Office of the Comptroller and Auditor General - Dirt Investigation - Chapter 8

Chapter 8 : Audit Examination of Financial Institutions

Background

In exercise of my power under Section 2 of the Comptroller and Auditor General and Committees of the Houses of the Oireachtas (Special Provisions) Act 1998, I appointed Aileen Barry a partner in the UK firm of Arthur Andersen to conduct examinations and investigations of the affairs, books of accounts and records of certain financial institutions relating to their deposits.

22 financial institutions were subject to such examination. These were as follows

ACC Bank plc

Allied Irish Bank plc

Allied Irish Bank Capital Markets plc

Allied Irish Bank Finance Limited

Ansbacher Bankers Limited

Bank of Ireland Limited

Barclays Bank plc

Citibank NA

Educational Building Society

First Active plc

Guinness & Mahon

ICC Bank plc

ICC Investment Bank Limited

ICS Building Society

Irish Permanent plc

Irish Nationwide Building Society

Irish Intercontinental Bank Limited

Lombard & Ulster Banking Limited

National Irish Bank Limited

TSB Bank

Ulster Bank Limited

Ulster Bank Markets Limited

In the case of financial institutions with branch networks the accounts of selected branches were examined. In the case of single office institutions or those who employed centralised filing systems the examinations were conducted centrally.

In all, examinations were conducted at 56 sites.

Purpose of the Examination

The principal objectives of the examination were to

establish if declarations were in place for non-resident accounts

verify the validity and timely completion of any non-resident declarations held in respect of an account

note factors which might indicate that an account holder classified as non-resident was, in fact, Irish resident

establish the basis upon which non-resident accounts had been reclassified as resident accounts

review closed accounts and determine the method of closure

verify that correctly completed declarations and accompanying documentation were held for accounts classified as exempt from DIRT on the basis of declarations made by companies, pension funds and charities

review the operation of Special Savings Accounts (SSAs), including the validity of account applications, compliance with the required notice period for withdrawals and appropriate restriction of the maximum account balance of ?50,000 and to verify that appropriate declarations were held in respect of such accounts.

In addition, assurance was sought on the validity of a sample of UK addresses given by persons who completed non-resident declarations.

Methodology

Each institution prepared data files which contained specified account information at pre-defined dates agreed with me. These files were then copied from their source computer systems to a standalone personal computer within each institution. Each file was then reconciled to its source system to ensure that all data being presented for sampling was complete and accurate.

With the assistance of personnel of financial institutions, the criteria required to extract individual account categories were identified. Samples were then extracted from these separate categories in accordance with an agreed work programme. However, this approach did not produce the required result for all institutions and in these instances work was undertaken in conjunction with the institution to extract an appropriate sample of data by an alternative method.

Where an institution had a number of accounts in a particular category below the recommended sample size, testing was performed on all accounts in that category.

Fieldwork - Branch Visits

In all cases, fieldwork was undertaken at the selected branch or, in the case of institutions with centralised records or a single office structure, at its head office. The storage and presentation of the documentation required for review purposes differed significantly from institution to institution.

In addition to the review of data, general discussions were held with institution personnel, usually the branch manager or internal audit staff, in order to ascertain standard procedures or to clarify specific issues.

Non-resident book - Declaration and Validity

Declarations were reviewed in order to ascertain how accurately and when they had been completed by account holders and any omission or exception was recorded.

In the final reporting of errors and omissions in the completion of declarations the following allowances were made

Any failure to repeat an address given in full on one part of the form was not regarded as an error.

The omission of the capacity in which the form was signed was not treated as an error, provided that the declaration related to the account holder and owner of the beneficial interest.

The omission of the interest beneficiary's country of residence was not reported as an exception, provided that the address given for the beneficial owner of the interest was sufficiently detailed that the country of residence could be identified clearly.

Declarations which were dated more than three months after the account opening date were not reported as an exception if the institution advised the account was DIRT paying or if no interest was paid prior to the declaration date.

Declarations signed by the account holder for a different account number from that included in the sample were accepted, if it was clear that both the accounts listed on the declaration and the sampled account were owned by the same individual.

All other anomalies in respect of the completion of the declaration have been reported as validity exceptions, either because they raise doubts as to the timeliness of the form's completion or raise a question mark over the non-resident status of interest beneficiaries.

Non-resident book - Irish Residency

Factors which may be an indicator of Irish residency were noted as risk indicators. The existence of a particular indicator or combination of indicators cannot be taken as definitive evidence of resident status. In highlighting the risk profile of a financial institution’s non-resident deposit book they serve to indicate the instances and general extent to which vigilance in the administration of such accounts is called for. In interpreting the indicators certain matters should be borne in mind.

While most financial institutions operate a ‘hold mail’ service and ’hold mail’ is a risk indicator of an Irish resident operating a non-resident account, there may be bona fide reasons for the non-residents not wishing to receive correspondence. Some of the reasons given for this were that

customers may prefer to deal with mail at a convenient time when they are in Ireland

account holders work away from home

confidentiality, particularly for customers residing in Northern Ireland.

Most financial institutions allow customers to operate ‘care of’ addresses. A ‘care of’ address is a risk indicator of an Irish resident operating a non-resident account. It was generally felt that the risk was heightened if this address was that of the branch and some institutions made it clear that they would prefer not to accept such addresses. However, there are potential bona fide reasons for such addresses, for example, customer travel or working arrangements.

To take account of these considerations the results note the number of accounts for which a financial institution believes it has sufficient evidence to confirm non-resident status.

In certain cases an account contained more than one reportable risk indicator. In these instances it has only been reported once. However, where an account had both declaration compliance errors and risk indicators, this account will be included in the results for both categories. The number of accounts containing both types of exception is also shown in the results.

Fieldwork - Subsequent Work

Following the completion of the branch review, a fieldwork report was prepared which listed all exceptions and anomalies noted for each category of account. In addition, for branches or institutions where reclassified and closed accounts were reviewed, a fieldwork report was prepared which listed the findings in respect of the reasons for reclassification and the method of account closure, where these could be identified.

These reports were then sent to the personnel identified by the institution as the appropriate person to receive such reports. Each institution or branch thereof was then given a period of a minimum of five working days to provide any additional documentation or clarification which would enable any apparent exceptions to be resolved e.g. by providing a copy of a declaration which could not be located at the time of the branch visit. The Auditor also requested that a copy of the fieldwork report be signed and returned with any response provided.

All institutions provided additional information which eliminated a number of apparent exceptions and a revised schedule of exceptions was the subject of further confirmation with the institutions. Institutions may have undertaken further research which, if additional time had been available, may have led to a reduction in the level of reportable accounts.

Examination Findings

The detailed findings in respect of each institution are set out in the chapters which report my findings on their administration of DIRT and non-resident exemptions. Set out below are the overall findings of the audit examination by the auditor appointed by me.

Documentary Compliance - Overall Findings

Non-Resident Declarations (Forms 37)

In general, the audit found that across all financial institutions, the level of missing non-resident declarations, at 1.5%, was low. However, this must be set against the quality of declarations held since some reportable deficiency existed in over a quarter of the forms. A reportable deficiency would, prima facie, render the declaration invalid and could at the instance of Revenue render the underlying account liable to DIRT. The results may be summarised as follows

Documentary Compliance - Declarations of Non-Residence (Form 37)

Numbers

Percentages

Sample

8,377

100%

No Declaration held

125

1.5%

Late dated Declarations

Undated Declarations

Declaration Validity Exceptions

1,235

429

498

14.8%

5.1%

5.9%

This pattern was reasonably consistent across all banking sectors. The results for each sector were as follows.

Documentary Compliance - Declarations of Non-Residence (Form 37)

Clearing Banks and their Subsidiaries

State

Institutions

Building Societies

and former

Building Societies

Other

Institutions

Total

Sample

3,851

1,117

2,553

856

8,377

No Declaration held

33

24

45

23

125

Late dated Declarations

Undated Declarations

Declaration Validity Exceptions

747

122

146

157

49

73

185

201

218

146

57

61

1,235

429

498

Total Declaration Exceptions

1,048

303

649

287

2,287

Percentage Exception

27%

27%

25%

34%

27%

Figure 8.1 sets out the documentary compliance results in graphical form in respect of the administration of non-resident accounts by the financial institutions examined.

Figure 8.1

Non-Resident Documentary Compliance

% Declaration Deficiencies

Proportion of Cases Where no Declaration Existed

Risk Profile of Non-Resident Book

Overall 18% of the non-resident book sampled had factors indicative of a risk of resident status. The prevalence of such factors varied from a level of 12 - 13% in the case of building societies and State banks to 35% in the case of the Other Institutions category.

Of the accounts with factors indicating a risk of resident status, nearly half also had deficiencies in the completion of their declarations.

Authenticity Risk Profile - Non-residents
Bank

Clearing Banks and

their Subsidiaries

State

Institutions

Building Societies

and Former

Building Societies

Other

Institutions

Total

Percentage

Sample

3,389

1,117

2,636

526

7,668

Risk Indicator
Irish Address

166

61

62

37

326

4.3%

PO Box or ‘Care of’ Address

133

31

92

85

341

4.5%

Hold Mail

290

23

75

59

447

5.8%

Transaction Profile

75

4

20

3

102

1.3%

Accounts with Liens

31

0

6

0

37

0.5%

Other

20

23

49

2

94

1.2%

Total accounts with risk indicators

715

142

304

186

1,347

17.6%

Percentage with risk factors

21.1%

12.7%

11.5%

35.4%

17.6%

Accounts also with documentary exceptions

349

60

131

95

635

47%

Accounts where financial institutions believe they hold appropriate evidence of non-residence

478

83

72

103

736

55%

Verification of Principal Place of Residence

Testing was undertaken in respect of a sample of addresses shown on declarations reviewed during the audit fieldwork. The principal place of residence of the interest beneficiary, provided by customers on the declarations of non-residence, were sampled and their validity checked. A sample of 214 UK addresses was compared to databases accessed through Experian, a customer credit reference service, which is derived from the register of voters and Post Office data sources, in addition to County Court records and credit listings.

The principal place of residence was classified as invalid where the address was not found on the Post Office address database or where the supporting databases indicated that the beneficiary had never been registered at the declared address in the period from 1986 to date.

Following this review 156 (73%) of the sampled addresses were classified as valid. In these cases the register of voters indicated that the customers were currently registered at the address or had been registered at some point since 1986 ( the earliest date for which records are held) and had subsequently moved away.

In regard to the other 58 declaration addresses sampled, which represent 27% of the total, the failure to identify the declarant as resident at that address, while not conclusive proof of the existence of bogus accounts, does indicate a strong possibility that the non-resident status may not be genuine for a proportion of the non-resident books of financial institutions.

Reclassifications

The Auditor noted from her enquiries that internal reviews of non-resident account status were undertaken at various times by individual institutions. These reviews generally resulted either in the receipt of an appropriate declaration or the reclassification of the account. It was noted that a number of accounts were subsequently returned to non-resident status when the relevant documentation was received.

The determination of the cause of reclassifications for a sample of accounts was inconclusive and no conclusions could be drawn in 38% of cases examined.

Factors which caused this included

The fact that the exact date of reclassification could not always be determined

Very few institutions were able to provide confirmation of the reason for reclassification of an account other than where it occurred as the result of a specific review. Very few institutions retain copies of customer correspondence and therefore no trail exists to verify whether account holders had notified the branch of a change of residence. Consequently, it was only possible to determine the date from which an account should have been reclassified for a very small minority of accounts.

It has been assumed that most changes occurred as a result of staff action, although only rarely could any conclusion be reached as to the event which prompted such action.

It was also noted that an account may be reclassified automatically as DIRT-paying as part of the normal closure procedure. Accordingly, a number of such accounts were automatically included in the sample to be tested even though the reclassified status arose as a result of closure. In addition, it would appear that some accounts may be reclassified by an institution after a specific period of dormancy.

Some institutions were unable to confirm the total number of reclassifications in each period.

In respect of the 62% of cases for which conclusions could be drawn the following was the cause of reclassifications effected in the period 1990 - 1998

Cause

Percentage of Cases

Bank Reviews - formal

Bank Initiated - ongoing vigilance

Customer initiated change

Revenue Inspection

Other

25%

24%

11%

1%

1%

Bank reviews culminating in reclassifications peaked in the period 1990-92. In this period such reviews accounted for two-thirds of all reclassifications.

Account Closures

Where available, the total number of accounts closed in each period was obtained. Risk indicators were included in the fieldwork report in respect of closed accounts where noted.

It was not possible to draw any conclusions from the method of closure noted due to the limited information available in respect of closed accounts. The date of closure was not found to be a relevant indicator as to the reason for closure. "Withdrawal" was the most common method, but no differentiation was made between cash or bankers' draft transactions. Closures marked as transfers could have been to any type of account.

It was noted that a number of institutions flag an account as closed after it has been dormant for a specific period of time and consequently the method of final withdrawal of the balance could not be ascertained.

Other DIRT Exempt Accounts

In the case of companies subject to Irish Corporation Tax, Pension Funds and Charities interest on deposits is exempt from DIRT on the basis of a declaration as specified in the Taxes Consolidation Act 1997. A sample of these declarations was examined with the following results.

Documentary Compliance - Other Exempt Accounts

Clearing Banks and

their

Subsidiaries

State

Institutions

Building Societies

and Former

Building Societies

Other

Institutions

Total

Percentage

Sample

518

225

114

240

1097

No Declaration

20

15

9

59

103

9%
Declaration Exceptions

82

41

32

59

214

20%
Total Exceptions

102

56

41

118

317

29%

At 9% the level of missing declarations was high.

Figure 8.2 sets out in graphical form the results of documentary compliance findings in respect of other exempt accounts for the financial institutions examined.

Figure 8.2

Documentary Compliance - Other Exempt Accounts

% Declaration Deficiencies

Proportion of Cases Where No Declaration Existed

Special Savings Accounts

A lower rate of DIRT is chargeable on SSAs. In order to open such accounts a declaration must be completed. The results of the audit of documentary compliance in the case of SSAs was as follows.

Documentary Compliance - SSAs

Clearing Banks &

their

Subsidiaries

State

Institutions

Building Societies

and Former

Building Societies

Other

Institutions

Total

Percentage

Sample

302

154

96

108

660

No Declaration

10

0

1

1

12

2%

Declaration Exceptions

6

6

10

13

35

5%

Total Exceptions

16

6

11

14

47

7%

The appointed Auditor found that, in general, compliance with the conditions attaching to such accounts, including notice periods and investment limits, was good. Most institutions had some form of systems control to monitor compliance with such conditions.

Interest Reporting

Before November 1993 when the Revenue gave a concession that a non-resident declaration (Form 37) would also serve to exempt interest on non-resident accounts from any requirement to report it to Revenue, it was necessary to hold a Form F for this purpose. The audit identified a sample of 3,639 non-resident accounts being operated prior to the concession and examined the extent to which Form Fs were held. Such forms were located in 43% of cases. The results are as follows

Interest Reporting Exemption
Documentary Exemption (Form F)

Clearing Banks & their

Subsidiaries

State

Institutions

Building Societies

and former

Building Societies

Other

Institutions

Total

Sample

2,143

392

818

286

3,639

No Form

642

376

818

221

2,057

Percentage Exception

30%

96%

100%

77%

57%

A number of institutions believed that the completion and retention of Form F was not required, and this view was apparently endorsed by the fact that interest reporting returns were never requested. There was also a widespread belief that the non-resident declaration replaced Form F or removed the requirement for Form F to be held. Building societies which had operated the Composite Rate Tax scheme prior to the introduction of the DIRT scheme, advised that they had received no guidance regarding the necessity to obtain the forms.

Affidavits

Institutions did not appear to make use of the affidavit, even when an account was opened by one individual on behalf of another.

General Findings of the Appointed Auditor

Declaration Form Layout

The various forms required were noted as being complicated in layout and likely to lead to completion errors, although there was a noticeable improvement in the completion of the new style forms which came into use in 1997. Variations from the prescribed declaration format were not reported unless there were any inadequacies in the information requested.

Foreign Currency Account Declarations

The legislation concerning foreign currency accounts changed twice following the introduction of DIRT in 1986 - the first time in 1991 (Section 11 Finance Act 1991) and the second time in 1993 (Section 3 Finance [No 2] Act 1992). The legislation with regard to foreign currency accounts can be viewed as ambiguous and it was not appreciated by the institutions until some time after 1 January 1993 that all foreign currency accounts (irrespective of the date on which they were opened) would require a Form 37 from 1 January 1993.

Attitude to Documentary Compliance

From site visits performed and discussions held with the institutions, it would appear there has been a widespread lack of understanding regarding the full compliance requirements for DIRT exemptions. In particular there is a belief that the eligibility as advised by the customer matters more than the holding of a written declaration evidencing that eligibility.

Misclassification of Accounts

The coding of exempt accounts was not uniform and led to a significant level (27%) of misclassifications. The samples tested were deliberately less than for non-resident accounts, on the basis of risk assessment and although only 9% represented missing declarations, we found 20% other exceptions (principally late). Compliance with the regulations governing Special Savings Accounts was generally good, only 2% missing and 5% other exceptions.

Auditor’s Suggestions

The Auditor noted that the results of this exercise were not surprising given her experience in acting on reviews of compliance of the tax deduction scheme for interest that operates in the UK. Historically, many UK deposit takers faced similar problems to those existing in Ireland, prior to the introduction of full Inland Revenue audit powers. Drawing on the lessons learned in the UK, the Auditor reflected that

Guidance Notes, drawn up by Revenue, perhaps after consultation with the industry, would be useful in clarifying Revenue expectations of compliance. The Guidance Notes might advise on procedures that an institution could adopt to validate an incomplete declaration. For example, date stamping all declarations on receipt, could be regarded as validating the date of a declaration received undated. A further example could be annotation by the institution of an account holder’s other account number(s) on the declaration.

Where risk factors appear or are known at the outset, the institution might be required to complete a "deposit taker’s certificate" regarding the declaration, or make a file note of the validating evidence held.

Institutions might be encouraged to appoint an officer responsible for compliance in this area and to implement a programme of training of staff.

Whilst the Revenue now have full powers to audit institutions’ compliance, knowing the importance to the financial community of customer confidentiality, institutions might be offered the opportunity of appointing, at their own cost, an independent auditor, approved by the Revenue, to audit their compliance, perhaps on a cyclical basis.

Analogous UK Scheme

The Auditor informed me, in respect of a comparable tax heading in the UK, that the following situation pertains

Current audits of the Tax Deduction at Source from Interest, (TDSI), scheme rarely find an error rate exceeding 5% and frequently full compliance is found. Inevitably human error gives rise to individual lapses, such as the odd undated declaration, or failure to take a new declaration on the occasion of a new name being added to the account.

However, when the audits first commenced in 1989-90, the findings were considerably different. Missing or old style declarations, (UK institutions could rely upon a pre-existing section 17 declaration from 1984 to 1988), frequently gave rise to error rates of 20% to 30%. Initially, the Inland Revenue accepted "care of" and Post Office box addresses as valid, but gradually tightened their interpretation of a valid address, such that now a description of the account holder’s residence is the only acceptable address.

If the percentage of missing, late or invalidly completed declarations exceeded 5% of the sample tested, the Inland Revenue would look to extrapolate the results across the whole population of deposits, unless the deposit taker undertook a complete review and offered a settlement based on actual liability. If the error rate was less than 5% and more than 50% of the errors were regarded by the Inland Revenue as "clerical errors", no extrapolation would be sought. Other factors, such as the Inland Revenue’s confidence in the statistical validity of the results, would also influence a decision on extrapolation.

If declarations cannot be located at the time of audit but are produced subsequently, the Inland Revenue would take into account the explanation for their initial absence in determining how long the institution might have to locate them. There may have been branch mergers, errors in archiving, misfiling, etc. In general, a time is mutually reached when the institution believes it is not cost-effective to search any longer.

The Inland Revenue seek tax on any interest paid on a deposit, if at the time of paying the interest, the deposit taker was not in possession of a validly completed declaration evidencing entitlement to gross interest. Even if the deposit taker subsequently obtains a declaration which confirms that the account has always been non-resident, the Inland Revenue would look to recover tax from the deposit taker, and this tax would then only be repaid on a claim by the account holder.

The Inland Revenue’s uncompromising policy regarding tax liability in the event of deposit takers’ failures has led to the present high level of compliance. The burden on institutions has been great and many have chosen to transfer large numbers of accounts to subsidiary companies located outside the UK, principally the Channel Islands, rather than bear the cost of compliance. However, some institutions have organised system controls successfully and maintain their full deposit base in the UK.

For information or problems with these pages please contact : webmaster@audgen.irlgov.ie

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