ASIAN DEVELOPMENT BANK Operations Evaluation Department

Transcription

1 ASIAN DEVELOPMENT BANK Operations Evaluation Department PROJECT PERFORMANCE EVALUATION REPORT FOR INDIA In this electronic file, the report is followed by the Management response.

2 Performance Evaluation Report Project Number: PPE: IND Investment Number: 7122-IND December 2005 IND: SBI DFHI Limited Operations Evaluation Department

3 CURRENCY EQUIVALENTS Currency Unit Indian rupee/s (Re/Rs) At Appraisal At Project Completion At Operations Evaluation (November 1995) (July 2001) (September 2005) Re1.00 = $ $ $ $1.00 = Rs Rs Rs ABBREVIATIONS ADB Asian Development Bank CCIL Clearing Corporation of India Limited DFHI Discount Finance House of India Limited FY fiscal year GDP gross domestic product IPO initial public offering FIRR financial internal rate of return NDS negotiated dealing system OEM Operations Evaluation Mission PCR project completion report PPER project performance evaluation report PSOD Private Sector Operations Department RBI Reserve Bank of India RRP report and recommendation of the President SBI State Bank of India SBIGL SBI Gilts Limited SBICAP SBI Capital Markets Limited SBI DFHI SBI DFHI Limited T-bill treasury bill (i) (ii) NOTES The fiscal year (FY) of the Government ends on 31 March. FY before a calendar year denotes the year in which the fiscal year ends, for example, FY2000 ends on 31 March In this report, $ refers to US dollars. Director General : B. Murray, Operations Evaluation Department (OED) Director : D. Edwards, Operations Evaluation Division 2, OED Team Leader : T. Ito, Evaluation Specialist, Operations Evaluation Division 2, OED Team Members : V. Ramos, Evaluation Officer, Operations Evaluation Division 2, OED R. Perez, Senior Operations Evaluation Assistant, Operations Evaluation Division 2, OED Operations Evaluation Department, PE-675

4 CONTENTS Page BASIC DATA EXECUTIVE SUMMARY ii iii I. INTRODUCTION 1 II. PROJECT DESCRIPTION 1 A. Background 1 B. Rationale 2 C. Deal Processing 2 D. Expected Outcomes 3 E. Progress Highlights 3 III. EVALUATION 4 A. Development Outcome 5 B. Investment Outcome 13 C. ADB s Effectiveness 14 D. Overall Rating 16 IV. ISSUES, LESSONS AND FOLLOW-UP ACTIONS 16 A. Issues 16 B. Lessons Learned 17 C. Follow-Up Actions 18 APPENDIXES 1. Evaluation Standards Statistical Data Chronology of Government Securities Market Reforms Key Infrastructure for the Government Securities Market Summary of Financial Statements of SBI Gilts Limited and SBI DFHI Limited Financial Internal Rate of Return on the Asian Development Bank s Investment in SBI DFHI Limited 32 Attachment: Management response The guidelines formally adopted by the Operations Evaluation Department on avoiding conflict of interest in its independent evaluations were observed in the preparation of this report. The fieldwork was undertaken by Susan Thomas (international consultant, debt market specialist) and Renuka Sane (research assistant) under the guidance of the mission leader. To the knowledge of the management of the Operations Evaluation Department, there were no conflicts of interest of the persons preparing, reviewing, or approving this report.

5 BASIC DATA Investment No IND: SBI DFHI Limited (formerly SBI Gilts Limited) Key Investment Data ($ million equivalent) As per ADB Actual Document Amount Approved 4.50 Amount Disbursed 4.19 Terms of Equity Investment: - No. of Shares 15,000,000 15,000,000 a - Par Value per Share Rs10.00 Rs10.00 b - Dividends Cash Cash - Divestment Mechanism Sale Sale Key Dates Expected Actual Fact-Finding Mar 1995 Appraisal 24 Jul 4 Aug 1995 Investment Negotiations Oct 1995 Board Approval 19 Dec 1995 Subscription Agreement 29 Oct 1996 Letter Agreement 29 Oct 1996 Disbursement 5 Mar 1997 Key Performance Indicators (%): Appraisal PCR PPER Financial Internal Rate of Return 12.0 Not computed 18.3 c Borrower Country Type of Business Project Title SBI DFHI Limited India Primary dealer in Government securities Investments in Debt Market Securities Mission Data No. of Missions No. of Person-Days Fact-Finding 1 24 Consultation 1 2 Appraisal 1 15 Negotiations 2 22 Investment Review 4 14 Operations Evaluation 1 28 ADB = Asian Development Bank, DFHI = Discount Finance House of India Limited, PCR = project completion report, PPER= project performance evaluation report, SBI = State Bank of India. a Since the merger of SBI Gilts Limited with DFHI Limited in FY2004, ADB has held 1,363,266 shares of SBI DFHI Limited. b Since the merger of SBI Gilts Limited with DFHI Limited in FY2004, the par value per share has been set at Rs100. c The assumptions in this calculation include (i) ADB s divestment on 31 March 2006, (ii) no increase in net worth and no dividend paid to ADB in FY2006, and (iii) the 30 June 2005 exchange rate of Rs43.52 = $1.00.

6 EXECUTIVE SUMMARY On 19 December 1995, the Asian Development Bank (ADB) approved an equity investment of Rs150 million ($4.5 million equivalent) in SBI Gilts Limited (SBIGL) (the Project), together with equity investments of Rs715 million ($21.2 million equivalent) in SBI Capital Markets Limited (SBICAP) and Rs145 million ($4.3 million equivalent) in SBICAP Securities Limited. SBIGL, the main focus of this project performance evaluation report, was incorporated on 7 March 1996 and became operational on 1 July ADB subscribed Rs150 million ($4.19 million equivalent at disbursement) on 5 March SBIGL s total paid-up capital was Rs1,000 million, of which 85% was contributed by the State Bank of India (SBI) and its affiliated institutions and 15% by ADB. The purpose of the Project was to support the formation of an accredited primary dealers network that would provide an impetus for active trading in Government securities. The expectation was that SBI would (i) reduce its shareholding in SBIGL to less than 50% within 4 years of the date of incorporation of SBIGL; and (ii) arrange for the divestment of ADB s shareholding at the same price at which SBI sold it shares. SBIGL was also to make a public offering of its shares within 4 years of the date of incorporation. At the time of investment approval, ADB assumed that the divestment of its shares would take place within 6 years of the date of incorporation. The financial internal rate of return on the ADB investment in SBIGL was estimated at 16.5% in rupee terms and 12.0% in dollar terms. The main project developments are as follows. First, the Government securities market grew steadily during fiscal year (FY) 1997 FY2004, but experienced a setback in FY2005. Second, SBIGL was profitable during FY1997 FY2003. Third, SBIGL merged with the Discount and Finance House of India Limited (DFHI) in FY2004 and was renamed SBI DFHI Limited (SBI DFHI). Following the merger, ADB owned 4.69% of SBI DFHI s total shares. Fourth, SBI DFHI was profitable in FY2004, but recorded a net loss in FY2005. Fifth, ADB received dividends every year throughout the investment period except FY2005. Sixth, divestment of SBI s and ADB s shareholdings and SBI DFHI s initial public offering (IPO) have not yet materialized. This report rates the development outcome of the Project as satisfactory for two main reasons. First, the investee company contributed to establishing and maintaining the primary dealer system, which in turn served the development of the Government securities market as originally envisaged. Second, without the presence of SBIGL and SBI DFHI, the absorption of primary dealers in the primary market, especially at the initial stage and during the FY2005 market downturn, would have been insufficient. This report rates the investment outcome of the Project as satisfactory. The Operations Evaluation Mission estimated the financial internal rate of return of this investment to be 18.3% in dollar terms, compared with the 12.0% estimated at appraisal, assuming that ADB would divest on 31 March As this estimate suggests, the investment return has been highly satisfactory. However, effective measures to mitigate the market risk and the exit risk are not in place, and thus the final investment outcome remains uncertain. This report rates ADB s effectiveness as partly satisfactory because even though ADB s investment was timely and complemented its public sector operations, the quality of ADB s work in terms of processing and monitoring fell short of expectations. ADB s demonstration role also appears to be diminishing. On the basis of the foregoing, the overall rating of this Project is satisfactory.

7 iv This report identifies four key lessons as follows: (i) (ii) (iii) (iv) ADB equity investments require a workable and viable exit mechanism. Planned IPOs are often not a workable mechanism, because they depend on the nature of the investee, on a sponsor, and on industry and market conditions. ADB cannot rely solely on IPOs, and therefore needs to pursue a put option incorporated in a letter agreement or shareholders agreements. To make such an option enforceable, the method of revaluating an investee at ADB s divestment should be carefully discussed during processing and stated in the agreements. If an investee company proposes a merger or acquisition, both the Private Sector Operations Department and the Office of the General Counsel should carefully examine the implications of such a proposal and give prompt feedback to the investee. ADB may need to pursue a put option if it cannot agree to the proposal. To increase clarity and predictability, the subscription agreement should contain a clause specifying the necessary procedures if the investee decides to pursue a merger or acquisition. The quarterly private sector investment management note should elaborate and update the status of projects that deviate significantly from their original scope and plan. Extra attention should be paid to monitoring such projects. In the case of this Project, the reluctance of the sponsor to divest its shareholding and of the investee company to undertake an IPO are major deviations. ADB approved this Project in conjunction with the investment in SBICAP, but the investment in SBICAP was monitored and evaluated separately. ADB headquarters administered this Project, whereas the India Resident Mission administered the investment in SBICAP, and a project completion report was prepared only for this Project. An assessment of the complementarity of the two investments is beyond the scope of this project performance evaluation report. To the extent possible the monitoring and evaluation of closely related projects sharing the same sponsor should come under the purview of one investment officer and a project completion report and project performance evaluation report should be prepared that cover all associated projects. Considering the diminishing demonstration effect of this Project, the accumulated net worth of SBI DFHI, the investment returns to date, and the limited prospects for privatizing SBI DFHI, ADB should develop an exit strategy from this investment taking market conditions into account. SBI, as the sponsor of this Project, can be a possible counterpart for sales of SBI DFHI shares held by ADB. Bruce Murray Director General Operations Evaluation Department

8 I. INTRODUCTION 1. This project performance evaluation report (PPER) assesses the Asian Development Bank s (ADB s) equity investment in SBI DFHI Limited (SBI DFHI, the Project), formerly SBI Gilts Limited (SBIGL). 1 In line with the good practice standards identified by the Evaluation Coordination Group of Multilateral Development Banks on Private Sector Operations, 2 this PPER focuses on three project dimensions: development outcome, investment outcome, and ADB s effectiveness. 2. An Operations Evaluation Mission (OEM) visited India from 8 to 20 September The OEM met with representatives of the Ministry of Finance, the Reserve Bank of India (RBI), the State Bank of India (SBI), SBI DFHI and other financial institutions participating in the Government securities 3 market, the International Monetary Fund, and the International Finance Corporation. This PPER incorporates the OEM s findings, the observations of relevant ADB staff, and a review of reports and documents related to this Project. The draft PPER was circulated to SBI DFHI and within ADB. Comments received were considered when finalizing the PPER. II. PROJECT DESCRIPTION A. Background 3. In the early 1990s, the Government of India initiated a comprehensive economic reform program that ADB supported through its Financial Sector Program 4 and Capital Market Development Program. 5 One of the aims of these programs was to develop the Government securities market. The Financial Sector Program supported the introduction of a treasury bill (T-bill) auction, whereas the Capital Market Development Program supported the introduction of a primary dealer system for T-bills and dated securities. Primary dealers were expected to ensure the orderly primary distribution of new issues of Government securities, to be market makers for Government securities that would increase liquidity in the secondary market, and to serve as conduits for RBI s open market operations. In November 1995, RBI licensed six primary dealers, 6 including SBI. 1 ADB Report and Recommendation of the President to the Board of Directors on Proposed Investments in Debt Market Institutions in the Republic of India. Manila. (Investment 7122: SBI Gilts Limited, approved on 19 December 1995, for Rs150 million [$4.5 million equivalent at approval].) 2 Evaluation Coordination Group of Multilateral Development Banks on Private Sector Operations, Working Group on Private Sector Evaluation MDB-ECG Good-Practice Standards for Evaluation of Private Sector Operations.; Walter I. Cohen and Associates, LLC Second Benchmarking Review of ECG Members Evaluation Practices for the Private Sector Investment Operations Against Their Agreed Good Practice Standards. Preparation of ADB s evaluation guidelines on private sector operations is under way. 3 Government securities in this report refer to securities issued only by the central Government. 4 ADB Report and Recommendation of the President to the Board of Directors on a Proposed Loan to India for the Financial Sector Program. Manila. (Loan 1208-IND, approved on 15 December 1992, for $300 million.) 5 ADB Report and Recommendation of the President to the Board of Directors on a Proposed Loan to India for the Capital Market Development Program. Manila. (Loan 1408-IND, approved on 23 November 1993, for $250 million.) 6 The initial six primary dealers were Securities and Trading Corporation of India Limited, Discount and Finance House of India Limited, SBIGL, ICICI Securities Limited, PNB Gilts Limited, and Gilts Securities Trading Corporation Limited.

9 2 4. ADB s policy dialogue on improving the efficiency of Government-owned banks focused on SBI, 7 its associate banks, and other affiliates, which are collectively referred to as the SBI group India s largest financial conglomerate. In response to a changing policy environment, the SBI group formulated a restructuring plan. To reorient securities businesses under the plan, SBI Capital Markets Limited (SBICAP) 8 was to spin off its brokerage activities to a new subsidiary, SBICAP Securities Limited. SBI planned to set up SBIGL as a primary dealer of Government securities. B. Rationale 5. SBI recognized the need to sponsor a primary dealer for Government securities not only for its investment banking interests, but also to support RBI s efforts to establish a Government securities market. A primary dealer was necessary to achieve integrated, full-service investment banking operations within the SBI group, complementing both SBICAP and SBICAP Securities Limited. The SBI group requested ADB to support its restructuring process through equity investments in SBICAP, SBICAP Securities Limited, and SBIGL. ADB agreed because (i) the proposed investments in the three institutions were in line with ADB s strategy and policy dialogue with the Indian authorities to develop India s capital markets, (ii) the participation of ADB would provide an impetus to the SBI group s privatization program, (iii) the backing SBI provided to these institutions would be solid, (iv) the SBI group had good management teams and a large pool of trained professionals in banking and financial services, (v) the proposed investments were expected to have a positive impact and a significant demonstration effect on other public sector financial institutions in India, and (vi) the management of SBI supported capital market reforms. Overall, this Project is a good example of ADB s private sector operations supporting broader policy dialogue undertaken by the public sector side of ADB. In addition to supporting the Government s reform program, the activities of the public sector side of ADB helped develop a better enabling environment for the private sector in the Government securities market. C. Deal Processing 6. Subsequent to the concept clearance in March 1995, ADB fielded processing missions between April and August On 19 December 1995, ADB approved its equity investment of Rs150 million (about $4.5 million equivalent) in SBIGL, together with equity investments of Rs715 million (about $21.2 million equivalent) in SBICAP 9 and Rs145 million (about $4.3 million equivalent) in SBICAP Securities Limited Since the 1980s, the Government has gradually sold SBI shares to individuals and institutions. As of 30 September 2005, RBI held 59.73% of SBI s shares. The other major shareholders included foreign institutional investors (11.86%); American Depository Receipts and Shares, and Global Depository Receipts (7.88%); Indian individuals (5.68%); Indian banks and other financial institutions (5.34%); and Indian corporate bodies, trusts, or partnerships (1.73%). 8 SBICAP was incorporated in 1986 as a wholly-owned merchant banking subsidiary of SBI. 9 ADB s equity investment in SBICAP was disbursed in January 1997, and this equity investment is still in ADB s portfolio. The quarterly report on private sector operations for the quarter ended 30 September 2005 gave this investment an RR-5 (substandard) risk rating and assessed its operational status as problems in full operation stage. A project completion report on SBICAP has yet to be prepared. 10 Even though SBI incorporated SBICAP Securities Limited in 1998, no securities trading took place because of the declining Indian economy. SBICAP, which was already overcapitalized at that time, took over instead and incorporated the originally proposed securities trading business into its existing legal structure and business operations. Eventually, ADB s equity investment in SBICAP Securities Limited was canceled on 17 September 1999.

10 3 7. SBIGL, which is the main focus of this PPER, was incorporated on 7 March 1996 and became operational on 1 July ADB finalized the Subscription Agreement with SBIGL and the Letter Agreement with SBI on 29 October Upon fulfillment of predisbursement conditions, for example, the adoption of an operating policy statement, the appointment of an independent external auditor, and the submission of relevant legal documents, ADB subscribed Rs150 million ($4.19 million at disbursement) on 5 March SBIGL s total paid-up capital was Rs1,000 million, of which 85% was contributed by SBI and its associate banks and subsidiaries and 15% was contributed by ADB. D. Expected Outcomes 8. The purpose of ADB s investment in SBIGL, as stated in the report and recommendation of the President (RRP) (footnote 1), was to support the formation of an accredited primary dealers network to facilitate active trading in Government securities. ADB considered its support to the Government securities market as beneficial to the debt market, because the yield curve for Government securities would serve as a benchmark for pricing debt instruments in other risk categories. The OEM concurs with this view. 9. The Letter Agreement stipulates that SBI will (i) reduce its shareholding in SBIGL to less than 50% within 4 years of the date of incorporation of SBIGL, and (ii) arrange for the divestment of ADB s shareholding at the same price at which SBI divests it shares. The subscription agreement stipulates that SBIGL will make a public offering of its shares within 4 years of the date of incorporation. At the time of investment approval, the divestment of ADB s shares from SBIGL was expected to take place 6 years from the date of incorporation. The financial internal rate of return (FIRR) on ADB s investment in SBIGL was estimated at 16.5% in rupee terms and 12.0% in dollar terms, based on a price-earning ratio 11 of 9 and the projected earning per share of Rs2.77 in the sixth year. 12 This calculation assumed that SBIGL would not pay any dividends to ADB during the entire investment period. 13 E. Progress Highlights 10. A project completion report (PCR) on the equity investment in SBIGL was circulated in July The PCR assessed project implementation as successful for several reasons. First, SBIGL maintained its business scope as originally envisaged, and its performance was satisfactory in terms of financial position, profitability, quality of management, and risk control systems. Second, SBIGL remained an active player in both the primary and secondary markets for Government securities. Third, ADB received dividends totaling $1.12 million from Fiscal Year (FY) through FY2000. The PCR considered that calculating the FIRR of this investment would be premature and provided limited information on its development outcomes, especially the performance of the Government securities market. 11. The PCR highlighted two outstanding issues. First, the reduction of SBI s shareholding to less than 50% and the listing of SBIGL s shares on the stock exchange within 4 years of the date of incorporation of SBIGL had not materialized. The PCR stated that ADB would enter into 11 The RRP did not indicate the basis of this price-earning ratio. 12 The RRP did not discuss the reason for this 2-year discrepancy between the company s initial public offering and ADB s divestment. 13 The RRP did not indicate the basis of the zero dividend scenario throughout the investment period despite the projected steady increase in profits. 14 ADB Project Completion Report on the Equity Investment in SBI Gilts Limited in India. Manila. 15 SBI DFHI s financial year corresponds to that of the Government.

11 4 discussions with SBIGL s management to develop a strategy for listing SBIGL in the next 2 years. Second, ADB had not discussed the divestment of ADB shares in SBIGL with SBI and SBIGL. The PCR recommended that ADB monitor and actively seek divestments in the coming years. 12. The OEM largely verified the PCR s findings. The Project s outcome can be summarized as follows. First, the Government securities market steadily grew during FY1997 FY2004 but experienced a setback in FY2005. Second, SBIGL was profitable during FY1997 FY2003. Third, SBIGL merged with the Discount and Finance House of India Limited (DFHI) in FY2004 and was renamed SBI DFHI. Fourth, SBI DFHI was profitable in FY2004, but recorded a net loss in FY2005. Fifth, ADB received dividends every year throughout the investment period except FY2005. Sixth, the two issues highlighted in the PCR remain unresolved, and the Project remains in ADB s portfolio. III. EVALUATION 13. The method of deriving performance ratings drew on the good practice standards (footnote 2). Table 1 shows the performance ratings for the three key dimensions and underlying indicators for the Project. 16 Appendix 1 defines the dimensions, indicators, and 4- point ratings scale referred to in Table 1. This chapter elaborates on the OEM s findings as the bases of the performance ratings for each dimension and indicator. The last section of this chapter provides the overall project rating. Table 1: Evaluation of Investment 7122-IND: SBI DFHI Limited Item Development Outcome Capital Market Development Business Success Private Sector Development Investment Outcome (return on equity) Unsatisfactory Partly Satisfactory Satisfactory Excellent ADB s Effectiveness Screening, Appraisal, and Structuring Supervision and Administration Role and Contribution ADB = Asian Development Bank, SBI DFHI = SBI DFHI Limited. Source: Operations Evaluation Mission. 16 The good practice standards suggest that the synthesis ratings for the three dimensions (development outcomes, investment outcomes, and ADB s effectiveness) reflect summary qualitative performance judgments based on the underlying indicator ratings. The standards assume that the synthesis ratings for the three dimensions are not a simple average of the indicator ratings.

12 5 A. Development Outcome 14. The good practice standards suggest that the rating of the development outcome of financial market operations be based on a project s (i) contributions to the development of the country s private sector and/or the development of efficient capital markets, (ii) contributions to the company s business success, (ii) economic viability, (iii) contributions to the country s living standards, and (iv) environmental sustainability. For this Project, the OEM identified the Project s contributions to capital market development 17 and private sector development and the investee s business success as the underlying indicators for the development outcome. The assessments of the Project s economic viability 18 and the contribution to the country s living standards 19 are integral to the assessments of capital market development in this evaluation. The OEM did not consider environmental sustainability to be relevant to this Project. 1. Capital Market Development 15. The OEM rated the Project s contribution to capital market development as satisfactory based on the assessments in paras a. Performance of the Government Securities Market 16. Table 2 shows the key performance indicators of the Government securities market, extracted from Appendix 2. The assessment of these indicators can be summarized as follows: (i) the outstanding stock of Government securities 20 steadily increased from 15% of gross domestic product (GDP) as of the end of FY1996 to more than 30% as of the end of FY2005; (ii) the average maturity of dated securities issued during FY2002 FY2005 was more than 14 years, compared with less than 6 years in FY1996; (iii) the maximum maturity of the securities reached 30 years, compared with 10 years in FY1996; and (iv) the market turnover of Government securities 21 increased from 52% of the outstanding stock of securities from FY1996 to 359% in FY2003, but dropped to 127% in FY2005. These performance indicators suggest largely satisfactory progress in achieving the Project s purpose with some reservations about the market setback in FY2005. This setback reflected the reversal in interest rates trends, which had constantly fallen during FY1996 FY2004. The ensuing paragraphs further assess these indicators and other relevant information. 17 This PPER assesses the Project s contribution to capital market development separately from its contribution to private sector development for the following reasons. First, development of the Government securities market, the primary purpose of this Project, might not immediately enhance private sector activities given the predominance of state-owned financial intermediaries and institutional investors in India. Second, the Government securities market might not immediately lead to a better corporate debt market. Third, privatization of SBI, a secondary purpose of this Project, could be considered as a separate issue from market development. 18 The good practice standards suggest that the economic viability of financial sector operations should be rated on (i) whether the subprojects financed are economically viable; (ii) whether the project has led to the use of economic viability criteria, or economic internal rate of return, in the intermediary s investment decisions; and (iii) whether the project has resulted in benefits to the economy. The first and second criteria are not relevant to this Project. The third criterion is covered in this evaluation under the assessment of capital market development. 19 The good practice standards suggest that a project s contribution to the country s living standards should be rated based on the project s economic benefits and costs to those who are neither its owners nor its financiers. This subject is covered in this report under the assessment of capital market development. 20 Government securities in India comprise T-bills and dated securities. Dated securities are issued at maturities greater than 1 year. Most dated securities bear coupons. T-bills are short-term instruments typically with a maturity of 92 days and 364 days. In the past, the Government also issued 14-day and 182-day T-bills. The OEM could not obtain consistent year-end data on the outstanding balance of 91-day T-bills, therefore Government securities here refer only to dated securities and 364-day T-bills. 21 Government securities here include all dated securities and T-bills with different maturities in circulation.

13 6 Table 2: Key Government Securities Market Indicators, FY1996 FY2005 FY1996 FY1999 FY2000 FY2003 FY2004 FY2005 Item Unit A. Performance Indicators 1. Outstanding Balance of Dated Securities % of GDP a Outstanding Balance of T-bills b % of GDP a Turnover Ratio of Government Securities c % Average Maturity of Dated Securities Issued During the Year Years Minimum and Maximum Maturities of Dated Securities Issued Years B. Memorandum Items 1. Gross Fiscal Deficit of the Central Government % of GDP a d 2. Weighted Average Cost of the Dated Securities Issued During the Year % Net Market Borrowing by the Central Government e % of the gross fiscal deficit Issuance of Dated Securities Rs billion ,760 1,870 1, Issuance of 91-day T-Bills Rs billion , Issuance of 364-day T-Bills Rs billion GDP = gross domestic product, FY = fiscal year, T-bill = treasury bill. a GDP is provisional for FY2003 and estimated for FY2004 and FY2005. b This refers to 364-day T-bills only. c See footnote 21. d e The fiscal deficit in FY2005 is an estimate. Market borrowing refers to dated securities and 364-day T-bills. Sources: Appendix 2; Center for Monitoring Indian Economy (memorandum items 4 6). 17. During FY1996 FY2003, the central Government budget deficits were sustained within the range of 4.9% to 6.5% of GDP, while the proportion of net market borrowing increased from 39.5% to 71.8% of the Government s budget deficit. This led to the increase in the outstanding stock of Government securities as a percentage of GDP. This trend changed when the Government embarked on a process of fiscal consolidation by enacting the Fiscal Responsibility and Budget Management Act in Pursuant to this act, budget deficits were reduced to around 4.5% of GDP in FY2004 and FY2005. In FY2005, international interest rates and international crude oil prices rose, contributing to an increase in inflation. This led to a tightening of domestic monetary policy and a rise in the yield of Government securities (Figure 1). In response, the Government increased nonmarket borrowing 22 for deficit financing in FY According to official statistics, nonmarket borrowing by the Government includes small savings, state provident funds, special deposits, reserve funds, and 91-day T-bills. The OEM could not obtain the breakdown of nonmarket borrowing in FY2005 so that it could do a detailed analysis of the outcome of the increased reliance on nonmarket borrowing by the Government.

14 7 Figure 1: The Zero Coupon Yield Curve as on 1 March ( ) Yield (%) Maturity (Years) Source: National Stock Exchange, Wholesale Debt Management Segment. 18. The significant increase in liquidity of Government securities during FY1996 FY helped price discovery as reflected in the finer bidding patterns emerging in auctions and the narrow spreads in the secondary market. The increased primary issues and secondary market transactions of Government securities, complemented by the diversification of debt instruments 24 (Appendix 3), led to the entry of additional private sector primary dealers (para. 19). However, the rise in interest rates during FY2005 resulted in a drop in the turnover ratio from 313% in FY2004 to 127% in FY2005 (Figure 2). This suggests that the development of a secondary market in Government securities remains below its potential. 25 The inadequate depth and breadth of the market relates to (i) the commercial banks and Life Insurance Corporation of India s ownership of the majority of dated securities in circulation; and (ii) the large portion of the holdings of dated securities by commercial banks and the Life Insurance Corporation of India being statutorily mandated investments. 26 Recently, cooperative banks, regional rural banks, mutual funds, and nonbank finance companies have entered the Government securities market, but their trading activities remain limited. 23 This partly reflected mutual funds active participation in the secondary market during this period in conjunction with the launching of fixed income schemes. 24 These included floating rate bonds introduced in 1995, capital indexed bonds introduced in 1997, and interest rate futures introduced in In contrast, liquidity in the stock market remained largely consistent across episodes of much higher fluctuations in the index price. 26 For this reason, a large portion of the market trading volume typically ends once the securities are placed with commercial banks and the Life Insurance Corporation of India. Consequently, of 140 to 170 dated securities in circulation, only 10 to 20 have significant liquidity. Little liquidity is available for dated securities that are more than 3 years old.

15 8 Figure 2: Secondary Market Turnover and Dated Security Yield (10-year) 10-year zero coupon security yield (%) Turnover ratio (%) Year Source: National Stock Exchange, Wholesale Debt Management Segment. b. Roles of Primary Dealers 19. India had largely achieved interest rate liberalization by the mid-1990s. This laid the foundation for the development of the Government securities market, which in turn formed the basis for the establishment of the yield curve of debt instruments. In 1996, 27 RBI took the key step of introducing primary auctions 28 at the issuance of dated securities. To facilitate primary auctions, RBI introduced the primary dealer system (described in the box) to assist price discovery and to develop the secondary market. Initially, licenses were given to six public sector financial institutions, including SBIGL. Subsequently, RBI issued primary dealer licenses to 12 additional financial institutions. Following the merger of SBIGL and DFHI in FY2004, 17 primary dealers were operating. Of these, eight are promoted by public sector banks and public financial institutions, six by foreign banks, two with the collaboration of foreign security houses, and one by local nonbanking finance companies. The extent of operational diversification varies across primary dealers The RBI started primary auctions of T-bills in Initially, RBI resorted to uniform price auctions, which since 1999 have progressively been replaced by multiple price auctions. 29 Tentative information obtained by the OEM suggested that five primary dealers, including SBI DFHI, had their revenues driven mostly by primary dealer business in Government securities.

16 9 Box 1: Primary Dealer System Primary dealers are responsible for ensuring the success of primary auctions of dated securities through a system of annual bidding commitments, annual success ratios, and underwriting of the auctions. RBI sets the annual requirement for bidding commitments for primary dealers based mainly on their capital size. The underwriting commitments, as the exclusive opportunity given to primary dealers, are decided separately prior to the actual auction for primary issuance. Primary dealers bid to underwrite various amounts at various commission rates. RBI decides on the actual allotment of the underwriting commitment by primary dealers after considering the commission sought, the amount that succeeded in the primary auction, and the permitted range for devolvement and RBI s underwriting. Primary dealers also bid in the primary issue auction and must fulfill their annual commitment across all the primary auctions conducted during the year. To ensure that bidding is not too defensive, RBI mandates a 40% success ratio. Unlike the underwriting auction, primary dealers compete with other auction participants, mainly RBI-registered financial entities, in the primary bidding auction. a Primary dealers are given access to the RBI repo market, the call money market, and the RBI liquidity adjustment facility for their funding and liquidity management. There is no underwriting procedure for T-bills, and for primary dealers, bidding commitments are fixed at a flat percentage of the notified amount for each auction (determined at the beginning of the year). a To enable small investors to participate in the primary auction, RBI introduced a scheme for noncompetitive bidding in January The scheme provides for the allocation of 5% of a notified amount at the weighted average of accredited bids. 20. Most people interviewed by the OEM acknowledged the contribution of the primary dealer system, complemented by other relevant reforms and market infrastructure development 30 (Appendixes 3 and 4), to the development of the Government securities market during the past decade. During FY2001 FY2004, primary dealers continuously absorbed more than half the Government securities issued in the primary market. Their turnover represented more than 20% of secondary market transactions. However, the primary dealers absorption of Government securities at primary auctions dropped to 28.5% in FY2005, reflecting their reluctance to hold securities in an environment of rising interest rates. As a result, primary dealers turnover of Government securities in the secondary market dropped to 17.7% in FY2005. The contracted activities of primary dealers are closely related to the accounting norm, requiring mark-to-market reevaluation of all securities in hand on primary dealers financial statements. 31 The lack of interest rate hedging instruments also explained the primary dealers prudent stance during the adverse market conditions. 32 According to OEM interviewees, most primary dealers, especially those specializing in Government securities trading, recorded financial losses in FY The major achievement in this area included the negotiated dealing system and the Clearing Corporation of India Limited, both operationalized for Government securities transactions in February The commencement of Government securities trading on the three stock exchanges (the National Stock Exchange, the Bombay Stock Exchange, and the Over-the-Counter Exchange of India) as of January 2003 was another major achievement. 31 A different accounting standard is applied to commercial banks in the revaluation of Government securities categorized as held to maturity. 32 Reflecting the volatility in interest rates, over-the-counter markets have emerged for interest rate swaps and forward rate agreements. Given the short-term nature of these instruments, the risks associated with holding Government securities can only be partially hedged.

17 10 c. Investee s Presence in the Market 21. Table 3 indicates that SBIGL (FY2001 FY2003) and SBI DFHI (FY2004 FY2005) constantly played a lead role in the primary market. Their secondary market transactions were limited to 2.9% to 4.6% of primary dealers total turnover during FY2001 FY2004, but increased to 8.8% in FY2005. Table 3: Primary Dealers Market Shares in Government Securities, FY2001 FY2005 (%) Absorption by Primary Dealers/ Total Amount Issued Primary Market SBIGL and SBI DFHI Absorption a / Absorption by Primary Dealers Primary Dealers Turnover/ Total Market Turnover Secondary Market SBIGL and SBI DFHI Turnover a / Turnover by Primary Dealers Year FY FY FY FY FY FY = fiscal year, SBIGL= SBI Gilts Limited, SBI DFHI = SBI DFHI Limited. Note: Given the use of multiple sources, this table is highly tentative. a FY2001 FY2003 for SBIGL and FY2004 FY2005 for SBI DFHI. Sources: Reserve Bank of India Report of the Internal Technical Group on Central Government Securities Market. Bombay; State Bank of India Annual Report Bombay; SBI DFHI data. 2. Business Performance of SBI Gilts Limited and SBI DFHI Limited 22. The OEM rated the investee s business performance as satisfactory 33 based on the key financial ratios of SBIGL (FY1997 FY2003) and SBI DFHI (FY2004 FY2005) (Table 4). 33 The good practice standards suggest that a project s contribution to a company s performance should be rated on the basis of its contribution to the financial intermediary.

18 11 Table 4: Key Financial Ratios of SBIGL and SBI DFHI, FY1997 FY2005 (%) Item FY 1997 FY 1998 FY 1999 FY 2000 FY 2001 FY 2002 FY 2003 FY 2004 FY 2005 Return on Average Assets (5.1) Return on Average Capital (9.4) Trading Income/Total Income a Interest Income/Total Income a Underwriting Fees/Total Income a Securities Held/Total Capital Underwriting/Total Bidding Success Noninterest Expense/Total Income Total Capital/Total Assets Current Assets/Short-Term Borrowing (1.1) (4.0) Market Shares in Primary Auctions Market Shares in Secondary Trading FY = fiscal year, SBI DFHI = SBI DFHI Limited, SBIGL = SBI Gilts Limited. Note: Given the use of multiple sources, this table is highly tentative. a The summary financial statements submitted by SBI DFHI Limited s management to the Operations Evaluation Department indicated that its total income in FY2005 was negative because of large trading losses, making financial ratios not applicable. Sources: Appendix 5; SBI DFHI data. a. SBI Gilts Limited (FY1997 FY2003) 23. SBIGL remained profitable during FY1997 FY2003, with a return on capital in the range of 18% to 27% from FY1998 (Table 4). During FY1999 FY2001, SBIGL maintained a relatively large portfolio of Government securities that amounted to more than 400% of capital, while increasing short-term borrowing in FY1999 FY2000 and decreasing current assets in FY2001. The entry of 12 additional primary dealers into the market from FY2000 led to SBIGL having a reduced market share in both primary and secondary markets. The new management that took over in FY2001 emphasized profits rather than turnover, resulting in a significant contraction in the secondary market share. As a result, during FY1999 FY2001, more than 85% of SBIGL s income came from interest income rather than from trading income. Subsequently, SBIGL progressively reduced its total holdings of Government securities from 487% of capital as of the end of FY2001 to 225% as of the end of FY2003, taking advantage of the secular downward trend in market interest rates. During this period, SBIGL was highly profitable, resulting in an increase in net worth from Rs1,817 million as of the end of FY2001 to Rs2,740 million as of the end of FY2003. Income from underwriting fees remained marginal throughout the period. b. Merger of SBI Gilts Limited with Discount Finance House of India Limited 24. In 1988 RBI, together with public sector banks and other financial institutions, formed DFHI, one of the oldest money market players in India outside the banking system. After obtaining its primary dealer license in 1996, DFHI shifted its operational focus to Government securities and became the second largest primary dealer in terms of net worth (as of the end of FY2003). In FY2003, SBI bought DFHI s shares from the Industrial Development Bank of India, the Unit Trust of India, and the Life Insurance Corporation. As a result, the proportion of SBI s a

19 12 shareholding in DFHI increased from 29.05% as of the end of FY2002 to 55.13% as of the end of FY2003. The merger of SBIGL with DFHI was a strategic decision by SBI as the sponsor of the two primary dealers. The High Court of Judicature in Bombay sanctioned the proposed amalgamation of SBIGL with DFHI in July The effective date of the merger was 23 April The shareholders of SBIGL were issued 1 share of Rs100 of DFHI for every 11 shares of Rs10 each that they held. This swap ratio was determined based on the independent valuation of shares conducted by two external auditing firms. By the end of March 2005, SBI owned 67.01% of the total of 29,090,906 shares, ADB owned 4.69%, 34 and other banks and financial institutions owned the rest. c. SBI DFHI Limited (FY2004 to date) 25. In FY2004, SBI DFHI recorded profits of 12.6% of average total assets, reflecting the downtrend in yields of Government securities. SBIGL s and SBI DFHI s operational data indicate that the merger resulted in an increased market share in the primary market, but not in the secondary market, in FY2004. The operational efficiency of SBI DFHI in FY2004, measured by the ratio of noninterest expense to total income, was comparable to that of SBIGL prior to the merger. As of the end of FY2004, SBI DFHI s total capital (as a percentage of total assets) was also at the same level as that of SBIGL as of the end of FY2003. The rise in interest rates in FY2005 resulted in a substantial trading loss of Rs1,795 million and a net loss of Rs940 million (equivalent to 5.1% of average total assets). During FY2005, SBI DFHI, like many other primary dealers, reduced its market share in the primary market. The information submitted by SBI DFHI indicated that other public sector primary dealers 35 recorded equivalent losses in FY2005. The management of SBI DFHI explained that the loss in FY2005 was a natural consequence of the drop in prices of Government securities, given that SBI DFHI had to fulfill a significant bidding commitment and achieve the success ratio required by RBI. The management expected improved performance in the first half of FY2006 compared with the same period in FY Private Sector Development 26. The OEM rated the Project s contribution to private sector development as partly satisfactory. While the Project and ADB s overall policy dialogue in the financial sector contributed to the entry of 12 private sector primary dealers, little progress has been made in privatizing the SBI group companies, including SBIGL and SBI DFHI. 27. During appraisal, ADB considered that equity participation would provide an impetus to the SBI group s privatization program, as stated in the RRP. The letter agreement and the subscription agreement stipulated as follows: Within 4 years from the date of incorporation of the Company, or at a later date as the Bank may otherwise agree, the Sponsor shall reduce its shareholding in the Company to less than 50%, and ensure that the combined share ownership in the Company of the Sponsor s majority-owned subsidiaries is reduced to less than 50% (which form of reduction in ownership may include divestment by the Sponsor and issue of new Shares in the Company) subject to obtaining all applicable governmental and other approval. (Paragraph 1 [e], Letter Agreement) 34 The Subscription Agreement ensures ADB s representation on the board of SBIGL as long as ADB holds 5% or more of the issued share capital of the company. With lower ADB shares after the merger, and no revisions initiated on the Subscription Agreement, ADB s representation on the SBI-DFHI s board is no longer ensured. 35 These included Securities Trading Corporation of India Limited, PNB Gilts Limited, and Gilt Securities Trading Corporation Limited. These primary dealers are similar to SBI DFHI in terms of operational scope.

20 13 The Company shall undertake, in consultation with the Bank, to make a Public Offering of its Shares at a price of not less than Rs10 for each share within 4 years from the date of the Company s incorporation or by a later date as the Bank shall otherwise agree. (Section [a], Subscription Agreement) 28. Notwithstanding the foregoing, SBI s management informed the OEM that it did not intend to reduce its shareholding in SBI DFHI to less than 50%, as a reduction in shareholding of subsidiaries through an initial public offering (IPO) is not in accordance with SBI s current corporate policy. Given that SBI DFHI is unlikely to require any additional capital in view of its high net worth, SBI DFHI s management understandably does not consider an IPO to be necessary at this juncture. 29. SBI has undertaken a gradual process of organizational reforms since the mid-1990s following the recommendations of an international consulting firm. As a result, the organization became flatter, forming specialized business units for each market segment, namely, small and medium enterprises, microcredit, rural credit, and so on. This has improved efficiency and expedited decision making. However, little progress has been made in privatizing SBI and its group companies. In view of market imperfections and the remaining weaknesses in the current primary dealer system that surfaced during the market setback in FY2005, the OEM justified continued public sector ownership in some primary dealers, including SBI DFHI, at least as a transitional measure. Assessing the justification for continued public sector ownership in SBI and other group companies is beyond the scope of this PPER (para. 45). 30. The entry of 12 primary dealers sponsored by private sector financial institutions as of FY2000 promoted competition in the Government securities market and contributed to the increased liquidity during FY2000 FY2004. People interviewed by OEM suspected that sophisticated risk management skills introduced by primary dealers sponsored by foreign financial institutions might have had some spillover effects on the local financial industry, possibly contributing to a reduction of systemic risks in the market. B. Investment Outcome 31. The OEM rated the investment outcome as satisfactory. 36 If effective measures to mitigate market risk and exit risk had been in place, the investment outcome could have been assessed as excellent. 32. The OEM used the net asset value method 37 to estimate the FIRR of the investment based on the following: (i) ADB received dividends every year during FY1998 FY2004 that totaled Rs184,909,000 (equivalent to about $4.0 million) and no dividend in FY2005; and (ii) SBI DFHI s net asset value was Rs9,581 million as of the end of FY2005. If ADB had divested all of its shares at the end of FY2005, the FIRR would be 24.4% in rupees and 20.0% in dollars (Appendix 6, tables A6.1 and A6.2). Using the same method and assuming that ADB would divest on 31 March 2006, the FIRR would be 18.3% in dollar terms (Appendix 6, Table A6.3). The assumptions in this calculation include no increase in net worth and no dividends paid to 36 The good practice standards suggest that the investment outcome should be rated based on the investment s gross contribution in relation to the corresponding at-approval standards for minimally satisfactory expected performance. 37 The net asset value method is based on the book value of the company s net worth.