Taxation and Regulation of the Financial Sector

Since the onset of the financial crisis, governments have been searching for policies to offset the costs of bailing out the financial sector, and to put in place structures to reduce the probability and severity of a future crisis. Several proposals have been considered, and enacted, for new taxes on banks and other financial institutions.

This note, based on a chapter of the 2011 EEAG Report on the European Economy presented on Feb. 22 in Brussels, considers some aspects of these proposals in the light of existing and new regulations of the financial sector. There are two alternative, though perhaps complementary, objectives for new taxes on the financial sector: i) raising revenue, and ii) inducing behavioural change that will make another crisis less likely.

According to the European Economic Advisory Group at CESifo (EEAG), a group of seven scholars from seven European countries who release their report on a yearly basis, the first objective could be backward-looking, to reimburse the public for the costs of support during in the financial crisis. But that raises several questions, not least what those costs were. Should costs include only specific support through bailouts, or the net effects on public finances after taking into account the wider economic effects of the crisis? It is unreasonable to expect the huge public deficits that have arisen in the wake of the crisis be reimbursed by the financial sector. And in any case, it is not easy to identify who were the beneficiaries of public support, and so who might be expected to reimburse the costs: after all, the point of the support was to contain social costs, and so everyone benefited, not just bank shareholders, creditors and employees.

A forward-looking perspective would be to raise revenue that could be used to build a resolution fund for the next crisis. Here there are again two possible approaches. One option would be the Financial Activities Tax (FAT) recently proposed by the IMF. At one extreme, this would have a narrow base, including only economic rents and the remuneration of very highly paid employees. To the extent that this remuneration could be capturing part of the economic rent, then such a tax would in principle be non-distorting. However, depending on revenue requirements, such a tax may require a high rate. If the tax were introduced cooperatively at a high rate in all countries, this may not be a problem. But the incentives for a single country to charge a high rate are weak, since it could induce banks to shift business elsewhere.

At the other extreme, all remuneration could be included in the FAT base. This would be similar to a tax on the value added, since value added is equal to total remuneration plus economic rent. Since this is a larger base, the rate could be lower. Such a tax would have different properties from the narrower version of the FAT. This version could be seen as a substitute for the absence of VAT in the financial sector, although the form of the tax would differ from VAT, and a number of technical details remain to be resolved. Either form of the tax could be introduced alongside existing corporation taxes.

A second approach based on a revenue-raising requirement is to levy a tax as an insurance premium, where the tax reflects the risk that an individual company will require support from the resolution fund and the amount of support that it would then require. Such a tax would be complex, however, since it would need to be based on factors that reflect the financial fragility of the company – its size and the degree to which it is systemically connected to the rest of the sector. The relevant factors are difficult to measure; moreover, they are exactly the factors that regulators should be considering in attempting to set rules to reduce the probability of future crises. This raises an important issue about the relative merits of taxation and regulation in affecting behaviour.

By its nature, such a tax could certainly affect the behaviour of banks and other financial companies, possibly inducing them to reduce leverage and the risk of their investments. However, these effects would depend on the form of the tax. Given that a system of regulation is already in place and is unlikely to be abandoned in favour of tax, then the effects of tax would depend on how it differed from regulation.

The effects of an insurance-premium type tax tie in closely with the second possible objective of a new tax in the financial sector: to make a future crisis less likely, by inducing banks and other financial companies to reduce leverage or to invest in less risky assets. One option for this objective is the Financial Securities Contribution (FSC) also proposed by the IMF. Basically this is a tax on the bank’s balance sheet that exempts the equity capital and insured liabilities but includes off-balance sheet operations. Several countries have either introduced, or announced that they plan to introduce, such a tax. While this tax is partly designed to raise revenue, it is also clearly intended to reduce leverage.

In principle, this tax could be a meaningful addition to the existing and proposed Basel regulations. The main focus of these regulations, at least as they reflect the aim of solvency, is the Tier 1 capital regulation, which requires banks to meet a minimum ratio of equity capital relative to a measure of risk-weighted assets. In measuring risk-weighted assets, loans to companies frequently have a weight of 0.5, loans to banks have a weight of 0.2, and loans to governments are not included in the definition of risk-weighted assets. So the majority of assets are weighted less than their total value. And government bonds in particular are effectively treated as risk-free. The sovereign debt crisis indicates that this is an implausible, and unjustified, position. Improvements in the calculation of the sum of risk-weighted assets seem advisable, but this is not the focus of this note.

The Basel III system will also include a minimum capital ratio, expressed as equity capital relative to total assets. This avoids the deficiencies of the risk-weighting system, though at the expense of not considering the risk of the bank’s assets at all. The FSC is similar in that the tax is based on the size of total liabilities without any regard to the risk of assets. If the FSC tax rate is high enough, it may therefore induce banks to reduce its liabilities relative to total assets, or equivalently to increase its equity capital relative to total assets – a similar effect to the minimum capital ratio.

Whether either of these measures would reduce the probability of a bank defaulting depends on exactly how it responds. One option would be to shrink its balance sheet through a reduction of its holdings of government bonds, whose risk is not measured by the Basel regulations, while maintaining its equity capital. This would increase the ratio of equity capital to all assets, while maintaining the same ratio relative to risk-weighted assets. Such a strategy would raise the average risk of its remaining assets, but this would be more than offset by the higher capital ratio, implying that the probability of default would fall. However, it is also possible that a bank could follow a different strategy that could increase the probability of default.

In any case, the FSC, like the minimum capital requirement, could be seen as an attempt to correct for the deficiencies of the Basel risk-weighting system. But if the problem lies with the regulatory system, a more straightforward way of dealing with it would be to reform the regulatory system, rather than to introduce a compensating tax. The tax system could then be used for a more obvious purpose, raising revenue in as non-distorting a way as possible.

• Download Chapter 5, "Taxation and Regulation of the Financial Sector", of the EEAG Report on the European Economy 2011 (pdf, 547 KB)

• CESifo Working Papers published in the past three months