Executive Summary
1. Introduction
2. Developments in the Management of State Government Guarantees
3. Fiscal Risk of Guarantees in Indian States
4. Methodology for Computing Fiscal Risk of State Government Guarantees
5. State Government Guarantees: Major Issue and Recommendations
References
Annexure
Appendix Tables

In the Eighth Conference of the Finance Secretaries of State Governments with the Reserve Bank of India (RBI), on May 26, 2001, it was noted that several States have taken initiative to fix a ceiling on guarantees pursuant to the recommendations of the Technical Committee on State Government Guarantees (February, 1999). However, the devolvement probabilities of various guarantees are not identical and consequently, all guarantees cannot be treated as uniform in terms of their fiscal impact. There is thus a need to evolve a methodology for classifying guarantees into categories having broadly similar fiscal impact so as to assess the fiscal risk arising out of guarantees in a more realistic and objective manner. This, in turn, would facilitate the fixing of ceiling on guarantees in a non-mechanistic fashion to better reflect the risk inherent in guarantees, and enable making better provisioning to cover these liabilities. In this backdrop, it was decided to constitute a Group to examine the fiscal risk of guarantees extended by the State Governments. The Committee consisted of finance secretaries from Andhra Pradesh, Bihar, Gujarat, Kerala, Maharashtra, Meghalaya, Uttar Pradesh and members from Ministry of Finance, Government of India and Planning Commission. The Chief General Manager in Charge of Internal Debt Management Cell, RBI was the convenor. The terms of Reference of the Group were

1. To analyse and classify the different types of Guarantees, including letters of comfort issued by States,
2. To arrive at a methodology for assessing the fiscal risk of each type of guarantee, and,
3. To review the Legislations/Acts and policies of financial institutions having mandatory provisions requiring guarantees.

The Committee had three meetings in Mumbai. The second meeting was followed by a meeting with the financial institutions.

2. The Group noted the developments in the management of guarantees by states vis-à-vis the recommendations of the Technical Committee. Seven states (Assam, Goa, Gujarat, Karnataka, Rajasthan, Sikkim, West Bengal) have in place either an administrative or statutory limit on guarantees with two more states (Kerala and Tamil Nadu) in the process of fixing a ceiling.

3. The Technical Committee of State Finance Secretaries on State Government Guarantees had observed that pre-emption through automatic debit mechanism runs the risk of resulting in insufficient funds for financing obligatory payments such as salaries, pensions and interest payments. In view of the recommendation of the Committee as well as after examining the past experience with automatic debits, it has been announced in April 2002 by RBI in its ‘Monetary and Credit policy for the year 2002-03’ to, as a general policy, dispense with such automatic debits in future where there are no legal or other compulsions, and to suggest amendments where there is a legal compulsion. RBI has also proposed to review all existing automatic debit mechanism in consultation with state governments and others concerned to dispense such mechanisms wherever feasible. In keeping with the statement in the policy the Group recommends that the automatic debit mechanism in case of repayments to NABARD both under the RIDF and outside the RIDF should be discontinued and treated as similar to any other off-market borrowing of the State Government from financial institutions.

4. On the issue of transparency, the Group was of the view that the states need to publish data regarding guarantees regularly, in the format recommended by the Group of Finance Secretaries (Annexure V) in the annual budget. To further improve transparency it is recommended that both the annual sanctions of guarantees and outstanding amount need to be disclosed in the state budget. Further, in order that a proper database is created for capturing all guarantees, a Tracking Unit for guarantees may be designated (in the Ministry of Finance) at the state level. In the interest of financial stability and transparency, in addition, states may disclose the information on default, invocation and payment performance.

5. The Group noticed the continuously rising trend of outstanding guarantees which grew at an average rate of about 16 per cent per annum during the period 1992-2001. In terms of sectoral distribution the power sector at 44.6 per cent was the largest constituent of outstanding guarantees. As for beneficiaries, banks at about 15 per cent and all-India financial institutions at about 25 per cent were significant as an investor class. Although financial institutions are increasingly becoming conscious about the importance of proper risk assessment of projects, some institutions continue to insist on guarantees, either because it is mandated in the Acts governing them or as a conscious policy. The Group studied such Act provisions/policies of major institutions and recommended that the need for extending guarantees in favour of central financing agencies owned by Government should be examined and even done away with. It is essential that the lending institution should undertake due diligence and examine the commercial viability of the project instead of relying on State government guarantee. Insistence on viability of projects and generation of adequate repayment capacity will push through reforms in the area of levying user charges and removal of subsidies so essential in the reform process. The Group also recommended that where guarantee is taken as credit enhancement, it should be reflected in reduction in the lending rate.

6. The Group studied the methodology for assessing fiscal risk of guarantee obligations prevailing internationally. While it recognized the importance of classifying guarantees in terms of their default probabilities, it concluded that in the Indian context this methodology would not accurately reflect the fiscal risk in guarantees. It suggested the following methodology.

1. One of the first steps in assessing the fiscal risk of guarantees is to clearly segregate those which are effectively in the nature of direct liabilities and report these separately and assess the risk of such guarantees at 100% viz. as equal to debt. Such guarantees should be clubbed with debt while assessing the debt profile of the State for all purposes. A large number of guarantees fell in this category. It was noted by the Group that the Ministry of Finance, Government of India has already adopted a similar approach in the discussions under the Medium Term Fiscal Reforms Programme. For such guarantees, which are more like debt, it is apparent that the repayment provision should be made in the budget itself. Ministry of Finance, Government of India, will have to assess this while finalising the Plan and borrowing programme of the State.

2. For the rest of the guarantees, measuring the fiscal risk is necessary. This would involve further classification of the projects/ activities as high risk, medium risk, low risk and very low risk and assigning appropriate risk weights. The assessment of risk will be done at the State level. For making such assessment there are various options that can be adopted by the States. Making use of credit rating of bonds is one option. Government of India has already instructed that all bonds issued in future with government guarantee will have to be compulsorily credit rated. While reiterating the importance of compulsory credit rating, it should be kept in mind that invariably the rating is enhanced because of the availability of guarantee or some structured payment mechanism. State Governments should, in such cases, assess the risk sans guarantee with the assistance of rating agencies. It is therefore necessary that for all such guaranteed bonds, a dual rating, with and without guarantee, should be obtained. The rating, without taking into account guarantee, could then be used for purposes of classifying into high risk, medium risk, low risk and very low risk. In respect of existing loans and bonds a similar exercise can be undertaken by the State Finance Departments.

3. Once the guarantees have been categorized into Very Low, Low, Medium and High risk categories, the finance departments of states will have to use their judgement to assign devolvement probability to each risk category, say 5% for very low risk, 25% for low risk, 50 % for medium risk and 75% for high risk. The translation of risk assessment to devolvement probability is essentially a matter of judgement and is best left to individual States. The devolvement probability could then be applied to the underlying liabilities which are guaranteed to estimate the guarantee devolvement obligation, which could then be added to debt service obligation to arrive at the annual fiscal burden of debt and guarantees.

4. The guarantee commissions charged by States do not bear much relation to the underlying risk and may not be sufficient to constitute the Guarantee Redemption Fund (GRF). Secondly, it is infeasible at the present stage to increase guarantee commission as most bodies in favour of whom guarantees are extended are also in the public sector. Therefore the Group recommends that at least an amount equal to 1 per cent of outstanding guarantees may be transferred to the GRF each year from the fisc specifically to meet the additional fiscal risk arising on account of guarantees. The guarantee commission collected could also be credited to this Fund.

5. The Group felt that increasing the existing rates of guarantee commission may not be practical as the projects will not be able to bear additional guarantees, although merit was seen in linking guarantee fee to the category of risk. It can be left to each state to decide whether it would like to charge guarantee fee according to risk category.

6. As per the Eleventh Finance Commission, States should aim to limit interest payments to 18% of revenue receipts in the medium term. This norm could be modified to include possible devolvement on account of guarantee obligations of the states on the basis of the methodology indicated above, and the total obligation should not exceed 20% of revenue receipts. This will automatically serve as one measure for capping guarantees. Many states may have currently debt service plus guarantee obligation in excess of 20 %. In their cases it is imperative to place limits on incremental guarantees in any year in relation to revenue receipts.

7. In order to have a norm in terms of debt sustainability the underlying guarantee liabilities can be mapped out and likely amount of devolvement could be estimated for future years. The total of such likely devolvement during the life of the guarantees could then be treated as normal debt and clubbed together with debt obligations. Together, the liability could be measured as a ratio of SDP to ensure that debt plus likely devolvement on guarantees during its life is sustainable and to ensure that guarantees are also captured in such measures. To refine such measures the sustainability can be worked out in terms of net present values and then measured as ratio of SDP.

8. It was also felt that apart from assessing the fiscal risk and making provisions, the State Government should also take administrative measures to discipline the state level undertakings whose borrowings are guaranteed and set up an arrangement whereby they make provisions to meet possible shortfalls in project earnings. The Group recommends one of the following two methods to be used at the discretion of the state governments.

• The borrowing SPV/PSU/Co-operative/Local body be made to set up escrow accounts with contributions from project earnings on a predetermined and regular schedule. In the event of the revenue of the project suffering for any reason, repayments to the guaranteed bond/loan holders could be made out of these accounts before resorting to state government guarantees.

• A proper valuation of the user charges may be made and they may be enhanced suitably to go into a contingency fund/provision in the books of the borrowing institution, to be accessed in case of shortfalls in revenue.

It was felt that while any one of these contingency measures is very essential, the actual choice of which alternative to adopt and the mechanics of such an arrangement are best left to the individual state governments.

In the Eighth Conference of the Finance Secretaries of State Governments with the Reserve Bank of India (RBI), on May 26, 2001, it was noted that several States have taken initiative to fix a ceiling on guarantees pursuant to the recommendations of the Technical Committee on State Government Guarantees (February, 1999). However, the devolvement probabilities of various guarantees are not identical and consequently, all guarantees cannot be treated as uniform in terms of their fiscal impact. There is thus a need to evolve a methodology for classifying guarantees into categories having broadly similar fiscal impact so as to assess the fiscal risk arising out of guarantees in a more realistic and objective manner. This, in turn, would facilitate the fixing of ceiling on guarantees in a non-mechanistic fashion to better reflect the risk inherent in guarantees, and enable making better provisioning to cover these liabilities. In this backdrop, it was decided to constitute a Group to examine the fiscal risk of guarantees extended by the State Governments. The Committee consisted of finance secretaries from Andhra Pradesh, Bihar, Gujarat, Kerala, Maharashtra, Meghalaya, Uttar Pradesh and members from Ministry of Finance, Government of India and Planning Commission. The Chief General Manager in Charge of Internal Debt Management Cell, RBI was the convenor.

2. The terms of Reference of the Group were

1. To analyse and classify the different types of Guarantees, including letters of comfort issued by States,
2. To arrive at a methodology for assessing the fiscal risk of each type of guarantee, and,
3. To review the Legislations/Acts and policies of financial institutions having mandatory provisions requiring guarantees.

The Committee had three meetings in Mumbai. The second meeting was followed by a meeting with the financial institutions.

3. After this Introductory Chapter, Chapter 2 reviews the development relating to the management of guarantees by the State Governments and other related issues. Chapter 3 discusses the nature of fiscal risk arising out of State government guarantees, the volume and distribution of guarantees issued by the states. The issue of guarantees to financial institutions has also been dealt with in this chapter. Chapter 4 discusses the methodology for computing the fiscal risk of state government guarantees. The recommendations are summarised in Chapter 5. There are four annexures to the Report. Annexure 1 contains a list of persons associated in the work of the Committee. Annexure II presents the format on Disclosure of guarantees proposed by the Core Group on Voluntary Disclosure Norms for State Governments (January, 2001). Annexure III presents the questionnaire relating to Guarantees given to the State Governments for the purpose of this Report. Annexure IV describes the international experience in managing guarantees. Annexure V gives a format for disclosure of Guarantees by State Governments while Annexure VI gives provisions of the Acts/Policies of financial institutions governing guarantees. A set of Tables on state finances is also appended to the report.

Recognizing the growing magnitude of guarantees issued by State Governments and their potential impact on the future fiscal position of the States, the need to have a policy on guarantees was felt in the meeting of Finance secretaries held in November 1997. In response to the request made by the Finance Secretaries of the States, the RBI constituted a Technical Committee on State Government Guarantees, consisting of selected State Finance Secretaries to examine the issue of State government guarantees in all its aspects. The Committee submitted its report in February 1999. The recommendations made by the Committee related to, inter alia, (i) Imposition of ceiling on guarantees, (ii) Selectivity in calling for and providing of guarantees, (iii) Greater transparency in the reporting of guarantees and standardisation of documentation, (iv) Guarantee fee and constitution of a contingency fund for guarantees, and (v) Monitoring and honouring of guarantees. The developments in the management of guarantees by state governments subsequent to the Report are delineated below.

2. Following the guidelines given in the Report of the Technical Committee on guarantees, Karnataka and Rajasthan (1999) Assam, Sikkim (2000), West Bengal (2001) have introduced ceiling on guarantees. The developments in different states relating to imposition of ceiling on guarantees is placed in Table 1.1

3. The issue of imposition of ceiling on guarantees is under active consideration in Tamil Nadu and Kerala. Tamil Nadu is planning to implement a ceiling on guarantees at 30 per cent of the revenue receipts of the second preceding year. In Kerala, a cap of 10 per cent of the Gross State Domestic Product (GSDP) is proposed for guarantees.

4. The Technical Committee had recommended that each State should set up a contingency fund or make some provision for discharging the devolvement on guarantees provided by them. The guarantee fees collected should be credited to the fund set up for the purpose. RBI had circulated guidelines in this regard in August 2001. Several States have taken steps to set up a Guarantee Redemption Fund and earmarked guarantee fees towards the Fund. Orissa may be considered a pioneer, having introduced a Guarantee Reserve Fund as far back as in 1969 and from 2002-03, a Guarantee Redemption Fund will replace the existing Guarantee Reserve Fund. The Government of Orissa has already earmarked Rs.20 crore for the Guarantee Redemption Fund. Gujarat has created a Guarantee Risk Fund with an initial provision of Rs.25 crore. This apart, guarantee fees received by the State are transferred to the Fund. Andhra Pradesh created a Guarantee Redemption Fund in 2001 and contributed Rs.12.1 crore to the corpus in 2001-2002. It has also decided to contribute 1 per cent of the guarantees outstanding against such corporations /institutions whose liability is not taken over directly or indirectly by the Government as at end-December of every year. This apart, all accretions by way of guarantee commission realised during the preceding year will be earmarked and transferred to the Guarantee Redemption Fund. The Governments of Karnataka and Rajasthan each have also set up a Guarantee Redemption Fund. Presently, some other states are also actively considering the proposal to set up a Guarantee Redemption Fund in their respective states.

5. The structure of Guarantee fee varies from state to state. International experience shows that guarantee fees do not have much relation to the risk profile of the loans guaranteed. Guarantees should be structured in a manner so as to minimize taxpayer exposure and to strengthen private performance incentives. Klein (1997) has suggested establishment of an agency, a guarantees commission, at arms-length from project promoters, which would help develop standardized guarantee products, facilitate learning across projects, reduce the need for state and municipalities to issue guarantees, allow the employment of competent staff to do so and limit taxpayer exposure in a relatively transparent way.

6. In the Indian context, no consistent association between the guarantee fee and the default probability is observed. In 1992, the Central Government decided the fee structure for guarantees. Borrowings under the market borrowing programme were to be charged a guarantee fee of 0.25 per cent per annum of the guaranteed amount whereas for guarantees covering public sector borrowings, a guarantee fee of 1 per cent was fixed. For other borrowings, the guarantee fee was fixed at 2 per cent.

7. The guarantee fee structure of selected states for which information has been obtained is shown in Table 2. The guarantee fees varied from 0.1 per cent to 2 per cent for different categories of projects.

8. In October 1998 RBI had introduced prudential requirements for guaranteed loans and investments. It advised banks that investment in State Government guaranteed bonds outside the market borrowing programme would attract a credit risk weight of 20 per cent. In case a guarantee is invoked but the bond has remained in default, a credit risk weight of 100 per cent is assigned. The enhanced risk weight applies to the guaranteed bonds of the defaulting entities. Furthermore, with effect from March 31, 2000, in respect of advances sanctioned against State government guarantee, if a guarantee is invoked and remains in default for more than two quarters, such loans will be categorised as doubtful assets and provision of 100 per cent to the extent to which the advance is not covered by the realisable value of the security has to be made. As regards the advances guaranteed by the state governments which stood invoked as on 31.3.2000, necessary provision is required to be made in a phased manner during 2000-01 to 2002-2003 with a minimum of 25 per cent each year.

9. Secondly, in case of infrastructure financing and financing of Special Purpose Vehicles (SPVs) against government guarantee, the RBI has advised the banks/FIs that while they are free to sanction term loans for technically feasible, financially viable and bankable projects undertaken by both public sector and private sector undertakings, they shall fully satisfy themselves that the projects financed by them have income generating capacity sufficient to service such loans. Further, banks/FIs should satisfy themselves that the project is run on commercial lines and that they do not run into liquidity mismatch on account of lending to such project. Further, in February 2002, the RBI stressed that state government guarantees may not be taken as a substitute for satisfactory credit appraisal and such appraisal requirements should not be diluted on the basis of any reported arrangement with the RBI or any other bank for regular standing instructions/periodic payment instruction for servicing the loans or bonds. As per the February 2002 Circular of RBI, while such public sector units may include SPVs registered under the Companies Act set up for financing infrastructure projects, it should be ensured by banks and financial institutions that these loans/investments are not used for financing the budget of the State Governments. Banks and financial institutions should undertake due diligence on the viability and bankability of such projects to ensure that revenue stream from the project is sufficient to take care of the debt servicing obligations and that the repayment/servicing of debt is not out of budgetary resources. Further, in case of financing SPVs, banks and financial institutions should ensure that the funding proposals are for specific monitorable projects other than those being implemented by State Governments in view of the fact that the borrowings of State Governments for budgetary purposes are met by banks and financial institutions by contributions to their approved market borrowing programmes.

10. Some recent initiatives at the state level to promote investment in infrastructure also raise the issue of fiscal risk of contingent liabilities. For instance, in the Andhra Pradesh Infrastructure Development Enabling Act (APIDEA), 2001, the Government has decided to guarantee receivables of infrastructure projects provided they are not collected directly from users. The Government will also provide off-take guarantees if it is the service distributor and is responsible for collection of user levies. In West Bengal, in the interest of the development of infrastructure, the Government has decided to continue to give performance guarantees. The West Bengal Ceiling on Government Guarantees Act, 2001 specifies that the ceiling on guarantees imposed by the Government will not be applicable for any loans raised by West Bengal Infrastructure Development Finance Corporation Limited.

11. Though privatization and autonomy to PSUs in the infrastructure sector is expected to reduce the need for financial support from Government, the Government in many cases is likely to substitute a contingent liability for a real revenue liability in the form of a variety of project oriented guarantees such as traffic guarantees in the case of toll roads. While there is a need to raise funds for infrastructure in view of long term growth considerations, there is also a need to look at the qualitative dimension of guarantees and the associated fiscal risk.

12. The issue of providing automatic debit mechanisms to assure payment of State Government dues by debit to RBI account has also come under close scrutiny at the Conference of State Finance Secretaries. Some state governments had earlier given instructions to RBI to debit their accounts on specified dates or in case of specified events to meet certain obligations. This issue had earlier been examined by the Technical Committee which had observed that pre-emption through automatic debit mechanism runs the risk of resulting in insufficient funds for financing obligatory payments such as salaries, pensions and interest payments. With increasing stress on the funds position of state governments and given the discipline of the WMA/Overdraft Regulation Scheme, the general consensus amongst the finance secretaries, Ministry of Finance, Government of India and Reserve Bank of India was that such automatic debit mechanism should not be allowed for any of the other bonds issued/guaranteed by the state governments other than the state development loans (SDL) issued under the approved market borrowing programme. In view of the recommendation of the Committee as well as after examining the past experience with automatic debits, it has been announced by RBI in its ‘Monetary and Credit Policy for the year 2002-03’, as a general policy, to dispense with such automatic debits in future where there are no legal or other compulsions, and to suggest amendments where there is a legal compulsion. RBI has also proposed to review all cases of existing automatic debit mechanism in consultation with state governments and others concerned to dispense such mechanisms wherever feasible.

13. In the case of account with RBI, apart from SDL, automatic debit mechanism is permitted for Rural Infrastructure Development Fund (RIDF) and for borrowings from NABARD. In keeping with the statement in the Monetary and Credit Policy the Group recommends that the automatic debit mechanism in case of repayments to NABARD both under the RIDF and outside the RIDF should be discontinued and treated as similar to any other off market borrowing of the State Government from financial institutions.

Transparency

14. A major constraint in analyzing the true fiscal position of States is the absence of a consistent and standard pattern of reporting data on guarantees. In the Conference of state finance secretaries held on June 12, 1999, a Core Group on Voluntary Disclosure Norms for State Governments was set up. The Group recommended disclosure of guarantees in a specific format (Format annexed - Annexure II). The statement would cover not only explicit guarantees but also letters of comfort and other structured payment arrangements which impinge on the budget of the States. Some State Governments disclose information on outstanding guarantees and guarantees issued by them in the State Budgets. In Andhra Pradesh, all contingent liabilities are disclosed in the State budget. Rajasthan discloses guarantees figures in Budget Document 4(8). West Bengal discloses all guarantees in the Budget Publication No. 6. Orissa had brought out a white paper on guarantees.

17. With a view to streamlining the process for borrowings by state governments in the overall context of fiscal consolidation, Government of India has implemented the following procedure. The Fiscal Reforms Unit of the Department of Expenditure, Ministry of Finanace, will work out the limits of maximum prudential debt of each state on a year-to-year basis and convey it to the Planning Commission. Within the allocation approved by the Planning Commission, necessary permission under Article 293(3) of the Constitution will be given to the states on application. Regarding borrowing by Special Purpose Vehicles (SPVs) the following procedure has been put in place.

1. For running concerns which are able to service the loans, prior permission under Article 293(3) may not be necessary.
2. States need to obtain prior permission under Article 293(3) for borrowings by Special purpose Vehicles which clearly have to be serviced by the state,
3. SPV borrowing which enters the Public Account of the state budget, and is further on lent to the Consolidated Fund of State, will not in the ordinary course be allocated,
4. All such state government guaranteed borrowing, through private placement or directly negotiated with financial institutions, will have to be compulsorily credit rated.

18. While broadly welcoming the new framework for borrowings by state governments, Reserve Bank had proposed that negotiated loans that enter the Consolidated Fund and constitute the direct liability of the State Governments should not be raised from banks for the following reasons.

1. Such borrowings dilute the discipline of having a market borrowing programme since they fall outside the ambit of the market borrowing programme.
2. The volume of market borrowing by the Special Purpose Vehicles compete with the approved market borrowing programmes of the state governments and lead to pressure on the interest rates.
3. By virtue of a guarantee, they also lead to crowding out of resources available to the private sector.

19. Reserve Bank therefore proposed that specifically approved financial institutions (other than banks) could subscribe to the negotiated loans. RBI also proposed that SPV loans which enter the Public Account of the state budget and are further on-lent to the Consolidated Fund should be made conditional to such borrowing being restricted to specifically approved financial institutions. Both the proposals have been endorsed by the Central Government and the list of such financial institutions is being finalized.

There is a growing recognition in the emerging literature on guarantees in different countries and research pioneered by the World Bank that the fiscal risk of sovereign/quasi-sovereign guarantees is growing, principally due to the hard budget constraint imposed on a number of sub national entities requiring them to mobilize funds through guaranteed loans and bonds. With the rise in the magnitude of contingent liabilities, the direct liabilities arising out of defaults have also been on the rise. This may be attributed to lack of proper feasibility analysis of projects by either the lender or the guarantor, lack of review and monitoring by lender organisations and overall inadequate appreciation of the fiscal implications of guarantees. It is thus crucially important that Governments must consider the expected value of future commitments while issuing guarantees. At the simplest level, this would require that the Government knows what guarantees it has issued and how much it might bear if the guarantees were invoked. This is done by estimating the default probability of guarantees. Valuation of guarantees enables decisions to be made on the basis of real rather than apparent costs and benefits. In other words, a clear methodology to analyse the fiscal risk should be followed by the State Governments. Since the fiscal risk of guarantees varies from project to project and also across sectors, a proper classification is needed to estimate the fiscal risk of guaranteees.

2. In the first part of this chapter the fiscal sustainability indicators of Indian States is presented. It shows that while the fiscal situation of the Central Government remained more or less stable over the last few years, the fiscal situation of the States has weakened. The second part looks at the magnitude and composition of guarantees issued by Indian States, and how it compares with other liabilities of the State Governments, including debt. The final part of the chapter deals with the specific issue of the policies related to guarantees followed by the financial institutions.

2. Gross Fiscal Deficit as a ratio to NSDP has risen significantly in many states over the last decade. (Appendix Table 1) While Centre’s deficit has remained more or less steady, States’ GFD showed sharp increases after 1998. More importantly, Revenue Deficit (RD) to GFD ratio has risen at both levels of Government indicating deterioration in the quality of fiscal deficit (Appendix Table 2). The nominal stock of domestic debt of the combined Government sector has been growing at about 16 per cent during the latter part of the 1990s(Appendix Table 3). The higher growth in domestic debt than in the nominal GDP growth has led to steady debt accumulation to 63.7 per cent of GDP by end March 2001 as compared with 58.3 per cent as at end-March 1996. The debt growth remained below the nominal GDP growth during the first half of 1990s but has generally exceeded the latter since then (Appendix Table 3). The need for reducing the fiscal deficit and debt to sustainable levels has led to the Central Government proposing a fiscal responsibility and budget management bill.

3. As on end-March 2001, the outstanding State government guarantees amounted to Rs.1,69,562 crore (8.9 per cent of the GDP) and is more than the outstanding market loans of the State governments. The potential liability on account of guarantees as and when they devolve constitute a higher fiscal risk which the States need to take cognizance of. In this context, prudent management of the fiscal risk arising out of guarantees is of utmost importance. Also any defaults on part of the State to service guarantees would affect the financial position of Banks/FIs. Hence, the fiscal risk of guarantees needs to be properly evaluated and policy action to contain such risks should be chalked out and implemented.

4. The freedom for borrowing by the States is limited due to a number of reasons. First, States are allocated borrowing limits under the market borrowing program and negotiated loans in consultation with the Planning Commission. Second, loans from Central Government are also pre determined as part of Plan assistance. Furthermore, there are limits on Ways and Means advances (WMA) and Overdraft. The magnitude of the public account as a source of funding the deficits is by and large beyond the control of the State Governments. This apart, the entire (earlier, eighty per cent) small savings collections are passed on to the States. To some extent States can influence the collections under small savings through aggressive small savings drive and special incentives have been given by some States to increase the small savings mobilised in the State. In sum, therefore, a hard budget constraint operates at the State level. The spillover in revenue expenditure and revenue deficit beyond the budgeted figure is accommodated through reduction in capital expenditure. Revenue receipts have remained stagnant on account of low user charges, losses of public sector enterprises, slow /negligible growth in tax revenues whereas revenue expenditure has increased sharply on account of recent pay revisions. Consequently, the phenomenon of increasing revenue deficit and lower capital expenditure has become endemic.

5. In order to meet the requirements of financing infrastructure and compensate for the decreasing capital expenditure, States have been increasingly resorting to issuing guarantees on behalf of pubic sector entities undertaking infrastructure investment and other developmental activities. In addition, there are assured payment arrangements which, even if not backed by explicit guarantees, represent a direct liability on the cash flow of the State. Hence, the rising guarantees and assured payment arrangements at the State level pose issues of sustainability of State finances much in the same way as rising level of debt poses at the Central level. The following table gives the state-wise outstanding guarantees

6. Outstanding guarantees grew from Rs.40,317 crore at the end of March 1992 to Rs.133,062 crore in 2000 and Rs.169,562 in 2001.

7. The distribution of guarantees among different sectors show that guarantees to infrastructure projects constituted a major share of guarantees at the State level. As per data of selected Indian states as on March 2001, outstanding guarantees of the State Governments covered financing of power sector ( 44.6 per cent) , road ( 0.7 per cent) and other infrastructure including irrigation ( 13.4 per cent). These apart, housing sector constituted 5.4 per cent whereas guarantees on behalf of cooperative banks was 11.0 per cent and State Financial Corporations was 4.5 per cent.

Financial Institutions and State Government Guarantees

8. On a rough approximation, of the total state government guarantees, banks account for about 15 per cent, all-India financial institutions about 25 per cent and others (including central public sector units in the power sector) about 60 per cent.

9. Financial institutions turn out to be significant beneficiaries of guarantees. Some of the institutions have statutory requirement for guarantees, but almost all the institutions ask for government guarantees routinely while sanctioning loans or for subscribing to bonds floated by state level undertakings. The policies followed by the various financial institutions with regard to state government guarantees are governed by statutory provisions and/or policies determined by the management of the respective institutions.

10. A major portion of Government guarantees in recent years has been given to facilitate the development of infrastructure including power, road transport, housing, urban development, cooperative sector and other welfare projects of the State Governments. Infrastructure financing by financial institutions like IDBI, IFCI, ICICI, SIDBI and IIBI has been quite substantial.

11. As at end-March 2000, the all India Development Financial Institutions (e.g., IDBI, IFCI, ICICI, SIDBI and IIBI) provided Rs.86,917 crore - Rs.65,769 crore towards power projects, Rs.15,246 crore towards telecommunication projects, Rs.5,902 crore towards development of roads, ports and bridges. Such large volume of financing of the infrastructure sector has been possible due to the active role played by the State as a facilitator, through guarantees and other forms of assurances.

12. The following Table gives the percentage distribution of the sector-wise guarantees outstanding and the defaults for banks and financial institutions.

1. The overall default ratio at 3.7 per cent is not very high.

2. A substantial proportion of guarantees have been given by State Governments to the Power sector (34.2 per cent of total outstanding guarantees by the Banks and all India Financial institutions) and the agriculture and cooperative sector (41.7 per cent).

3. The highest amount of default is in the power sector. It accounts for more than half (52.8 per cent) of all defaults on guarantees and has a 15.1 per cent default ratio. The guarantees given to this sector bears the highest fiscal risk and to a certain extent this has been addressed by the Ahluwalia Committee. There is a continuing need to look at the power sector guarantees more closely in order to assess their impact on the medium term fiscal situation.

4. The default ratio is highest in the Industry sector (39.2 per cent) where a large number of textile, sugar and jute mills are running at a loss and unable to meet their principal and interest payment obligations. However, the industry sector got only 7.3 per cent of guarantees and thus not as significant as the power sector.

14. Thus, guarantees to power, industry, infrastructure and welfare projects should be closely monitored. While issuing fresh guarantees to these sectors, the State Governments need to properly assess the fiscal risks involved.

15. Almost all the state guaranteed bonds of SPVs or other SOEs are issued in the market through the private placement route. Many bonds are neither rated nor listed. In view of the growing size and importance of the private placement market (which constitutes more than 90% of the corporate debt market) which is growing at 45% annually over the last 6 years, and the high exposure of banks to this market (82% of banks' non-SLR investments is through the private placement route), RBI had constituted the "Working Group to Evolve a Framework for Collecting and Sharing by Banks/Financial Institutions of Information on Private Placement of Debt" which submitted its report in February 2002. On the basis of the recommendations of this Working Group, RBI has formulated draft guidelines (for all non-SLR investments, including private placements) to be issued to banks, a few important ones of which are as follows. To ensure the quality of issues RBI has instructed banks to invest only in issues rated by a credit rating agency and where the rating is current and valid. Investment in non-SLR securities may be made only in listed securities, or where the issuer has undertaken, in the offer document, to list the issue within three months of the issue. Further, banks are expected to make their own internal credit analysis and rating. To ensure end-use of funds, banks have been asked to satisfy themselves on the intrinsic viability of the project and undertake due diligence on the issuer as well as the project. It is also expected that the minimum disclosures in the offer document to be finalised by banks under the aegis of IBA/FIMMDA would include disclosure on end use of funds and an undertaking to provide self-certification on end-use to the debenture trustees and investors on a half-yearly basis.

16. The Act/Legislation governing policies of financial institutions related to guarantees is provided in Annexure VI. A brief description of the statutory provisions and practices followed by various financial institutions is given below.

17. National Bank for Agriculture and Rural Development (NABARD) : Clause (a) of sub-section (3) of Section 21 of the National Bank for Agricultural and Rural Development Act, 1981 provides that the National Bank may at its discretion provide by way of refinance loans or advances - to any State Co-operative Bank if the loan or advance is fully guaranteed for repayment of principal and interest by the Government. Thus it is clear that guarantee is mandated for such loans. Also there is no differentiation on interest rates on the basis of guaranteed and non-guaranteed loans. NABARD's policy is to insist on guarantees only in the case of non-scheduled banks. A beginning has been made during the current financial year particularly in the case of Gujarat to accept Government securities in lieu of Government guarantee.

18. National Co-operative Development Corporation (NCDC) : Clause (e) of sub-section (2) of Section 9 of the National Co-operative Development Act, 1962 provides, inter alia, that

19. From the above, it is clear that guarantees are mandated for NCDC loans to co-operative societies. However, for granting loans and advances directly to the national level co-operative societies and other co-operative societies with their scope extending beyond one State, the Corporation may grant loans without guarantee of the State Government/Central Government. To enable the Corporation to grant loans to (State) co-operative societies without the guarantee of the State/Central Government, the provisions in this regard of the National Development Co-operative Societies Act may have to be amended suitably. It is informed by NCDC that the Act is being amended to relax insistence on government guarantees. A committee has been set up to improve the appraisal and monitoring system. The current practice in NCDC is that about 80-90 per cent of the projects are appraised strictly where Government guarantees are not considered as a substitute. There have been occasions when NCDC refused to sanction projects that had government guarantee if they were not found viable. It is also informed that NCDC has not invoked any guarantee so far.

20. Life Insurance Corporation (LIC) : Clause (e) of sub-section (2) of Section 6 of the Life Insurance Corporation Act, 1956 provides that the Corporation is empowered to make advances or lend money upon the security of movable or immovable property or otherwise. In Life Insurance Corporation Act, 1956 there is no provision mandating guarantee by the State Governments while granting loans and advances to infrastructure/State Government sponsored/run project.

21. The policy followed by LIC can be broadly summarized as follows -

1. Loans to SEBs and SRTCs are generally granted on the security of mortgage/hypothecation of assets. However, in some states non-plan loans in power sector are granted under Government guarantee, guarantee being considered as superior to mortgage of assets. In some cases borrowers themselves have offered government guarantee as security instead of mortgaging assets.
2. Loans for water supply and sewerage sectors are granted for financing water supply and sewerage schemes of local bodies. Quantum of LIC's loan depends on cost of the scheme and is generally 25% to 66.7% of the cost, depending on cost. The remaining cost is borne by the local body/State Governments. Water supply being the primary responsibility of the State Governments and as balance cost over and above LIC loan has to be borne by the Government, guarantee from the State Government is sought for.
3. LIC has an obligation to invest 25 per cent in Government Bonds and Government guaranteed securities. LIC insists on a rating of at least AA and, in the absence of rating, Government guarantee is necessary. Investments are approved only on the basis of viability of the projects which ensures revenue streams. Government guarantee is required in the schemes relating to social sector like Housing, Road Transport, Power sector and Water. There is a large proportion of Government guarantee in respect of power corporations, infrastructure and state development corporations. The default rate on government guarantees at 16.1% is rather high.

22. National Housing Bank (NHB): NHB sanctions loans for projects only on the basis of viability. For example, in the State of Himachal Pradesh, no projects were sanctioned on the basis of guarantee and in the case of Rajasthan, out of 16 projects only one project was sanctioned against the guarantee. Appraisal standards are not diluted.

23. Housing and Urban Development Corporation (HUDCO) : HUDCO is governed by the Companies Act, 1956. There is no mandate in its Act for guarantees and the Board of HUDCO lays down policies for obtaining/dispensing with state government guarantees for loans to infrastructure/government sponsored projects.

24. These institutions are owned by the Central Government. It may not be appropriate for the Central Government owned institutions to obtain State Government guarantees from state level institutions. It is therefore imperative that the Central financial institutions (especially those owned by Central Government) assess projects solely on basis of viability and do not ask for guarantees. The Group is of the view that there is no need for extending guarantees in favour of central financing agencies owned by Government and hence such guarantees may be done away with. It is essential that the lending institution should undertake due diligence and examine the commercial viability of the project instead of relying on the State government guarantee. Insistence on viability of projects and generation of adequate repayment capacity will push through reforms in the area of levying user charges and removal of subsidies so essential in the reform process. Where guarantee is taken as credit enhancement, it should be reflected in reduction in the lending rate.

Chapter 4
Methodology for computing Fiscal Risk of State Government Guarantees

International Evidence

Ever since the rapid growth of guarantees in the US in the 1970s and the consequent fiscal strain in the 1980s, Governments have become conscious about the fiscal risk that is embedded in the issuance of guarantees by the State. The concern was mirrored in the development or adoption of a wide range of techniques, both sophisticated as well as simple, that came to be applied to the valuation of such risk. Such risks represent the future costs of commitments made now or earlier.

2. Expected losses can be calculated using a wide variety of techniques. In developed countries, like US, fiscal risks are estimated using option-pricing models. In this process the issuance of guarantees are equated to put options, and the deviation in the underlying asset values are used to estimate the value of guarantee. (Sosin,1980). The measurement through option pricing models is data dependent and it requires historical annual returns to measure business risk and the traded value of the equity to determine the subsidy component of guarantees (assuming zero guarantee fees). Besides the option pricing model, it is possible to estimate the value of guarantees by using market based methods which are relatively straightforward in technique (often called "Rule of Thumb" techniques, Mody and Patro,1996). One approach in this category, used by Metron (1990) is to estimate the implied value of loan guarantees as the difference between market price of the security and a comparable default free security (Treasury Bonds) of similar maturity. The approach requires an active secondary market that prices the security and that the guarantee cover is not partial. Partial guarantees complicate the valuation process, which then needs option pricing models.

3. Market-based measurement of the implicit guarantee value is also feasible when comparable instruments with or without guarantees exist or where market values of a security exist before and after guarantee. (Hsueh and Kidwell, 1988), Quingley and Rubinfield, 1991). In cases where a government has issued a large number of similar guarantees for many years and has recorded information on defaults, the expected cost of the guarantees can be estimated using actuarial techniques. It is also possible to use econometric modeling or simulating outcomes based on multiple scenarios with different probabilities.

4. Valuing a government’s guarantees and other contingent liabilities has important advantages over simply noting maximum exposures. By calculating the expected cost of guarantees, the government and its observers can more easily compare guarantees with cash subsidies. When guarantees are not valued, a government may choose to provide a guarantee instead of a subsidy even when the former is more costly, because the costs of the guarantee are hidden and may be borne by a future administration. When guarantees are valued, decisions are more likely to be made on the basis of real, rather than apparent, costs and benefits. (Thobani, 1999).

5. While calculating contribution to hidden deficit (a flow concept) and outstanding public liabilities (a stock concept) , Brixi et al (2000) provide a methodology for the calculation of risk weights of guarantees. For the calculation, each guarantee and its underlying project are assessed. Projects are ranked according to their risk. Accordingly, the default risk of each guarantee is estimated. Risk-adjusted amount (hidden subsidy) of state guarantees is calculated by multiplying the loan amount for which a guarantee was issued and its default risk. Instead of assigning quantitative values of risks, risks were divided into four categories (Very High, High, Medium and Low).

6. For each of these categories default probabilities are attached based on perception. With the help of such classification, risk adjusted amounts of guarantees issued each year and the claims paid from the budget on guarantee defaults each year were calculated. Brixi et al also provide projected fiscal strain estimates based on such risk assessments. Such forecasting may be useful.

7. Measuring the fiscal risk of guarantees is beset with difficulties as many of the States do not report guarantees in the budget document. Also, as the secondary market for securities with State Government guarantees is not sufficiently deep, market prices of these securities are not available. While discussing the parameters for fixing the ceiling on guarantees, the Technical Committee suggested reduced weight for guarantees while calculating the true fiscal obligation of the States. Normally, the weight for guarantees should be based on the historical experience of devolvement of guarantees on the State Governments. The Report suggested that each State could make its own assessment of the default probability. However, as a rule of thumb, the Report suggested converting guarantees into debt equivalent by using a factor of 1/3 and then fixing a ceiling at 50% of NSDP for outstanding debt and guarantees (1/3), with some state specific modifications. The Committee was in favour of sufficient flexibility to each State Government to choose the most appropriate parameter while ensuring transparency in respect of all the parameters. The approach was simple, but it could not differentiate among risky and sound investments that were guaranteed by the State Governments. It was recognised that an improved method of estimating the fiscal risk has to be evolved and related to some extent with the guarantee fee structure. A classification of guarantees was required to arrive at the true fiscal costs of guarantees.

8. A realistic method of classifying guarantees is according to their default probabilities. However, in the current Indian context this was not considered a viable alternative. Before the middle of the 1990s both the volume and the devolvement of guarantees were low and manageable for the state governments. There has been a spurt of guarantees thereafter and the devolvement magnitude of these guarantees would only be making itself apparent now. Thus the past default data would not be a reasonable predicator of the default probability of the recent guarantees, and to that extent, are likely to underestimate the fiscal risk. More sophisticated methods like option pricing, actuarial techniques etc. require a reliable and substantial database which is lacking. Besides, for states to use any valuation and classification technique easily, the emphasis has to be on simplicity and uniformity.

9. The Group debated at length to arrive at a functional classification of guarantees based on the risk profile of these guarantees. It was felt that the starting point is a uniform and complete disclosure of all guarantees as recommended earlier. This would lead to transparency in the fiscal operations of the State Governments.

1. One of the first steps in assessing the fiscal risk of guarantees is to clearly segregate those which are effectively in the nature of direct liabilities and report these separately and assess the risk of such guarantees as 100% viz. as equal to debt. Such guarantees should be clubbed with debt while assessing the debt profile of the State for all purposes. A large number of guarantees fell in this category. It was noted by the Group that the Ministry of Finance has already adopted a similar approach in the discussions under the Medium Term Fiscal Reforms Programmes. Repayment provisions for these guarantess, which are in the nature of 100% risk should be made in the budget itself. Ministry of Finance, Government of India, will have to assess this while finalising the Plan and borrowing programme of the State.

2. For the rest of the guarantees, measuring the fiscal risk is necessary. This would involve further classification of the projects/ activities as High risk, Medium risk, Low risk and Very Low risk and assigning appropriate risk weights. The assessment of risk will be done at the State level. For making such assessment there are various options that can be adopted by the States. Making use of credit rating of bonds is one option. As stated earlier, Government of India has already instructed that all state government guaranteed borrowing will have to be compulsorily credit rated. While reiterating the importance of compulsory credit rating, it should be kept in mind that invariably the rating is enhanced because of the availability of guarantee or some structured payment mechanism. State Governments should, in such cases, assess the risk sans guarantee with the assistance of rating agencies. It is therefore necessary that for all such guaranteed bonds, a dual rating, with and without guarantee, should be obtained. The rating, without taking into account guarantee, could then be used for purposes of classifying into high risk, medium risk, low risk and very low risk. In respect of existing loans and bonds a similar exercise can be undertaken by the State Finance Departments.

3. Once the guarantees have been categorized into Very Low, Low, Medium and High risk categories, the finance departments of states will have to use their judgement to assign devolvement probability to each risk category, say 5% for Very Low risk, 25% for Low risk, 50 % for Medium risk and 75% for High risk. The translation of risk assessment to devolvement probability is essentially a matter of judgement and is best left to individual States. The devolvement probability could then be applied to the underlying liabilities for which guarantees have been extended and arrive at the debt service obligation plus guarantee devolvement obligation to arrive at the annual fiscal burden of debt plus guarantees.

4. The guarantee commissions charged by States do not bear much relation to the underlying risk and may not be sufficient to constitute the Guarantee Redemption fund (GRF). Secondly, it is infeasible at the present stage to increase guarantee commission as most bodies in favour of whom guarantees are extended are also in the public sector. The Group recommends that at least an amount equal to 1 per cent of outstanding guarantees may be transferred to the GRF each year from the fisc specifically to meet the additional fiscal risk arising on account of guarantees. The guarantee commission collected could also be credited to this Fund.

5. As per the Eleventh Finance Commission, states should aim to limit interest payments to 18 per cent of revenue receipts in the medium term. This norm could be modified to include possible devolvement on account of guarantee obligations of the States on the basis of the methodology indicated above, and the total obligation should not exceed 20% of revenue receipts. This will automatically serve as one measure for capping guarantees. Limits could similarly be placed on incremental guarantees in any year in relation to revenue receipts especially for States where guarantees outstanding are already at very high levels.

6. In order to have a norm in terms of debt sustainability the underlying guarantee liabilities can be mapped out and likely amount of devolvement could be estimated for future years. The total of such likely devolvement during the life of the guarantees could then be treated as normal debt and clubbed together with debt obligations. Together, the liability could be measured as a ratio of SDP to ensure that debt plus likely devolvement on guarantees during its life is sustainable and to ensure that guarantees are also captured in such measures. To refine such measures the sustainability can be worked out in terms of net present values and then measured as ratio of SDP.

7. The Group felt that increasing the existing rates of guarantee commissions may not be practical as the projects will not be able to bear additional guarantees, although merit was seen in linking guarantee fee to the category of risk. It can be left to each state to decide whether it would like to charge guarantee fee according to risk category.

10. It was also felt that apart from assessing the fiscal risk and making provisions, the State Government should take administrative measures to discipline the state level undertakings whose borrowings are guaranteed and set up an arrangement whereby they make provisions to meet possible shortfalls in project earnings. Among the many measures suggested, the Group recommends one of the following two methods to be used at the discretion of the state governments.

1. The borrowing SPV/PSU/Co-operative/Local body be made to set up escrow accounts with contributions from project earnings on a predetermined and regular schedule. In the event of the revenue of the project suffering for any reason, repayments to the guaranteed bond/loan holders could be made out of these accounts before resorting to state government guarantees.

2. A proper valuation of the user charges may be made and they may be enhanced suitably to go into a contingency fund/provision in the books of the borrowing institution, to be accessed in case of shortfalls in revenue.

It was felt that while any one of these contingency measures are very essential, the actual choice of which alternative to adopt and the mechanics of such an arrangement are best left to the individual state governments.

The Group’s recommendations are circumscribed by the need to limit guarantees so as to contain fiscal risk while ensuring that financing of development and infrastructure is not impeded in the States. The underlying focus of the recommendations is, therefore, on

• the overall fiscal sustainability,
• the need to ensure that the risk of devolvement is minimized through due diligence, proper appraisal and follow up,
• proper assessment of the fiscal risk arising on account of guarantees, and
• limiting fiscal risk through sustainable limits on debt plus guarantee servicing linked to revenue receipts where guarantee servicing is taken as the risk of devolvement assessed on the basis of credit rating sans guarantee

2. The recommendations of the Group are given below.

1. While noting that there are a large number of guarantees in regard to liabilities which were clearly intended to be met out of the budgetary resources, the Group took the view that such guarantees should be identified separately and treated as equivalent to debt. Such guarantees should be transparently included, reported and disclosed as indirect debt in the debt profile of the State to be monitored by the GOI and the States as part of overall debt for purposes of ascertaining sustainability of the State Government.

2. The Group was of the view that states need to publish data regarding guarantees regularly, in a uniform format (Annexure V) in the annual budget. To further improve transparency it is recommended that both the annual sanctions of guarantees and outstanding amount need to be disclosed in the state legislature.

3. In order that a proper data base is created for capturing all guarantees and monitor their underlying liability, a Tracking Unit for guarantees may be designated (in the Ministry of Finance) at the State level.

4. A large proportion of guarantees is in favour of all-India financial institutions like NABARD, NCDC, HUDCO, LIC, etc., which are owned by the Central Government. Since in the Indian federal context, there is an implicit underwriting of States’ borrowings by the Centre, such guarantees amount to the Centre guaranteeing itself. The Group therefore, recommends that Acts/policies of these central financial institutions should be amended/rationalized so that guarantees are not routinely insisted upon while extending loans. The Group felt that the need for having such guarantees should be examined and even done away with. It is essential that the lending institution should undertake due diligence and examine the commercial viability of the project instead of relying on the security of government guarantee. Insistence on viability of projects and generation of adequate repayment capacity will push through reforms in the area of levying user charges and removal of subsidies apart from ensuring financial stability. Where guarantee is taken as credit enhancement, it should be reflected in reduction of lending rate.

5. Consequent to the announcement in the Monetary and Credit policy for the year 2002-03 to dispense with automatic debit mechanism, the Group recommends that the automatic debit mechanism in case of repayments to NABARD both under the RIDF and outside the RIDF should be discontinued and put on the same footing as any other off-market borrowing of the State Government from financial institutions.

6. While working out the methodology for assessing the fiscal risk of guarantees other than those which can be considered as direct liability, fiscal risk can be measured as follows. Projects/ activities need to be classified as high risk, medium risk, low risk and very low risk and assigned appropriate risk weights. The assessment of risk will be done at the State level. For making such assessment, States can make use of credit rating of bonds sans guarantee. These ratings can then be used for assigning devolvement probability which could then be applied to the underlying liabilities for which guarantees have been extended to arrive at the debt service obligation plus guarantee devolvement obligation representing the annual fiscal burden of debt plus guarantees.

7. The Group recommends that at least an amount equal to 1 per cent of outstanding guarantees may be transferred to the GRF each year from the fisc specifically to meet the additional fiscal risk arising on account of guarantees.

8. As per the Eleventh Finance Commission, states should aim to limit interest payments to 18 per cent of revenue receipts in the medium term. This norm could be modified to include possible devolvement on account of guarantee obligations of the States on the basis of the methodology suggested by the Group, and the total obligation limited to 20% of revenue receipts. This will automatically serve as one measure for capping guarantees. Limits could similarly be placed on incremental guarantees especially for States where guarantees outstanding are already at very high levels.

9. In order to have a norm in terms of debt sustainability the underlying guarantee liabilities can be mapped out and likely amount of devolvement could be estimated for future years. The total of such likely devolvement during the life of the guarantees could then be treated as normal debt and clubbed together with debt obligations. Together, the liability could be measured as a ratio of SDP to ensure that debt to SDP ratio is sustainable and to ensure that guarantees are also captured in such measures. To refine such measures the sustainability can be worked out in terms of net present values and then measured as ratio of SDP.

10. The Group felt that increasing the existing rates of guarantee commissions may not be practical as the projects will not be able to bear additional guarantees, although merit was seen in linking guarantee fee to the category of risk. It can be left to each state to decide whether it would like to charge guarantee fee according to risk category.

11. It was also felt that apart from assessing the fiscal risk and making provisions, the State Government should also take administrative measures to discipline the state level undertakings whose borrowings are guaranteed and set up an arrangement whereby they make provisions to meet possible shortfalls in project earnings. The Group recommends one of the following two methods to be used at the discretion of the state governments.

• The borrowing SPV/PSU/Co-operative/Local body be made to set up escrow accounts with contributions from project earnings on a predetermined and regular schedule. In the event of the revenue of the project suffering for any reason, repayments to the guaranteed bond/loan holders could be made out of these accounts before resorting to state government guarantees.

• A proper valuation of the user charges may be made and they may be enhanced suitably to go into a contingency fund/provision in the books of the borrowing institution, to be accessed in case of shortfalls in revenue.

It was felt that while any one of these contingency measures are very essential, the actual choice of which alternative to adopt and the mechanics of such an arrangement are best left to the individual state governments.

1. Brixi, Hana Polackova, Hafez Ghanem and Roumeen Islam, 1999. Fiscal Adjustment and Contingent Government Liabilities, Policy Research Working Paper No. 2177, World Bank
2. Brixi, Hana Polackova, Sergei Shatalov and Leila Zlaovi (2000), Managing Fiscal Risk in Bulgaria, World Bank
3. Hsueh, Paul L. and David S. Kidwell, 'The impact of a State Bond Guarantee on State Credit Markets and Individual Municipalities', National Tax Journal 41: 235-245.
4. IDBI, Development Banking in India, 1999-2000
5. Klein, Michael, 1997, Managing Guarantee Programs in Support of Infrastructure Investment, World Bank.
6. Lewis, Christopher and Ashok Mody, 1997. 'The Management of Contingent Liabilities: A Risk Management Framework for National Governments' in Timothy Irwin, Michael Klein, Guillermo Perry and Mateen Thobani, eds., Dealing with Public Risk in Private Infrastructure.
7. Mody Ashoka, and Dilip Patro (1996), 'Valuing and Accounting for Loan Guarantees', World Bank Research Observer, Vol 11, No. 1, February, 119-42
8. Polackova, Hana, 1998. Contingent Government Liabilities: A Hidden Risk for Fiscal Stability, Policy Research Working Paper No. 1989, World Bank.
9. Quigley, John M. and Daniel L. Rubinfeld. 1991. ' Private Guarantees for Municipal Bonds: Evidence from the After Market.' National Tax Journal December:29-31.
10. Sosin, Howard B. 1980. 'On the Valuation of Federal Loan Guarantees to Corporations'. Journal of Finance 35: 1209-1221.
11. Thobani, M., 1999, Private Infrastructure, Public Risk, Finance and Development, March, World Bank.

1. Finance Secretary of Andhra Pradesh - Shri S. K. Arora
2. Finance Secretary of Bihar- Shri M. N. Prasad
3. Finance Secretary of Gujarat- Shri S. G. Mankad
4. Finance Secretary of Kerala- Smt. Sudha Pillai
5. Finance Secretary of Maharashtra- Shri A. K. D. Jadav and J. S. Sahni
6. Finance Secretary of Meghalaya- Shri P. J. Bazeley
7. Finance Secretary of Uttar Pradesh- Shri B. M. Joshi
8. Convener of the Group- Smt. Usha Thorat, CGM-in-Charge, Internal Debt Management Cell, Reserve Bank of India
9. Advisor, Planning Commission, Government of India- Dr. N. J. Kurien
10. Joint Secretary, Ministry of Finance, Government of India- Dr. R. Bannerji

1. Please detail the different types of outstanding guarantees as at the end of March 2001 in the following format -

(Rs. Crore)

Sector	Guarantees		Letters of Comfort		Other arrangements including structured payment arrangements
	Loans	Bonds	Loans	Bonds	Loans	Bonds
1. Power
2. Road Transport
3. Ports
4. Urban Development And Housing
5. SFCs
6. State Co-op.Banks
7. Other Infrastructure of which Irrigation
8. Others



2. If available, please provide data on guarantees in the following broad categories –

Broad categories of Guarantees extended by the State Government

(Rs. Crore)

S. No	Type of Institution	Amount Guaranteed during 2000-01	Amount Guaranteed outstanding as on March 31, 2001
1	Running Concerns/Firms/Undertakings
2	Securitised dues of SLUs/PFCs
3	Borrowings by SPVs serviced by the State Budget

3. Please give a broad categorisation of the beneficiaries of outstanding guarantees as on March, 2001 (banks, FIs, Central government institutions etc.,)

Beneficiaries of Outstanding Guarantees as on March 31, 2002

(Rs. Crore)

S. No.	Institutions	Outstanding Guarantees
1	Banks
2	IDBI, ICICI, IFCI
3	HUDCO
4	LIC
5	NCDC
6	PFC
7	NHB
8	REC
9	NABARD
10	Others

4. What is the fee charged for issuance of guarantees and what is being done with this revenue? Is there any relation between the fee structure and the probability of guarantees being devolved on the State Government?
5. Is there a Guarantee Redemption Fund or such similar provision to address the future liabilities that may arise on account of guarantees? What in your view should be the formulation to fix the contributions to such a fund ?
6. Are the contingent liabilities disclosed transparently in the State Budget?
7. Whether you have or are planning to have a ceiling on guarantees – if so, what is the criteria for fixing the ceiling?
8. What is the present policy in giving guarantees – are there any priorities for any specific sectors. If so which are these?
9. What steps would you suggest that should be undertaken to minimize the risk of devolvement under various types of guarantees?

The international experience in managing the fiscal risk associated with guarantees is crucial to understand the issues that have arisen worldwide while dealing with guarantees and other contingent liabilities.

Governments throughout the world tend to provide guarantees to private investors in a variety of activities. Prominent among such guarantees are deposit insurance for bank depositors as well as pension or social security insurance. Otherwise, guarantees for housing, agriculture, students, exports and public corporations tend to dominate the picture in OECD countries. Existing data suggest that total guarantee exposure may amount to some 15 - 20 per cent of GDP or more than a quarter of gross debt in OECD countries. This does not capture implicit guarantees. During the 1980s OECD export credit agencies incurred huge losses and the Government as guarantor had to bear part of it. This experience has prompted a slow change in guarantee management procedures. Most prominently, the United States has instituted more transparent accounting principles for its guarantee operations under the 1991 Credit Reform Act.

To manage the contingent liabilities in a judicious manner , the Australian Government has issued a Finance Circular 1997 on Potential Liabilities and Losses of the Government. This Finance Circular is the major policy document in this area, and aims to answer common questions on the management and reporting of contingent liabilities. The Finance Circular includes:

• definitions of indemnities, guarantees and letters of comfort (all of which give rise to contingent liabilities) and their use;
• the types of risks covered by indemnities and guarantees;
• the importance of these instruments in the context of risk management;
• the basis of authority for issuing guarantees and indemnities;
• the difference between issuing these instruments and spending public money;
• the importance of seeking appropriate authority for contracts which contain indemnities as well as spending proposals;
• the need to seek legal advice on relevant instruments;
• the need to record all contingent liabilities in agency registers; and
• the need to monitor, report and review all contingent liabilities.

1. Gross Fiscal Deficit as a Ratio to NSDP
2. Fiscal Indicators: Centre and States
3. Growth of GDP, Growth of Debt and Average Cost of Debt: Indian States

Note: Data for 2000-01 for Arunachal Pradesh, Bihar, Himamchal Pradesh, Jammu & Kashmir, Karnataka, Madhya Pradesh, Maharashtra, Orissa, Punjab, Sikkim and Uttar Pradesh is estimated by projecting the NSDP based on the growth rate in 1999-00. For Goa, projections for the years (1998-99, 1999-00 and 2000-01) is based on the growth rate of 1997-98. For Mizoram and Nagaland, projections for the two years (1999-00 and 2000-01) is based on the growth rate of 1998-99.