by Kristin Gulbrandsen
I am very glad to see all of you here. The Governor welcomed you this morning by stating how important the topic of this conference is to our work at Norges Bank Financial Stability, and how central it is to the reform of financial regulation. I shall take this opportunity to make a few comments about one specific part of the reform process, namely the question of whether there should be a special tax regime for banks.
There is currently a lively international discussion on the taxation of banks. The backdrop is the huge costs that many governments have incurred by providing assistance to their banking industries. It is now commonplace to state that the private sector debt problem is being transformed into a public sector debt problem. At the same time, many of the banks that received government support have quickly returned to profitability and have been able to pay dividends and bonuses to their shareholders and managers. It is widely seen as unreasonable and unfair that government support should be used in this way. In many countries the governments are seeking to recover the support costs by introducing new taxes on the financial sector. I should note that this argument is less relevant in Norway, because the government costs of the financial crisis have been quite small here.
The proposed taxes can be divided into four categories. There are Tobin taxes on financial transactions, bonus taxes, supplementary taxes on profits, and taxes on the liabilities of financial institutions. The reason given for most of these proposals is the need to recover government support. Proposed tax rates are calibrated with that in mind. Some of the taxes are explicitly meant to be temporary. Even for more permanent taxes there is very little discussion of the longer term consequences.
One long term motivation for some of the taxes is to build resolution funds that can be used for resolving future financial crises. The idea is that the financial industry should pay the expected resolution costs up front, to protect government finances. The critics respond that the very existence of a fund will strengthen market expectations that financial institutions will be rescued by the government, and thus amplify the moral hazard problem. My view would be that this discussion is a red herring: There is hardly any evidence that the existence of a resolution fund has any importance on the cost of banks’ market funding.
The relevant question we need to pose is: what will the long term incentives of a new tax be? Can we hope to mitigate the moral hazard problem in the financial sector by some form of taxation? In other words: is there a Pigovian case for a new tax?
The banking sector provides very important social functions; it runs the payment system, it offers safe vehicles for financial saving and it offers credit. Our macroprudential concern is that these functions shall not be interrupted. For this reason governments are imposing deposit insurance arrangements, and consequently they are implicitly guaranteeing that the systemically important activities of the banks will be maintained. These guarantees create a well-known moral hazard problem, which is the topic for this conference.
The banking industry does not get these guarantees for free. They pay fees to the deposit insurance system. They are subject to more regulatory interferences than most other industries: Bank supervision is more intrusive than the supervision of any other industry. There is extensive special regulation for the banking industry. There are for instance very specific capital requirements, which will be made substantially tougher in the near future. It is still the case that the banking industry holds very little equity as compared to other industries. This is part of the moral hazard problem because the downside of risky activities is limited for the shareholders. The top management of banks has a similar asymmetrical remuneration function, because bonuses can rarely be reclaimed.
The banking industry can in normal times easily fund risky activities through market borrowing. The bank creditors have high expectations that they will be protected if a bank becomes distressed. This is reflected in the two sets of credit ratings offered by the rating agencies; the standalone rating of a financial institution is normally several notches below the rating that takes expected government support into account. This translates into lower costs for market funding than the bank would have had without the implicit government guarantee. These lower costs constitute an implicit subsidy, and are the most directly identifiable effect of the guarantee.
Andrew Haldane of the Bank of England has recently estimated that the implicit subsidy to large British banks is of the same order of magnitude as their net profits. Our own estimates for large Norwegian banks give a similar result. The subsidy is large enough to be economically important. It can be reduced if better resolution regimes are installed: In a press release last year Moody’s warned that many large banks would be downgraded if the “living will” idea significantly reduced the likelihood of government rescue. But it seems unlikely that the subsidy can be completely eliminated. The subsidy means that the banking industry will tend to be larger than is socially optimal: There is a clear case for reclaiming the subsidy through taxation.
A recent paper by Shin and Shin arrives at the same result, but from a different starting point. They find that the share of market funding on a bank’s balance sheet is a good indicator for the probability that it will fall into distress. I believe this result makes sense: It is in line with our general understanding of banking crises as the result of the overextension of credit. Rapid increases in credit cannot be funded by increased deposits, and must instead be based on borrowed money. Empirically we can see that the use of market funding increases during credit booms. Shin and Shin therefore propose a tax on market funding as a countercyclical measure: A tax will make the funding of rapid credit growth more expensive and hopefully limit the size of future credit booms.
This is a tax proposal that looks appealing to me; it can correct a market imperfection and it can work as a stabilizing instrument. In the Norwegian context, I would in particular consider a tax on market funding as relevant for calibrating the fee to our deposit insurance system. Member banks now pay a fee that is proportional to their guaranteed deposits, with some modest graduation depending on their capital adequacy positions. Norges Bank recommended two years ago that the risk graduation should be increased, and that more reliable risk indicators be introduced. We proposed at the time to use loan growth during the previous three years as an indicator. The research I just referred to and our experiences from the financial crisis tell us that using the extent of market funding as a risk indicator could be a very good alternative. It would perhaps even be easier to implement in practice.
Sweden has already decided to impose a similar tax. Their tax base is the total liabilities of banks, with deductions for guaranteed deposits and equity. The remaining liabilities are close to representing market funding. The Swedes will coordinate this liability tax with their deposit insurance fees. The governments of Germany and France have also proposed similar taxes.
What I have been presenting to you now is just one example of how the research on the recent crisis is useful for our thinking about regulatory reform. We are in fact very dependent on input from the research community to improve our understanding of how the financial system works. I would therefore encourage you to continue extracting lessons from the experiences we have just been through, and to share these lessons with us.
Good luck with the rest of the conference.