This judgment text has undergone conversion so that it is mobile and web-friendly. This may have created formatting or alignment issues. Please refer to the PDF copy for a print-friendly version.
AQP
v
Comptroller of Income Tax
[2011] SGHC 229
| Case Number | : | Income Tax Appeal No 1 of 2010/Y |
| Decision Date | : | 17 October 2011 |
| Tribunal/Court | : | High Court |
| Coram | : | Tay Yong Kwang J |
| Counsel Name(s) | : | Nand Singh Gandhi (Allen & Gledhill LLP) for the Appellant; Julia Mohammed (Inland Revenue Authority of Singapore) for the Respondent. |
| Parties | : | AQP — Comptroller of Income Tax |
[LawNet Editorial Note: In the appeal to this decision in Civil Appeal No 139 of 2011, the Court of Appeal, on 16 January 2013, ordered the proceedings to be remittted to the Income Tax Board of Review. See [2013] SGCA 3.]
17 October 2011 | Judgment reserved. |
Tay Yong Kwang J:
Introduction
1 This is an appeal concerning the question whether losses caused to a company by a fraudulent director (“the Ex-MD”) are deductible for income tax purposes under section 14(1) of the Income Tax Act (Cap 134, 2008 Rev Ed) (“the Act”). The Ex-MD had misappropriated company funds and caused the appellant to incur a loss of $12,272,917 (“the Loss”). A few years after the Loss, the appellant lodged a claim with the respondent seeking relief for the Loss under section 93A of the Act. The respondent rejected the appellant’s claim.
2 After hearing the parties and considering the evidence and the submissions, I now dismiss the appeal. I will set out the reasons for my decision after the outline of the relevant facts and the decision from which this appeal arises.
Facts
Background
3 The appellant is a company incorporated in Singapore. It began as a sole proprietorship established by the founder, Mr B (“the Founder”) in 1956. It was initially involved in the trading of re-conditioned bearing products before becoming distributors of bearings and seals. In 1973, the sole proprietorship was incorporated as a company. On 10 November 1995, it was listed on SESDAQ. On 2 February 1998, it was upgraded to the SGX Main Board.
4 Among the children of the Founder who took over the management of the appellant was the youngest son Mr C (“the Ex-MD”). The Ex-MD entered into a service agreement with the appellant to hold office as Managing Director for three years commencing on 20 October 1995. The service agreement was renewed for a further term of three years from 20 October 1998. At all material times, he also served as a member of the Board of Directors.
5 On 1 December 1999, the Ex-MD was dismissed as both Director and Managing Director for misappropriation of the company’s funds as revealed by the investigations of the Commercial Affairs Department (“CAD”). The Ex-MD was charged and tried in the District Court.
The conviction of the Ex-MD
6 In Public Prosecutor v Kwek Chee Tong in DAC 48461/99 (“PP v KCT”), the District Judge (“the DJ”) convicted the Ex-MD of 24 charges of criminal breach of trust under section 409 of the Penal Code (Cap 224, 1985 Rev Ed) and sentenced him to a term of nine years in prison.
7 During the trial, it was revealed that the Ex-MD’s modus operandi was to make out false purchase orders to the appellant’s suppliers for the purchase of bearings, the appellant’s stock in trade. Based on these false purchase orders, several cheques were issued to him or his nominees on his claim that he had advanced money from his personal account to the appellant to make the purchases. He also falsely claimed that he had made loans to the appellant’s customers for their purchases against which he reimbursed himself from the appellant’s funds.
8 In reality, the Ex-MD used the misappropriated company funds to pay for his gambling debts and for personal use (at [151]). The DJ found at [284] that:
The evidence revealed that no one questioned [the Ex-MD’s] instructions. There was total trust reposed in the accused by virtue of his senior management position. In evidence he said that he need not have to tell anyone about, and (as a fact) he had complete control over, the usage of [the appellant’s] funds. He had access to millions of dollars as [the appellant] was a public listed company at the material time.
9 The DJ also considered the following aggravating factor at [288]:
There was another aggravating factor: the blatant way in which the monies were siphoned from the company, some in the form of cash cheques, others as company cheques, made out directly in the names of or given to, junket operators, or individuals with no trade dealings with [the appellant]. In most instances the company’s funds were used to repay the accused’s gambling debts and feed his gambling habit. The effect on the victim company (as well as the financial market) would be an erosion of confidence.
The actions leading to the current proceedings
10 After the Ex-MD’s misappropriation came to light, the appellant made provisions for doubtful debts of $12,410,141 inclusive of the Loss in its statutory accounts under Extraordinary items for the year ended 31 December 1999. However, no claim for deduction for the Loss was made for the Year of Assessment 2000 (“YA 2000”).
11 In 2003, the appellant instituted legal proceedings against the Ex-MD for the amount misappropriated and obtained judgment against him. However, recovery proved fruitless and the Ex-MD was subsequently adjudged bankrupt.
12 On 15 December 2005, the appellant lodged an “error or mistake” claim for the Loss under section 93A of the Act with the respondent. By a letter dated 1 December 2008, the respondent made a determination that relief could not be granted “as there is no error or mistake within the meaning of Section 93A of [the Act]”.
13 The appellant then filed a Notice of Appeal to the Income Tax Board of Review (“the Board”) on 5 December 2008. This was followed on 30 December 2008 by the Petition of Appeal seeking an order (among other things) to direct the respondent to grant relief to the appellant under section 93A of the Act and to allow the Loss as a deduction for YA 2000 under section 14(1) of the Act.
The decision of the Income Tax Board of Review
14 The Board referred to the English approach in Curtis (HM Inspector of Taxes) v J & G Oldfield, Limited (1925) 9 TC 319 (“the Curtis test”) and came to the conclusion that the Loss sustained by the appellant was not deductible under section 14(1) of the Act. The appellant’s appeal was therefore dismissed.
15 In applying the Curtis test to the facts of the present case, the Board took particular notice of the power or control the Ex-MD had wielded in the appellant at [17]:
Besides being the Managing Director the Ex-MD was a substantial shareholder. Both directly and through his family company D Pte Ltd his interest or deemed interest ranged between 12.8% to 14.9%. This is a substantial shareholding. As a substantial shareholder and Managing Director, and based on the evidence, the Ex-MD was in the language of Rowlatt J supra “in a position to do exactly what he likes”.
16 The Board reasoned that the Loss was not deductible since the Ex-MD was “in the same position as the managing director in the Curtis case”. The Board further held that the Curtis test (as understood and applied by the Board) is consistent with subsequent cases such as The Roebank Printing Company Limited v The Commissioners of Inland Revenue (1928) SC 701, 13 TC 864 (“The Roebank case”) and Bamford (HM Inspector of Taxes) v ATA Advertising Ltd (1972) 48 TC 359 (“Bamford”).
17 The Board then went on to opine on the issue of whether the appellant’s omission to set off the Loss in YA 2000 could qualify as an “error or mistake” under section 93A(1) of the Act, if the Loss were deductible. The Board held that the omission by the appellant was not “due to oversight” as “it was a decision made after due consideration that the Loss was not an allowable deduction under section 14 of the Act” (at [28]). However, the Board agreed with the appellant’s argument that “[i]f the decision was a mistake it was one of law and still a mistake falling within section 93A of the Act”. Nonetheless, the Board held that this was an academic point as it had already decided that the Loss was not deductible.
The issues on appeal
18 Two issues arise on appeal:
(a) Did the Board err in holding that the Loss incurred by the appellant was not wholly and exclusively incurred by the appellant in its production of income under section 14(1) of the Act?
(b) Did the Board err in holding that an erroneous opinion or a grossly negligent error, such as a mistake of law, can constitute an “error or mistake” under section 93A of the Act?
Deducibility of losses incurred as a result of defalcation
The Curtis test
19 Section 14(1) of the Act stipulates the deductions a taxpayer is allowed to make against taxable income:
For the purpose of ascertaining the income of any person for any period from any source chargeable with tax under this Act (referred to in this Part as the income), there shall be deducted all outgoings and expenses wholly and exclusively incurred during that period by that person in the production of the income...
20 Section 15(1)(b) reinforces section 14(1) by emphasizing that any other types of disbursements or expenses “not being money wholly and exclusively laid out or expended for the purpose of acquiring the income” are not allowed as deductions.
21 The statutory framework thus emphasizes that the touchstone for deductibility of expenses is that they must necessarily be incurred for the specific purpose of the production of income. The Court of Appeal in Pinetree Resort Pte Ltd v Comptroller of Income Tax [2000] 3 SLR(R) 136 stressed (at [47]) that section 14(1) “requires a nexus between the incurrence of the expense and the production of income.” [emphasis added]
22 The general requirement of a “nexus” in the relevant statutory provisions, however, does not in itself shed light on how losses sustained by a company due to the defalcation of an employee is to be treated. When would losses resulting from defalcation have sufficient nexus to be considered “wholly and exclusively incurred ... in the production of the income”?
23 In the seminal case of Curtis (HM Inspector of Taxes) v J & G Oldfield, Limited (1925) 9 TC 319 (“Curtis”), the managing director of the company was, up to the time of his death, in control of its business in wines and spirits. An investigation after his death revealed that many payments and receipts not relating to the company’s business but to his private affairs had passed through the company’s books. The amount was accordingly written off as a bad debt. A dispute then arose between the company and the Revenue as to whether the said loss had “arisen in the course of a company’s trading” and therefore tax-deductible. The General Commissioners allowed the company’s claim and the Revenue appealed.
24 The High Court reversed the General Commissioners’ decision and agreed with the Revenue that the loss was not a trading loss and thus not deductible from the company’s profits for income tax purposes. The relevant portion of Rowlatt J’s judgment, or the Curtis test, (at 330-331) is reproduced here:
When the Rule speaks of a bad debt it means a debt which is a debt that would have come into the balance sheet as a trading debt in the trade that is in question and that it is bad. It does not really mean any bad debt which, when it was a good debt, would not have come in to swell the profits. What the Commissioners have been misled by, in my judgment, quite clearly is this. They have allowed themselves to act under the impression that they were taxing the Company on what the Company in a loose way had made and secured. In point of law they were engaged in assessing the profits of the Company’s trade, not of the Company itself but of the Company’s trade, and I have to consider whether there is the least ground for supposing that losses of these sums resulting in this bad debt were losses in the trade. I quite think ... that if you have a business (which for the purposes of to-day at any rate I will assume) in the course of which you have to employ subordinates, and owing to the negligence or the dishonesty of the subordinates some of the receipts of the business do not find their way into the till, or some of the bills are not collected at all, or something of that sort, that may be an expense connected with and arising out of the trade in the most complete sense of the word. But here that is not this case at all. This gentleman was the Managing Director of the Company, and he was in charge of the whole thing, and all we know is that in the books of the Company which do exist it is found that moneys went through the books into his pocket. I do not see that there is any evidence at all that there was a loss in the trade in that respect. It simply means that the assets of the Company, moneys which the Company had got and which had got home to the Company, got into the control of the Managing Director of the Company, and he took them out. It seems to me that what has happened is that he has made away with receipts of the Company dehors the trade altogether in virtue of his position as Managing Director in the office and being in a position to do exactly what he likes.
25 In the present case, the crux of the dispute between the appellant and the respondent is as to how the Curtis test is to be understood and applied. The appellant argued that in Curtis, the “real reason” why the loss was not deductible was because the defalcations took place outside the company’s trading or income-earning activities, since the money had already been earned and was simply passed through the company’s books into the pocket of the director.[note: 1] The respondent, however, understood the Curtis test as prohibiting the deduction of losses incurred by a director who was “in a position to do exactly what he likes”.[note: 2]
26 Since both parties have submitted extensively on foreign authorities in support of their arguments, it is necessary that I first set out how various Commonwealth jurisdictions have dealt with the issue of the deductibility of losses sustained as a result of an employee’s defalcation. I will then discuss the approach which I think our courts should adopt.
The approach in different jurisdictions
United Kingdom
27 In The Roebank case, the managing director was entitled to a commission on the company’s profits at the end of each year and was authorised by the directors to draw sums during the year in anticipation. He drew in excess of commissions due. The company then claimed tax deduction on the basis that these payments were made as remuneration of an employee for the purpose of and in connection with its business and therefore deductible in arriving at the profits of the company for income tax purposes.
28 The General Commissioners applied the Curtis test and refused the deductions claimed. The company’s appeal to the Court of Sessions was dismissed. In dismissing the appeal, Lord President Clyde expressed his understanding of the Curtis test at 875-876 in the following manner:
“If it were legitimate for us to make an inference of fact from the facts stated in the Case, I should be disposed to think that the managing director was using the Company as his banker. But it is, I think, enough to say that the financial indulgence shewn to him by the Company went far beyond anything which could be justified by, or could be consistent with, any trading interest of the Company. If so, then the loss resulting from the granting of that indulgence cannot form a legitimate or proper deduction in the ascertainment of the Company’s trading profits.”
29 Both Curtis and The Roebank case were applied much later in Bamford. In Bamford, the taxpayer’s trade consisted of advertising and publicity work. It had a board of three members including a Major A.E. Newnham (“Newnham”) who held a 50% shareholding in the company. The financial matters were left in the hands of Newnham. In 1966, Newnham misappropriated £15,000 by drawing a cheque on the account of the tax payer in favour of another company which he also controlled. In the tax payer’s accounts for the following year the £15,000 was written off as a bad debt. However, the tax payer was assessed on the basis that the £15,000 was not deductible.
30 The tax payer appealed. The General Commissioners decided in favour of the tax payer. However, on appeal by the Inspector of Taxes, Brightman J reversed the decision of the General Commissioners on the following grounds at 368:
Mr. Beattie [counsel for tax payer] submitted that there is no logical distinction to be drawn between petty theft by a subordinate employee and massive defalcation by a director. In my view there is a distinction. I can quite see that the Commissioners might find as a fact that a £5 note taken from the till by a shop assistant is a loss to the trader which is connected with and arises out of the trade. A large shop has to use tills and to employ assistants with access to those tills. It could not trade in any other way. That, it seems to me, is quite a different case from a director with authority to sign cheques who helps himself to £15,000, which is then lost to the company. I find it difficult to see how such a loss could be regarded fairly as “connected with or arising out of the trade”. In the defaulting director type of case, there seems to me to be no relevant nexus between the loss of the money and the conduct of the company’s trade. The loss is not, as in the case of the dishonest shop assistant, an incident of the company’s trading activities. It arises altogether outside such activities. That, I think, is the true distinction. In my view, the decision in the Roebank case did not depend on the absence of fraud. I consider that it is a decision which I should follow.
31 In his learned judgment, Brightman J appears to have settled once and for all what the “the true distinction” is when applying the Curtis test. However, as will be evident in the cases in the other Commonwealth jurisdictions, there is still much to be said on what “the true distinction” really is between defalcations which are deductible for income tax purposes and those which are not.
Australia and New Zealand
32 The High Court of Australia also referred to Curtis in The Commissioner of Taxation (New South Wales) v Ash (1938) 61 CLR 263 (“Ash”). In Ash, the taxpayer was a solicitor who had to pay off certain debts which were incurred due to the fraudulent misappropriation of a former partner. Applying Curtis, Latham CJ highlighted what in his opinion was the salient fact that the defalcating partner had been “acting within his authority as a partner [and had] received moneys as a principal for the firm” (at 274). This led the learned CJ to hold that the losses sustained by the taxpayer were not deductible, because (at 278):
You cannot treat ... the depredations of a partner as if they were the peculations of an office boy ... The partner was a proprietor, and whilst all must sympathize with the taxpayer and deplore the wrong done to him by this partner it is impossible to treat that wrong as a characteristic incident of the carrying on of his profession...
33 Ash has been cited with approval in DR & P v Comptroller of Income Tax (2001) 5 MSTC 293. The Singapore Income Tax Board of Review understood Ash as establishing the proposition that if the fraud was perpetrated by a partner as a partner, no deduction would be allowed for the loss sustained due to the fraud.
34 When faced with the misappropriation by a book-keeper (as opposed to a partner) of a solicitor taxpayer, the New Zealand court in Commissioner of Taxes v Webber [1956] NZLR 552 (“Webber”) had no difficulty distinguishing Ash. Turner J held that “[t]he case in which income is actually received and then misapplied by the proprietor of a business is expressly distinguished ... from the case where a payment has to be made for a regular and almost unavoidable incident of the business in question” (at 561). The book-keeper had made false representations to the taxpayer leading to the latter advancing a sum of money to the former. In Turner J’s opinion, this type of loss represents “that kind of casualty, mischance or misfortune which is a natural or recognized incident of the carrying on of this solicitor’s kind of practice” (at 562) and was therefore held to be deductible.
35 In W G Evans & Co Ltd v Commissioner of Inland Revenue [1976] 1 NZLR 425 (“Evans”), the defalcator, who held 100 out of 10,000 shares and was employed as the company’s secretary, was subsequently appointed a director. The controlling shareholder, Mr E, was away from office most of the time and had the habit of forwarding cheques and accounts to the defalcator for banking and payment. Unknown to Mr E, the defalcator was siphoning money from the company for five years. Casey J held that the loss sustained due to the defalcations was a deductible loss.
36 In distinguishing Curtis, Casey J held that the defalcator “was in nothing like the position of the managing director in Curtis’s case who could exercise an overriding control to take the money independently of the company’s normal operations. Although a director, [the defalcator] was nothing more than the company’s accountant so far as these defalcations were concerned” (at 432). Importantly, it was in this context that Casey J eventually concluded at 435:
The fact that he was also a director, shareholder and officer of the company does not alter the fact that he misappropriated the money while dealing with it as part of the company’s activities, and not by the exercise of overriding power or control outside those activities altogether, as did the sole managing director in Curtis’s case. The risk of such defalcations was inherent in the operations of the company carried on by necessity in this way, and accordingly the resulting loss is fairly incidental to the production of the assessable income and is deductible. [emphasis added]
Canada
37 Similar to Evans, the Curtis test was also distinguished in the Canadian case of Cassidy’s Ltd v MNR 89 DTC 686 (“Cassidy”). In Cassidy, the defalcator, a chartered accountant, was the secretary-treasurer of the company and was responsible for accounting procedures. He was subsequently promoted to vice-president and general manager. It was revealed that the defalcator had issued cheques from the taxpayer company (“Packer”) to his own bank account, ostensibly for the company’s operations but in reality for payment of his personal debts.
38 In allowing these losses to be deductible, the Tax Court of Canada first decided against the proposition that “simply because a theft or defalcation was committed by a senior employee, the losses resulting from such a commission are not deductible by the employer in computing income for tax purposes” (at 691). From its understanding of the Curtis test and subsequent cases, the court abstracted the general principle that “case law distinguishes not so much as to the level of the employee in the employer’s hierarchy but how and at what stage in the income-earning process the money is stolen or embezzled” (at 692).
39 In determining “how and at what stage in the income-earning process the money [was] stolen or embezzled”, the court made the following observation at 693:
[The defalcator] commenced defrauding Packer in 1978 when he was controller and before his appointment in October, 1981, to vice-president and general manager of Packer. When he began his theft of funds he did not have “complete discretion” over the operations of Packer and was responsible to [the president of Packer]. Even after his appointment as a general manager he was obliged to report monthly to, and follow rules determined by, his superiors. Packer’s business was controlled by [the defalcator’s] superiors. This is not a case like Curtis ... where the person defrauding his employer was in sole control of the employer’s business.
40 Following from the above observation, the Canadian Tax Court concluded that because the defalcator “did not defalcate money qua director or general manager of Packer” (at 693), the losses sustained by Packer were tax-deductible.
41 Just a few months later, the Federal Court of Canada sought to apply Cassidy to the facts before it in Parkland Operations Limited v The Queen 90 DTC 6676 (“Parkland”). The company had an operating line of credit with a bank and a “signature card” had to be signed to draw down the credit. The court found as a finding of fact that it was understood that two signatures – of the six shareholders – were required on the “signature card”: one from either of two ‘primary’ shareholders and the other from any of the four ‘new’ shareholders. However, illegitimate withdrawals were soon made by two of the ‘new’ shareholders without the knowledge or approval of the rest.
42 The court referred to Evans and Cassidy in coming to its decision that the losses incurred by the two defalcating shareholders were deductible. The court held at 6680:
[The defalcating shareholders] may have been minority shareholders for Parkland, but they misappropriated the funds in question not in the capacity as shareholders, but rather as thieves, with neither the knowledge nor consent of the other shareholders. They misappropriated the money while dealing with it in the course of the company’s activities, and not by exercising some overriding control over the funds which existed outside of those activities...The taxpayer is entitled to the same relief where a minority shareholder, oblivious to the plans or desires of the other shareholders, misappropriates funds as he would be where a senior employee was the thief. [emphasis added]
43 It must be noted that the fact that the defalcating shareholders had acted “with neither the knowledge nor consent of the other shareholder” is only relevant because of the specific factual matrix in Parkland. The control given to the defalcating shareholders with regard to their wrongful act had been circumscribed right from the start. Parkland was not a case where the defalcating shareholders could draw down the credit as and when they liked and then use the money for their own illegitimate purposes. The fact that permission from other shareholders had always been required before the drawing down of credit adequately explains why the defalcating shareholders were held not to have “[exercised] some overriding control over the funds” which they had misappropriated.
India
44 The Curtis test has also been extensively discussed and applied in India. In Badridas Daga v The Commissioner of Income-Tax (1958) 45 AIR 783 (“Badridas”), the defalcator was an agent of the appellant and was conferred extensive powers of management that included the authority to operate bank accounts. While acting under such authority, the agent withdrew money from the bank and used it for the discharge of his personal debts. The Supreme Court of India held that the losses sustained by the appellant were tax-deductible by discussing and then distinguishing Curtis.
45 The court first distinguished Curtis on the basis that the nature of the defalcating act with regards to the trade was different in the two cases (at 788):
Now, in [Curtis] ... the trading has ceased and [the] subsequent operations on the bank account are not incidental to the carrying on of the trade. But here, we are dealing with a banking business, which consists in making advances, realising them and making fresh advances, and for that purpose, it is necessary not merely to deposit amounts in banks but also to withdraw them. That is to say, a continuous operation on the bank account is incidental to the conduct of the business.
46 Interestingly, the court also distinguished Curtis on the basis that the power or control the defalcating employee possessed was different in the two cases (at 789):
[In Curtis, the defalcator’s] withdrawals would be more like a partner overdrawing his account with the firm than an agent embezzling the funds of his employer, and it could properly be held that such overdrawing has nothing to do with the trading activities of the firm ... When once it is established that [the defalcating agent] had authority to operate on the bank accounts, and that he withdrew the moneys in the purported exercise of that authority, his action is referable to his character as agent, and any loss resulting from misappropriation of funds by him would be a loss incidental to the carrying on of the business. [emphasis added]
47 The two different bases on which Curtis was distinguished above have led to two different conceptual views of the Curtis test, as evident in the appellant’s and respondent’s submissions in the present case. Is the Curtis test premised on the nature of the defalcator’s act with regards to the trade (i.e. whether it occurred as part of or wholly outside the trading activity), or is it premised on the power or control of the defalcator with regards to the defalcating act (i.e. whether the defalcator committed the act qua director)?
48 In Sassoon J David & Co (P) Ltd v Commissioner of Income-tax [1978] 98 ITR 50 (“Sassoon”), the defalcator was a director in sole charge of the firm’s business. He withdrew a sum of money from the firm’s bank account for his private purposes. The Income Tax Tribunal held that because the defalcator’s withdrawals were “more like a partner overdrawing his account with the firm”, the overdrawing had nothing to do with the trading activities of the firm and was thus not tax-deductible.
49 The High Court of Bombay reversed the decision of the Tribunal and held that such losses were deductible. Curtis was, once again, distinguished on the two different bases as in Badridas. The court first held that Curtis could be distinguished because the defalcator had “[withdrawn] the amount as if the withdrawal was for the purpose of the business of the company and then pocketed it” (at [23]). This parallels the “first basis” in Badridas where the court inquires whether the defalcator acted as if his act was part of the trade, or wholly outside it. However, the High Court of Bombay also stressed that the defalcator “could [not] be said to be in the position of a proprietor” with regards to the misappropriation (at [37]). This parallels the “second basis” in Badridas where the court inquires whether the defalcating employee could commit the defalcation only because of the overriding power or control given to him.
South Africa
50 The approach in South Africa is much more straightforward as the Curtis test is not followed there. The South African courts have held that all money lost by way of embezzlement cannot be deducted for income tax purposes because such losses are not incurred in the production of the company’s income (Lockie Bros Ltd v Commissioner for Inland Revenue (1922) 32 SATC 150 (”Lockie Bros”); Income Tax Case No. 298 (1934) 8 SATC 58).
51 It can be thus be said that the South African courts have interpreted the “nexus requirement” in its equivalent statutory provision – “incurred in the production of income” – very strictly and literally. As the “nature and scope of the [employee’s employment] was to manage the company’s business and not steal its funds” (Lockie Bros at 155), losses resulting from defalcation, by whomever and wherever committed, are not deductible for income tax purposes.
The appellant’s submissions
52 The appellant put forward three main arguments in support of its position that the Loss it has sustained should be deductible:
(a) The Commonwealth cases support the appellant’s understanding of the Curtis test – i.e. whether the misappropriation occurred in the course of the appellant’s normal income-earning activities and not outside them;
(b) The Board’s understanding of the Curtis test – i.e. whether the defalcator was “in a position to do whatever he likes” - is undesirable both legally and on policy grounds;
(c) Even if the Board’s understanding of the Curtis test is to be applied, the Ex-MD was not “in a position to do whatever he likes”.
The respondent’s submissions
53 In response, the respondent submitted:
(a) The Commonwealth cases have established the proposition that losses arising from embezzlement by a director or partner are not deductible;
(b) The appellant’s understanding of the Curtis test should not be adopted lest financial indulgence is seen as being condoned;
(c) The Ex-MD of the appellant was not merely an employee but was a person with overriding power or control similar to the defalcating director in Curtis.
The approach to be adopted in Singapore
54 Having surveyed the various approaches in different Commonwealth jurisdictions, I have come to the view that the correct understanding of the Curtis test was applied by the Board. The correct understanding of the Curtis test, in my view, is as follows: Did the defalcator possess an “overriding power or control” in the company (i.e. in a position to do exactly what he likes) and was the defalcation committed in the exercise of such “power or control”? If so, the losses which result from such defalcations are not deductible for income tax purposes.
55 I recognise that this test which I adopt is much narrower in allowing for the deductibility of defalcation losses as compared to the test submitted by the appellant. I find the appellant’s understanding of the Curtis test undesirably wide and I do not think that it should be applied in Singapore.
The Commonwealth cases are, on the whole, in greater support of the “overriding power or control” test
56 Admittedly, the Commonwealth cases discussed earlier appear to be somewhat inconclusive as to what the correct understanding of the Curtis test is. Both parties in the present case were able to find support for their respective understanding of the Curtis test in most of the cited cases. As my analysis of the Indian cases show (see above at [44] – [49]), some cases clearly employ both approaches (of the appellant and the respondent in the present case) in distinguishing Curtis.
57 Nonetheless, putting the Indian cases aside, I am of the view that the UK, New Zealand and Canadian cases provide greater support for what I have outlined as the correct understanding of the Curtis test – “Did the defalcator possess an “overriding power or control” in the company and was the defalcation committed in the exercise of such “power or control?” – as compared to that put forth by the appellant.
A. Why the UK authorities do not support the appellant’s position
58 In discussing the UK cases subsequent to Curtis, the appellant cited The Roebank case for the finding that the defalcations committed there “were not connected with the company’s trade”[note: 3] but missed the importance placed by Lord President Clyde on (at 876):
“...the financial indulgence shewn to [the defalcator] by the Company [which] went far beyond anything which could be justified by, or could be consistent with, any trading interest of the Company. If so, then the loss resulting from the granting of that indulgence cannot form a legitimate or proper deduction in the ascertainment of the Company’s trading profits.” [emphasis added]
59 As for Bamford, the appellant concluded that “the true distinction” as articulated by Brightman J was between a loss arising as an incident of the company’s trading activity and a loss arising altogether outside such activities.[note: 4] In my opinion, the appellant’s interpretation of Brightman J’s holding is too general to be of any practical use and it does not do justice to the nuanced context in which the learned judge saw the need to draw a “true distinction”. It will be recalled that Brightman J said (at 368):
I can quite see that the Commissioners might find as a fact that a £5 note taken from the till by a shop assistant is a loss to the trader which is connected with and arises out of the trade. A large shop has to use tills and to employ assistants with access to those tills. It could not trade in any other way. That, it seems to me, is quite a different case from a director with authority to sign cheques who helps himself to £15,000, which is then lost to the company. [emphasis added]
60 From the emphasized portion of the quote above, it seems that Brightman J was compelled to draw a distinction between different defalcations because of the necessity of certain trade practices (e.g. a large shop having to employ assistants) which, in his opinion, came with the natural and inevitable risks of an employee’s defalcation. The learned judge then held that such defalcations were to be treated as “arising as an incident of the company’s trading activities”, as compared to a “defaulting director type of case”. This nuanced context is, in my view, critical in understanding the “true distinction” drawn in the Curtis test (see below at [71] – [75]).
B. Why the New Zealand authorities do not support the appellant’s position
61 The appellant was also of the opinion that Evans supported its submission that the “true test” should simply be whether the defalcation occurs in the course of the normal income-earning activity of the taxpayer company.[note: 5] In my opinion, the appellant was too quickly drawn to Casey J’s conclusion that (at 435):
The fact that [the defalcator] was also a director, shareholder and officer of the company does not alter the fact that he misappropriated the money while dealing with it as part of the company’s activities, and not by the exercise of overriding power or control outside those activities altogether, as did the sole managing director in Curtis’s case.
62 This is because reading Casey J’s conclusion in a vacuum only begs the question as to what it means for the defalcator to have “misappropriated the money while dealing with it as part of the company’s activities”. Crucially, Casey J had held earlier (at 432) that the defalcator in Evans:
...was in nothing like the position of the managing director in Curtis’s case who could exercise an overriding control to take the money independently of the company’s normal operations. Although a director, [the defalcator] was nothing more than the company’s accountant so far as these defalcations were concerned. [emphasis added]
63 Reading Casey J’s judgment in totality, the operative phrase which the appellant’s case hinges on – “misappropriated the money while dealing with it as part of the company’s activities” – cannot be interpreted so widely as to cover any defalcation so long as it was committed as if it were part of the company’s legitimate activities. The court in Evans had held that the misappropriation was “part of the company’s activities” only because the defalcator was “nothing more than the company’s accountant so far as these defalcations were concerned”. For that reason, the precise power or control which the defalcator exercised in committing the defalcation is similarly the crucial factor in New Zealand.
C. Why the Canadian authorities do not support the appellant’s position
64 Once again, Cassidy might appear at first glance to be in support of the appellant’s submissions, given the Canadian court’s explicit rejection of the proposition that “simply because a theft or defalcation was committed by a senior employee, the losses resulting from such a commission are not deductible by the employer in computing income for tax purposes” (at 691).
65 However, the Canadian court also held that the defalcator in Cassidy did not have complete discretion over the taxpayer’s operations. The business was controlled by the defalcator’s superiors and the situation in Cassidy was “not a case like Curtis ... where the person defrauding his employer was in sole control of the employer’s business” (at 693).
66 Therefore, while the possibility exists that a defalcation committed by a senior employee such as a director could be deductible, the critical question is always, as the Canadian court accurately articulated, whether the director did “defalcate money qua director” (at 693). If he did so, the loss is not deductible. If he did not – e.g. the defalcators in Evans and Cassidy - the loss is deductible. This is consistent with the test I outlined earlier – did the defalcator possess an “overriding power or control” in the company and was the defalcation committed in the exercise of such “power or control”?
67 The appellant also cited Parkland for the proposition that any defalcation committed which is “oblivious to the plans and desires of the other shareholders” should be deductible.[note: 6] But as mentioned in the earlier discussion of Parkland (see above at [43]), the lack of approval of the other shareholders in Parkland was relevant in that case only because it went towards proving that the defalcating shareholders did not possess overriding power or control over the funds misappropriated. Otherwise, it is never a relevant inquiry, for the purposes of income tax deductibility, whether the defalcating act had “caused harm” to other shareholders or directors, as the appellant submitted.[note: 7]
68 Thus, it appears to me that most of the Commonwealth cases have understood the Curtis test as drawing the critical distinction, for the purposes of income tax deductibility, between defalcations by a mere employee and defalcations by a person who “in virtue of his position [of power or control was] in a position to do exactly what he likes” (Curtis at 331). This is the approach that has in fact been applied in Singapore (see DR & P v Comptroller of Income Tax (2001) 5 MSTC 293), and in my opinion, should continue to do so.
The approach to be adopted is legally sound and justifiable in policy
A. The appellant’s objections
69 Notwithstanding the supporting authorities, the approach which I intend to apply has been challenged by the appellant as being “difficult, if not impossible, to apply as guidance to future cases”.[note: 8] The appellant has argued that the “overriding power or control” test is vague and imprecise[note: 9] as compared to its understanding of the Curtis test (i.e. did the loss arise in the course of the company’s normal trading activities?).[note: 10] The appellant then concluded its critique with a probing question – “why should deductibility be governed by the degree of control or the size of shareholding?”
70 I respectfully disagree with the appellant’s arguments. In the paragraphs that follow, I will attempt to show the reasons why I think the approach I intend to apply is both legally sound and justifiable in policy.
B. The deductibility of losses resulting from defalcations is a common law exception
71 I refer to my earlier observation (see above at [22]) that neither section 14(1) nor section 15(1)(b) of the Act explicitly addresses the issue of whether losses which result from defalcations should be deductible. The general requirement that there must be a “nexus” between the alleged expense and the production of income also does not explain how and why defalcation losses ought to be treated as having a “nexus” with the production of income. Strictly speaking, and as actualised in the South African courts, one could maintain that all losses which result from defalcations should not be deductible since there is no connection at all between a defalcation and the production of income.
72 The Curtis test should therefore be understood as a common law exception developed by the courts to render certain defalcation losses as having sufficient “nexus” with the production of income such that they could be deductible for income tax purposes. This common law exception was perhaps developed in tandem with the rise of larger partnerships and companies in the early 20th century. It is essentially a judicial recognition of the increasingly difficult task for the large, modern firm to have to monitor every activity of its lower echelon employees and to ensure that all their acts necessarily contribute to the “production of income”.
73 The difficulty of doing so must have, at some point, prompted the courts to recognise, as in Curtis (at 330), that:
...if you have a business (which for the purposes of to-day at any rate I will assume) in the course of which you have to employ subordinates, and owing to the negligence or the dishonesty of the subordinates some of the receipts of the business do not find their way into the till, or some of the bills are not collected at all, or something of that sort, that may be an expense connected with and arising out of the trade in the most complete sense of the word.
74 The rendering of certain defalcation losses as being tax-deductible must therefore be properly understood as a common law exception prompted by the realities of the commercial world to make sense of a statute which, on plain reading, simply did not provide for the treatment of such losses. Seen as an exception (rather than as a right) granted to the taxpayer, it may be that the taxpayer is in no position to argue, as submitted by the appellant, that there must be a “clear and precise” test to constitute proper “guidance to companies and investors”.[note: 11]
75 Indeed, commercial morality demands that the taxpayer be guided solely by its duty to take all reasonable efforts to ensure that defalcations by its employees are kept to a minimum. The Curtis test seeks to alleviate the hardship of the taxpayer by granting tax-deductibility to certain defalcation losses but it cannot and should not provide a warped incentive for firms to allow certain employees to do as they please and then claim that they had “expected” the losses which result from the employees’ defalcations to be tax deductible.
C. The need to deter companies from leaving the powers of their directors and shareholders unchecked
76 Seen in this light, the ‘narrower’ understanding of the Curtis test which has been adopted in the present case is clearly justifiable. The distinction between losses resulting from the defalcations of lower echelon employees and those who possess an “overriding power or control” in the firm are justified on two policy grounds.
77 First, the distinction drawn by the Curtis test as understood by the Board is as an extension of sympathy to large firms in view of their inability to keep all their employees, especially those of the lower echelon, in check. Second, and more importantly, it functions as a form of deterrence to firms that do not provide adequate checks on employees who possess such “overriding power or control” that they are able to incur great financial and social damage. As the respondent correctly submitted, deduction should not be given in the latter scenario lest financial indulgence is seen as being condoned.[note: 12]
78 Indeed, a firm which disregards its social duty and allows a director or shareholder who has “overriding power or control” over millions of dollars to “do exactly what he likes” is justifiably penalized when the defalcation losses caused by the said director or shareholder are considered an expense which is not tax-deductible. This hopefully answers the appellant’s question – “why should deductibility be governed by the degree of control or the size of shareholding?”
D. The Curtis test as being prescriptive of commercial practice
79 In arguing against the Board’s understanding of the Curtis test, the appellant also submitted that it should not be faulted for granting the Ex-MD such “overriding power or control” because this was “the way the appellant conducted its business”, and that “run in this way, [the appellant’s] business was successful”.[note: 13] This appears to me to disregard the deterrent nature of the Curtis test.
80 As a deterrent tool, the Curtis test only makes sense if it is not only descriptive but also prescriptive of commercial practice. This prescription is none other than the recognition that it is a highly dangerous and irresponsible practice for companies to leave the power of their directors or shareholders unchecked. For that reason, the fact that “[the appellant’s] business was successful”[note: 14] in granting the Ex-MD overriding power or control is irrelevant in the light of the prescriptive force of the Curtis test. The non-deductibility of such defalcation losses is the risk a company like the appellant must be made to undertake if it intentionally chooses, whether for commercial reasons or otherwise, to leave the power of certain directors or shareholders unchecked.
E. The appellant’s position on the Curtis test is not viable
81 Since the “overriding power or control” test is solely a question of fact, the appellant has a point in submitting that such a test could result in uncertainty. However, I am of the view that the appellant’s understanding of the Curtis test is clearly not a viable substitute.
82 According to the appellant’s understanding of the Curtis test, the court should simply ask whether “the loss arose in the course of the company’s normal trading activities”. In application, this means that as long as the fraudulent transactions, for example, arose out of dealings with suppliers or customers, such defalcations must be deductible since “dealings with suppliers or customers” are part of the company’s normal trading activities. The appellant submitted that this approach would lead to greater certainty than the one applied by the Board.
83 With respect, the appellant’s test may have undesirable consequences. The effect of such a test may be to encourage firms to turn a blind eye towards the power or control wielded by their employees. This is because so long as their employees, especially the high-level ones, “dress up” their defalcations as if they were part of the company’s activities, the losses which they incur would be tax-deductible. In other words, the more sophisticated the fraud, the better it is for tax purposes. Such a result is completely antithetical to the historical, legal and policy grounds which I have expressed earlier that underlie the Curtis test.
What the test of being “in a position to do exactly what one likes” amounts to
84 The appellant also argued that even if the “overriding power or control” test should apply, the Ex-MD was not in a position to do exactly what he liked and the Board’s comparison with Curtis was therefore not apposite.[note: 15] The appellant took pains to point out that legally, the Ex-MD had no power to defraud the company and the fact that the appellant’s Board of Directors removed him as soon as his misconduct was discovered is clear evidence that he was not in a position to do exactly what he liked.[note: 16]
85 In my view, the appellant is mistaken in assuming that the Curtis test is merely a test of the legal arrangements within a firm. If the appellant’s definition of being “in a position to do exactly what one likes” is adopted, the effect would be not much different from the approach which I have rejected – i.e. did the loss arise in the course of the company’s normal trading activities? A company could easily draw up legal provisions which purportedly govern what its directors can or cannot do, but be nonchalant in checking whether its directors abide by those provisions.
86 Therefore, the test of whether the defalcator was “in a position to do exactly what he likes” should be a test of the factual arrangements within the company. Thus, while it would be rare to find listed companies like the appellant[note: 17] legally giving its directors unbridled power, the Curtis test should be applied to pierce through such legal facades and penalize companies which factually did give its directors unjustified overriding power or control. Only when applied as such would the deterrent purpose of the Curtis test stated earlier (see above at [77]) be fulfilled.
Application of the Curtis test to the facts
87 Having agreed with the Board’s understanding of the Curtis test, I am also of the view that the Board correctly applied the test to hold that the Loss in the present case should not be tax deductible. While the Ex-MD was not a sole proprietor as in the Curtis case, it cannot be seriously doubted that the Curtis test can extend to other high-level employees who similarly possess overriding power or control.
88 In the present case, the Ex-MD did possess an overriding power or control in the company and the defalcations were in fact committed in the exercise of such power or control.
89 The Board correctly relied on the finding of fact by the DJ in the conviction of the Ex-MD (PP v KCT at [284]):
The evidence revealed that no one questioned [the Ex-MD’s] instructions. There was total trust reposed in the [Ex-MD] by virtue of his senior management position. In evidence he said that he need not have to tell anyone about, and (as a fact) he had complete control over, the usage of [the appellant’s] funds.
90 This is clear evidence that the Ex-MD did possess an overriding power or control such that he could do exactly what he liked. The “total trust reposed in the [Ex-MD]” by the appellant factually gave the Ex-MD overriding power or control as he “need not have to tell anyone about ... the usage of [the appellant’s] funds”.
91 The defalcations were also committed by the Ex-MD in the exercise of his overriding power or control. In other words, the defalcations were committed by the Ex-MD “qua director” (Cassidy at 693). In fact, it was probably only due to his “overriding power or control” as a trusted director that he could execute, as the DJ found (PP v KCT at [288]):
the blatant way in which the monies were siphoned from the company, some in the form of cash cheques, others as company cheques, made out directly in the names of or given to, junket operators, or individuals with no trade dealings with [the appellant]. In most instances the company’s funds were used to repay the [Ex-MD’s] gambling debts and feed his gambling habit.
92 Therefore, because the Ex-MD did possess an overriding power or control in the company and the defalcations were in fact committed in the exercise of such power or control, I dismiss the appeal and uphold the Board’s decision. The Loss due to the Ex-MD’s defalcations does not qualify for deduction under section 14(1) of the Act.
The definition of “mistake” in section 93A(1) of the Act
Background of the dispute
93 Given that my holding on the first issue disposes of the appeal, there is strictly speaking no need for me to deal with the second issue submitted by the respondent. Nonetheless, I will give reasons to explain why I am of the view that the Board’s definition of an “error or mistake” in section 93A of the Act is preferable to that of the respondent’s.
If any person who has paid tax for any year of assessment alleges that an assessment is excessive by reason of some error or mistake in the return or statement made by him for the purposes of the assessment or, where he is exempted from liability to furnish a return under section 62(2), in the notice of assessment served on him, he may, at any time not later than 6 years (if the year of assessment within which the assessment was made is 2007 or a preceding year of assessment) or 4 years (if the year of assessment within which the assessment is made is 2008 or a subsequent year of assessment) after the end of the year of assessment within which the assessment was made, make an application in writing to the Comptroller for relief. [emphasis added]
95 The appellant’s application for the deduction of the Loss due to the Ex-MD’s misappropriation (incurred in YA 2000) was made only on 15 December 2005. Therefore, the appellant also had to, on top of arguing that the Loss it incurred is deductible, justify its omission in claiming for the deduction in the appropriate year.
96 Before the Board, the appellant first argued that it had omitted to claim the deduction earlier “due to an oversight”. This was rejected by the Board as the Board found that the appellant’s decision was better characterized as one “made after due consideration that the Loss was not an allowable deduction under section 14 of the Act”.
97 Nonetheless, the Board was of the opinion that “[i]f the decision [by the appellant] was a mistake it was one of law and still a mistake falling within section 93A of the Act”. In its conclusion, the Board also held that it did not find favour with the respondent’s argument that “error or mistake under section 93A of the Act does not include an erroneous opinion or grossly negligent error, but ‘must be genuinely due to ignorance or inadvertence’”.
98 Although the Board clarified that this point is merely academic in the scheme of things, the respondent argued in its submissions that the Board erred in rejecting the respondent’s argument that error or mistake under section 93A of the Act must be genuinely due to ignorance or inadvertence.
99 Interestingly, the respondent did not submit on whether it agrees that a mistake of law can constitute an “error or mistake” under section 93A(1). The Board disagreed with the respondent’s definition of “error or mistake” as it was of the view that such a definition was too narrow and could not cover a genuine mistake of law.
A genuine mistake in law can constitute an “error or mistake” under section 93A(1)
100 I agree with the Board’s decision that an “error of mistake” under section 93A(1) should not be confined to merely “ignorance or inadvertence” and should be wide enough to cover genuine mistakes of law. There appears no good reason why genuine mistakes of law (for instance, having received wrong advice from professionals on what expenses are deductible or where the courts have re-interpreted what was thought to be established law) are any less deserving of relief than genuine mistakes of facts (such as computational or arithmetical errors). In both scenarios, the taxpayer’s “error or mistake” pertains solely to the computation of its taxable income and not some other extrinsic considerations.
101 It is on this basis that the Board rightly, in my opinion, distinguished the Hong Kong case of Extramoney Ltd v Commissioner of Inland Revenue [1997] 2 HKC 38 (“Extramoney”). In Extramoney, the taxpayer had deliberately attributed profits of another company within the group to itself in order to benefit either itself or other companies within the group structure but that turned out (at least in the view of the liquidators) to be an erroneous judgment call. Chan J held that the “error” claimed did not constitute an “error” under the Hong Kong equivalent of section 93A(1). The learned judge then so opined (at 50):
I do not think that a deliberate act in the sense of a conscientious choice of one out of two or more courses which subsequently turns out to be less than advantageous or which does not give the desired effect as previously hoped for can be regarded as an error within [the equivalent of section 93A(1)].
102 Indeed, if there was any “error” made by the taxpayer in Extramoney to speak of, it was an error with regards to the commercial advantage of attributing the said profits to itself. I agree with the learned judge in Extramoney that such errors are clearly outside the ambit of section 93A(1), which provides only for an “error or mistake in the return or statement made by [the taxpayer]”. It must follow then that genuine mistakes or errors made by the taxpayer in the filling up of the return or statement falls within section 93A(1) and this could have been caused by a genuine mistake of law instead of fact.
A caveat when dealing with mistakes of law
103 There is, however, an important caveat to the general rule established above that a genuine mistake of law can entitle a taxpayer to relief. The caveat is found in section 93A(3) of the Act:
No relief shall be given under this section in respect of an error or mistake as to the basis on which the liability of the applicant ought to have been computed when the return or statement was in fact made on the basis of or in accordance with the practice of the Comptroller generally prevailing at the time when the return or statement was made.
104 It would appear from section 93A(3) that even if the taxpayer was operating under a genuine mistake of law, the taxpayer would not be entitled to relief if the Comptroller, at the material time, was also operating under the same mistake. While not canvassed by the respondent, it seems to me that the respondent could have easily proven that the “practice of the Comptroller generally prevailing” in YA 2000 was to deem losses caused by the defalcations of directors non-deductible under the Act (see DR & P v Comptroller of Income Tax (2001) 5 MSTC 293). It follows that the appellant would not have been able to obtain relief because of section 93A(3) as well. Nonetheless, since both parties did not submit on this point, I do not propose to say anything more on this issue.
105 Thus, while I agree with the Board that a genuine mistake of law is still a mistake falling within section 93A of the Act, the appeal still fails as the appellant has not succeeded on the first issue.
Conclusion
106 In conclusion, the appellant is unable to claim relief under section 93A(1) of the Act because the Loss sustained was not deductible under section 14(1). I therefore dismiss this appeal with costs.
[note: 1]Appellant’s Case at [25] – [26].
[note: 2]Respondent’s Case at [30] – [31].
[note: 3]Appellant’s Case at [28].
[note: 4]Appellant’s Case at [29].
[note: 5]Appellant’s Case at [35].
[note: 6]Appellant’s Case at [57].
[note: 7]Appellant’s Case at [57].
[note: 8]Appellant’s Case at [64].
[note: 9]Appellant’s Case at [72].
[note: 10]Appellant’s Case at [69].
[note: 11]Appellant’s Case at [78].
[note: 12]Respondent’s Case at [83].
[note: 13]Appellant’s Case at [73].
[note: 14]Appellant’s Case at [73].
[note: 15]Appellant’s Case at [62] – [63].
[note: 16]Appellant’s Case at [62] – [63].
[note: 17]Appellant’s Case at [61].
Copyright © Government of Singapore.
This judgment text has undergone conversion so that it is mobile and web-friendly. This may have created formatting or alignment issues. Please refer to the PDF copy for a print-friendly version.
Version No 0: 17 Oct 2011 (00:00 hrs)