It would be a reform move towards greater transparency if, in addition to cash based numbers, the Indian Budget were to be presented on an accrual basis to account for contingent liabilities. In India, there has been a tendency to look for cash-neutral ways to meet unanticipated current expenditure by issuing long-term government bonds. Clearly, this results in inter-temporal inequity since the lower costs are availed now and the expenses are pushed out into the future.
In the last several years, as the Indian growth story has taken hold, we are again not paying sufficient attention to asset-liability mismatches in public finance. For instance, the focus is shifting away from provisioning for defined benefits pensions for government and armed forces personnel. In recent months, the spotlight has been on the unsustainibility of continued issuance of oil bonds to meet the under-recoveries of the oil-marketing companies (OMCs), namely IOC, BPCL and HPCL. The stocks of oil bonds, in residual maturity buckets of 4 years and below, 5-10 years and 13 or more years, are shown in the table. These oil bonds are bullet bonds, i.e. the principal is not amortised over the life of the bonds and interest is paid semi-annually. Effectively, there is an asset-liability mismatch since no specific assets are being held against these oil bond liabilities.
About 77 per cent of oil bonds issued carry maturities longer than 13 years and the longest-maturity securities expire on March 29, 2026. If the OMCs retain the bonds on their balance sheets, there would be considerable duration risk. Alternatively, if the oil securities are redeemed prior to maturity these are bought by insurance companies such as LIC at discounts of 7 per cent or higher.
In comparison, even though the bonds are similar in nature, less attention has been paid to FCI and fertiliser bonds. During 2006-07, the government issued 16-, 18- and 20-year maturity securities to FCI amounting to Rs 16,200 crore. Separately, on December 7, 2007, the government issued 15-year maturity bonds of Rs 3,890 crore to 22 fertiliser companies.
The history of the Indian financial sector is dotted with instances of disasters, which occur when investors are offered assured returns without ensuring prudent asset-liability management (ALM). If an assured return of, say, 8 per cent per annum is guaranteed, the promoter of such an investment scheme has to ensure that the corresponding fixed-income assets yield 8 per cent plus management fees. In the event of a duration mismatch between the assured-return liabilities and the associated assets, additional risk capital needs to be allocated. It is widely understood that the familiar borrow short, lend long practice in banking has to be backed by risk capital. It is less well accepted in India that asset management activities require adequate risk capital to cover at least three standard deviation disaster scenario events and provide for fat tail distributions.
The lack of uniform enforcement of sound ALM practices in the Indian financial sector leads to periodic "scams". The individual and institutional investors who are adversely affected manage to get the attention of regulatory authorities and such incidents usually end up with palliative measures at the expense of tax-payers. For instance, the 1992 Harshad Mehta episode included the surreptitious use of overnight bank funds to take long positions in the stock market. This extreme form of asset-liability mismatch is solvent as long as stock-market valuations keep going up. Later in 1996-98, the CR Bhansali, Nedungadi Bank, Canara Bank mishaps and then in 2001 the Madhavpura Cooperative Bank "accident" happened because, among a number of reasons, equity investments were funded by debt liabilities. As stock values eroded and the underlying liabilities matured there was a run on the financial institution.
In the real sectors, for example, power, fertilisers and hydrocarbons, mismatches between government revenues and liabilities are not viewed from an ALM perspective. This is because there can be no watershed disasters with the government as the back-stop. However, basic ALM practices should be observed since indefinite subsidies can foster, apart from pushing the burden of adjustment on to future governments, fatally flawed asset-liability mismatches much in the same manner as in the financial sector.
More From This Section
There has been a steady moving away from administered interest rates, exchange rates and insurance premiums in the Indian financial sector. In contrast, there is less of a consensus about administered prices in the real sectors. The conventional wisdom is that before relaxing controls in the financial sector we should get prices right in the other sectors. Given the opening up of the financial sector, it is time that India did more to track and appropriately account for "assured returns" in real sectors.
With the constraints imposed by the FRBM Act, oil bonds are not included in the Budget numbers. In the interest of fairness across time, the principal amounts of these bonds should amortise equally over their maturities. This would spread the fiscal burden uniformly over the life of these securities on successive governments. Oil bonds are not granted SLR status but in other respects these securities are similar to regular government bonds with principal repayment at maturity. Long-term investors such as insurance companies favour bullet bonds since they prefer to limit reinvestment risk on their books. Consequently, if accrual based budgeting were to be adopted, the repayment burden on the exchequer could be accounted for over the life of government bonds without altering the principal redemption characteristics.
As of now, about 40 per cent of OECD countries have adopted accrual based budgeting. However, according to OECD and IMF surveys it is difficult to arrive at consistently meaningful accrual based numbers for government budgets. It follows that in India it will take time to train a core group of professionals. Further, there is no unanimity in the literature that accrual based accounting is appropriate for all government expenditure. Johan Christiaens and Jan Rommel writing in Financial Accountability and Management (February 2008) state that the "economic information provided by accrual accounting is (more) relevant for the business like (parts of) governments". All issues considered it would be a progressive step if cash based Indian Budget numbers are supplemented by accrual based figures. In due course, state governments could also be encouraged to present their budgets on a cash and accrual basis.