India's banks hold stressed assets of around Rs 9 lakh crore, including over Rs 4 lakh crore in acknowledged bad debts. The bulk of these bad loans is held by public sector banks. A very substantial portion of the stressed and non-performing assets has also arisen from the infrastructure sector. Worryingly, over 50 per cent of bank credit to industry is to infrastructure, with the power sector alone accounting for 22 per cent of outstanding industrial credit. But banks are hardly the ideal vehicles to lend to infrastructure projects since asset-liability mismatches are guaranteed. Most bank borrowings have short- to medium-term tenures, with a preponderance of demand deposits and short-term deposits. The longest tenures amount to five years or so. Meanwhile, infrastructure projects are capital-intensive and have a long gestation period. There may be zero or negligible revenue accruing for the first few years. In the Indian context, with endemic delays, projects are often of even longer gestation. Even if a project eventually gets underway and generates returns, banks (and other Indian lenders including specialised infrastructure non-bank financial corporations) cannot afford to wait that long. India does not possess lenders with the requisite deep pockets, such as pensions and reinsurance companies. Banks have tried to control the risks of infrastructure lending by setting prudential sector limits and creating consortiums. But the asset-liability mismatches remain and the systemic risks are not mitigated by being spread out. The quantum of required funding is large enough to have badly affected most public sector banks. Attempts to create structures such as take-out financing have also not proved efficacious. A take-out structure implies that a loan will be taken over by a new lender after it is held for several years by the first lender. In this, the first lender continues to face huge risks.
Read more from our special coverage on "INFRASTRUCTURE"
The ideal instruments for financing long-gestation, capital-intensive projects are long-tenure bonds. A bond may be subscribed to by anybody, without fear of asset-liability mismatches - because, in theory, bonds can be sold on. However, that implies an active secondary bond market with the requisite liquidity to price a bond accurately and sell it on. Unfortunately, India does not possess an active secondary market in bonds. The major players in the bond market consist of foreign portfolio investors or FPIs (who are constrained by preset limits), banks and a few debt mutual funds. The primary market consists largely of government debt and a large proportion is held to maturity. Imparting liquidity to the secondary market would require policy measures to encourage trading and to attract more players including high-net worth individuals and overseas FPIs. Among other things, this has to involve further changes to the methodology by which bank portfolios are marked to market, in order to induce more activity. The current situation is after all unsustainable. Banks cannot continue to operate with such an overhang of asset-liability mismatches and such massive non-performing assets. Nor can financing for the infrastructure sector be throttled because banks have hit their lending limits. Developing a more active secondary market in bonds should be a matter of priority and urgency because it can address both those issues at the same time.