How will the creation of a 'bad bank' affect India's banking sector?

The bad bank's form would be that of a centralised asset-restructuring company.

The creation of a 'bad bank' could accelerate the resolution of stressed assets in India's banking sector, but it may face significant logistical difficulties and would simultaneously require a credible bank recapitalisation programme to address the capital shortfalls at state-owned banks, says Fitch Ratings.

Here's more from Fitch:

India's banks have significant asset-quality problems that are putting pressure on profitability and capital, as well as constraining their ability to lend. Fitch expects the stressed-asset ratio to rise over the coming year from the 12.3% recorded at end-September 2016. The ratio is significantly higher among state-owned banks. Asset-quality indicators may be close to their weakest levels, but the pace of recovery is likely to be held back by slow resolution of bad loans.

A bad bank that purchases stressed assets and takes them to resolution was featured in the government's latest Economic Survey, and in a speech by a senior Reserve Bank of India official. Its most likely form would be that of a centralised asset-restructuring company (ARC).

Its proponents believe it could take charge of the largest, most complex cases, make politically tough decisions to reduce debt, and allow banks to refocus on their normal lending activities. Similar mechanisms have previously been used to help clean up banking systems in the US, Sweden, and countries affected by the Asian financial crisis in the late 1990s. Senior European policymakers have recently discussed the prospect of a bad bank to deal with NPLs in the EU.

Fitch believes that a bad bank might provide a way around some of the problems that have led Indian banks to favour refinancing over resolving stressed loans. For example, large corporates often have debt spread across a number of banks, making resolution difficult to coordinate.

The process would be simplified if the debt of a single entity were transferred to one bad bank. This could be particularly important in India's current situation, with just 50 corporates accounting for around 30% of banks' stressed assets.

Several small private ARCs already operate in India but they have bought up only a very small proportion of bad loans in the last two years, as banks have been reluctant to offer haircuts on bad loans even where they are clearly worth much less than their book value. This is, in part, because haircuts invite the attention of anti-corruption agencies, making bank officials reluctant to sign off on them. Reduced valuations also increase pressure on capital.

A larger-scale bad bank with government backing might have more success. However, it is unlikely to function effectively without a well-designed mechanism for pricing bad loans, particularly if the intention is for the bad bank to be run along commercial lines and involve private investors.

One estimate from the Economic Survey suggests that 57% of the top 100 stressed debtors would need debt reductions of 75% to make them viable. Banks would need capital to cover haircuts taken during the sale of stressed assets, and the bad bank would most likely require capital to cover any losses incurred during the resolution process.

Fitch estimates that the banking sector will require around US$90bn in new total capital by FY19 to meet Basel III standards and ongoing business needs. This estimate is unlikely to be significantly reduced by the adoption of a bad-bank approach, and could even rise if banks are forced to crystallise more losses from stressed assets than we currently expect.

We believe that the government will eventually be required to provide more than the USD10.4bn that it has earmarked for capital injections by FYE19 - be it directly to state-owned banks or indirectly through a bad bank.