The economics of bank recapitalisation bonds
The recent issue of recapitalisation bonds by the government is a step in the right direction. Recapitalisation is a tried and tested tactic and has been successfully replicated in many countries, including India, in the past.
With regard to the critiques of bank recapitalisation, there is indeed justification for the use of public funds. This is because the benefits of recapitalizing banks are greater than the costs of broad disruption in the real economy which show up, for instance, in the decline in bank lending. However the benefits from such an exercise are often difficult to quantify as they largely relate to avoiding disruptive effects that are qualitative in nature.
The Rs1.35 trillion package in itself seems largely adequate going by the ministry of finance estimates. The latest figures shown in the ministry presentation put the increase in non-performing assets (NPAs) from financial year (FY) 2015 till June 2017 at Rs4.55 trillion. The increased provisioning in the time period 2014-15 to 2017-18 is pegged at Rs3.79 trillion. So the banking sector recapitalisation to the tune of Rs2.11 trillion seems sufficient for tackling the problem of stressed assets. Additionally, after provision for bad assets and on a conservative basis, even if 50% remains as a residual for growth capital of banks, the multiplier impact may be significant.
India had used this tool before in the FY1986 to FY2001 period, wherein the government recapitalized public sector banks (PSBs) with a total amount of Rs20,446 crore. It borrowed the amount from the banks and issued special non-marketable securities, which were, however, later converted into marketable securities or perpetual bonds. The banks subscribed to these bonds and there was no cash outgo from the budget during the year of recapitalisation. However, it was an addition to the public debt. Consequently, the real impact on the budget was only when interest was paid by the government to the banks on the securities held by the latter. In total, interest paid by the government to the banks on special securities worked out to be Rs7,888 crore or 0.07% of the gross domestic product (GDP) per annum on average. During the period, however, the banks also paid Rs15,222 crore as dividend to the government working out to 0.04% of GDP on average. So, the net impact on fiscal was only 0.03% of GDP.
If the structure of the 1990s was adopted, it would not alter any fiscal maths in the current scenario. An interest cost of Rs8,000 crore per year is only 0.07% of GDP and 1.6% of the total interest payment on revenue expenditure of the government. The impact on public debt is minimal at only 0.8% of GDP.
Internationally, governments have recapitalized banks through various means. These can be mainly divided into direct capital infusion, issuance of public debt into banks either as a swap for bad assets or unrequited, and by assuming the bank’s liabilities.
In fact, such capital infusion has been done directly in many cases like the Maiden Lane LLCs created in the US to bail out AIG (American International Group) in 2008. In many cases, like Argentina (1994-95), Finland (1991), Hungary (1994), etc., long-term loans were extended to banks. Often, such bonds were given as a swap for bad loans, though that is not the case in the recapitalisation programme announced in India. Unrequited bond issues, similar to what seems to have been announced in India, have taken place in many cases like Poland (1993-94), Hungary (1993-94), Latvia (1994) and Ghana (1990). Internationally, such bond issuances have seen the government keep principal payments out of the budget, while interest payments were included in the expenditures (though Chile in 1984 did not record even the interest payments).
There is generally little to gain by issuing such bonds through an agency rather than the government itself, except in the case where the agency may have the necessary infrastructure, thereby separating bank restructuring costs from other government activities. The government may thus look into the P.J. Nayak committee recommendations, according to which the government may set up a Bank Investment Company (BIC).
What about the maturity and interest rate structure? In most such cases, recapitalisation bonds have had long maturities as early repayment would lead to a burden on government finances. Short maturities are also undesirable as a large amount of debt, if rolled over, could create uncertainty in markets. A range of maturities may also be desirable to establish a yield curve.
Keeping a market-related rate makes sense, but whether such an option could be the ideal one in the current circumstances in the Indian context is highly debatable. The argument for a higher market yield is that in the current situation when the banks are underperforming, a lower than market rate may not help the bank in generating a healthy interest margin that is already under pressure given that interest rates are on a declining path. This could thus defeat the very purpose of recapitalisation.
The counterargument could be that if such bonds are covered under held-to-maturity (HTM), these would not be sensitive to interest rate risk and hence will not attract any capital charge. These could be then a clear attractive option for banks. Keeping such bonds under HTM would also solve the typical problems of such dated maturities like mismatch, loss exposure and higher risk premium. In hindsight, we believe that the duration of such bonds should not be too long. This will avoid the likely lack of matching long-term liabilities if any and also to mitigate any market volatility in valuation of long-dated bonds.
The moot question is what will be the impact on yields? The initial reaction of the market has been a bit negative, possibly reflecting the uncertainty. The good thing is that AT1 (additional tier-1 bonds) rates have declined by as much as 150 basis points on expectations of a stronger bank post recapitalisation. Further, as the bonds are cash-neutral and can be categorized under HTM over and above the permitted cap, the current liquidity conditions being benign, the demand for new investments would be sustained. Both these factors will have a sobering impact on bond yields as well.
Ultimately, this recapitalisation will lead to an improvement in the government’s finances as it would also be able to sell its stake in public sector banks at much higher valuations (market capitalization went up Rs1.2 trillion in a single day). Even on the demand side, some banks who were not investing their extra cash into debt securities due to capital shortage may now be able to do so instead of placing them with the Reserve Bank of India’s reverse repos.
C. Venkat Nageswar and Soumya Kanti Ghosh are, respectively, deputy managing director (global markets) and group chief economic adviser, State Bank of India. These are their personal views.
Comments are welcome at firstname.lastname@example.org