Search interesting materials

Tuesday, October 09, 2012

Should government capitalise public sector banks?

by Harsh Vardhan.

What would you say if someone was borrowing money at 8% and investing it to earn around 3%? "Uninformed!", "financially illiterate!" or even outright "foolish"! And yet this is what our government has been doing with trillions of rupees over the last many years and has committed to continue to do so in the future. The process by which this is done is called capitalisation of public sector (PS) banks. Such capitalization is not only a bad idea economically as it puts enormous stress on the government resources, but also one which affects that behavior of banks and hence the robustness of the whole banking sector.

Commercial banks need capital to grow. Capital adequacy requirements ask all banks to keep a minimum amount of shareholder capital in proportion to their balance sheet size. Currently, in India this requirement is 9% of "risk weighted assets" of banks. Roughly, it means that banks are expected to have equity capital which is 9% of their commercial loans.

As banks grow their business, their risk weighted assets also grow. This means that banks has to increase their capital base in line with the growth of their loan book. Such increase in capital can come from exactly two sources - retained profits that are added to the capital base, or fresh infusion of capital from shareholders (old or new). In India, given the overall profitability of the banks (~1.1% return on assets) and the amount of dividend that they pay (~20%) of post-tax profits, banks do not have enough retained profits to support their business growth. Therefore, every now and then, they go to shareholders to raise fresh capital.

PS banks pose a peculiar challenge for the government. Being the majority owner of these banks and having committed to stay the majority owner, government has to infuse capital into these banks proportional to its ownership stake. Since the government wants to maintain its ownership at 51%, it has to supply atleast 51% of the fresh capital that PS banks need. RBI governor Dr. Subbarao, in a recent speech, said that the capital infusion by government into PS banks over the next decade will be of the order of Rs.0.9 trillion. I have read estimates of other analysts where this number is as high as Rs.2.50 trillion. These estimates depend on the assumptions one makes about a number of factors - the rate of growth of banks (which in turn depends on the growth of the overall economy), the profitability of banks, their dividend policy, their ability to raise other forms of capital (especially tier II capital), regulatory requirements on capital, etc. No matter how you estimate it, the number is very large. In other words, the government will be compelled to invest a very large amount of capital into PS banks over coming years.

Why is this a problem? Let’s look at the some simple public finance issues. India is in a deep fiscal crisis, and it is not easy to find trillions of rupees to put into PS banks. If such resources were injected into PS banks, it is not conducive to healthy public finance, since these injections are not a good deal for the government. The Indian government currently borrows long term money at over 8%. The dividend yield on PS banks shares has been between 2% and 3% over the last decade. This means that the government earns between 2% and 3% on its investments in PS banks. There is a 5% “negative carry” or loss that government bears on these investments.

A private investor also earns a low dividend yield from investing in PS banks, but can benefit from capital gains - a potential increase in the value of shares which the investor can obtain when she sells the shares. Government has never sold shares of PS banks (except when it initially listed some banks) and will not do so if it has to maintain majority ownership which is its stated policy. Hence, for the government, the financial analysis of a proposal to put money into PS banks should hinge on a comparison between the flow of dividends versus the cost of borrowing.

Capitalisation of PS banks is, thus, bad for government finances. It's a double whammy! On the one had government has to raise vast resources to be invested into banks and then carries a loss of around ~5% on these investments year after year.

Ownership and behavior of banks

Government capitalisation of PS banks is not just a fiscal challenge. It also impacts the competitive dynamics of the banking industry. Most privately owned banks are under constant scrutiny of investors and analysts. When they go to external investors for raising capital, they have to satisfy these investors on number of critical aspects of the business - profitability and its sustainability, efficiency of capital use, quality of management team, cost efficiency, etc. In other words, private banks face a market test; they do not get capital for free. Only well run private banks get equity capital that is required for growth.

None of these questions get asked when government puts capital into a PS bank. One has never heard a senior government official commenting on the Return on Asset (RoA) or Return on Equity( RoE) of PS banks. The decision to put capital into PS banks is treated as a mechanical and administrative decision. This absence of a market test has systemic consequences. PS banks have ~70% share of the Indian market. When the majority owner is asking no or very few questions on performance, and is assuring an almost unlimited supply of capital, these banks have little incentive to improve financial metrics such RoA and RoE. This hurts the overall banking industry. For example, PS banks can underprice loans compared to their private sector peers. Such behavior would migrate the whole business to lower returns. It is hard for a private bank to be profitable when facing rivals that are not concerned about return on capital.

Misplaced obsession with majority ownership

The source of this whole capitalisation issue is the government's obsession with retaining majority (over 51%) ownership of PS banks. This is often explained in terms of the need to maintain the "public sector character" of these banks. While there may be a separate debate on whether we need to maintain public sector character for all the 25 plus PS banks, the fact is that the government does not need majority ownership to achieve this objective.

  All PS banks are not companies under the Companies Act. The notion of 51% giving majority control is enshrined in the Companies Act. PS banks were created under the Nationalisation Act (SBI has its own SBI Act). The Nationalisation Act provides the government untrammelled control over these bank. While it does prescribe 51% government ownership in the PS banks, the control of government is independent of the level of its ownership. Furthermore, there is a limit of a 5% (10% with prior approval of the RBI) stake owned by any single shareholder in all banks. There is no chance, therefore, of any external shareholder acquiring control in these banks. Even relatively minor changes to the functioning of PS banks require approval of the parliament. Where is then the question of diluting the public sector character if the government ownership were to drop to, let's say 26%, which is the threshold for "significant" minority stake in a company?

In the long run, therefore, it makes no sense for the government to commit itself to the capitalisation of PS banks. Precious government resources can be better deployed in critical areas (such as power transmission and distribution) where private capital on large scale is hard to come by. In the medium term, it can use tactical measures such as merging banks where it has significantly high ownership with those where the ownership is already down to 51%. But these tactics will not solve the issue structurally. The only long term solution is to give up the majority obsession, explain to all the stakeholders the fallacy of this obsession and the resulting pressure on public finance, build a political consensus to enact necessary legislative changes and then dilute down to a reasonable level.


  1. RoE is not dividend yield.

    1.1% RoA * capital ratio of 9% = 100/9*1.1% = 12% return, at the very least. (All assets don't have a risk weighting of 100%, so actual leverage is higher).

    Your entire case is built on a trivial oversight.

    1. Ritwik - there is a point to what Harsh is saying. If you own a stock that you do not forsee selling at all then you cash returns are just the dividend yields, even though your MTM returns will be higher. But what's the point of money that is yours but not allowed to touch. Maybe Govt can borrow against those shares.

      However I also agree that to say there is a negative carry is stretching the argument a bit farther (but not too far)

    2. The accounting profit of the government is converted to cash not through sale, but through monetization.

      A government is never cash constrained in its own currency.

  2. Excellent points to which I will add one more. PS banks are essentially asked to undertake off-balance-sheet development functions for the government. As you demonstrate, the cost of those functions is quite difficult to measure. Very bad fiscal practice.

    Should you include capital gains in the calculation of govt ROI? Probably, since eventually - though maybe not soon - the govt will have to bow to reality and sell of some portion.

    I have heard that the real reason the govt is so loathe to drop below 51% shareholdings is that below that point employees lose their "public sector" status and benefits. Do you know if it's true? If so, it would certainly help explain what seems like an otherwise indefensible policy. If so, we should discuss it, so the public can openly debate whether the recapitalization and negative carry cost of PS bank ownership is worth protecting the benefits of however many PS bank employees.

  3. Whether or not there should be a public stake in banks should be based on arguments other than profitability. Profitability is irrelevant to the question.

    Why does any govt need to have a stake in banks? The main argument I've heard in the Indian context has been that it was not profitable for banks to expand in rural areas, and such unprofitable ventures were best supported by the public balance sheet. Can we have a debate on this? Can the govt not achieve the same goal by incentives for private banks instead?

  4. The RoE for PSU banks ranges from 20%+ ( Dena Bank, Bank of Baroda, PNB) on the higher side to 11% (Bank of Maharashtra, Indian Overseas Bank, Oriental Bank) with the only outlier Central Bank having a meagre 5% RoE.

    Thus the notion of taking dividend yield as RoE is wrong since the Govt can monitize the bank stakes through a PSU bank ETF and offering units to retail Indians as was proposed sometimes back. Thus there are ways to unlock value here.

  5. Dividend yield is not the whole and sole measure of returns. PSU banks market cap is up 10 times in 10 years, so money invested 10 years ago is yielding 20-30% a year and might yield even more in future. Arguments need to be more robust. When discussing RBI or Fed's policy, P&L on balance sheet is not an important argument e.g., economic stability is. Severe conflict of interest and faulty allocation of resources because of Government ownership due to reasons like Social justice goals and corruption are reasons enough for government to move out of banking.

  6. Is IDBI Bank a Public Sector Bank or Private one? Is it desirable to have more IDBI like shareholding structure?

    1. Also Axis Bank... SUUTI, LIC and other insurance companies hold 37% stake, but there is no government intervention in the management. RBI also classifies Axis bank as a private one.

  7. Your arguement has merit if the government were a financial investor. I suppose nationalisation of banks was not done with an intent of ROEs. But if we were to calculate the cumulative value of financial delivery to the rural sectors (which the private banks can/will never achieve), I believe the 'negative' return as you put it, is worth the investment. After all, except for a brief period in the 90's, PS banks have never stopped making money. No bank has become unviable till date.

    Your point on governance is also not valid as all PS banks are today listed. Agreed, there is a lot to be done on governance improvement (to split Chairman and MD to start with). Pity that the Institutional minority shareholders do not care to highlight these issues. Also issues like bank consolidation and efficiency leave a lot to be desired.

    Having said this, I agree that GoI does not need 51% stake in each bank to exercise control. The concept of a Bank Holding Company has been mooted in the past, which would reduce GoI's financial burden, while allowing it to maintain the desired ownership.In my opinion, banks should be the last sector to be privatised given its importance in an socio-capitalist economy like ours.

  8. For everyone's benefit... I am not confusing RoE and dividend yield. The point I am making is that RoE, especially accounting RoE, is irrelevant. What the government earns is dividend independent of what the RoE is. The point of capital appreciation is factually true but is again not relevant for the government which has not sold a single share and hence has not realised any gains. Its unlikely to do so if it wants to hold on the majority. The key point is all in the incremental capital infusion will happen at the dividend yield of 3% when the government's borrowing cost is 8%.

    1. Imagine a government that does absolutely nothing with its borrowing. Public debt only backs the money supply and pays interest to risk free holders. Bank deposits pay, say 4% and government debt backing it 5%. Is this government commercially unsustainable, with a cash flow of -5%?

      Of course not.

      There is the little matter of currency sovereignty and seigniorage.

      A government that makes an accounting profit of 4% on its investment is outdoing every other shareholder because nobody else's borrowing costs are a low as 8%. The incremental capital infusion is forever sustainable, even without the sale of a single share. A government with currency autonomy is never cash flow constrained in its own currency.

      Points about govt control and public choice issues with any public sector enterprise notwithstanding, the financial case you make is weak and not macroeconomically informed.

  9. The conclusion in the article makes sense but it is making unnecessary arguments. Banks were nationalized to ensure that credit is made available to all the deserving sections of the economy during the times of scarity of credit resources..etc etc... Over the period of time, this objective was broadly achieved but other issues like profitability, politicization, "directed" investments etc became more relevant. Now conditions have changed vastly and we talk of return to shareholders, financial inclusion and raising equity on the stock market. What we need to work on is 1-elimination of government holdings in PSU but Government to keep one bank as a fully owned government bank for whatever government want to do with it in terms of financial inclusion, access to risk free deposit etc. Government can make it as big as it wants to be and Parliament will have control over it ultimately. It could even be converted from Postal Office Saving Bank structure into a full fledged commercial bank. Thus, government plays a developmental role as ratified by the Parliament. 2- Allowing all other banks free play in terms of access to capital market and some of them may grow bigger, some of them may die a natural death etc and having strong prudential regulatory control over them through RBI or a unit specifically for Banking control and supervision 3- Support development of specific sectors like SMEs or export oriented units through specific schemes for which cost of such promotion can be specifically budgeted.

    Once we agree to this architecure, most of the arguments though sensible, become irrelevant. Idea of converting PSU bank shares into a ETF structure is interesting and should be seriously looked at. Keeping public sector charcter of banks merely to help employees enjoy the benefits could be one of the considerations and if indeed so, employees should be given choice to move over to the government owned bank.

    1. Why do you need 1? Financial inclusion can be achieved through private banks, no? Risk free deposits can be achieved by deposit insurance (which is at 1 lakh now and needs to be raised), no? What is it that the govt needs to do, that it cannot do with private banks?

    2. The idea is very simple. Government does what it wants to do transparently and allows free play to the private sector banks with a much simpler regulatory architecure. Depsoit insurance has a cost in terms of moral hazard and monitoring and hence, suggested framework avoids it. Private banks too can participate and act as banks which will accept "risk free deposit" on behalf of government and hand over the proceed to the government. They can also voluntarily accept as a part of their corporate social repsonsibility objectives like financial inclusion. We need a separate "Payment and Settlement Corporation of India" which will act as central clearing institution which will charge for services it renders. Once we have this architecure in place, banking regulation for capital adequacy purposes will become simpler and much less costly.


Please note: Comments are moderated. Only civilised conversation is permitted on this blog. Criticism is perfectly okay; uncivilised language is not. We delete any comment which is spam, has personal attacks against anyone, or uses foul language. We delete any comment which does not contribute to the intellectual discussion about the blog article in question.

LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.