Office of the Comptroller and Auditor General - Dirt Investigation - Chapter 27

Chapter 27 : AIB Group Internal Auditor’s Report 1991


The public concern about the scale of bogus non-resident accounts emanated from an article published in the Sunday Independent of 5 April 1998 and the Magill articles of October and November 1998. The source of the information contained in those articles appears to be a document sent by the AIB Group Internal Auditor, Mr Anthony Spollen to the Audit Committee in March 1991 and other material associated with the follow up to that document.

The document in question covers a number of issues which were causing Mr Spollen concern. The extract from the document dealing with the issue of bogus non-resident accounts in the Republic of Ireland is reproduced at Appendix H but in essence it states

the Taxation Department of the bank estimates that at least 60% of non-resident accounts are bogus to a value of ?600 million

the DIRT payment which should have been made to Revenue in October 1990 in respect of bogus accounts is in the region of ?10 million for the six month period

there is a contingent liability (probably close to ?100 million) for DIRT arrears back to 1986

the absence of any documentation for a large number of accounts has been the subject of comment and report in many internal audit reports

the Group Taxation Department’s statement that a determined effort was made to eliminate all bogus non-resident accounts by April 1990, was not true

there was no amnesty from Revenue in respect of arrears of DIRT.

Senior management of the bank did not accept the validity of these assertions and ultimately the matter was put to the Group Audit Committee of the bank in the wider context of the independence of the Group Internal Auditor and his other specific concerns. In a report dated 27 May 1991, the Group Audit Committee supported the senior management’s view of the matter. The relevant extract from the report states

"Non-Resident Accounts in the Republic of Ireland ("Form F Deposits")

8.1 The Internal Auditor submits - (a) that he has frequently commented on the inadequacy of documentation for a large number of accounts; (b) that the situation is critical; and (c ) that the exposure of the Group is substantial, viz. some ?10 million for April - September 1990 and "probably close to ?100 million [since 1986]" for which some form of provision will have to be made in the accounts. He further alleges that the Group Taxation Department’s assurances that a determined effort was made to eliminate all bogus non-resident accounts by April 1990 was "clearly not the case"

We recognise that the situation is a very difficult one for all concerned and is general throughout the industry. We have considered the views of senior management and in particular their on-going dialogue with the office of the Chief Inspector of Taxes and we noted that the Financial Director has informed the Central Bank of the continuing position. We accept that (a) in para 8.1 above is the case but that (b) is a view not shared by senior management. With respect to (c), the appropriate provision will be determined in due course by the Audit Committee on advice submitted to them."

A provision in respect of a DIRT liability was not included in the annual accounts.

The bank claims that Mr Spollen’s expressed concern about bogus non-resident accounts may have been prompted by his dissatisfaction with a proposal to transfer him elsewhere in the Group. Mr Spollen denies that his actions were influenced by the proposed transfer.

There is clearly a conflict on a number of key points between what is contended by Mr Spollen and the bank’s position. In the following paragraphs I will set down the basis for the opposing positions by reference to the documents I have examined and the oral evidence I have heard.

"60% of Non-Resident Accounts are Bogus"

Mr Spollen

In documents examined, Mr Spollen stated in a memorandum dated 6 March 1991 to Mr Brian Wilson, Group General Manager, that "for the purpose of calculating the DIRT which should have been deducted in the half year to October 1990, Jimmy O’Mahony and the Financial Control people have assumed that 60% of all accounts designated non-resident are bogus". In his submission to the Audit Committee dated 10 March 1991, he stated that the 60% was an estimate by the Taxation Department of AIB. In oral evidence, Mr Spollen confirmed that to the best of his recollection Mr O’Mahony was the source of the 60% estimate.

Mr Jimmy O’Mahony

I have seen no document signed or attributed to Mr O’Mahony which records 60% as an estimate of bogus non-resident accounts in AIB.

In oral evidence, Mr O’Mahony stated "I have no recollection of mentioning to Mr Spollen a figure of 60% as representing the proportion of non-resident declarations which were not genuine. If I did mention it, it would have been an off-the-cuff remark, and not an attempt to precisely measure the percentage. I had no way of measuring this percentage and neither did Mr Spollen. In my humble opinion, I do not think that anyone could, no matter what yardstick one applied to determine what the problem was."

Mr Don Walsh (then Head of Group Internal Audit - Ireland)

In oral evidence, Mr Walsh stated the 60% bogus rate was communicated to him by telephone by Mr John Clifford, a member of his internal audit staff who was involved in discussions with Mr O’Mahony or some other member of the Taxation Department or other department. Those discussions, in which Mr Walsh had earlier participated on 4 March 1991 with Mr Clifford, were part of the proceedings of a committee which was set up to come up with an action plan going forward on the DIRT problem. At that particular meeting, it had been decided that a calculation of the potential liability of AIB would be done for the earlier DIRT payment for the tax year 1990-91 and the 60% assumed rate of bogus non-resident accounts was used in performing that calculation.

Mr Walsh stated in evidence to me that after receiving the phone call from Mr Clifford (on 5 March 1991)

"I would have gone directly; I would have considered this significant information; I would have conveyed that immediately to him (Mr Spollen)".

The evidence shows that Mr Spollen didn’t come up with the 60% estimate of bogus accounts. The estimate was supplied to him. The only material point of contention is the extent to which the 60% could be used as a reasonable basis for estimating the number of bogus non-resident accounts.

On the basis of the estimates of reclassifications performed in the first half of 1991 as set out in Chapter 25, it would appear that the percentage by value of accounts incorrectly designated as non-resident for DIRT purposes was of the order of 46%-48% in February 1991.

"The DIRT payment which should have been made to Revenue in October 1990 in respect of bogus accounts is in the region of ?10 million for the six month period"

The basis for this statement would appear to be Mr Walsh’s oral notification to Mr Spollen of the information given to him by telephone by Mr Clifford on 5 March 1991.

According to Mr Walsh’s evidence to me

"He [Mr Clifford] supplied me with the non-resident resources base that was being used. He supplied me with a 60% default assumption that was being used. He supplied me with the interest rates and the retention tax rates. It broke down into a figure of ?2,500,000 for branches and a figure of ?6,900,000 for the RDC totalling ?9,400,000. I conveyed that to the Group Internal Auditor (Mr Spollen) together with the unreconciled difference."

Mr Walsh clarified in evidence that the ?2,500,000 represented the DIRT underpayment in respect of six months interest on branch deposits and the ?6,900,000 represented the DIRT underpayment in respect of eleven months interest on deposits with the RDC.

He also explained that the reference to the "unreconciled difference" arose because the calculation used figures sourced directly from the retail deposit centre and these differed to an amount of ?160 million with the statistics provided by the Technology Audit section. The bulk of the difference was represented by sterling deposits introduced by Republic of Ireland branches. Effectively, the total value of the non-DIRT paying Irish pound non-resident accounts was ?772 million approximately. The ?9.4 million calculation was based on that figure.

In summary, there is agreement that a calculation was made to cover possible liability for DIRT underpaid in the 1990-91 tax year. However, while the documents prepared at the time by Mr Spollen refer to the liability viz. "in the region of ?10 million", as being in respect of all non-resident deposits for the half year to October 1990, the bank contends that a higher unadjusted figure was used by Mr Spollen viz. the Technology Audit figure, and that the ?9.4 million figure includes DIRT on eleven months interest on non-resident accounts at the retail deposit centre rather than six months interest. More importantly, the bank contends that no such liability arises as they had an amnesty for the past from Revenue provided they fully regularised the position by 30 June 1991 (per the report dated 25 March 1991 from Mr John Keogh, Group Financial Director in response to Mr Spollen’s letter of 10 March 1991 to the Group Audit Committee).

"There is a contingent liability (probably close to ?100 million) for DIRT arrears to 1986"

According to documents which I have examined, the basis of Mr Spollen’s contention that there was a contingent liability of ?100 million appears to be statistics supplied by the Technology Audit Section of the bank relating to "Branch Deposits and Branch Introduced Deposits in Retail Deposit Centre (Combined) - Republic of Ireland". These show that, at February 1991, the total number of non-resident deposits was 87,660 with a total value of ?982,989,409. These figures were used for the calculation of the DIRT unpaid of ?10 million for the half year to 30 October 1990. The calculation of the ?100 million appears to derive from multiplying that ?10 million by the number of half years between April 1986 and April 1991 i.e. 10.

Mr Spollen stated in evidence to me that he understood that the statistics regarding the number and value of non-resident accounts at February 1991 would have been produced by the "financial control and taxation people" and that he would have applied the 60% estimate of bogus non-resident accounts to those figures viz.

60% of 87,660 = 53,000 approximately

60% of ?983 million = ?600 million approximately

Mr Spollen also stated in evidence to me that the estimate was never intended as a precise calculation. He pointed out that, while it did not take account of changes in volumes of accounts or interest and DIRT rate changes in the intervening years, it erred on the conservative side if anything. In this regard, he referred to a comment allegedly made by Mr Jimmy O’Mahony that 75% of the non-resident accounts in the RDC was "funny money" and that the position at February 1991 should have reflected the work that was carried out prior to that date on reclassifying bogus non-resident accounts as DIRT paying.

The bank’s views on Mr Spollen’s calculation of the contingent liability of ?100 million are as follows

"It is not clear that the extract from Mr Spollen’s submission dealing with DIRT was ever intended by him to be a calculation of DIRT due. For this and other reasons outlined below AIB has contended that it is quite wrong to use it as an accurate basis for calculation.

The figures appear to be based on an interrogation of the computer systems in February 1991 which produced a report highlighting the number and value of non-resident deposits designated by reason of having either a Form F (Declaration under Section 175 of the Income Tax Act 1967) and/or Form 37 (Declaration under Section 37 of the Finance Act 1986) ‘flag’. The calculation appears to have been:

(i) ‘Bogus’ non-resident accounts numbers: 87,660* x 60% = 53,000

(ii) ‘Bogus’ non-resident account values: 982,989,409* x 60% = ?600m

(iii) Outstanding DIRT liability: ?10m** x 2 x 5 years = ?100m

* As per totals on the computer interrogation

** ?10m appears to be an estimate of DIRT on the interest based on the above values for a half year.

This is not a reliable basis to determine the DIRT at issue for the following reasons:

Volume, interest rate and DIRT rate changes

In grossing up the liability over five years, no account was taken of volume, interest or DIRT rate changes.

Number of Accounts

The figure of 87,660 for the total number of accounts included 15,782 accounts in the RDC which had nil balances.

Foreign Currency Accounts

Both the total number and value of accounts apparently used in the calculations (87,660 accounts to the value of ?982,989,409) appear to include an element of foreign currency accounts. These were outside the scope of DIRT at that point in time and, consequently, should not have been included. The total value of non-resident accounts at the time was actually higher and the value of non-resident Irish pound accounts was lower.

DIRT Status of Accounts

The figures appear to assume that the ‘non-resident’ flag denoted the existence of a Form 37 and that, in each case where such a flag was present, the account was exempt from DIRT. This was not so as the presence or absence of a ‘non-resident’ flag did not determine the DIRT status of an account. A separate DIRT flag controlled this.

The non-resident status of an account and its DIRT status are not necessarily synonymous. For instance, some non-residents did not complete declarations. While such customers will be classified (correctly) as non-residents, they would be liable to DIRT in the absence of a Form 37. Of a total of 74,614 accounts designated non-resident in the Branch Network as at March 1991, a total of 47,976 was designated liable to DIRT.

Level of ‘bogus’ accounts

This figure results from an application of an estimated 60% default level to the figures extracted in order to establish the level of ‘bogus’ non-resident accounts from which DIRT was not being deducted. Mr Spollen did not arrive at this estimate himself and says so in the documents in question. He attributed this estimate to the then Group Taxation Manager, Mr O’Mahony who has no recollection of giving such an estimate. To the extent that any such estimate was given by him, he has stated that it would have been on the basis of ‘what if’ the default was at this level. That being the case, he has stated that it would have been with a view to prompting management to respond to the Revenue initiative of February 1991. We are unable to establish any basis in fact for the 60% estimate and do not believe it has any.

An examination of this issue undertaken in late 1998 found that by end March 1991 there were no more than c.4,000 potentially ‘bogus’ non-resident accounts with balances in the order of ?110m by comparison with the alleged figure of 53,000 accounts to the value of ?600m.

In his letter of 6 March, 1991 Mr Spollen stated that 17,000 accounts had no declarations of non-residence. This has no basis in fact. Internal Audit files indicate that evidence from 20 Branches was reviewed to establish recent audit trends in the months prior to January 1991. On the basis of this, it was concluded that there was an average document default level of 25%. However, this could range from minor completion errors to missing declarations. There is no evidence to substantiate the figure of 17,000 accounts with no declarations."

"The absence of any documentation for a large number of accounts had been the subject of comment and report in many internal audit reports"

It is clear from a review of internal audit reports up to 1991 that there was a high degree of documentary non-compliance detected since the introduction of DIRT in 1986. There was an awareness at senior management level that this was the case. In a letter dated 9 October 1989 from Mr Henry O’Brien, Head of the Inspection Unit in Ireland, to Mr Spollen he stated

"I feel decidedly uncomfortable about the position in the Republic of Ireland. When DIRT was first introduced, there was a move to clean up the position, but then word went around the branch system that the Revenue would not exercise their rights to review the position at branches, and progress more or less ceased.

Following a discussion with Jimmy O’Mahony, we carried out a 100% sample at a number of branches earlier this year and the outcome was disconcerting.

In general, there is not a major problem in Dublin or the East Coast area, but from West Cork to Donegal, the position is bad in a large number of branches.

If a decision was taken to put our house in order forthwith, there would be implications for the bank, by virtue of (a) higher interest rates payable on resident deposits, and (b) a probable loss of deposits. On the other hand, the bank is acting improperly at best in paying interest gross when the required forms are not held and it could be contended that we are not exercising due diligence in this area. There is a school of thought that 1992 will fix everything, but I don’t believe this.

I would like to see a gradual move towards putting our house in order and I propose to discuss this with Pat O’Mahony when we have our meeting shortly - also with Jimmy O’Mahony".

Mr Spollen passed the letter on to Mr Keogh on 10 October 1989 with a note that said inter alia "and you will find it interesting".

Documentary non-compliance was still a problem in late 1990. In a memorandum dated 16 October 1990 from Mr Walsh to Mr P. J. O’Mahony, General Manager Branches, enclosing regional summaries of audit findings in branches in the Republic of Ireland for the half year to September 1990 he stated that "the summaries reveal shortcomings in control standards across all regions with some variation but with the high risk areas of ....Form 37/Form F.......featuring frequently as requiring improvement and increased management attention."

In the annual presentation by the Group Internal Audit Department to the Group Audit Committee in December 1990 a reference was made to a comprehensive review of Form Fs having been undertaken in both Britain and Ireland (12 staff were involved in this assignment) and that the RDC had non-compliance with Form F procedures.

In his report dated 25 March 1991 to the Group Audit Committee, Mr Keogh referred to that presentation to the Group Audit Committee. He stated

"The findings of that review (viz. The comprehensive review of Form Fs in both Britain and Ireland) were not stated and no issues arising from it were flagged for the Audit Committee as giving cause for concern".

Mr Culliton, Chairman of the Group Audit Committee at the time, stated in evidence to me that references to documentary non-compliance on non-resident accounts were omitted from or, at least, did not have that significance attached to them in the Internal Auditor’s presentation to the Audit Committee in December 1990. He stated that it was his practice during his term as Chairman of the Audit Committee to enquire of the Internal Auditor on the occasion of the presentation of the report whether or not there were any issues that he wished to bring to the attention of the Audit Committee and the Committee was assured that there were no such matters.

Mr Spollen stated in evidence to me that in December 1990 he had no feel at all for the extent of the problem. In response to my enquiries as to why, in the face of overwhelming evidence from internal audit reports, he did not flag the matter as a serious issue for the Audit Committee prior to 1991, he stated that there was an apparent failure on the part of internal audit to follow up their findings on documentary non-compliance with branch managers. He stated that his awareness of the scale of the problem dated from his meeting with Mr Walsh at the end of January 1991 in the context of being informed of there being a problem with the bona fides of a third of the non-resident accounts in Northern Ireland.

In summary, there is no doubt that the problem of extensive documentary non-compliance on non-resident accounts was identified and reported by Internal Audit in the years up to 1991. Senior Management in the persons of Mr Keogh, Mr P. J. O’Mahony and Mr Jimmy O’Mahony were made aware of the serious nature of the problem prior to 1991. Mr Spollen was also made aware of the problem but did not appear to appreciate the gravity of the situation. Consequently, he did not report it as a serious issue to the Group Audit Committee.

"The Group Taxation Department’s statement that a determined effort was made to eliminate all bogus non-resident accounts by April 1990, was not true".

The bank’s efforts to eliminate bogus non-resident accounts would have come to Mr Spollen’s attention primarily from (a) a letter dated 6 February 1991 from Mr Brian Wilson and (b) a memorandum dated 12 February 1991 from Mr Walsh.

(a) "These initiatives were worked out following discussions which I had with Pat O’Mahony, General Manager - Branches and Information Technology, who had already been tackling the issue with his management team during last year."

(b) "Documentation standards for non-resident deposits in the Ireland Division have been inadequate and following representation from Group Internal Audit, Senior Executive Management have put a strong message into the system over the last 12/18 months that matters were to be rectified".

The question of what constitutes a determined effort is subjective. While it is clear that some effort was being made, it wasn’t determined enough because it wasn’t successful. The fact that such an extensive reclassification took place in 1991 is a testament to the failure of any such earlier efforts.

"There was no amnesty from Revenue in respect of arrears of DIRT".

The matter of whether there was or wasn’t an amnesty from Revenue is one of the key issues of the investigation being undertaken by me. The previous chapter of the report addresses this issue.

For the purposes of this section of the report, it is appropriate to consider the basis of Mr Spollen’s concern. In a memorandum of 28 February 1991 headed "Comments on the content of the Letter from Investigation Branch of the Revenue Commissioners dated 15 February 1991, and the Group Internal Audit approach" from Mr Spollen to the members of the Group Internal Audit Management Committee he states:

"He [the Senior Inspector in Investigation Branch] also stated that any cases discovered prior to 30/06/91 will be subject to DIRT payments to be negotiated at the Investigation Branch and that such payments will not be subject to penalty or publication.

What period will this payment cover ?

If over 1 year, does the bank have a liability ?

What is the position re: Accounts closed either this year or previous years which were not bona-fide Non-Resident accounts ?"

In a memorandum dated 28 February 1991 from Mr Jimmy O’Mahony to Mr Spollen he states

"Your request for a further letter from the Revenue Authorities confirming that an amnesty is being afforded to A.I.B. and all other Financial Institutions for the period prior to 6th April 1990 cannot be complied with. You should understand that a situation such as this is extremely delicate. To seek what you require would mean legislation and this is not desirable for this type of issue. We are being offered a way out of a very difficult situation which I believe will meet the needs of both the Revenue and the other Financial Institutions without it being publicised - this being very much in the interest of the bank."

In a letter dated 6 March 1991 to Mr Brian Wilson, Mr Spollen states

"Jimmy O’Mahony states that we cannot get confirmation that an amnesty is being offered as this would require legislation which he feels is not desirable for this type of issue.

We cannot therefore say that we do not have a huge potential liability ....(If we calculate the D.I.R.T. for the year to 31/3/91 on these bogus accounts as being close to ?20m....and if the problem has been improving significantly over the past years as was indicated to the Investigation Branch it does not need me to say to you that the potential liability is very large indeed.)

It is not clear from the letter if the DIRT payment to be negotiated with the Investigation Branch is for 1 year or more than that."

Mr Spollen took the Revenue letter of 15 February 1991 at face value and was not reassured by the contents of Mr Jimmy O’Mahony’s memorandum in the absence of any written evidence from Revenue to the contrary.

In summary, Mr Spollen saw the need to calculate a contingent liability in respect of DIRT arrears in the absence of any evidence of a Revenue amnesty while senior management saw no need for such a calculation as they believed that they had a Revenue amnesty.

Circumstances in which seriousness of problems came to Mr Spollen’s attention

There is a conflict of evidence as to the source of the original estimate of bogus non-resident accounts in AIB in the Republic of Ireland attributed by Mr Spollen to Mr Walsh.

Mr Spollen

In a letter dated 28 January 1991 to Mr Brian Wilson, Mr Spollen states

"He [Mr Walsh] currently estimates that the size of the Form F problem in the Republic of Ireland could run into hundreds of ? millions".

In a memorandum dated 30 January 1991 to Mr Walsh, Mr Spollen states

"Please quantify immediately the amount of false Form F money on the books of the Group. The estimate (?350m - ?400m) which you have given me is frightening."

Mr Spollen stated in evidence to me that to the best of his recollection what happened was that Mr Walsh came to him and said "We have a problem in Northern Ireland" and he responded per the evidence

"So I said to him presumably: ‘What is the situation in the Republic?’ and he said that this could run into hundreds of millions. I would have said to him: ‘Well, hang on, hundreds of millions; what does this mean?’ and he said: ‘Well, it could be 350 or 400'. That would have scared me for one reason -- because Walsh always understated problems."

Mr Walsh

In regard to the letter of 28 January 1991 from Mr Spollen to Mr Wilson, Mr Walsh stated in evidence to me

"In late January 1991 the group internal auditor in a discussion with me asked for my estimate of the default level of non-resident accounts in the Republic of Ireland. The impression has been given that I raised this with the group internal auditor. That is not the case, he raised it with me in late January 1991. We had a reasonably long discussion. I told him what our procedures were from an audit point of view. I told him that it would be, as far as I could see, impossible to come up with an estimate or to quantify what the situation might be. And the only way that could be achieved in my view -- and I said this to him -- was to interview individual branch management. That would involve a huge exercise in visiting branches. The discussion was left at that. A few days later the group internal auditor showed me a copy of the letter which you have outlined. I was quite taken aback at the mention of hundreds of millions of pounds because I have no recall at that time of having given any such estimate. I said this to the group internal auditor. I also made a point of correcting it with the recipient of the letter a short time later."

In regard to the memorandum of 30 January 1991, Mr Walsh stated in evidence to me

"This was an even more specific estimate. Again, I had no recall at that time of having given such an estimate. I noted at the time that the estimate hadn’t been sent anywhere else. But I was quite surprised at the time as to his need to put this in writing to me including such an estimate considering we were working in adjoining offices."

He also stated that he told Mr Spollen that the estimates he was quoting were not his (Mr Walsh’s) but that

"the conversation was overtaken by his request as to would I quantify the overall situation, which I had already verbally explained to him was not on. That it couldn’t be done without visiting branch management. When I put this to him he said that he would get technology audit to do it. He went out directly to the technology audit people and gave them his instructions. He made it clear to me that I was to respond to him when that exercise was concluded."

The contents of this page were last updated on 26/09/03