Parliamentary Commission for Banking StandardsSupplementary written evidence from HM Treasury

Thank you for your letter of 30 January following up the issues we discussed at the Commission’s Panel on tax, audit & accounting on 24 January. You requested further information on the following aspects of an Allowance for Corporate Equity (ACE):

the cost of introducing an ACE;

the treatment of equity supporting the foreign operations of UK banks and the UK operations of foreign banks; and

options for offsetting the costs of an ACE.

Cost of Introducing an ACE

An ACE has an inherent cost in terms of CT revenues associated with allowing an additional deduction for equity. The average annual cost to the Exchequer of an ACE is estimated to be as follows:

£1 billion if available to all banks.

£0.8 billion if available only to banks subject to the UK Bank levy.

It should be noted that these estimates are based on latest projected estimates of shareholder equity, and make no allowance for any behavioural change that may be triggered by the introduction of an allowance. The estimates also do not take into account the impact of losses being carried forward by the sector- all other things being equal, this would imply any existing losses would be exhausted over a longer time period, though this may be influenced by other factors, such as overall profitability. The full set of assumptions underlying these calculations is set out in the annex.

Estimates of the cost of an ACE that only applied to “new equity” are particularly sensitive to assumptions about the behavioural impact. It is very likely that creating a distortion between new and old equity would drive banks to pay off existing equity and replace it with new capital. So while we estimate that, in the absence of any behavioural response, the cost of an ACE limited to “new equity” would be around £100 million p.a. (£80 million p.a. if limited to Bank Levy paying banks), it is very likely that the actual annual costs would quickly approach the £1 billion figure set out above for an ACE for all banks’ equity.

As discussed in previous evidence an ACE limited to the banking sector is likely create some definitional issues at the perimeter which would need to be resolved. Moreover, to defend against legal challenge, we would need to be able to clearly demonstrate that there were financial stability benefits that justified such preferential treatment for the banking sector. Even if this could be successfully managed, it is still likely that an ACE limited to banks would trigger other businesses—both those who operate in the same markets as banks and those in other unrelated sectors- to call for similar treatment. Inevitably, introducing the ACE more widely would substantially increase the cost.

Treatment of Equity Supporting the Foreign Operations of UK Banks and the UK Operations of Foreign Banks

The application of an ACE to the foreign operations of a UK bank would be a policy decision that would need to be legislated for in a Finance Act, but will to some extent depend on the way in which the supporting equity is raised:

Where the UK bank operates in a foreign jurisdiction through a branch (referred to in tax treaties and legislation as a “permanent establishment”), this is a part of the UK bank rather than a separate legal entity, and so would not hold capital in its own right. The ACE could either be applied to any equity distributions in the UK bank’s solo accounts, or alternatively only to capital allocated to UK operations- existing corporation tax rules for banks operating through branches already include provision for allocating capital between them.

Where the UK bank operates in a foreign jurisdiction through a subsidiary, and the equity is raised directly by the UK parent bank, this equity will be shown in the bank’s solo accounts, and the ACE could be applied to any equity distributions in the UK bank’s solo accounts.

Where the UK bank operates in a foreign jurisdiction through a subsidiary, and the equity is raised by the foreign subsidiary itself, this would be reflected in the consolidated accounts of the banking entity as a whole, but as we currently tax UK companies on the basis of their solo accounts, following this logic would mean any distributions arising from this would fall outside the remit of the ACE.

Similarly the treatment of the UK operations of a foreign bank might depend whether those operations are conducted via a permanent establishment or a subsidiary company of the foreign bank. Where the foreign bank operates through a permanent establishment the permanent establishment will not hold any equity in its own right. However, the permanent establishment will see allocated to it part of the equity capital of the bank as a whole, and therefore there may be a case to allow an ACE on that equity. Note the cost estimates provided above do not include the application of an ACE in this way.

For Corporation Tax, we follow OECD principles in imputing an amount of equity (“free capital”) to permanent establishments. This process aims to reflect the funding structure that the permanent establishment would have if it was a separate, independent entity conducting the same business in the same location. It has the effect of restricting the amount of interest that can be deducted in the permanent establishment’s a computation to help ensure an arm’s length measure of profit is subject to tax in the relevant territory. However, this process simply deems part of the funding to be equity and in reality no distributions will be paid. For an ACE, we would therefore need to consider whether these rules would need to be adapted.

Where a foreign bank operates through a UK subsidiary, the subsidiary would be required to hold its own regulatory capital as required by the Regulator. As a result the ACE would apply to any distributions the subsidiary makes. This would potentially advantage foreign banks operating through subsidiaries in the UK and create an incentive for foreign banks to change their current structure to benefit from an ACE. This would be a factor which could further increase the costs of an ACE.

Offsetting the Costs of an ACE

You asked what increase in the rate of the Bank Levy would be necessary to offset the cost of an ACE, and we estimate that this would be in the region of five basis points—ie this would imply an increase from the current full rate of 0.130% to a new rate of 0.180%. As with the estimates of the cost of an ACE, this estimate includes no assumptions about behavioural impacts, which again might be significant, and could reduce the competitiveness of the UK as a location for international banking activity.

Offsetting the costs of the ACE through a higher rate of corporation tax may be a more logical approach. However, this would still create winners and losers at individual business level, particularly in light of any losses carried forward in the sector, and therefore the behavioural response is unclear. In addition, as you know, the Government is committed to making substantial reductions to the main rate of corporate tax, as this will reduce the cost of new investment and therefore incentivise activity across the economy.

It is worth noting that those who have advocated introduction of an economy wide ACE, such as the Mirrlees Review, have tended to argue the costs should be met through an increase in consumption taxes such as VAT. However, in the case of an ACE limited to the banking sector, it would not be practical to offset the costs by increasing VAT paid by the banking sector, due to the nature of VAT exemption (as explained in HMG’s previous written evidence). In addition, even if an operable solution could be found, changes in the VAT treatment of the banking sector could not be made without amendment to existing EU VAT directives.

Annex

ASSUMPTIONS UNDERPINNING THE COST ESTIMATES OF A BANKING SECTOR ACE

The following assumptions have been utilised in arriving at the estimates.

i. The underlying liability base proposed by ACE is approximated by projected estimates of shareholder equity held by the financial sector.

ii. Equity is the projected net stock position after internal generation, further issuance and buy-backs. The level of equity projected is invariant to how it is generated.

iii. Stock of equity assumes that financial institutions adhere to Basel III regulatory requirements and timetable. No slippage is assumed.

iv. Stock of equity is derived from projected risk-weighted assets (RWAs) for the sector multiplied by the required level of equity needed to adhere to Basel III regulatory requirements. Estimates of the RWA are a combination of external research estimates as well as those derived by HMRC/HMT.

v.The above estimates are “raw” Exchequer impacts—they do not take into account the impact of losses being carried forward by the sector nor of the future projected profitability of individual entities or the sector. The final Exchequer impact from this measure is a function of many endo- and exogenous factors beside the above measure meaning it is not possible to hypothecate the impact of this measure to the speed at which losses will be potentially eroded.

vi. The level of RWA, and therefore the level of equity that is used to support it, is assumed to be homogenous in nature across the sector and institution-by- institution.

vii. Following from vi., it is assumed that there is a homogenous relationship between the level of RWA and Total Assets of a firm. The split of equity between UK and non-UK operations is done by using the split of total assets between the UK and non-UK operations (HMRC estimates).

viii. The modelling approach is based on the methodology used for other costings ratified by the OBR—for example, Basel III. It is a “top-down” macro approach in which total assets of the banking sector, as per advise from HMT, are limited to four times expected GDP. The level of equity is then subsequently calculated to support this under Basel III; therefore, the model does not identify/apportion assets or the level of corresponding equity required to individual for non-UK headquartered financial institutions. It is also silent on the source of equity by institution. A lack of data for individual entities, as well as timing and resource constraints, has prevented the ability of HMRC to construct a “bottom-up” model that would calculate the Exchequer impact of individual entities, and then aggregate them to an overall Exchequer impact.

ix.The expected risk-free rate has been taken from UK yield curve as (5 February 2013; source: bloomberg.com). The Exchequer impact is very sensitive to the risk-free rate used: using a yield of 30-year gilt produces an annual average cost to the Exchequer of around £1,500 million per year over the forecast period, whereas a yield using the 5-year yield results in an expected Exchequer cost of around £450 million, with the spread between these two yields being 237bp. That is for every 1bp increase in the risk-free yield results in a £4 million cost to the Exchequer. The above yield estimates represent an un-weighted average of the likely Exchequer impact from using 5-year, 10-year and 30-year benchmark UK conventional gilt yields as of 5 February 2013.

x.As banks have a risk-premium, their borrowing costs would be greater than the UK risk-free rate meaning that if a representative borrowing rate is used to determine the Exchequer impact of the introduction of ACE, then the above costs are likely to be a lower bound estimate. The same would likely be the case if a return on equity (RoE) of equity was employed estimate the expected Exchequer impact.

xi.Note this is for the banking sector only—it excludes non-financials and the insurance and assurance industries. If these were included, the overall cost of ACE would be substantially higher.

8 April 2013