RBI policy: Extra haircut on rated bonds to put poorly run states in a spot

The Reserve Bank of India’s proposal to offer one percentage point extra haircut on margins for liquidity operations for rated state bonds may dampen prospects for poorly run states, say experts.

In its June monetary policy, the central bank said initial margin for rated SDL mortgaged with RBI for liquidity operation would be in the range of 1.5 to 5.0 per cent, depending on maturity basket. Unrated SDLs would have to take a haircut of 2.5 to 6.0 per cent. But the market is divided whether really the states would want to get rated to start with.

Also, the move to value state development loans, or state bonds, at market prices, instead of a uniform 25 basis points above government securities (G-secs), may push banks to book higher treasury losses and appetite for bonds issued by states would likely reduce.

So far, banks have been booking valuation gains on the bonds. For example, if a bond is purchased at G-sec plus 60 basis points, and is valued at G-sec plus 25 basis points, the bank gets to book a valuation gain of 35 basis points on the bonds purchased, boosting its profit. Now, if they have to value those at market prices, there would be no gain and there is a likelihood of a loss too.

“If valuation-linked gains are not there, the incentive to invest in these bonds would lessen in a rising interest rate scenario,” said Soumyajit Niyogi, associate director, India Ratings and Research.

Generally, state bonds are rarely traded and banks keep them under their held-to-maturity category, where they don’t need to be valued in line with the market rates.  

“While the RBI cannot push states to get rated, it is offering incentives to banks as investors to convince states to get rated. This will improve the transparency and fiscal discipline of states,” Niyogi said.

Not all states would want to get rated, but the better ones can ask for a finer pricing on their loans.

“Only those well-run states that have fiscal deficit, lower debt and low interest cost would voluntarily get rated. Once they get a rating of say, AAA, they can then tell the market to treat them on a par with the central government and offer a rate similar or closer to the equivalent maturity G-secs,” said N S Venkatesh, a bond market expert and former bank treasurer.

“States that don’t have a good fiscal discipline won’t be interested in getting rated, but then the investors won’t buy their bonds too. So, slowly, the market forces will make them go for a rating,” said Venkatesh.

But not everyone shares this view.

At present, all states command the same kind of yields in auctions as these bonds are guaranteed by the government and are highly secured instruments.

Now, if the states are getting rated, the spread between themselves will rise. And therefore, the spread between the G-secs and SDLs would vary widely. Poorly run states would likely have to pay a heavy coupon for their loans, say experts. Still, their rates would likely be lower or on a par with well-rated non-banking companies, as state loans are eligible for calculation of Statutory Liquidity Ratio (SLR), or the share of deposits that banks will have to invest in bonds.

“A state bond is an SLR security and has a risk-free weightage. Why would it get a rating anything below AAA? The difference in spread between two SDLs is illiquidity premium and not credit-risk premium. This rating business is a non-starter,” said the head of treasury at a private sector bank.

According to bankers, SLRs constitute about 20 to 30 per cent in a bank’s bond portfolio. Most of these though are earmarked for held-to-maturity basket as secondary market trading is next to nothing for SDLs. However, if ratings start, a secondary market may develop as foreign portfolio investors (FPIs) would want to invest in these bonds.

What's next?


  • Banks may push states to get their bonds rated
  • Ratings may mean poorly managed states would pay higher coupons
  • Best-rated states may get to borrow at near G-sec levels
  • Market-based valuation will wipe out valuation gain for banks
  • Without valuation gain, SDLs may become unattractive
  • Lack of demand for SDLs would be good for G-secs