Our response to HMRC consultation on strengthening the code of practice on taxation for banks


The Building Societies Association represents mutual lenders and deposit takers in the UK including all 45 UK building societies.  Mutual lenders and deposit takers have total assets of over £375 billion and, together with their subsidiaries, hold residential mortgages of £245 billion, 20% of the total outstanding in the UK.  They hold more than £250 billion of retail deposits, accounting for 22% of all such deposits in the UK.  Mutual deposit takers account for 31% of cash ISA balances.  They employ approximately 50,000 full and part-time staff and operate through approximately 2,000 branches.

The mutual sector

Customers of mutual institutions who have a savings account, or mortgage, are members and have a right to vote. Mutuals are not proprietary companies and not therefore driven by external shareholder pressure to maximise profits to pay dividends.

It is worth point out that building societies may not raise more than 50% of their funds from the wholesale money markets or put less than 75% of their lending into residential mortgages. They are not allowed to take risk positions in commodities, currencies and derivatives.

A key feature of the mutual sector is its diversity. Mutuals vary from a handful of large, national businesses to the vast majority of local and regional societies. Their knowledge of local markets and their understanding of individual customers' needs means they offer something different in the marketplace.  This quality is valued highly by many of their members.

Executive summary

No “strengthening” of the code’s application to the mutual sector is necessary or desirable. Some extension of the code to other parts of the financial services sector, or indeed outside it, may be desirable. The proposals are likely to have no effect, or only a very negligible effect, on the ”tax behaviour” of mutuals such as building societies which is already of a very high standard.  Strengthening the code will simply add more administration and bureaucracy for the sector with minimal tax effect.

We consider these proposals to be an unnecessary change to a process that appears to have been working effectively and in a manner that was considered acceptable to all parties in the formal banking sector.  In particular, the ”light touch” approach to small and medium sized building societies which has been taken to date is both sensible and practical. Rather than compel building societies to sign up again unconditionally to the code or extend the full code to small and medium sized societies, either directly or indirectly, we consider that HMRC should be targeting those parts of the financial services sector and indeed other sectors with regrettable and longstanding histories of non-compliance and avoidance.

There is no reason why the code of practice cannot be extended to such large, complex and multi-national organisations; the wording even reflects this.  Most building societies who have adopted the current code have simple structures and purely domestic operations. Absolutely nothing will be achieved by making these institutions sign up again to the code.  It is a waste of time and resources for our members and also, in our view, for HMRC. It strikes us as overtly political with no tangible benefits either to our members or to HMRC.    We agree with other bodies that it is wholly inappropriate for HMRC to seek to strengthen its own powers in relation to the code, particularly when the formal banking sector has been demonstrably compliant with the code.  These proposals are erode the separation of powers between the judiciary and executive, and undermine the maintenance of a stable and predictable tax system. 

The legal uncertainty about how the code will operate in future creates significant risks for banks and building societies.  Without amendment, it is hard to see how they can unconditionally adopt/ re-adopt the code without effectively putting themselves in a compromised legal position.


The Building Societies Association is most grateful to HMRC for engaging with the BSA and other stakeholders on this important issue.  In particular, we welcome officials’ willingness to talk to groups of our members.  We note from a speech given by a senior official at a meeting in June[1], and the follow-up HMRC summary, that HMRC regards the current code as successful with avoidance by banks at an all-time low.  It has seen a positive change in banks' behaviour and has told select committees and the media so. 

We wish to state categorically that the Building Societies Association has no wish to promote or defend tax avoidance.  We therefore understand why the government wants this changed behaviour in major banks to carry on.  Clearly, we support such action.   But we continue to have grave reservations about the need for any strengthening of the code’s application to the mutual sector. We consider that, viewed as a sector, the “behaviour”’ of building societies as regards their tax obligations is exemplary and, given the nature of societies and their ”ownership” structure, much less prone to the aggressive type of behaviour that feeds the tax avoidance industry. We challenge HMRC to produce the evidence, either in risk ratings or otherwise, that suggests societies are purveyors or wide-scale purchasers of tax avoidance schemes. We do not think that, compared to any other sector of the economy, building societies fare unfavourably in this regard. The case for the application of the code to mutuals (as opposed to other parts of the financial services sector) was weak to start with; the case for ”strengthening” its application to the sector is ill considered and badly made.

Furthermore, we take issue with the view that there are “no obvious downsides” from non-adherence to the code. The ”obvious downsides” from not adhering to the code would be seen in the ”risk rating” of an institution and in the level of HMRC scrutiny afforded to such an institution. If HMRC is genuinely concerned that ”behaviour” may change as more institutions use up losses and become taxpaying again, an ”obvious upside” of a truly ”voluntary” code is that it will allow HMRC to focus on those non-adhering institutions where ”behaviour” has not genuinely changed and towards which HMRC can focus risk resources much more easily and cost effectively.

Much of the code is built on the current disclosure regime. It is a general and voluntary guideline on taxation practice, reflecting good governance and the need for a direct, open and transparent relationship with HMRC which we support.  As pointed out by others[2], its wide and flexible language is not meant to be used to prescribe detailed rules or impose binding conditions.  The consultation makes clear that the code – though not the guidance or protocol - remains unchanged.  But the code has clearly been written with complex, international banks in mind; it largely assumes they are all handled by HMRC’s large business service and therefore have a customer relationship manager (only three of our members have one).   

Furthermore, the change which is proposed in the draft legislation to define “banks” by reference to the application of bank levy seems to narrow the scope of the code.  It will certainly catch banks; it will also catch small and mid sized building societies but will now be unable to catch much larger financial services institutions or ”shadow banking” organisations involved in avoidance behaviour. This definition appears to be designed to list those institutions that have not chosen to adopt the code.

For example, part 1 of chapter 1 of the supplementary guidance note dated 9 December 2009 answered the question, “What do you mean by banks?”  It concluded that the code would extend to the “banking type activities…of predominantly non banking groups”. This would include, for example, a large insurance group involved (say) in structured finance avoidance.  However, a “predominantly” insurance group satisfying the test at paragraph 13(1)(a) and (4) of Schedule 19 FA 2011 would now seem to fall entirely outside the code in relation to its “banking type” activities. This is ill-considered, poorly devised and deeply unfair.

With perhaps one exception, the large majority of building societies would not be regarded by their members or the public at large as equivalent to ”banks”. Additionally, in the minds of the public,[3] the sector does not carry the stigma of “greedy banks” that have been rescued by the government and now “owe” the public in return. In this respect, the code is not clear, objectively fair or proportionate in its aims and ambit and there is no case for ”strengthening” its application to the sector.

We question what benefit would be gained from any extension and the imposition of the burden of full compliance on the mutual sector.  We are yet to see what benefit, for example, a full-blown written tax strategy would bring to a building society. In our view this would be disproportionate to the tax risk profile and scale of mutuals’ operations.

HMRC’s proposed arrangements fail to adequately distinguish between non-compliance and non-adoption, a most important difference.  As the BBA has pointed out, smaller institutions for which the code has little relevance could be unjustifiably tainted by association, with no legal recourse or right of appeal for damage to their reputation by implication.  Publication of a list of non-adopters – most of which are only required to adopt section 1 so not the code per se – calls into question the meaning of the world “voluntary”.

The attachment of a statutory sanction to a “voluntary” code appears to apply the penalty but afford none of the protection of a statutory regime. The arbitrary nature of the penalty is compounded by the fact that, as presented at paragraph 3.9 of the consultation, “[as] the code will remain voluntary there will be no statutory right of appeal against HMRC’s view of a bank’s compliance with the code and decision to name such a bank”. This is unfair[4].

Finally, we share others’4 reservations about the proposed requirement for banks and building societies to unconditionally adopt or re-adopt the code.  We understand many institutions have expressed their commitment to the code in side letters; in some cases HMRC encouraged this and even drafted the side letter.  These side letters performed an important role in setting out the manner in which institutions proposed to ensure that they complied with the code.  They did not amount to attempts to dilute the impact of the code.  The letters also served a useful purpose of allowing an institution to explain its approach to compliance and thus gain comfort that HMRC viewed its governance model as code compliant. Without the greater clarity provided by these side letters it would not have been possible for a number of institutions to sign up to the code in the first place.

Q.1 We welcome respondents’ views on whether requiring smaller banks to only adopt Section 1 of the Code remains a tenable approach under the strengthened Code?

Before discussing whether smaller institutions should be “required” (although the code is supposedly voluntary) to adopt all or part of the code, there are a couple of far more important, wider issues that need to be addressed.  The first is the scope of the code.  Why has the code of practice been introduced only for (part of) the formal banking sector when those in “shadow banking” are not excluded?  Why is the code applicable to just to some of the banking sector?  There are clear, longstanding and blatant tax abuses by multinational businesses in other sectors, sectors which are not subject to a code. 

The second issue is the use of the bank levy as the criterion for adoption of the code.  The levy was introduced to change the behaviour of banks with regard to balance sheet risk.  This is rather different to taxation.  We refer back to April 2009, when a few months’ ahead of the first consultation paper on the code, officials told BSA staff that building societies were not going to be included.  This adds to the impression that using the bank levy as a measure is merely a convenience but, for the reasons given above, it does not seem to work particularly well.  We would prefer to see a more straightforward definition, for example, one based on FSMA principles or used in other parts of the UK Taxes Act.

Therefore, our concern with the code remains the assumption that the “banks” in question are large and multinational with a separate and probably sophisticated tax function. The majority of building societies do not have this function or even dedicated tax staff.  They outsource services relating to corporation tax reporting and iXBRL submission.  In a few cases, societies may not even have a dedicated ”finance director”.  The application of the code of practice needs to be proportionate to the business risk of the bank or building society.  The vast majority of building societies only offer simple savings and retail lending products. They do not have the capability or appetite to be users, promoters or funders of tax avoidance. Therefore anything other than adoption of section 1 only for these institutions would seem disproportionate to the business risk.  These societies operate a mainstream and uncomplicated approach to employee remuneration at all levels and maintain a straightforward savings and predominantly residential mortgage product set which is not sensitive to tax.  They do not invest members’ resources to facilitate complex avoidance schemes, rather they endeavour to manage their tax affairs consistently with their understanding of intentions of parliament and the spirit of HMRC regulations.

The point is that whereas the banking sector mostly comprises large, sophisticated institutions with some smaller institutions, the mutual sector has a large number of smaller institutions, with only perhaps a few that would be comparable at all to banks in terms of their need for and use of sophisticated tax resources. The obligations that the full code aims to impose are disproportionate and meaningless to the vast majority of building societies. 

The application of the full code would result in unnecessary cost for the sector. For the vast majority of smaller and mid-sized building societies, having a written tax strategy and tax code of governance in place would be a pointless piece of administrative ”red tape”. It would be particularly hard to fathom when the thrust of the government’s business aims generally is towards reducing unnecessary red tape and bureaucracy for smaller and medium sized enterprises.  For the very small handful of mutuals that do have such sophisticated resources, the existing measures and safeguards, such as the current code, risk assessment processes and the disclosure regime are clear, proportionate and sufficient.  In addition, HMRC’s Jim Harra’s statement on 24 June 2013, backed up by HMRC’s own meeting notes, confirms that all is working well.  The relationship should be two-way: building societies deserve the trust of HMRC unless there is a reason for that to be forfeited.

Our understanding of section 1 compliance in this context remains broadly that explained in 2009, namely self-assessment not subject to audit or challenge.  But letters sent to our members as late as 6 August 2013, ten calendar days before the consultation ends and during the holiday period, suggest HMRC has a different interpretation, one that is not included in this consultation.  The letter provides a “brief summary” – one full side of A4 – that sets out HMRC’s wide-ranging and detailed expectations of an adequate tax governance strategy and policy for smaller institutions.   We are not sure if this means small institutions that choose to adopt the full code or small institutions that choose to sign up to section 1 only.  We assume it means the former.  If the latter, this guidance differs to the discussions we had with officials who had suggested a slight, general amendment[5] to an existing document.   Furthermore, this letter – if indeed referring to future section 1 compliance - could be seen as a “back door” move to make smaller institutions such as the majority of building societies adopt something akin to the full code and one that is wholly disproportionate to their risk profile.  Early clarification of this issue would be most appreciated.

Other regulators, notably the Prudential Regulation Authority and the Financial Reporting Council, have recognised the lower risk nature of smaller mutuals such as the majority of building societies and devised alternative but genuinely proportionate regimes for them.  Examples in prudential regulation are the simplified liquidity regime and the standardised approach to credit risk which both use a standard set of measures and avoid model-based approaches, and the provision of a template for recovery and resolution plans.  The FRC has allowed smaller financial institutions such as building societies to use the new UK financial reporting framework acknowledging that full international financial reporting standards – mandatory for quoted institutions – is burdensome, less relevant and costly.  Both regulators understand that while uniformity might at some level seem desirable it runs the risk of producing skewed results and disproportionately high costs for the institutions concerned.  While one institution may be a hundred times the size of another, the costs of implementing change are not similarly a hundred times more; they are often much less, meaning a greater cost burden proportionately for the smaller institution. 

We question why HMRC would consider naming smaller institutions - including the majority of building societies - that have not adopted, or complied with, the code.  All they were invited to adopt was section 1 of the code.  That was considered to be a proportionate response.  They have not adopted the code per se. 

Q2. Views are welcomed from respondents on the proposed timetable for adoption/re-adoption.

The timetable is unrealistic. If HMRC does not intend to report before 30 September 2013 on its plans in relation to the responses to this consultation, this will leave insufficient time for board papers to be prepared, submitted and considered and board meetings held before the planned production of the list of signatory banks at the pre-budget report (which usually takes place in November or early December). There should be more time allowed for organisations to consider the detailed changes following the consultation process and before any ”list of signatories/non-signatories” is produced.

It should be noted that timescales for adoption of the full code would be significantly longer than for section 1 only.

Q.3. HMRC welcomes comments and views on the proposed approach set out to revise the Governance Protocol on Code compliance and whether the proposals provide the necessary assurance safeguards around the naming of non-compliant banks.

We have a number of serious concerns: 

  1. The first is the timing of these proposals.  It could be argued that HMRC is not trying to rectify a problem but trying to harness the current media interest in tax.  It is aware that many institutions are currently reporting tax losses; it fears that should the economy recover ie the institutions become taxpayers again, some might be inclined to enter into “code red” transactions.  There is no evidence for such thinking (or for the underlying situation) among building societies.  Therefore the move to “strengthen” the code is premature.
  2. HMRC states it is seeking compliance only.  But it has the power - not the duty - to name institutions.  That was neither transparent nor fair.  HMRC could use the threat of naming in any dispute.  It therefore seems disingenuous to refer to the adoption of the code as a “voluntary commitment” (section 2.3) when the failure to commit is punished by public censure. Furthermore, the fact that, at section 3.9, there is “no statutory right of appeal against HMRC’s view of a bank’s compliance with the code and decision to name such a bank” as the code is voluntary acts as an incentive not to commit to the code (see (5) below).  If the publication of code compliance and non-compliance is to go ahead, independent oversight with an appeal process is critical.
  3. If the code is only applicable to (a limited section of) the banking sector how can its application be described as “achiev[ing] outcomes that are even-handed across different customer groups” (first bullet point, 3.12).
  4. The criteria for non-compliance seem to be vague.  Under equivalent provisions for disclosure of the details of individual tax payers there has to have been a “relevant tax penalty” s94 (2) Finance Act 2009.  This would indicate that a clearly understood process has taken place to determine that the taxpayer has supplied deliberately inaccurate or false information; there has been a deliberate failure to notify or a deliberate issue of a false invoice.  Furthermore there are clear materiality trigger points with regard to public disclosure.  The right to confidentiality is therefore lost only when taxpayer wrongdoing has been fully established by due legal process.  This does not seem to be in place for code disclosure.  “Non-compliance” could happen merely because HMRC and an institution disagree over the interpretation of words whose meaning will never be tested in court.
  5. Publicly naming an institution for non-compliance when there is ongoing litigation could impact adversely on due legal process.
  6. There will be many situations in which the spirit of the tax law and/or the intentions of parliament will not be clear. In particular, there are likely to be circumstances where the transaction or arrangements being entered into and their tax treatment would never realistically have been considered by parliament at the time the applicable legislation was enacted. This may be of particular concern when considering whether the interaction of different parts of the tax code together produces a result which was never intended by parliament. In such circumstances it will be very difficult for an officer of a bank or building society (even one who is professionally advised) to be certain that an institution’s activities are within the spirit of the tax law and not contrary to the intentions of parliament[6].
  7. The naming of institutions that “HMRC considers have not complied with their Code commitments” is too vague and open to abuse.  While the process is to be overseen by the Tax Assurance Commissioner, his remit is defined in the “Litigation and Settlement Strategy” the objective of which is the achievement of “the best possible return for the Exchequer” and “maximising revenue flows” (section 8).  This does not give a taxpayer certainty.  It could give rise to actions against HMRC for defamation or a detrimental impact on business name, goodwill etc.
  8. Likewise “the intentions of parliament” benchmark requires qualification.  Which parliament is meant? EU (particularly for VAT), UK, Scottish parliament etc?  We continue to have reservations about the code’s suggested method for discerning the intention of parliament.  Defining the purpose of established law can be very difficult. The quantity of HMRC guidance is ample proof of this. Even in law, not everyone agrees.  Some have already suggested that HMRC is seeking to sidestep the legislature and courts by imposing on institutions its own view of what is correct and potentially requiring tax to be paid that need not be under the law.  Proposed changes to the guidance in this consultation appear to reflect this.  An example is the threat to apply the code retrospectively when it was written to be prospective. Manipulation of the code in this way could lead to the code being operated on a subjective, unpredictable and inconsistent basis.  Our members’ experience to date is that it has not happened – but that might not be the case in a “strengthened” application of the code (or in further changes to the guidance).
  9. There appears to be a conflict between the 2009 Taxpayer Charter on keeping taxpayer information confidential and the naming powers to be enacted.  At the first annual HMRC stakeholder meeting on 18 July 2013, HMRC officials made great play of taxpayer confidentiality.  Why is the banking sector different?
  10. There appears to be little information on what would be disclosed in respect of those institutions considered non-compliant.

Q.4. Do these proposals offer sufficient transparency for the public around how the rules will operate?

Please see above.

Q.5. We also welcome views on whether any other enhancements should be considered at this time to the Governance Protocol.

As we have said elsewhere, we would like to know why the banking sector, in fact, only a subsection of the sector, should be singled out in this manner particularly when they have been demonstrably compliant while multinationals in other sectors appear to be avoiding tax rather more obviously. This seems to be a deflection of HMRC resources away from where they are really needed. HMRC has already publicly stated that the code is working well.  Banks and building societies deserve the trust of HMRC unless there is a reason for that to be forfeited.

Q.6. We would welcome views from respondents on whether the examples set out below provide a sufficient degree of guidance of the types of transactions, or patterns of transactions or other behaviours that would lead to HMRC concluding that a bank is not complying with its Code commitments?

Please see our answer to question 3.  We would also direct HMRC to example 3 in section 3.25: a “serious lack of internal governance” is that “transactions are undertaken before Code Compliance is considered”.  Banks undertake numerous transactions which have a significant range of materiality, complexity and novelty.  Even those with a sizeable tax department cannot assess every transaction and must focus their attention on material or unusual transactions. This should not constitute a “serious lack of internal governance”.

Q.7. Do respondents consider this to be an appropriate descriptor for transactions within the ambit of the GAAR?

The examples given do not address the position where a transaction may be undertaken to address a defect in the law.  For example, the tax scheme for offshoring debt[7] to enable tax relief to be given on a paid basis was used to give equivalence to s343 ICTA88 / s940A CTA10 protection while changes to the tax legislation were under discussion following the “Butterfill” Act.  This would meet the “intentions of parliament” trigger but unreasonably so.

Q.8. Do respondents agree that this definition will result in appropriate coverage by the Code?

According to the consultation document, a transaction is a potentially abusive transaction even if the GAAR Advisory Panel concludes that the transaction in question represents a reasonable course of conduct so that, in its view, no actual challenge under the GAAR is appropriate.  This is clearly unsatisfactory as it gives no clarity or certainty to the taxpayer and leads to abuse of power by HMRC.

Q.9. Do respondents agree that the legislation as drafted covers the issues set out in this Consultation Document appropriately?


Q.10. Are there any other matters that respondents would like to see covered in the legislation?

We would like to see the code properly extended to all parts of the financial services sector where abusive behaviour may still exist and indeed this should be extended beyond the financial services sector to those multinational enterprises that are clearly not paying their fair share of corporation tax but might do if the code were applied to them equally. Fundamental change is required in the ”tax behaviour” of certain enterprises outside the financial services sector; this should be the priority rather than ”tinkering” with the code to achieve what is likely to be only an inconsequential impact on the existing behaviour of banks and building societies.

Q.11. HMRC would also be grateful for any detailed drafting points that respondents might have on the draft clauses.

The above concerns are too fundamental to be corrected by changes in drafting.

[1] Jim Harra, director general for business tax at HMRC, at a “town hall” meeting on 24 June 2013.

[2] Reproduced from British Bankers’ Association response on strengthening the code of practice.

[3] For example, see research publication, “Mutuals: better customer service” www.bsa.org.uk/publications/industrypublications/customer_service12.htm

[4] Reproduced from British Bankers’ Association response on strengthening the code of practice.

[5] This was suggested as a few lines in an existing document roughly that tax would be considered when new products were introduced to ensure the tax outcomes were in line with the commercial outcomes. 

[6] Argument also put forward by the City of London Law Society, 25 September 2009.

[7] This offshoring debt example was a tax avoidance disclosure arrangement that enabled the refresh of tax losses.  After the “Butterfill” legislation was passed (allowing mergers in the mutual sector), a number of areas of tax law were identified as not giving mutuals the same tax rights as other trading entities.  In particular, there was no equivalent to a s343 / s940A transfer (transfer of trade and assets).  This meant that if a mutual which had trading losses merged with another mutual the trade losses would be forfeited.  This problem was highlighted to HMT and HMRC but there was a delay in enacting the amendments to the tax legislation – hence the above tax scheme was used to protect the tax losses until the changes came into effect.