. Distorted Competition and Consumer Protection..................................................... 27 G. Costs of Current Regulatory Arrangements............................................................. 28 4. CONCLUSION AND POLICY IMPLICATIONS...........................................29 A. Summary of Findings............................................................................................... 29 B. Policy Implications................................................................................................... 31 REFERENCES ...................................................................................................37 APPENDIX I: BCP ­ CRITERIA FOR ASSESSING CONSOLIDATED SUPERVISION ...................................................................................................39 APPENDIX II: FINANCIAL CONGLOMERATION BY SECTOR (2003) ............41 APPENDIX III: MARKET SHARES OF FINANCIAL CONGLOMERATES BY SECTOR (1990-2003) ........................................................................................45 2 Supervision of Financial Conglomerates: The Case of Chile 1. Introduction The objectives of this paper, which is drawn from the recently completed Financial Sector Assessment Program (FSAP) work in Chile undertaken by a joint World Bank- IMF mission, are threefold: · to describe the presence and characteristics of financial conglomerates in Chile (Section 2) · to assess the extent to which the vulnerabilities they introduce to the financial system are mitigated by existing oversight arrangements, and at what cost (Section 3) · to propose appropriate high-level modifications to the regulatory framework1 that the Chilean authorities might want to consider in light of the findings (Section 4). Chile represents an interesting case study because it already has a relatively strong supervisory framework, as well as ample publicly available information that can be used for analytical purposes. In common with other developing countries, however, its financial system lacks an integrated supervisory agency or a lead regulator that could (at least theoretically) better respond to the challenges arising from financial conglomerates. The analysis and recommendations will take the current supervisory structure as given and will not consider other possible institutional set-ups2. Given the tight time constraints and breadth of this exercise, an in-depth analysis of each vulnerability has not been attempted ­ where appropriate, future avenues of exploration in the form of a research agenda are mentioned. The paper does not cover the historical rationale and evolution of financial conglomerates, the theoretical arguments for and against their existence, or any current regulatory problems in individual financial sectors that are not 1As traditionally defined, regulation refers to the set of rules and standards that govern the operation of financial institutions, while supervision refers to the oversight/monitoring of the application of those rules and standards. For the purposes of this paper, the two terminologies are used interchangeably. 2Please see De Luna Martinez J. and Rose T. A. (July 2003), Organization for Economic Co-operation and Development (2002), and Abrams R. K. and Taylor M. W. (2000) for a review of alternative institutional arrangements and the experience of integrated financial sector supervision. 3 directly related to financial conglomerates. Finally, it is important to clarify that international best practice in consolidated supervision3 will only be discussed (and prescribed) to the extent that it is deemed an appropriate response to identified problems. 2. Overview of Financial Conglomerates in Chile A. Definitions The definition of financial conglomerates is by now well established in international regulatory circles. The Joint Forum4 defines such entities as "any group of companies under common control whose exclusive or predominant activities consist of providing significant services in at least two different financial sectors (banking, securities, insurance)". By contrast, the EU's Financial Groups Directive, scheduled to take effect in 2005, defines a group as a financial conglomerate if at least 40% of its business is financial and at least 10% or EUR 6 billion of its financial business is in each of the insurance and the combined banking/investment sectors. In view of the characteristics of the Chilean financial system, two major expansions to the Joint Forum definition have been made in this paper: · given the importance of pension funds or AFPs (assets of US$49 billion, or 62% of GDP at end-2003), pensions are included here as an additional financial sector5 · given the prevalence of mixed-activity conglomerates (see below), the above definition has been modified to include the financial sub-groups of such conglomerates, as long as they have an important financial sector presence. The revised definition that is used in this paper is therefore: "any group of companies under common control whose activities include the provision of significant services in at 3Interested readers should consult Basel Committee on Banking Supervision (July and November 2001). 4The Joint Forum was established in 1996 under the aegis of the Basel Committee on Banking Supervision, the International Organization of Securities Commissions (IOSCO) and the International Association of Insurance Supervisors (IAIS), in order to take forward the work of the Tripartite Group on a range of issues relating to the supervision of financial conglomerates. 5A distinction needs to be made between compulsory pensions (handled exclusively by AFPs) and voluntary pensions that can be offered by other market participants as well. Since AFPs also account for much of the latter (tiny in size) business, it is assumed here that the pensions sector consists solely of AFPs. 4 least two different financial sectors (banking, securities6, insurance, pensions)". Moreover, in order to ensure the maximum possible coverage for analytical purposes, a broad definition of what constitutes "significant" was adopted: most groups present in more than one financial sector were therefore categorized as financial conglomerates7. Chile does not currently have a legally embedded definition of financial conglomerates, while its legal characterization of related parties is not sufficiently comprehensive. The Securities and Insurance Superintendency (SVS) defines an `economic group' in Securities Market Law 18.045 as "...a group of entities that display the kind of links in their ownership, administration, or credit liability that lead to a presumption that the economic and financial action of the members thereof is guided by the group's common interest or is subordinate thereto, or that there are common financial risks in the credits granted to them, or in the acquisition of securities issued by them". Articles 96-102 of the same law specify the characteristics and describe the obligations of such groups in greater detail. Using this definition, there were 92 `economic groups' as of March 2004, up from 12 in 1988. Even though the definition is sufficiently broad to be considered comprehensive, it excludes two important types of companies: · non-registered (closed or foreign-listed) companies that could be part of a group ­ for example, Antofagasta is an important holding company of the Luksic group, but it is domiciled abroad and is therefore excluded from the SVS classification of the Luksic `economic group' · companies where more than one controller has major shareholdings ­ for example, since AFP Habitat is controlled by an investment company jointly 6Securities activities typically include the provision of securities brokerage and fund management services, as well as own account trading. For the purposes of this paper and given the relatively small size of the mutual fund segment (US 6 billion as of end-2003), securities brokerage is used as the proxy for activities in this financial sector. It should be noted that: (1) the majority of securities market players belong to banks and are therefore already included in any consolidated banking group statistics; (2) there is substantial over- the-counter (OTC) securities activity that is directly handled by the banks themselves. 7The precise meaning of "significant" (i.e. both the choice of proxy ­ assets, equity, income or capital ­ and threshold percentage) has been purposefully kept vague at this stage, given the need for additional analysis in order to be able to come up with an appropriate working definition for Chile. Given the presence of mixed-activity conglomerates, it is likely that such a definition would include similar criteria to the EU. 5 owned by Citibank and the Chilean Chamber of Construction, it is not included in either of the two `economic groups'. The absence of a comprehensive legal definition of related parties has not hindered financial sector regulators in their application of prudential norms. Both the SVS and Pensions Superintendency (SAFP) use the `economic group' definition in their application of related investment limits for their respective supervised entities. Because those entities are obliged to invest primarily in listed securities8, the limitations of the definition that are mentioned above do not generally arise. By contrast, the Bank and Financial Institutions Superintendency (SBIF) uses its own, more encompassing operational definition for connected lending limits and in its analysis of important bank groups, major shareholder changes, and/or proposed change of ownership deals. This is because article 84 of the General Banking Law gives the SBIF great leeway in determining what constitutes a connected party. B. Presence in the Chilean Financial System Financial conglomerates are dominant in Chile and are consequently significant for its financial system. In that respect, Chile is not different from its Latin American neighbors in the extent to which conglomerates dominate the financial, and indeed the non-financial, system9, . In fact, the concentration in the financial sector partly reflects 10 the structure of the Chilean corporate sector11, although the presence of foreign financial institutions has helped to partly ameliorate domestic concentration. As can be seen in Tables 1 and 2, the insurance sector constitutes the only exception with regard to the penetration of financial conglomerates. Appendix II provides a more detailed breakdown of the incidence of financial conglomeration by financial sector for 2003. 8Exceptions include operationes de libre disponibilidad (for AFPs) and mutuos hipotecarios (for insurance companies). 9The substantial presence of mixed-activity conglomerates in the financial system is a characteristic that often sets apart the financial systems of many emerging markets from those of developed countries. 10For a general description of the presence and characteristics of economic groups in Chile, see Agosin M. R. and Pasten E. H. (May 2003), Lefort F. and Walker E. (October 1999/April 2000a and b), and Majluf N., Abarca N., Rodriguez D. and Fuentes L. A. (April 1998). 11In addition to financial conglomerates and a few domestic `monolines', many of the remaining financial institutions actually belong to a mixed-activity conglomerate ­ for example, Cruz del Sur (Angelini group), Larrain Vial (Larrain group), and Banco Falabella (Solari and Del Rio group). 6 Table 1: Presence of Financial Conglomerates in Chilean Financial System (2003) As proportion (%) of each Insurance Banking Securities Pensions** sector* (life and non-life) Belonging to a financial 98% 91% 40% 97% conglomerate*** Belonging to foreign financial 41% 47% 13% 80% conglomerate Belonging to domestic `pure' 34% 23% 9% 1% financial conglomerate Belonging to domestic conglomerate 23% 21% 18% 16% with mixed activities**** Source: Own analysis based on data provided by SBIF, SVS and SAFP * Percentages are based on bank assets (banking), securities turnover by stock brokerage companies/corredores de bolsa (securities), direct premiums (insurance) and AFP assets under management (pensions) for 2003. ** AFP Habitat is assumed to be controlled by Citibank even though it is jointly owned with the Chilean Chamber of Construction. *** Large foreign and domestic groups that are primarily active in one Chilean financial sector (e.g. AIG, Cruz del Sur/Angelini, Zurich Financial Services) are not considered financial conglomerates. **** Only conglomerates with non-negligible mixed activities are included, e.g. the Yarur and Security groups are excluded because their non-financial activities are very small in comparison to the total. Table 2: Five Largest Firms by Financial Sector in Chile (2003) Ranking Insurance Banking Securities**** Pensions by Size** (life and non-life) Santander Citibank ING Provida/BBVA 1 (ffc*) (ffc) (ffc) (ffc) Banco de BBVA Chilena Consolidada/Zurich Habitat*** 2 Chile/Luksic (ffc) (foreign-owned) (ffc) (dmc*) Estado Santander Santiago Consorcio Financiero Cuprum/Penta 3 (state-owned, dfc*) (ffc) (dmc) (dmc) BCI/Yarur Bice/Matte AIG Santa Maria/ING 4 (dfc) (dmc) (foreign-owned) (ffc) BBVA Banchile/Luksic Cruz del Sur/Angelini Summa Bansander 5 (ffc) (dmc) (domestic-owned) (ffc) Source: Own analysis based on data provided by SBIF, SVS and SAFP * ffc: foreign financial conglomerate; dfc: domestic 'pure' financial conglomerate; dmc: domestic mixed-activity conglomerate ** Ranking is based on bank assets (banking), securities turnover by stock brokerage companies/corredores de bolsa (securities), direct premiums (insurance) and AFP assets under management (pensions) for 2003. *** AFP Habitat is assumed to be controlled by Citibank even though it is jointly owned with the Chilean Chamber of Construction. **** There is some overlap between the banking and securities sectors, since consolidated results for the former include securities operations (only for bank holding companies).In addition, there is substantial OTC debt trading activity that is directly conducted by the banks themselves. 7 The importance of financial conglomerates in the Chilean financial system has remained remarkably stable over the last 5 years. As can be seen in Diagram 1, following the increasing inroads into Chile by foreign financial institutions in the 1990's, a very high and stable level of financial conglomeration has prevailed in the last few years12. Pure `monoline' players are rare, and are mostly concentrated in specialized activities like insurance (e.g. Chilena Consolidada/Zurich Financial Services, Allianz, MetLife) or investment banking and brokerage (e.g. Larrain Vial, IM Trust). Appendix III provides a more detailed breakdown of conglomeration by financial sector for 1990-2003. Diagram 1: Market Shares of Financial Conglomerates by Sector (1999-2003) 100% 90% 80% 70% Sector Banking 60% Securities 50% 40% Insurance Financial 30% Pensions of % 20% 10% 0% 1999 2000 2001 2002 2003 Source: Own analysis based on data provided by SBIF, SVS and SAFP Note: Each company's market share is based on bank assets (banking), securities turnover by stock brokerage companies/corredores de bolsa (securities), direct premiums (insurance) and AFP assets under management (pensions). C. Typical Structure Chilean financial conglomerates can be categorized along three important dimensions for policy purposes: · ownership (domestic vs. foreign) · activities (mixed vs. `pure' financial) · banking presence (include bank or not) 12The only noteworthy fluctuation involved the insurance sector and can be attributed primarily to merger- and-acquisition activity (exit of AXA and Santander, purchase of Aetna by ING, entry and subsequent collapse of Inverlink etc.). 8 As can be seen in Table 3, domestic financial conglomerates are either mixed13 or purely financial in nature, while foreign ones are purely financial14. With a few exceptions, most financial conglomerates also have a banking presence, which typically represents their dominant activity in the financial sector. Geographical expansion (domestic vs. international), which forms another potential dimension, is not worth categorizing separately because of their limited foreign presence15. Table 3: Classification of Major Financial Conglomerates in Chile (2003) Name of Ultimate Insurance Non- Controller Banking* Securities* (life and Pensions* Financial non-life)* Activities Luksic Mining, (Banco de Chile) 17% 9% 1% transport, food, manufacturing Yarur Negligible (BCI) 11% 4% 2% (viticulture) Matte Forestry, (Banco Bice) 3% 9% 3% electricity and real estate Saieh (Corpbanca) 6% 4% 4% Garces, Leon and Vicuna (Consorcio 2% 8% Real estate, Financiero) mining, bottling (investment Delano and Lavin (just granted and mutual Real estate, (Penta) banking 6% 16% license) funds but no health stock broker) Grupo Security Negligible (various individuals) 2% 8% 2% (travel, real estate) Banco del Estado (state-owned) 16% 7% (small life Santander 22% 12% insurance 11% company) BBVA 7% 14% 1% 32% Citibank 3% 14% 24%** ING 9% 13% Health insurance 13In the sense of belonging to a larger group that includes significant non-financial activities. 14Minor exceptions apply in both cases: the Yarur and Security groups have some limited non-financial activities, while ING also offers health insurance. Since these activities are relatively minor when compared to the groups' overall size, they have all been classified as `pure' financial conglomerates. 15The most significant foreign financial sector presence of Chilean financial institutions currently involves equity participation in foreign pension funds (Mexico, Peru, Colombia, Ecuador and El Salvador) by AFPs Provida and Santa Maria. In addition, a few Chilean banks have branches and/or rep. offices in the United States (New York or Miami), although they are primarily used to serve Chilean clients. 9 Source: Own analysis based on data provided by SBIF, SVS, SAFP and Feller Rate * Percentages are market shares of each entity by financial sector. They are based on bank assets (banking), securities turnover by stock brokerage companies/corredores de bolsa (securities), direct premiums (insurance) and AFP investments (pensions) for 2003. ** AFP Habitat is assumed to be controlled by Citibank even though it is jointly owned with the Chilean Chamber of Construction. In the case of domestic financial conglomerates, the corporate structures used typically include holding companies, controlled via investment vehicles by the ultimate owner (individual and/or family). If the conglomerate owns a bank, activities such as mutual funds management, stock and insurance brokerage, factoring and leasing tend to come under it16, otherwise they typically come directly under the holding company or (more rarely) under the insurance company. Foreign financial conglomerates follow the organizational structure of their parent companies, although they still need to comply with Chilean prudential regulations, i.e. the local subsidiary of the foreign bank cannot own an insurance company or an AFP. Diagram 2: Stylized Structure of Domestic Mixed-Activity Conglomerate Ultimate Owner Direction of equity investment Responsibility for regulatory oversight Investment Vehicle (can be one or several) XYZ SVS (only if registered with it) Various non-financial All companies below sector companies can come under the bank Holding Company (directly or via or directly under the (can be one or several) intermediate holding company; investment vehicles) leasing and factoring operations can also be part of the bank SBIF SAFP itself (no subsidiaries Life Insurance P&C Insurance needed); most companies Bank (BHC) Company Company AFP (e.g. mutual funds, brokerage) can also be owned by an insurance company Leasing Factoring Securitization Insurance Securities Mutual Funds Company Company Company Brokerage Brokerage Management SBIF (only if part of BHC) SVS 16 Over the last few years, factoring and leasing operations are increasingly being subsumed into the bank. 10 Given the pyramidal and fairly transparent (at least on the financial sector side) corporate structure of most domestic financial conglomerates, both control and cash flow rights are typically exercised via large shareholdings. This is also reflected in the composition of the Board of Directors: a comparison of Board members across the holding company and its major financial sector subsidiaries indicates the presence of the same one or two individual(s) from the controlling family. D. Current Supervisory Arrangements Regulation and supervision of the financial system take place on a sector-specific basis. SBIF is responsible for bank supervision; SVS covers insurance companies and brokers, securities markets entities (stock brokers, mutual/investment fund managers) and all registered `open' companies; SAFP only covers AFPs. Only the Bank Holding Company (BHC) structure is explicitly recognized under Chilean law (Ley General de Bancos) for consolidated supervision purposes. In that case, the bank forms the parent entity and the SBIF becomes responsible for consolidated prudential requirements, although supervisory authority for securities subsidiaries remains with the SVS. As can be seen in the previous diagram, permissible subsidiaries and affiliates under a BHC structure are essentially restricted to the banking and securities sectors, the only exception being insurance brokerage. By contrast, there are no prudential requirements for holding companies that control firms from different financial sectors (including banks/BHCs). To the extent that such entities are publicly registered and form part of an `economic group', they are subject to standard SVS oversight with a few additional reporting (but virtually no prudential) requirements. SBIF also requires submission of audited information of the financial statements by all shareholders owning more than 10% of the bank's equity. In addition, it collects, via the bank itself, information about the ultimate shareholders, although it is questionable whether that information by itself is sufficient for an adequate assessment (see next section). Moreover, SBIF (as well as SVS and SAFP) has no on-site inspection 11 authority and cannot impose any supplementary reporting or prudential requirements to firms that are parallel to or above its regulated entity. In the case of foreign-owned/controlled financial conglomerates, responsibility for consolidated supervision rests with the relevant home supervisor. Reciprocal arrangements for information exchange take the form of a Memorandum of Understanding (MOU). On the banking side, Chile has already signed formal MOUs with the USA, Spain and Argentina; on the securities side, Chile has signed MOUs with 22 countries covering cooperation, technical assistance, exchange of information and consultation between regulatory bodies of different jurisdictions. There are no MOUs with foreign regulatory (as opposed to social security) authorities on the pensions side. Cooperation between the domestic supervisors has historically been informal and narrow in scope, but this is slowly changing. The Superintendents have cooperated on an informal basis for several years in drafting prudential regulations that are of common interest (e.g. voluntary pensions rules), while they are also participating in the Capital Markets Committee17 (together with the Central Bank) and the Risk Rating Commission18 (together with AFP representatives). An informal Comité de Superintendentes (SBIF, SVS and SAFP) was set up in 2001 with the objective of improving regulatory coordination, and will become official under the proposed Capital Markets II legislative reform. The Superintendents have used this forum to voluntarily exchange information and experiences, discuss topics of common interest and analyze the impact of the introduction of new sector-specific regulations. Some elements of cooperation at an operational level have also recently been undertaken: · joint on-site inspections by SBIF and SVS of some brokers where there is an overlap in jurisdiction because they belong to a BHC · creation of ad hoc working groups to study issues of common interest, such as external auditors and mutual fund management fees charged to AFPs. 17The Capital Markets Committee is an informal body that was established in the early 1990's in order to coordinate financial policy issues and facilitate the introduction of new financial legislation, such as the proposed Capital Markets II legislative reform. 18The Risk Rating Commission was established in order to classify risk and approve instruments for purchase by AFPs. 12 The Comité has no permanent staff; the Technical and Executive Secretaries come from the ranks of SBIF, while the Superintendent of banks acts as the coordinator of its activities. A few high-level individuals from each Superintendency are the designated contact persons for each working group. Strict firewalls exist between related companies in the banking/securities, insurance and pensions sectors. As can be seen in Table 4, firewalls (broadly defined) take the form of prudential limits on permissible activities, ownership of other companies, connected lending/investments and prohibitions on shared infrastructures/client bases. In the case of banks, since most limits are currently applied on a consolidated BHC basis, they do not apply for their subsidiaries that are mostly securities entities ­ so the two sectors are generally much more integrated with each other. A major effect of the imposition of these firewalls in the 1980's has been to reduce banks' importance for domestic mixed-activity conglomerates both organizationally (i.e. banks moved from the center to the edges of the group structure) and as a source of funding. Table 4: Summary of Major Firewalls by Financial Sector Firewall Type Banking Securities* Insurance Pensions Those related to Only allowed to offer securities insurance products. Management of pension intermediation, buying Allowed to purchase funds, and the granting Permissible Only those defined under and selling securities for consumer loans and and administration of the activities article 69 of LGB trading and other mortgages in the benefits established in DL activities authorized by context of their 3,500 SVS (article 27 LMV investment 18.045). operations. No explicit legal Can launch subsidiaries, Global limit for equity prohibition other than with the prior permission participations (subsidiaries) the requirement that of SAFP, to offer services and fixed assets of 100% of any additional related to social security capital. Banks can only form activities be abroad (e.g. AFP subsidiaries for activities "complementary" to management, custody, defined in article 70 of LGB, the core insurance collection of subject to meeting minimum Forbidden to participate business (article 4, contributions, benefits capital adequacy, and in other activities DFL 251). In management and payment Ownership management and financial besides article 27 of practice, the SVS has etc.). limits strength rating requirements. LMV, because their authorized Can also launch Operation of subsidiaries to activity is limited to subsidiaries in subsidiaries with the sole be supervised either by SBIF securities. regulated financial purpose of managing the or by SVS. The subsidiaries businesses such as social security resources cannot invest in other mutual and of their own or of other companies without the prior investment fund domestic AFPs. approval of the SBIF and the management, stock Authorized to invest in up investment cannot exceed 5% and insurance to 7% of the subscribed of paid-in capital. brokerage, and credit shares of a securities cards. deposit company. 13 Firewall Type Banking Securities* Insurance Pensions Definition of related party Under article 84 of LGB, all Definition of related is that of `economic on- and off-Balance Sheet party is that of group'(article 96 of SML credits to a party (irrespective `economic group' 18.045). of whether it is related to the (article 96 of SML Debt securities: Global bank) cannot exceed 5% of 18.045). limit of 3% of the assets capital, which can be raised to Limit on fixed of a related company and 30% depending on the type of income and shares 5% of the issue. guarantee specified in the Brokers must comply issued or guaranteed All securities of a related law. with minimum capital by companies in an company: 1% of a fund's The sum of all credits to requirements. For that `economic group': value. different related parties purpose, the calculation 7.5% of technical All securities of all cannot exceed the bank's of net capital excludes provisions and related companies capital on a consolidated intangible assets owed solvency margin. (including direct and Connected basis. by affiliated parties Limit on fixed indirect investment): 5% exposure All credits to related parties (Norma de Carácter of a fund's value. General N°18 SVS). income and shares Investment fund shares limits must be at arm's length. issued or guaranteed Great discretion given to In case brokers are by a related company and amount of committed SBIF to determine related organized as and its subsidiaries: payments (related fund parties, defined as owning at corporations, any 5% of technical administrator): 5% of least 1% and 5% of bank transaction with provisions and shares issued and equity for individuals and affiliated parties must solvency margin. subscribed. firms respectively. be done at arm's length. Limit on fixed Mutual fund shares Bank Directors' interests in (Article 89 LSA) income and shares (related fund other companies are issued by a related administrator): 5% of presumed to constitute related company abroad: issued shares. lending by association. 2.5% of technical Shares of related Lending to bank employees provisions and open/banking and limited to 1.5% of capital solvency margin. financial/real estate (except mortgages). companies: 2%/0.5%/5% of company's subscribed shares respectively. According to article 27 of LMV, if brokers are organized as corporations, their directors must accomplish certain qualification standards established by Article 26 of LMV. Directors cannot also be `Executives' (managers, in the Board of a deputy managers, and all "sociedad other individuals with the administradora de Specific integrity ability to represent the fondos de terceros" requirements firm or take important (article 169 LMV) or an included in insurance relevant decisions) in Board of Only non-executive Directors anonymous society or law, in addition to other financial companies Directors allowed (article 49 of LGB) its related companies. general rules for cannot also be AFP (article 36 LSA). incorporated Directors (article 156 of Article 44 of Law companies (ley de DL 3,500). 18.046: The company sociedades No legal restriction for an cannot make a contract anónimas). AFP Director to be a or negotiation when Director in other related there is a companies as well. director's interest involved without the pre-approval of the board. A director's interest is presumed when one of the parties is a corporation where he is also the director. 14 Firewall Type Banking Securities* Insurance Pensions According to article 16 of LGB, entities or individuals that control the bank (based on the Securities Market Law) and also own more than There is no explicit Risk Rating Commission 10% of its shares, must law/regulation on established to classify risk provide information to SBIF Joint SVS ­ SBIF outsourcing or and approve instruments on their financial situation, Circular N° 960 sharing of functions for purchase by AFPs. although that information regulates banks activity, and systems ­ in Pricing vector for Other cannot be more than what stock, securities and practice, some back valuation purposes comments SVS requires for anonymous money exchange. office activities are determined by SAFP open societies. In Chile, physical currently allowed to itself. Prohibition on sharing persons or legal persons be shared. Prohibition on sharing distribution network (unless can be brokers without General law on client database and via normal market contract) distinctions. sharing of private infrastructure with related and client database (law on information also financial companies. sharing of private applies. information) with related companies outside the bank holding company structure. Source: SBIF, SVS and SAFP * Firewalls for the securities sector refer solely to stock brokerage companies. The existence of regulatory firewalls combined with historically high profitability in most financial sectors has contributed to the management of major financial entities within each conglomerate typically on a stand-alone basis. This means that management control is exercised at the level of each entity, subject to an overall group direction dictated by the main controller(s) principally via participation in the relevant Board of Directors. On the basis of interviews with executives from such entities, the major group- level synergies that have been identified are limited to common branding, certain elements of risk management (only by foreign entities) and strategic outlook. Holding company involvement is typically minimal ­ given that such companies are often non- operational in nature, they mostly act as a coordination and reporting mechanism. The only exception involves entities that form part of a BHC, where common management structures and sharing of infrastructure (distribution networks, information technology systems) and client databases are allowed between the bank and its banking/securities subsidiaries. In conclusion, there is no comprehensive consolidated supervision framework for financial conglomerates in Chile. Different parts of the same group are supervised by different regulators, whose level of cooperation, coordination and information-sharing (though improving) is still informal and limited in scope. The silo-based nature of current 15 supervisory arrangements is reflected in the non-compliant status of the criteria underlying Principle 20 (see Appendix I) of the recent Basel Core Principles (BCP) assessment19. In that respect, Chile does not appear to be different than most other countries, since only 35% of them (based on a sample of 60 countries that had until recently participated in an FSAP) are actually partly or fully compliant with best practice in this area. 3. Assessment of Vulnerabilities and Costs of Regulation A. Typology of Vulnerabilities Introduced by Financial Conglomerates Five types of vulnerabilities, created by the presence of financial conglomerates, have been identified and are assessed in this section. Some of these reflect new types of prudential risks introduced by financial conglomerates (e.g. contagion), while others merely exacerbate pre-existing ones (e.g. inconsistent regulatory treatment). Although some of these practically overlap or tend to occur in tandem, it would be useful to treat them as conceptually distinct in order to determine the extent to which they are relevant for Chile. Moreover, given the limited presence abroad of Chilean financial conglomerates, the impact of the identified vulnerabilities as they apply across national jurisdictions is not examined. Table 5: Typology of Financial Conglomerate Vulnerabilities Type of Vulnerability Description Complexity and non-transparency of financial conglomerates creates potential Risk aggregation, for excessive concentration of exposures (intra-group or external), inadequate complexity and capitalization (excessive leverage or multiple gearing) and financial contagion transparency to the bank stemming from problems in other parts of the group Different regulatory regimes and lack of supervisory coordination create Inconsistent regulatory opportunities for regulatory arbitrage (e.g. in minimum capital requirements), treatment and insufficient reduce the effectiveness of overall supervision and raise costs of the failure coordination resolution process of a conglomerate 19For a detailed description of the BCP framework, see Basel Committee on Banking Supervision (September 1997 and October 1999) 16 Incomplete regulatory coverage of financial conglomerate entities or markets Incomplete coverage can result in arbitrage opportunities and imprudent concentration of exposures Psychological contagion (e.g. a run on a group bank) can originate from Contagion problems in other parts of a group that are perceived by the market to involve the bank by association (e.g. via common branding) A financial conglomerate can use informational advantages, cross-subsidies Distorted competition and/or off-market price transactions to undercut competitors in specific and consumer protection financial market segments and to exploit clients (conflicts of interest) It is important to clarify up-front that, due to systemic concerns, greater weight is placed on those vulnerabilities that affect the banking sector as opposed to those that affect other financial sectors (e.g. securities20, insurance). The reason is that the banking sector has traditionally been exposed to problems whose consequences have led to significantly greater social costs than problems in other sectors. This is also reflected by the different regulatory philosophies and practices21: · banking regulation is primarily based on the avoidance of systemic risk · securities and investment regulation is generally based on investor protection and enhancing the efficiency of markets via maximum transparency and disclosure · insurance regulation focuses on the solvency of individual companies and protection of policyholders, although generally without any systemic dimensions. As a result, the severity of the problems generated by any identified vulnerabilities, as well as the necessity for remedial measures, is greater for the banking sector. This also has implications for the regulation of financial conglomerates with a banking presence vis-à-vis those without it. B. Risk Aggregation, Complexity and Transparency There is little prima facie evidence to support the existence of excessive (when compared to the limits) intra-group exposures between financial sector entities of the 20To the extent that securities firms and mutual funds already come under a BHC structure, this distinction is not relevant for them. 21In addition to secondary goals such as investor protection, efficiency enhancement and broader social objectives. See Basel Committee on Banking Supervision (November 2001), Allen F. and Herring R. (2001), and van Lelyveld I. and Schilder A. (November 2002). 17 same Chilean conglomerate. As can be seen in Table 6, the strict firewalls previously mentioned have reduced the possibility of intra-group financial contagion originating from the group bank/BHC22, which (given its size) typically dominates all other financial sector entities as far as financial conglomerate controllers are concerned. Table 6: Intra-Group Exposures for Banking and Securities Sectors (2003) Financial Conglomerate Intra-group Exposure* Luksic 12.5% Yarur 9.2% Matte 5.4% Saieh 20.2% Consorcio Financiero N/A Penta N/A Grupo Security 31.4% Santander 19.4% BBVA 7.7% Scotia 6.8% Citibank 1.8% ING N/A SYSTEM AVERAGE 12.4% Source: SBIF. * Percentages represent related party (i.e. intra-group) exposure as a function of capital on a consolidated basis for the group bank (banking and securities) for 2003. Note: Banking and securities are put together because the limits apply on a consolidated BHC level, i.e. they include both the bank and its securities subsidiaries (if any). In addition, the possibility of systemic problems stemming from excessive exposures within the financial part of a conglomerate is reduced by the fact that: · domestic consolidated supervision already takes place for BHCs, which typically comprise both the banking and securities sectors in most financial conglomerates 22The equivalent (publicly unavailable) numbers for the insurance and pensions sectors are also low, although care should be taken with the different definition of related parties used for these sectors. For example, average system-wide intra-group exposure consisted of 0.4% of combined total assets and 0.5% of total investments for insurance companies and AFPs respectively. 18 · foreign financial conglomerates (which constitute around 40% of the Chilean banking system) are already supervised on a consolidated basis by their respective home-country supervisors · domestic financial conglomerates that are dominant in the banking and securities sectors (e.g. Luksic, Yarur, Saieh) do not have a significant presence in the insurance and AFP sectors, and vice versa (e.g. Concorcio Financiero, Penta). Of course, an important implicit assumption of this line of argument is that there is an appropriate level of supervision in each financial sector23, as well as adequate coordination and cooperation between the different supervisors (see below). There are, however, potential risks above the BHC level that are not adequately monitored or protected under current supervisory arrangements. These stem primarily from the non-regulated, non-financial part of a domestic conglomerate ­ for example, it is impossible currently to accurately assess: · true group-level solvency beyond simple accounting measures ­ to prevent the incidence of excessive leverage24 · group-level concentration of external exposures (direct and indirect) by company/sector ­ to prevent significant weakening of the `source of strength' (and thereby inability to inject more capital in the bank) because of problems in specific group counterparties/sectors · dependence of bank funding (liabilities) on the rest of the group ­ to prevent a funding crunch in the bank because of withdrawals from other group members. SBIF's solution to this problem has been to piece together and monitor the major conglomerates' structure and related equity positions via the collection of information from the bank itself and from its major direct shareholders, as well as from reporting of relevant credit exposures by all banks in the `Central de Riesgos' (risk database). 23For example, as the Inverlink case has shown, in the absence of immediate and effective intervention by the supervisors, it is extremely difficult to stop a desperate group from raiding available funds from its financial sector entities. 24Excessive leverage can occur when a parent issues debt (or other instruments not acceptable as regulatory capital in the downstream entity) and downstreams the proceeds to a dependent in the form of equity or other elements of regulatory capital. 19 However, this is only a second-best option since it relies primarily on backwards-looking accounting statements and indirect sources25, does not capture companies without domestic bank lending and has limited enforcement capabilities. In the case of financial conglomerates that have no banking presence, even this instrument is unavailable for SVS and SAFP. The situation is aggravated by the absence of two other factors: · the concept of a financial holding company (FHC), which would allow the separation of financial and non-financial activities within a conglomerate, as well as the better protection/isolation of the former from the latter · proper consolidation of financial statements, given the different accounting conventions that make it impossible to properly consolidate a bank at holding company level. Since only the equity method (i.e. one line item under the heading "investment in affiliates/associates" for the amount corresponding to the share of equity controlled) is used to consolidate its subsidiaries, the holding company's overall gearing ratio could be distorted. Finally, the counter-argument that financial conglomerates are actually `safer' because of diversification is not supported empirically. In a recent study26 of risks within financial conglomerates, it was shown that diversification effects are greatest within a single risk factor, decrease at the business unit level, and are smallest (incremental benefits as low as 5%-10%) across financial sectors, depending on the business mix. This would imply that most of the diversification benefits are derived at the level of the individual bank/BHC, given that it is already exposed to most types of risks27. Even in the case of financial conglomerates that belong to groups with mixed activities, the relatively high concentration of the Chilean economy (given its size and structure) and its corporate sector imply that there must be few benefits from diversification. In addition, a 25For example, in the case of the Luksic conglomerate, the cash flow and solvency position of Quinenco (holding company of Banco de Chile) worsened in 2002, but the performance of Antofagasta (the other major ­ and parallel to Quinenco ­ holding company) improved. However, the issue of whether one holding company would provide support to the other, as well as whether any adjustments need to be made to the accounting statements in order to better reflect economic reality, cannot be addressed properly. 26See Kuritzkes A., Schuermann T. and Weiner S. (November 2002). 27A related issue, which is not discussed here, is whether those diversification benefits are adequately captured by the existing minimum regulatory capital requirements ­ and whether the bank is therefore appropriately capitalized. However, the point to emphasize here is that risks are not necessarily diversified away when reaching up to higher levels in the conglomerate structure. 20 conglomerate is also subject to contagion risk, which was not considered in the above study (see below). C. Inconsistent Regulatory Treatment and Insufficient Coordination There appear to be relatively few opportunities and no evidence of significant regulatory capital arbitrage within a financial conglomerate. Such opportunities currently stem primarily from the following two factors: · certain credit activities (e.g. leasing, factoring and credit cards) can actually sit under/within or next to a BHC, so one potential incentive for financial conglomerates to choose the latter option would be to avoid regulatory capital requirements. Given the relatively small size (when compared to the bank) of such institutions, however, such opportunities are currently limited28. · loan sales/securitizations from the bank to group entities with less penalizing capital requirements (e.g. insurance company) is another possibility. However, such transactions are currently non-existent apart from mortgage securitizations29. A primary reason could be that banks currently have a sufficient level of capital in excess of minimum regulatory requirements, which does not force them to securitize their loans in order to improve their capital adequacy position. Incidents of accounting and valuation arbitrage also do not appear to be taking place. In theory, the different accounting and valuation rules in different financial sectors30 allow for the possibility of arbitrage opportunities. In practice, however, the existing firewalls on the type (arm's length pricing) and extent of related party transactions, as well as the fact that group companies are run on a stand-alone basis, have helped to mitigate this possibility ­ although it remains a concern. 28In addition, there are other factors that discourage such companies from operating next to the bank as opposed to under/within it ­ for example, the difficulty of easily sharing costs/distribution channels in such situations, and of obtaining funding (group bank funding would be subject to the related lending limits). Moreover, the return to the shareholders might be different because of the distinct tax treatment of profits in different entities ­ see Australian Prudential Regulation Authority (1998). 29The broader issue of credit risk transfer from the banking to the insurance and AFP sectors is not considered here. 30The SBIF, SVS and SAFP issue their own rules for their respective regulated entities, which complement and often supersede the generic accounting rules established by the College of Accountants in Chile. 21 However, the current implicit assumption by the Chilean regulators that financial conglomerates can be treated as a loose collection of independently-managed entities may not hold going forward. The functional supervision that has characterized the Chilean financial system is increasingly becoming more institution-oriented as a result of the `blurring of boundaries' between financial sector participants. This refers to the increasing incidence of overlapping activities, risks and products that has been happening both internationally31 and in Chile ­ for example: · insurance companies are currently allowed to directly originate consumer loans and mortgages, and to launch mutual fund and credit card subsidiaries · voluntary pensions are provided by banks, insurance companies and AFPs · annuities will be allowed to be sold by banks (in addition to insurance companies) under the proposed Capital Markets II legislation · all types of financial institutions participate in the growing securities and OTC derivatives markets. As a result, the incentives for regulatory arbitrage could increase in the future. Anecdotal evidence suggests that controllers of Chilean financial conglomerates have historically viewed each financial sector in its own right, both because of the firewalls and because they were generally profitable stand-alone businesses. As the ability to maintain high profitability (particularly for banks) erodes due to intensified competition32, controllers will increasingly be motivated to identify and leverage synergies among their group entities. Those synergies, which include the ability to arbitrage across regulatory domains, will be maximized if there is insufficient regulatory cooperation and coordination. Examples would include: · moving assets to entities with lower regulatory capital charge or preferred accounting and valuation rules 31It is worth noting that the overlap at the level of financial products is usually greater than what is implied by simply comparing the presence of financial institutions in different financial sectors. This is because new hybrid products are constantly evolving, potentially combining banking, securities and insurance components ­ see van der Zwet A. (2003). 32Causal factors include the on-going relaxation of regulatory restrictions, as well as market developments such as increasing internationalization, product innovation and technological advancements. 22 · using non-regulated entities to bypass exposure restrictions · taking excessive leverage within the conglomerate in order to maintain the group's equity participation in growing financial sector businesses. Even though the level of cross-border supervisory cooperation appears satisfactory, coordination among domestic regulators is still informal and potentially insufficient. On the cross-border front, several formal Memoranda of Understanding (MOUs) have been signed by SBIF and SVS with foreign supervisors. On the domestic front, although the proposed Capital Markets Reform law will formalize the Comité de Superintendentes, there is currently no MOU or even informal agreement that clearly describes/assigns the roles and responsibilities of each supervisor33. This is a potentially problematic arrangement in normal but especially in emergency situations, where there is an obvious need to act in a coordinated and prompt manner. In addition, the supervisory framework lacks the concept of a `lead regulator', whose main objective would be to monitor the behavior and assess the solvency of a financial conglomerate in its entirety (to the extent allowed by existing prudential arrangements). The apparent lack of sufficient regulatory coordination was noted in the case of Inverlink (see box below), where at least one commentator34 states that better information sharing among supervisors could have prevented some of the worst abuses that took place and detected the problems earlier. Box 1: The Inverlink Case Founded in the early 1990's, Inverlink grew rapidly from a stock brokerage firm into a full-fledged financial conglomerate with pension fund (AFP Magister), life insurance (Le Mans) and other non-financial concerns. After being accused of share price manipulation in 2001, the group was encouraged by the SBIF to retract its banking license application, while the Central Bank refused it authorization to deal on the foreign exchange market. At the beginning of February 2003, Inverlink's brokerage arm was revealed to have been buying market- sensitive information from the personal secretary of the Central Bank president. Faced with a run on its mutual funds caused by the accusations, the group resorted to selling certificates of deposit (CDs) worth 33Although a "Joint Resolution" was signed by Comité participants upon its establishment in 2001 (and augmented by a "Protocol of Activities" in 2003), it is neither comprehensive nor sufficiently detailed. 34See Larrain C. (spring 2003). 23 around US$100 million from CORFO that it illegally held. It appears that Inverlink had been using them (in addition to client assets) as collateral for short-term trading for several years. It also borrowed substantially from its insurance subsidiary, resulting in the latter's subsequent insolvency and collapse. After being alerted to it, CORFO reported the theft in March ­ by that time, the CDs had changed hands and were held mostly by local mutual funds. This triggered a dispute as to their ownership, with the holders maintaining that they had acquired the CDs in good faith and the government (at least initially) insisting that CORFO would not lose any money. Over the course of the next three days, the mutual fund industry lost over US$1.5 billion (around 27% of its assets) as nervous investors unloaded positions; this led to knock-on effects on the prices of other secondary market securities. Faced with a paralyzed market and mounting fears of a broader liquidity run, the government subsequently agreed to lift the stop payment on the CDs until the courts determine their legal ownership ­ a process that is still on-going. In addition, it reached an agreement with SVS and the banks (owners of most mutual funds) so that the latter would guarantee the immediate repayment of the stolen instruments and would take any court-mandated loss. This stabilized the financial markets, which have since returned to normalcy. The scandal led to the resignation of the CORFO Executive Vice President, Central Bank President and SVS Superintendent, as well as to the arrest and trial of Inverlink senior management. It was also Chile's first (albeit limited) market confidence crisis since the early 1980's, raising questions about cross-sector contagion, inadequacies in market infrastructure, insufficient internal controls and ineffective surveillance. Finally, as the Inverlink case has shown, the failure resolution process for a financial conglomerate is complex and problematic. Not only are current intervention and insolvency rules different for each financial sector, but they ignore the complex reality of conglomerate failures, such as: · the resolution of intra-group exposures · time needed to implement bankruptcy procedures, which can adversely affect the going concern value of certain financial sector entities · priority for `pecking order' of creditors at different group entities and levels · treatment of the same types of contracts (e.g. derivatives) by counterparties of group companies in different financial sectors, which might have potentially different valuation, accounting and investor protection rules 24 · the spillover/impact from unwinding of (potentially large) positions on markets with participants from different financial sectors The patchwork of bankruptcy laws and procedures that are currently in place is therefore unlikely to lead to an efficient resolution of a bankrupt conglomerate. D. Incomplete Coverage Given the current supervisory set-up, incomplete coverage is primarily a concern for two types of entities: unregulated financial companies and non-financial affiliates of mixed-activity conglomerates. In the context of Chile, the former primarily involves the credit card activities of department stores that have recently obtained a banking license (e.g. Falabella, Almacenes Paris and Ripley), which dominate the credit card industry35. These activities remain outside the scope of SBIF oversight, but now become relevant because of the bank presence. To the extent that the credit card company funds itself from the rest of the group (as opposed to non-group bank lending and securitization of receivables), then concerns arise for financial contagion propagating through the group. In addition, the `silo'-type regulation and relatively little coordination among supervisors may lead to incomplete surveillance in certain markets. The participants in these markets (primarily securities and derivatives) cut across those silos, so the need for even closer coordination is essential in order to ensure adequate monitoring (e.g. via sufficient data collection and sharing) and to minimize the incidence of regulatory arbitrage (e.g. via valuation and accounting arrangements). For example, there is anecdotal evidence to suggest that the level of information sharing and surveillance in OTC derivatives is still lacking, impeding the development of an adequate, holistic view of this market segment. 35It is estimated that as many as 15.5 million credit cards (out of a total of 17.8 million in Chile) in 2003 were issued by non-banks and did not therefore carry the Visa, Mastercard or American Express logos. Of these, the largest three issuers were Falabella (3 million), Almacenes Paris (2.7 million) and Ripley (2.2 million). 25 E. Contagion Psychological contagion (as opposed to financial contagion) exists independently of any firewalls. As mentioned earlier, the existence of firewalls in Chile essentially implies that group financial sector entities (bank, insurance company or AFP) would essentially `die within their walls', so that financial contagion does not spread from one to the other. However, empirical evidence from other countries suggests that the mere association (or common branding36) of a bank with a conglomerate and its controller is sufficient to create the perception in the market that their fortunes are linked37. Moreover, financial problems or reputation scandals in the non-financial parts of the conglomerate (e.g. fraud as in Parmalat's case) could result in a run on the group bank by depositors38, even if it appears to be sufficiently protected by the firewalls39. This is one reason for the empirically frequent support of ailing group entities by their related bank, even if the latter is under no legal obligation to back them. Contagion therefore places a great onus on banking regulators to monitor (to the extent possible) the financial health of the entire group/controller, while avoiding to give the perception that they have become responsible for it (moral hazard). In addition, contagion is exacerbated by capital market illiquidity and (potentially) by the existing firewalls in Chile. On the capital markets side, the inability to off-load positions in listed companies by conglomerate controllers would preclude fast injections 36In general, Chilean conglomerates have tended to brand their financial sector companies differently than their non-financial ones. Given that it is not currently an obligation under the law or regulatory norms, there are some exceptions to this trend ­ for example, Grupo Security (same name for real estate, travel agency and financial sector companies) and Falabella/Paris/Ripley (same name for bank and department store). 37Of course, as has been noted to us by some market participants, the other side of this coin is that clients are attracted by the stability projected by the presence of a large controller `backing' a financial institution. 38Given the large proportion of `risk-sensitive' depositors (i.e. firms and institutional investors) in the banking system, this type of liquidity risk is even greater in Chile than in other countries. It is exacerbated by the legal obligation of AFPs to withdraw their funds from any bank whose solvency drops below 10%. 39Two instruments that could be used to mitigate the damage to the banking system in such an eventuality are the BCCh's lender of last resort (LOLR) facility and its legal obligation to provide a 1000% guarantee on all sight deposits. The former would allow the BCCh (with the previous opinion of the SBIF) to grant emergency credit for up to 90 days to banks facing transitory liquidity problems, while the latter would ensure that all sight deposit account holders are repaid in the event of bank liquidation. However, both instruments are primarily mitigating (as opposed to preventive) in nature ­ the issue here is how to best ensure from a regulatory perspective that instances of contagion are minimized. 26 of liquidity40. International experience also suggests that regulatory firewalls might actually worsen the situation by preventing the sharing of available resources across the group41. F. Distorted Competition and Consumer Protection The incidence of conflicts of interest and distorted competition has not been investigated in sufficient detail to assess its importance. It is extremely difficult to attribute any perceived lack of competition or `excess' concentration to financial conglomerates per se, primarily because such groups have always played an important role in Chile's financial system. Anecdotal evidence of limited intra-group exposures42, monitoring of intra-group transactions and stand-alone management of group companies, appears to suggest that such occurrences might be limited. Future research on this topic could focus on the following behavioral indicators: · complaints stemming from securities underwriting, investment research and distribution channels (bank, mutual fund etc.) within the same conglomerate · solicitations to clients of one financial company (e.g. insurance company) from other conglomerate companies (e.g. bank), indicating the sharing of confidential client information · evidence of off-market price transactions or other contracts (e.g. services) · significant overlaps in participation of Directors and senior managers of financial conglomerate companies in other group entities · conglomerate predatory behavior (e.g. via under-pricing and cross-subsidies) in specific financial sectors, deliberately aimed at eliminating competitors 40For example, in the case of the Luksic conglomerate, Quinenco controlled a minority position of around 5.7% in a telecoms company (Entel) that could presumably be sold in the Santiago stock market to cover any short-term liquidity shortfalls. However, the market impact costs of such a transaction would be huge, given that Luksic's US$70 million position at the time was much greater than the average daily turnover of Entel (around US$2-3 million during 2003). 41"Firewalls between the regulated subsidiaries and the rest of Drexel Burnham Lambert Group [DBLG] did not persuade the market that the regulated subsidiaries would not be brought down by problems in their affiliates and parent; but the firewalls did exacerbate DBLG's liquidity problem by limiting the group's access to the resources of the regulated subsidiaries"; taken from Allen F. and Herring R. (2001). 42At least between financial sector companies and between financial and non-financial companies within a conglomerate. The extent of intra-group exposures between non-financial companies themselves has not been investigated. 27 · `cornering' of the market for specific security issues due to large (intentional or unintentional) purchases by a financial conglomerate via its bank, mutual fund, stock broker, insurance company and AFP. G. Costs of Current Regulatory Arrangements In order to properly evaluate the effectiveness of the current regulatory regime with respect to financial conglomerates, its risk mitigation capacity (described above) needs to be weighed against its costs. Those costs come in two different forms: costs to financial conglomerates and societal costs at-large. The former refers to inefficiencies in the operations of conglomerates as a result of the existence of strict firewalls, while the latter refers to broader costs incurred by society as a whole. Since (to our knowledge) neither of these costs has been studied or quantified to-date, a few suitable topics for investigation are highlighted below that could form part of a broader research agenda on the pros and cons of existing regulation. Inefficiencies to financial conglomerates created by regulatory firewalls come in the form of opportunity and operating costs, and manifest themselves in either reduced revenue or increased costs. Opportunity costs arise when a regulation prevents a financial institution from (directly or indirectly) engaging in profitable activities ­ for example, various types of economies of scale and scope that cannot be leveraged between group financial companies, such as shared infrastructure, functions, products, distribution networks and client databases. Operating costs arise from requirements that financial institutions perform certain functions, such as reporting to regulators, providing disclosure to customers and meeting certain operating standards. The duplication of some of these functions (e.g. reporting) increases such costs significantly. On the other hand, it should be noted that some of these activities would have been performed even in the absence of regulations, so the relevant measure is the incremental costs brought about by the existence of such regulations43. Both opportunity and operating costs are currently obscured because of historically good margins in most financial sectors. One potential 43See Elliehausen G. (April 1998). 28 way to estimate some of these costs in the case of banking would be to look at the experience of the removal/relaxation of certain firewalls by the General Banking Act of 1997, in terms of the organizational response of financial conglomerates (e.g. many banks decided to `absorb' their leasing and factoring operations) and the associated impact on their revenues and costs. Discouragement of innovation and risk-taking in financial services could represent the main societal costs of the firewalls. The dependence of bank profitability on traditional lending, as well as the relative absence of well-developed capital market instruments (e.g. structured securities, derivatives) and non-bank financial institutions (e.g. commercial finance companies) for a country with Chile's economic track record and credit rating, could be considered as partial evidence of the detrimental effect of firewalls on financial system innovation. However, additional investigation would be required to substantiate such anecdotal claims. 4. Conclusion and Policy Implications A. Summary of Findings The analysis of the previous section suggests that there do not currently appear to be significant weaknesses in the Chilean financial system stemming from the presence of financial conglomerates. This is primarily due to the influence of three factors: · strict firewalls, forcing a conglomerate's financial sector entities to be managed on a stand-alone basis · high profitability across most financial sectors (particularly for banks), obviating the need for controllers to identify and leverage potential synergies among their group entities · major presence of foreign players in all financial sectors, which are subject to consolidated supervision by their home-country supervisors. The risk mitigation capacity of existing regulatory firewalls has implied some costs that have not yet been quantified. Those costs come in two different forms: inefficiencies 29 in the operations of financial conglomerates due to the inability to leverage synergies, and broader societal costs. The former have been obscured to-date because of historically good margins in most financial sectors, while the latter are difficult to discern empirically; both of them merit further research and quantification. However, a number of vulnerabilities stemming from the presence of financial conglomerates have been identified and could potentially be important going forward: · risks above/parallel to the BHC level stemming primarily from the non-regulated, non-financial part of a conglomerate are not adequately monitored or protected against under current supervisory arrangements · regulatory cooperation via the Comité is still at an early stage (evidenced by the absence of an MOU or other formal arrangements) and incomplete in scope · the failure resolution process in individual financial sectors does not take into consideration the possibility of a financial conglomerate collapse, which can potentially greatly increase the costs involved · the group bank (particularly in the case of domestic mixed-activity conglomerates) remains exposed to psychological contagion because of the market perception of association with the conglomerate/controller. Although the existing framework has served the system fairly well until now, it is questionable whether this can continue unchanged. As the ability to maintain high profitability (particularly for banks) erodes due to intensified competition, controllers of financial conglomerates will increasingly be motivated to leverage synergies from their presence across financial sectors. In addition, growing inter-linkages in the financial system that go beyond the issue of financial conglomerates motivate the need for closer supervisory coordination. There are two main causal factors for this phenomenon: · structural characteristics of the system ­ in particular, the inter-dependencies between large institutional investors and banks, as well as the dominant presence of financial conglomerates 30 · blurring of boundaries between sectors, driven by the on-going relaxation of regulatory restrictions44 and market developments (competition and technological innovation). Both of these factors raise important cooperation issues not previously addressed (see Diagram below), which must now be dealt with appropriately by the supervisors. Diagram 3: Issues Requiring Cross-Sector Supervisory Cooperation ISSUE EXAMPLE(S) Surveillance of macro- `Wiring' of system (inter-sector linkages via common financial issues instruments and markets) Crisis management `Fire drill' (contingency planning) icm arrangements Resolution of conflicts of competence in event of failure of a financial conglomerate Avoiding supervisory gaps Responsibility for monitoring and disclosing activities in securities and derivatives markets Syste and overlaps `Hybrid' products Monitoring of financial Analysis of risk profile of financial conglomerates conglomerates More Competition within and Stimulating competition in AFP sector by (potentially) opening across financial sectors it to other sectors Consistency and integrity of valuation in different entities for same financial instruments Level playing field Consistency of capital adequacy treatment for same instrument in different sector entities B. Policy Implications The analysis and identification of potentially important vulnerabilities from the presence of financial conglomerates strongly indicate that the current silo-based approach would need to migrate toward a more consolidated supervisory framework. This would involve closer regulatory coordination, better surveillance mechanisms and improved control policies for financial conglomerate risks. At the same time, the on-going process of selective relaxation of existing firewalls could continue, albeit focusing on those 44For example, the General Banking Act of 1997 removed previously existing limitations on banks (e.g. the right to carry out factoring, custody, foreign fund management, securitization, and insurance brokerage via subsidiaries) that were imposed by the 1986 Banking Law. Moreover, investment limits for AFPs have been progressively relaxed by the SAFP since the late 1980's. 31 specific measures that are shown to be more efficiency- and competition-enhancing in relation to any potential risks that they introduce. These objectives underlie the reform agenda (described below) that Chile might want to consider in light of its situation. A broad agenda for reform consisting of three types of measures could be undertaken: · `horizontal' measures ­ to improve regulatory coordination and cooperation · `vertical' measures ­ to improve consolidated risk monitoring and control of financial conglomerates · selective relaxation of firewalls within financial conglomerates ­ to increase conglomerate efficiencies, in parallel with the introduction of `vertical' measures. This series of measures reinforces the rationale for a move from a rules-oriented to a risk- based supervisory approach that has been gradually taking place in Chile, particularly in banking. By necessity, detail on the measures has been kept at a minimum ­ specific fine- tuning and timing considerations merit further investigation necessitating the involvement of a working group comprised of all relevant parties. Proposed `horizontal' measures do not aim to expand the scope of existing regulation, but simply to improve regulatory coordination and information exchange. Because they generally do not depend on changes to existing legislation, they could be relatively easier to implement within a shorter time framework: · Strengthen the mandate of the Comité de Superintendentes via the creation of a formal MOU that clearly outlines the roles and responsibilities of each supervisor. These could include: information gathering and exchange consultation and mutual assistance on policy changes monitoring of markets and entities problem resolution, including a definition of procedures during normal and emergency situations conflict resolution process that specifies the ultimate arbiter in case of disputes between members. 32 Examples of MOUs can be found from a number of other countries and can serve as a useful starting point. In general, such MOUs contain broad, commonly shared principles (as opposed to detailed arrangements), which allows for greater flexibility and reduces the need for constant modifications. However, the potential limitations of MOUs in a civil code country like Chile are worth investigating. · Agree on the selection criteria and responsibilities of a chairman among the supervisors to oversee coordination during emergency and non-emergency situations. The Joint Forum45 provides a useful starting catalogue of possible coordination elements. For example, a separate coordinator can be appointed for each specific financial conglomerate, or an overall coordinator can be selected on a permanent or rotating basis among participating supervisors. · Establish a dedicated small but competent secretariat in order to provide supporting analytical firepower, which could be drawn from the ranks of the participating supervisory agencies for a period of time. · Set up working groups to investigate topics of common interest such as: the potential for harmonizing certain rules (e.g. fit and proper tests and valuation), monitoring and minimizing regulatory arbitrage or other distortions across financial sectors, and collecting and sharing data for surveillance of markets in which there are participants from various sectors (e.g. derivatives, securities). · Establish an exchange program between staff in different regulatory agencies · Set up a clear process for the resolution of conflicts of competence in the failure of a financial conglomerate with presence in different financial sectors · Permit consultation and exchange of information/views between regulators in `fit and proper' assessments concerning new license applications or change of ownership situations · Develop `what if' contagion scenarios involving the collapse of a financial conglomerate for assessing policy implications, setting priorities and deriving appropriate and well-coordinated contingency plans. 45See Basel Committee on Banking Supervision (July 2001). 33 Proposed `vertical' measures rely on changes to the existing legislation, and aim to improve consolidated risk monitoring and control of financial conglomerates. They could be thought of as longer-term policies that will need to be carefully studied before being implemented46, and are the following: · Embed the concept of a `financial conglomerate' in legislation, clearly defining its characteristics. Determining such a concept should involve the participation of all relevant financial sector oversight agencies as well as the tax and accounting authorities, and would be based on an assessment of the need for a more comprehensive supervision of complex and systemically important groups47. · For those groups that fall under the financial conglomerates definition, mandate the creation of an explicit, domestically-incorporated financial holding company (FHC) that would control all of their financial sector activities48. This would, on the one hand, clearly separate the financial part of a mixed-activity conglomerate and, on the other hand, provide a common parent entity for all financial sector regulators to focus on. In practice, this measure would not generally imply large adjustments to the existing corporate structures of most financial conglomerates, since many of them are already organized in a similar fashion49. · Establish a `lead regulator' for FHCs with a clear responsibility and accountability mandate, who can be selected based on criteria such as the relative importance of the group's participation in different financial sector segments and on systemic concerns. This structure would supplement the existing sector- specific supervisory arrangements, which would not change. 46In addition to the Joint Forum documents mentioned previously, the EU's Financial Groups Directive could also serve as a useful reference point and benchmark for such measures ­ see Financial Services Authority (October 2003). A recent comparison of alternative supervisory approaches can be found in Half C. (April 2002). 47As an example, see DeFerrari L. M. and Palmer D. (February 2001) for an overview of the US supervision framework for Large Complex Banking Organizations. 48Financial sector activities would be those defined by current legislation, i.e. those that come under the supervision of SBIF, SVS or SAFP. It should also be noted that the FHC can be defined (if deemed necessary) as an operating or a non-operating company. 49Foreign financial conglomerates are already generally structured in such a way for management (if not for legal) purposes. In addition, some of the most important domestic mixed-activity conglomerates (e.g. Luksic, Saieh and Matte) already have holding companies that could, with relatively few adjustments, serve the FHC purpose (i.e. LQ Inversiones Financieras, Corp Group Banking and Bicecorp respectively). 34 · Strengthen the ring-fencing of the financial conglomerate from the rest of the group ­ for example, by prohibiting common branding between the financial and non-financial parts of a group, and by requiring different Board Directors. · Introduce appropriate consolidation rules for FHCs, including the treatment of subsidiaries and equity participations50. · Mandate the adoption of a single auditing firm for the entire financial conglomerate in order to strengthen the external auditor's capacity to review the financial operations in their entirety. · Introduce supplementary reporting and disclosure requirements at FHC level (e.g. solvency and liquidity positions) and enable the `lead regulator' to directly request information from the ultimate controller(s). · Introduce prudential regulations for FHCs, such as supplementary licensing (fit and proper standards), reporting requirements and prudential standards (e.g. consolidated solvency standards51 and large exposures rules). Finally, the introduction of the above measures (particularly the `vertical' ones) could be complemented by the review and possible relaxation of existing firewalls within financial conglomerates. To the extent that financial conglomerates are deemed to be adequately supervised, the regulators could evaluate the pros and cons of existing intra- conglomerate firewalls and determine whether it would be appropriate to simplify and/or remove some of them in order to improve efficiencies, while ensuring that competition within specific sectors is not negatively impacted. There are significant differences in strategy and operations between running a financial conglomerate and running a collection of stand-alone financial entities. Allowing a move from the latter toward the former (which might, in any case, be forced on regulators by market developments) would permit the realization of potentially significant efficiencies and could foster competition across financial sectors. Examples of selective relaxation measures would be to: 50Several alternative consolidation techniques exist ­ see, for example, the Basel Committee on Banking Supervision (July 2001) and McDonald R. (1998). 51This is, incidentally, also a requirement under Basel II, i.e. the application of minimum capital adequacy standards at a consolidated level that reaches up to the financial holding company of a banking group. 35 · permit certain financial sector companies within an FHC to offer services (perhaps initially via subsidiaries/affiliates) traditionally provided by different financial sectors, e.g. allow a bank to offer AFP-type compulsory pensions · remove restrictions on shared functions and infrastructure between different financial sector companies, e.g. allow a Group Treasury at the FHC level. 36 References Abrams R. K. and Taylor M. W. (2000), "Issues in the Unification of Financial Sector Supervision", IMF Working Papers WP/00/213. Allen F. and Herring R. (2001), "Banking Regulation versus Securities Market Regulation", Wharton Financial Institutions Center 01-29. Agosin M. R. and Pasten E. H. (May 2003), "Corporate Governance in Chile", Central Bank of Chile Working Papers No. 209. Australian Prudential Regulation Authority (1998), "Risk and Capital Management Conference", proceedings papers can be downloaded from http://www.apra.gov.au/Policy/Risk-and-Capital-Management-Conference.cfm. Basel Committee on Banking Supervision (September 1997), "Core Principles for Effective Banking Supervision", Bank for International Settlements. Basel Committee on Banking Supervision (October 1999), "Core Principles Methodology", Bank for International Settlements. Basel Committee on Banking Supervision (July 2001), "Compendium of documents produced by the Joint Forum", Bank for International Settlements. Basel Committee on Banking Supervision (November 2001), "Core Principles: Cross- Sectoral Comparison", Bank for International Settlements. DeFerrari L. M. and Palmer D. (February 2001), "Supervision of Large Complex Banking Organizations", Federal Reserve Bulletin. De Luna Martinez J. and Rose T. A. (July 2003), "International Survey of Integrated Financial Sector Supervision", World Bank Policy Research Working Paper 3096. Elliehausen G. (April 1998), "The Cost of Bank Regulation: A Review of the Evidence", Board of Governors of the Federal Reserve System. Financial Services Authority (October 2003), "Financial Groups", Consultation Paper 204. Half C. (April 2002), "Evolving Trends in the Supervision of Financial Conglomerates", Harvard Law School International Finance Seminar mimeo. 37 Kuritzkes A., Schuermann T. and Weiner S. (November 2002), "Risk Measurement, Risk Management and Capital Adequacy in Financial Conglomerates", Wharton Financial Institutions Center 03-02. Larrain C. (spring 2003), "Supervisión Financiera: Lecciones del caso Inverlink", Centro de Estudios Públicos No. 92. Lefort F. and Walker E. (October 1999/April 2000a), "The Effects of Economic and Political Shocks on Corporate Governance Systems in Chile", Revista ABANTE, Vol. 2 No. 2, pp. 183-206. Lefort F. and Walker E. (October 1999/April 2000b), "Ownership and Capital Structure of Chilean Conglomerates: Facts and Hypotheses for Governance", Revista ABANTE, Vol. 3 No. 1, pp. 3-27. Majluf N., Abarca N., Rodriguez D. and Fuentes L. A. (April 1998), "Governance and Ownership Structure in Chilean Economic Groups", Revista ABANTE, Vol. 1 No. 1, pp. 111-139. McDonald R. (1998), "Consolidated Supervision of Banks", Bank of England Handbooks in Central Banking No. 15. Organization for Economic Co-operation and Development (2002), "Supervision of Financial Services in the OECD Area". van der Zwet A. (2003), "The Blurring of Distinctions Between Financial Sectors: Fact or Fiction?", De Nederlandsche Bank Occasional Studies Vol. 1 No. 2. van Lelyveld I. and Schilder A. (November 2002), "Risk in Financial Conglomerates: Management and Supervision", De Nederlandsche Bank Research series Supervision No. 49. Chilean Sources of Information Ministry of Finance (http://www.minhda.cl/) Central Bank of Chile (http://www.bcentral.cl/) SBIF (http://www.sbif.cl/) SVS (http://www.svs.cl) SAFP (http://www.safp.cl) 38 Appendix I: BCP ­ Criteria for Assessing Consolidated Supervision Principle 20: An essential element of banking supervision is the ability of the supervisors to supervise the banking group on a consolidated basis52. Essential Criteria 1. The supervisor is aware of the overall structure of banking organizations (i.e. the bank and its subsidiaries) or groups and has an understanding of the activities of all material parts of these groups, including those that are supervised directly by other agencies. 2. The supervisor has a supervisory framework that evaluates the risks that non-banking activities conducted by a bank or banking group may pose to the bank or banking group. 3. The supervisor has the legal authority to review the overall activities of a bank, whether the activities are conducted directly (including those conducted at overseas offices), or indirectly, through subsidiaries or affiliates of the bank. 4. There are no impediments to the direct or indirect supervision of all affiliates and subsidiaries of a banking organization. 5. Laws or regulations establish, or the supervisor has the authority to impose, prudential standards on a consolidated basis for the banking organization. The supervisor uses its authority to establish prudential standards on a consolidated basis to cover such areas as capital adequacy, large exposures and lending limits. 6. The supervisor collects consolidated financial information for each banking organization. 7. The supervisor has arrangements with functional regulators or individual business vehicles within the banking organization group, if material, to receive information on the financial condition and adequacy of risk management and controls of such business vehicles. 52Supervision of the banking group on a consolidated basis goes beyond accounting consolidation. It implies that there is a group-wide approach to supervision whereby all risks run by a banking group are taken into account, wherever they are booked. It is important to note that both accounting consolidation and consolidated supervision are key aspects of the supervision of banking groups. 39 8. The supervisor has the authority to limit or circumscribe the range of activities the consolidated banking group may conduct and the overseas locations in which activities can be conducted; the supervisor uses this authority to determine that the activities are properly supervised and that the safety and soundness of the banking organization is not compromised. Additional Criteria 1. For those countries that allow corporate ownership of banking companies: · the supervisor has the authority to review the activities of parent companies and of companies affiliated with the parent companies, and utilizes the authority in practice to determine the safety and soundness of the bank; · the supervisor has the authority to take remedial actions, including ring-fencing, regarding parent companies and non-bank affiliates concerning matters that could impact the safety and soundness of the bank; and · the supervisor has the authority to establish and enforce fit and proper standards for owners and senior management of parent companies. 40 Appendix II: Financial Conglomeration by Sector (2003) Table 7: Financial Conglomeration of Banking Sector (2003) DOMESTIC FC MARKET SHARE - TOTAL ASSETS* ULTIMATE OWNER(S) / MAIN FINANCIAL FOREIGN BANK MAINLY MIXED- BY ASSETS* (CLP MILLION) SHAREHOLDER(S) CONGLOMERATE (FC) FC ACTIVITY 22% 10,429,646 Banco Santander-Chile Santander X X 17% 8,228,465 Banco de Chile Luksic X X 16% 7,340,685 Banco del Estado State-owned X 11% 5,107,461 BCI Yarur X 7% 3,181,676 BBVA BBVA X X 6% 2,703,286 Corpbanca Saieh X 3% 1,499,362 Banco del Desarrollo Norte Sur (Catholic Church) X X 3% 1,524,416 Scotiabank Sud Americano Bank of Nova Scotia X X 2% 1,145,299 Banco Security Several (no single controller) X 3% 1,203,593 BICE Matte X X 3% 1,500,710 Citibank N.A. Citigroup X X 2% 1,071,272 BankBoston N.A. Fleet X X 1% 430,936 ABN AMRO Chile ABN AMRO X X 0% 230,446 Falabella Solari - Del Rio 1% 255,403 Dresdner Dresdner X X 0% 214,834 Conosur Del Rio 0% 168,063 Banco Internacional Furman 0% 224,446 HSBC Chile HSBC X X 0% 95,648 HNS Banco Ergas 0% 55,668 Ripley Calderon 1% 351,751 Deutsche Bank Chile Deutsche X X 0% 210,401 JP Morgan Chase JP Morgan Chase X X 0% 34,464 Do Brasil S.A. Banco do Brasil 0% 34,026 Bank of Tokyo-Mitsubishi BTM 0% 15,976 Banco de la Nacion Banco de La Nacion 0% 15,753 Monex Ergas Benmayor 100% 47,273,686 Source: Own analysis based on data provided by SBIF Note: Assets are defined as "activo contable total menos: las cuentas de ajuste y control del pasivo, operaciones a futuro del pasivo y el canje", while profit is defined as "utilidad del ejercicio". 41 Table 8: Financial Conglomeration of Securities Sector (2003) Source: Own analysis based on data provided by SVS Note1: Market share is defined as proportion of securities turnover by stock brokerage companies (`corredores de bolsa') both on and off the Santiago stock exchange ­ other securities market players (`agencias de valores', mutual and investment funds, and bank trading desks) are not considered. 42 Table 9: Financial Conglomeration of Life Insurance Sub-Sector (2003) FINANCIAL DOMESTIC FC MARKET SHARE - TOTAL PREMIUMS MARKET SHARE - TOTAL AUM ULTIMATE LIFE INSURANCE COMPANY CONGLOMERATE FOREIGN FC MAINLY MIXED- BY PREMIUMS (CLP MILLION) BY AUM (CLP MILLION) OWNER(S) (FC) ACTIVITY 0% 3,586,607 0% 2,821,316 ABN AMRO ABN AMRO X X 2% 30,304,400 0% 13,289,017 ALTAVIDA Santander X X 1% 12,835,604 0% 11,422,390 BANCHILE Luksic X X 8% 109,493,112 8% 807,831,535 PRINCIPAL Principal 1% 17,967,485 0% 19,630,461 BBVA BBVA X X 1% 15,385,706 0% 38,107,720 BCI Yarur X 5% 68,689,127 4% 376,044,131 BICE VIDA Matte X X 0% 0 1% 56,772,751 CAJA REASEGURADORA Mapfre 1% 19,532,593 0% CARDIF BNP Paribas 6% 77,853,600 6% 541,464,520 CHILENA CONSOLIDADA Zurich 1% 13,318,449 1% 106,096,134 CIGNA VIDA Cigna 11% 148,430,805 16% 1,531,660,145 CONSORCIO NACIONAL Consorcio X X 7% Chilean Chamber of 101,506,446 8% 801,057,137 CONSTRUCCION Construction 0% 570,096 3% 310,528,777 CN Life Consorcio X X 3% 36,686,649 3% 264,805,067 CRUZ DEL SUR VIDA Angelini 4% 56,521,463 3% 315,288,298 EUROAMERICA VIDA Clarke 0% 1,439,789 0% 5,896,096 HUELEN CAEP 4% 61,000,775 4% 386,016,041 INTER RENTAS AIG 3% 40,103,163 1% 138,592,841 INTERAMERICANA VID. AIG 12% 160,714,718 12% 1,148,179,567 ING VIDA ING X X 1% 14,169,420 0% 13,264,340 MAPFRE VIDA Mapfre 10% 135,214,633 9% 896,605,702 METLIFE Metlife 0% 6,850,669 1% 99,749,678 MUT DE CARABINEROS Mutual 1% 8,145,532 0% 47,013,148 MUT. EJERC. Y AVIAC. Mutual 1% 11,408,249 1% 65,639,271 MUTUAL DE SEGUROS Mutual 3% 39,084,912 3% 246,715,484 OHIO NATIONAL Ohio National 6% 77,714,089 5% 460,232,105 PENTA Penta X X 1% 17,122,253 1% 52,151,496 PREVISION VID. Security X 1% 20,201,694 3% 259,457,144 RENTA NACIONAL VID. Errazuriz 6% Corp Group & Mass 80,516,796 8% 736,252,657 VIDA CORP X Mutual 0% 9,072 0% 42,152,253 VITALIS Consorcio X X 100% 1,386,377,906 100% 9,794,737,222 Source: Own analysis based on data provided by SVS 43 Table 10: Financial Conglomeration of Non-Life Insurance Sub-Sector (2003) DOMESTIC FC MARKET SHARE - TOTAL PREMIUMS MARKET SHARE - TOTAL AUM (CLP ULTIMATE FINANCIAL FOREIGN NON-LIFE INSURANCE COMPANY MAINLY MIXED- BY PREMIUMS (CLP MILLION) BY AUM MILLION) OWNER(S) CONGLOMERATE (FC) FC ACTIVITY 3% 18,355,799 1% 3,463,212 ABN AMRO ABN AMRO X X 3% 16,544,013 5% 11,576,630 ACE Cigna 6% 41,095,178 6% 13,676,458 AGF ALLIANZ Allianz X X 4% 23,459,241 6% 14,587,694 BCI Yarur X 14% 91,652,092 10% 23,467,823 CHILENA CONSOLIDADA Zurich 1% 6,723,458 2% 5,591,630 CHUBB DE CHILE Chubb 2% 9,793,994 2% 4,889,387 CONSORCIO NACIONAL Consorcio X X Swiss Re and 1% 6,223,624 2% 4,222,729 CREDITO CONTINENTAL Deutsche Bank 17% 109,622,838 8% 18,100,174 CRUZ DEL SUR Angelini 0% 421,150 1% 2,410,044 HUELEN CAEP 3% 22,642,121 4% 8,625,501 ING ING X 9% 57,449,347 12% 27,681,124 LA INTERAMERICANA AIG 6% 6% Penta X X 40,953,973 13,925,034 LAS AMERICAS 1% 9,299,204 2% 5,010,300 LE MANS-ISE Various families 5% 34,840,258 4% 9,901,763 MAGALLANES Various families 1% 3,500,951 2% 4,090,933 MAPFRE GARANTIA Y CREDITO Mapfre 10% 62,181,863 4% 9,201,527 MAPFRE SEGUROS GRALES Mapfre 0% 758,060 2% 4,535,901 MUTUALIDAD DE CARABINEROS Mutual 1% 6,162,946 2% 4,346,552 RENTA NACIONAL Errazuriz Royal and 7% 46,124,560 7% 16,011,663 ROYAL & SUNALLIANCE SunAlliance 3% 17,368,729 2% 4,198,407 SECURITY PREVISION Security X 0% 2,379,166 2% 5,317,919 SEGUROS DE CREDITO COFACE Coface 3% 21,772,379 9% 22,283,263 SEGUROS GRALES CARDIF BNP Paribas 100% 649,324,944 100% 237,115,668 Source: Own analysis based on data provided by SVS Table 11: Financial Conglomeration of Pension Fund Sector (2003) DOMESTIC FC MARKET SHARE - BY TOTAL AUM (CLP ULTIMATE OWNER(S) / MAIN FINANCIAL FOREIGN AFP MAINLY MIXED- AUM MILLION) SHAREHOLDER(S) CONGLOMERATE (FC) FC ACTIVITY 32% 9,351,446 Provida BBVA Pensiones Chile S.A. X X 24% 7,099,202 Habitat Citibank & Chilean Chamber of Construction X X 16% 4,702,050 Cuprum Empresas Penta S.A X X 13% 3,725,907 Santa Maria ING S.A. X X 11% 3,352,198 Summa Bansander Santander Chile Holding S.A. X X 3% 785,005 Planvital Inversiones y Asesorias Los Olmos S.A. (BSI) 2% 490,144 Magister Inverlink Holding de Inversiones S.A. X 100% 29,505,951 Source: Own analysis based on data provided by SAFP Note: AFP Habitat, in which Citibank and the Chilean Chamber of Construction have equal shareholdings, is assumed to be controlled by the former. 44 Appendix III: Market Shares of Financial Conglomerates by Sector (1990-2003)53 Diagram 4: Share of Foreign Financial Conglomerates by Sector (1999-2003) 90% 80% 70% or 60% Sectla Banking 50% Securities nanci 40% Insurance Fi Pensions of 30% % 20% 10% 0% 1999 2000 2001 2002 2003 Diagram 5: Share of Domestic `Pure' Financial Conglomerates by Sector (1999- 2003) 40% 35% 30% 25% Banking Securities 20% Insurance 15% Pensions 10% 5% 0% 1999 2000 2001 2002 2003 Diagram 6: Share of Domestic Financial Conglomerates Belonging to Mixed- Activity Groups by Sector (1999-2003) 30% 25% 20% Banking Securities 15% Insurance Pensions 10% 5% 0% 1999 2000 2001 2002 2003 53Source: Own analysis based on data provided by SBIF, SVS and SAFP. Each company's market share is based on bank assets (banking), securities turnover by stock brokerage companies/corredores de bolsa (securities), direct premiums (insurance) and AFP assets under management (pensions). AFP Habitat is assumed to be controlled by Citibank even though it is jointly owned with the Chilean Chamber of Construction. 45