How bond gains can recapitalize banks
Indian banks are sitting on a gold mine. They hold ridiculous amounts of government bonds, and three years of falling interest rates have increased the value of these bonds by about 10%. However, most of the bonds are held under the statutory liquidity ratio (SLR) rules, so they are mostly held at book value; their rise in market value does not show on the banks’ books. SLR reform holds out the tantalizing prospect of a bonanza that could help finance recapitalization.
Rates have fallen about 2.5 percentage points since the beginning of 2014. Bond prices rise when rates fall, so bondholders have done well. Banks hold about half of all Central and state government bonds, not counting Treasury Bills, according to Reserve Bank of India (RBI) data. Most of that is held under the SLR scheme. The RBI allows banks to hold SLR assets at book value, meaning they are valued at the original price paid, not the price at which they could be sold in the moment.
In the current environment, banks could be sitting on as much as a 10% unrecognized gain in their SLR portfolios. This would vary from bank by bank, depending on their holdings. To a small extent they can opportunistically mark bonds to market when their value rises, so some measure of the 10% gain may already be captured.
Because banks hold such a large share of all bonds, one can approximate the portfolio of the entire banking sector by looking at all outstanding bonds. Banks buy the bulk of their holdings at issuance. The difference between face value and current market value therefore represents the loss or gain to the bank. The Clearing Corporation of India Ltd (CCIL) shows that difference to be about 10% of the face value for Central government bonds. Gains on state development bonds are probably only slightly less, though they hardly trade.
If my assumptions are correct, banks are sitting on as much as Rs2.3 trillion in unrecognized capital gains. The SLR ratio has come down gradually from 24% in 2012 to 20.5% of net demand and time liabilities (NDTL), but the phased reduction plan has reached its end. Further aggressive reductions in the SLR could give banks a potentially large opportunity to rebuild their finances.
The European Central Bank—unintentionally—took this route to help European banks using a different set of mechanisms. It first financed major bank purchases of bonds via the long-term refinancing operation (LTRO). It subsequently dropped interest rates, raising the value of those bonds. Most recently it has begun a massive quantitative easing (QE) programme, buying the bonds back from banks so they realize the gains.
The new Economic Survey makes a strong case for rescuing the ailing public sector banks (PSBs). It advocates a bad bank structure and reform of PSB lending practices. Recapitalization is an essential component of a broader set of much needed action.
The funding gap is massive. Credit rating agency Fitch determined they would need Rs6 trillion from 2017-19. Credit rating agency Icra estimated the banks may be able to raise another Rs0.95 trillion from the capital markets. The new budget does not change the government’s current plans to allocate Rs0.7 trillion to the task. This leaves PSBs short by at least Rs4 trillion, or about 0.8% of GDP for each of the next three years. Even if it had the will, the government cannot stretch to bridge that gap without blowing out its budget targets.
Of course, the SLR itself has a strong connection with the budget. Despite its name, the SLR has almost no prudential rationale. It represents financial repression, ensuring a captive market to lend to the government.
Marking the banks’ SLR holdings to market—even without allowing banks to sell the holdings—could put a serious dent in demand for government bonds. If future bond sales become much tougher, the interest cost to the government will rise. Further, if bond prices fall, that will hit the bank balance sheets, defeating the very point of the measure.
Reassuringly, banks hold well more than required. The average SLR holdings over the past year amounted to 6.7% of NDTL more than required, and demonetization almost doubled that at year-end. All of the excess is marked to market, so it is not self-evident that demand for bonds would collapse with SLR reform. Former RBI governor Raghuram Rajan made the point that cutting SLR at a time when credit demand is low would help prevent a flight of money from government bonds.
As part of a broader package, a clean-up of the balance sheets and revitalization of the PSBs could facilitate greater credit demand from industry. In that case, without banks’ captive demand for government bonds, interest costs might rise. Banks may face some capital losses too. But that only happens in the happy scenario of a return to healthy financial intermediation.
Further, the SLR’s financial repression imposes a major distortion on India’s yield curve. SLR reduction improves government and corporate bond market liquidity. Along with other measures to broaden participation in the bond market, it could prove a strongly positive move for bonds.
SLR reform also has the benefit of being a somewhat market-based recapitalization scheme. The almost Rs1 trillion the government has already spent over the past decade recapitalizing banks has perversely gone to the worst-performing PSBs. As the Economic Survey points out, incentives must change to encourage better performance. The promise of a Banks Board Bureau has not come to fruition.
This new borrowing dynamic would need to be managed carefully, but it could even serve as a credible commitment device for the government’s downward path of fiscal deficits. If SLR reform helps resolve the banking crisis quickly and advances financial market development, these complications are the kind India can afford.
Russell A. Green is the Will Clayton fellow for international economics at Rice University’s Baker Institute for Public Policy, and a former US treasury attaché to India.