Commissioner of Taxation v Mercantile Mutual Insurance (Workers' Compensation) Ltd

[1999] FCA 351

FEDERAL COURT OF AUSTRALIA

Commissioner of Taxation v Mercantile Mutual Insurance
(Workers Compensation) Ltd [1999] FCA 351

INCOME TAX – deductions – insurance – whether deductions allowable under s 51(1) for amounts taxpayer estimates it will be required to outlay to settle insurance claims reported but not settled in the year of income and insurance claims incurred but not reported in the year of income or whether deduction limited to provision made in accounts reflecting the present value of future claims.

Income Tax Assessment Act 1936: s 51(1)

Commissioner of Taxation v Manufacturers’ Mutual Insurance Limited (1931) 31 SR (NSW) 575 followed
Ballarat Brewing Co Ltd v Federal Commissioner of Taxation (1951) 82 CLR 364 cited
Texas Co (Australasia) Ltd v Federal Commissioner of Taxation (1940) 63 CLR 382 followed
RACV Insurance Pty Ltd v Commissioner of Taxation [1975] VR 1 cited
Commercial Union Assurance Co of Australia Ltd v Federal Commissioner of Taxation (1977) 32 FLR 32 cited
Federal Commissioner of Taxation v James Flood Pty Ltd (1953) 88 CLR 492 cited
Nilsen Development Laboratories Pty Ltd v Federal Commissioner of Taxation [1981] 144 CLR 616 referred to
Coles Myer Finance Ltd v Commissioner of Taxation (1993) 176 CLR 640 referred to
Australia and New Zealand Banking Group Ltd v Commissioner of Taxation (1994) 48 FCR 268 cited
Commissioner of Taxation v The Myer Emporium Ltd (1986-7) 163 CLR 199 cited
Commissioner of Taxation v Orica (1998) 154 ALR 1 cited
Burrill v Commissioner of Taxation (1969) 67 FCR 519 referred to
Commissioner of Inland Revenue v Mitsubishi Motors New Zealand Ltd [1996] AC 315 cited
New Zealand Flax Investments Ltd v Federal Commissioner of Taxation (1938) 61 CLR 179 cited

Ogilvy & Mather Pty Ltd v Commissioner of Taxation (1990) 95 ALR 663 approved
Commissioner of Taxation v Raymor (1990) 24 FCR 90 referred to
JRowe & Son Pty Ltd v Federal Commissioner of Taxation (1971) 124 CLR 421 cited
W Nevill & Co Ltd v Commissioner of Taxation (1937) 56 CLR 290 referred to
Commissioner of Taxation v Energy Resources of Australia Ltd (1996) 185 CLR 66 cited
Federal Commissioner of Taxation v Australian Guarantee Corporation Ltd (1984) 2 FCR 483 referred to
Commissioner of Taxation v Woolcombers (WA) Pty Ltd (1993) 47 FCR 561 cited
Merchant v Commissioner of Taxation [1999] FCA 49 referred to

COMMISSIONER OF TAXATION v

MERCANTILE MUTUAL INSURANCE (WORKERS COMPENSATION) LIMITED
NG 963 of 1998

COMMISSIONER OF TAXATION v
MERCANTILE MUTUAL INSURANCE AUSTRALIA LIMITED
NG 964 of 1998

HILL, SACKVILLE AND HELY JJ
1 APRIL 1999
SYDNEY

IN THE FEDERAL COURT OF AUSTRALIA

NEW SOUTH WALES DISTRICT REGISTRY

ON APPEAL FROM A SINGLE JUDGE OF THE FEDERAL COURT

BETWEEN:	COMMISSIONER OF TAXATION                   NG 963 OF 1998 Appellant
AND:	MERCANTILE MUTUAL INSURANCE (WORKERS COMPENSATION) LIMITED Respondent
BETWEEN:	COMMISSIONER OF TAXATION                   NG 964 OF 1998 Appellant
AND:	MERCANTILE MUTUAL INSURANCE AUSTRALIA LIMITED Respondent
JUDGES:	HILL, SACKVILLE AND HELY JJ
DATE OF ORDER:	1 APRIL 1999
WHERE MADE:	SYDNEY

THE COURT ORDERS THAT:

1.          The appeal be dismissed

2.          The appellant pay the respondent’s costs of the appeal.

Note:    Settlement and entry of orders is dealt with in Order 36 of the Federal Court Rules.

IN THE FEDERAL COURT OF AUSTRALIA

NEW SOUTH WALES DISTRICT REGISTRY

ON APPEAL FROM A SINGLE JUDGE OF THE FEDERAL COURT

BETWEEN:	COMMISSIONER OF TAXATION                   NG 963 OF 1998 Appellant
AND:	MERCANTILE MUTUAL INSURANCE (WORKERS COMPENSATION) LIMITED Respondent
BETWEEN:	COMMISSIONER OF TAXATION                   NG 964 OF 1998 Appellant
AND:	MERCANTILE MUTUAL INSURANCE AUSTRALIA LIMITED Respondent

JUDGES:	HILL, SACKVILLE AND HELY JJ
DATE:	1 APRIL 1999
PLACE:	SYDNEY

REASONS FOR JUDGMENT

HILL J:

1. Income tax is an annual tax.  It exists to finance the operations of government year by year.  But divisions of time are of necessity arbitrary.  Hence assessable income or allowable deductions may be required to be allocated to the one year or the other, especially where the overall transactional outcome spans the artificial division into years which annual tax accounting requires.  The present appeal illustrates the difficulties which are inherent in this allocation in the context of insurance.

2. The appeal is from the judgment of a judge of this Court (Foster J) which allowed  appeals brought in the original jurisdiction of the Court from the disallowance by the Commissioner of Taxation of objections by Mercantile Mutual Insurance (Workers Compensation) Ltd (“MMIWC”) and Mercantile Mutual Insurance (Australia) Ltd (“MMIA”) (which companies are herein compendiously referred to as “Mercantile Mutual”) against assessments made  by the Commissioner in respect of the year of income 1989 for MMIWC  and the year of income 1992 in respect of MMIA.  Each of the companies had adopted a substituted accounting period ending on 30 September in lieu of the year of income ending on the preceding 30 June.

3. MMIWC is engaged in the business of writing employers’ liability insurance, including workers’ compensation insurance.  MMIA is engaged in the business of writing public liability and compulsory third party motor vehicle insurance.  Each is a wholly owned subsidiary of Mercantile Mutual Holdings Limited.  There are appeals not yet heard which relate to other subsidiaries raising the same issues as the present appeals.  The parties have agreed that the decision in the present case will resolve the issues of principle in each of the other appeals.

4. A characteristic common to the insurance business carried on by each of the respondent companies  is the time which may elapse between the event which gives rise to a claim by a policy holder and the time when the claim is paid.  In the case of some claims the happening of the event may be reported and the claim paid all in the one year. In such cases, no question of allocation of the claims to an income tax year arises. However, in many cases there may be a considerable interval of time between the reporting of a claim and its ultimate pay out.  In many cases, indeed the claim may be paid over many years in the future.  Workers’ compensation claims are an example.  Such insurance is often denominated as “long tailed insurance” to mark this fact.

5. The present appeal concerns two classes of case. The first is where the event insured against occurs in the year of income, is reported in that year but is not settled before the end of the year of income. That class of case is referred to as claims reported but not settled. The second class of case concerns claims where the event insured against occurs in the year of income but is not reported in that year and comes to be settled after the year of income. That class of case is referred to as claims incurred but not reported (“IBNR”). Both classes of cases raise the same question. It is not in dispute that in each case the insurer has incurred a liability entitling it to a deduction under s 51(1) of the Income Tax Assessment Act 1936 (“the Act”).  What is in dispute is the manner in which that liability should be calculated.

   The case law background

6. The Act proceeds on the basis that the subject matter of the tax, “taxable income”, falls to be calculated by determining the assessable income and deducting therefrom all “allowable deductions”: s 48(1) and ss 6(1), 17. Relevant to the present appeal allowable deductions include what may compendiously be referred to as the business outgoings made deductible under s 51(1) of the Act which provides, relevantly:

   “All losses and outgoings to the extent to which they are incurred in gaining or producing the assessable income, or are necessarily incurred in carry on a business for the purpose of gaining or producing such income, shall be allowable deductions…”

7. Where claims are reported, but not settled, it has now been accepted for some considerable time that a taxpayer is entitled to a deduction in the year in which the event occurred giving rise to the claim, the deduction being given, of necessity for an estimate. So much was decided under the provisions of s 19(1)(a) of the Income Tax (Management) Act 1928 (NSW) in Commissioner of Taxation v Manufacturers’ Mutual Insurance Limited (1931) 31 SR (NSW) 575.

8. Ferguson J, with whose judgment Street CJ and James J agreed, said at 585:

   “…there is no legal obligation in the sense of a cause of action immediately enforceable against the company; but from a practical business point of view, and that is the point of view from which these questions should be regarded, it is just as certain that some money will have to be paid in respect of pending claims as in respect of claims which have gone to judgment.  The amount is uncertain, but apparently it is susceptible of more or less accurate estimate,  Any statement of the affairs of the company professing to show the result of the year’s operations, which neglected to take into account of this liability, would be grossly inaccurate and misleading.  In my opinion, therefore, it is an obligation standing on the same footing as an actual expenditure, which the company is entitled to deduct as a loss or outgoing actually incurred in producing the assessable income, subject of course to any necessary future adjustment.”

9. The passage set out above was cited with approval by Fullagar J in Ballarat Brewing Co Ltd v Federal Commissioner of Taxation (1951) 82 CLR 364 at 370. The latter case was not concerned with insurance. It concerned a deduction for discount or rebates granted to customers subject to the fulfilment of conditions including payment by the customer of the purchase price, where such payment had not yet been made by the end of the year of income, but the rebate was virtually certain to be given. To not have taken the rebate or discount into account would have produced the result that the trading result would have overstated the profit of the company.

10. Certainly, after Texas Co (Australasia) Ltd v Federal Commissioner of Taxation (1940) 63 CLR 382 was decided, it became clear that in a case where the amount deductible was an estimate, later revision of that estimate when the real figure was ascertained would give rise to either an increased deduction where the estimate was too high or assessable income where it was too low.  Speaking in the context of liabilities in overseas currency affected by exchange variations, Dixon J said at 465-6 in that case:

    “During any given accounting period the profit or loss made by the taxpayer’s operations must be ascertained by a comparison between its position at the beginning and at the end, based upon estimates of value and upon the accrual of debits and credits.  But discrepancies between the liabilities carried into the period and the cost of defraying them must come into the comparison as an actual reduction or increase of the profit or loss otherwise produced by the comparison, provided always that the liability is one belonging to an income account and that the loss ought not for other reason to be referred to capital. For where liabilities are not fixed in their monetary expression, whether because of contingencies or because they are payable in foreign currency, a difference between the estimate and the actual payment must be born as a business expense, and where the continuous course of a business is divided for accounting purposes into closed periods it is a reduction of the net profit, which otherwise would be calculated for the period.”

11. The question whether claims incurred but not reported gave rise to an allowable deduction was not agitated in Australia until RACV Insurance Pty Ltd v Commissioner of Taxation [1975] VR 1. Accountants had, in that case, given evidence that it was necessary in drawing up accounts to bring to account in the same year as the revenue from a particular transaction the expenditure or anticipated losses “directly related to that revenue”. Menhennitt J of the Supreme Court of Victoria, discussed the evidence of one expert accountant, Mr Buckley, in the following passage at 14:

    “…applying that principle he believes that it is proper and necessary to bring to account by way of debit to the profit and loss account in the year in which you bring a certain premium income resulting from contracts of the kind of compulsory third-party insurance estimates of the amounts likely to have to be paid out on account of claims which have been notified and also the best estimate that can be made of claims that have not yet been notified so that they will be matching the claim against the revenue.”

    His Honour continued:

    “In my view not only is this in accordance with the accepted accountancy principles, but it is also what s51(1) of the Income Tax Assessment Act contemplates.”

12. Hence in RACV it was held that an insurer was entitled to claim a loss or outgoing as having been incurred when the event occurred which imposed upon the insurer a liability to indemnify the policy holder. The fact that no notice had been received of the claim was held to be irrelevant. On the question of the quantum of the deduction, Menhennitt J said at 17:

    “In its annual accounts the taxpayer included a statistical estimate of $1,320,000 upon the basis I have referred to.  Mr Sawkins the actuary said that in his opinion this was a reasonable estimate in the circumstances of the taxpayer at that time.  By the time it lodged its return the taxpayer had made estimates case by case of the claims by then reported and arising out of events happening in the year ended 28 February 1971 and this is the amount it claimed in its return, namely $1,420,424.  I am satisfied on the evidence that this amount represented the total of reasonable estimates of the taxpayer’s liability in respect of claims of which it in fact had notice by that date.”

13. Although the RACV case had held that IBNR claims gave rise to a deduction thrown against the earned premiums for the year, a deduction for an estimate of such claims was disallowed by the Commissioner to Commercial Union Assurance Co of Australia Ltd on the basis that the policies required notice of the occurrence of the event to be given either forthwith or as soon as possible, and accordingly the insurer had no legal liability to policy holders who did not comply with the condition.  The evidence was that the invariable practice was to pay claims in full notwithstanding noncompliance with the condition.  It was held in Commercial Union Assurance Co of Australia Ltd v Federal Commissioner of Taxation (1977) 32 FLR 32 that the taxpayer was entitled to the deduction claimed. It was not a prerequisite to deductibility that the insurer be under a legal obligation to pay enforceable by law. It sufficed that the taxpayer had in the year of income been subjected to the liability and had thus incurred it, cf Federal Commissioner of Taxation v James Flood Pty Ltd (1953) 88 CLR 492 at 506.

14. In passing it may be noted – it is relevant to a submission made by the Commissioner in the present appeals – that Newton J of the Supreme Court of Victoria said at 43, dealing with the argument that the non-fulfilment of the condition requiring notice led to the conclusion that no amount had been incurred at the time the event insured against occurred:

    “It would, in my opinion, be wrong to regard them as contracts to pay money to the insured, or for his benefit, upon the reporting to the insurer within a specified time of the occurrence of an event.  On the contrary they were insurance contracts subject to a condition precedent as to notice, and since it was certain that the condition as to notice would never be relied on by the insurer, the liability of the insurer was ‘incurred’ within the meaning of s 51 upon the happening of the event insured against.”

15. Of greater importance to the Commercial Union decision and the present appeals, however, is that in holding the amount claimed by the insurer to be deductible his Honour spoke of treating the deduction as being thrown against the premiums of the year. 

16. The judgment makes it clear that, in principle, there is no difference between claims reported but not settled and claims incurred but not reported. In both cases the insurer has incurred expenditure which is an allowable deduction.  The case held too that subsequent revision of estimates allowable gave rise to further deductions or assessable income as the case may be.  On this matter his Honour said at 48-49:

    “In my opinion it must now be accepted that the words ‘the assessable income’ in s 51 mean assessable income of the taxpayer generally without regard to division into annual accounting periods: see AGC (Advances) Ltd  v Federal Commissioner of Taxation (1975) 132 CLR 175 at 185-189, 195-199. … Hence, in my opinion, just as an estimated provision for outstanding claims, including claims incurred but not reported, arising from insurances current in one year, is an allowable deduction under s 51 in calculating the insurer’s taxable income for that year ... so an increase in the estimate at the end of any subsequent year for any of those claims which are then still outstanding, is an allowable deduction in calculating the insurer’s taxable income for that subsequent year,  The increase in the estimate constitutes a new loss or outgoing which was incurred for the first time in the year when the revised estimate was made… I may add that any other conclusion would give to s 51 an operation which would be quite out of accord with practical realities, having regard, inter alia, to inflation. …No doubt provision for outstanding claims ought ideally to be made out of the premiums earned from the insurances out of which the claims arose.  But where an original estimated provision which was deducted from those premiums is found on a later revised estimate to be insufficient, the amount of the deficiency ought to be deducted from the premiums earned in the year when the revised estimate is made, if the insurer’s accounts are to give a true and fair view of the profit of the insurer for that year.”

17. The judgment of Newton J in Commercial Union has been referred to with approval in the High Court in Nilsen Development Laboratories Pty Ltd v Federal Commissioner of Taxation [1981] 144 CLR 616 and Coles Myer Finance Ltd v Commissioner of Taxation (1993) 176 CLR 640.

18. A full court of this Court in Australia and New Zealand Banking Group Ltd v Commissioner of Taxation (1994) 48 FCR 268 held that the same result should follow where the taxpayer was not an insurance company but a self insurer. Special leave to appeal from this decision was refused: “The Legal Reporter” Vol 15, No. 18, 14 Nov 1994 p SL4. In that case the Commissioner argued not merely that the deduction of an estimate of claims was restricted to insurance companies, but also that no reasonable estimate was possible, or that if it was the amount of the estimate there made was unreasonable. In response to that argument I said (and the other members of the Court, Northrop and Lockhart JJ agreed at 280-281):

    “The concept of ‘estimate’ does not involve arbitrarily seizing upon any figure.  What is involved is the formation of a judgment or opinion based upon reason.  That judgment or opinion must necessarily be made bona fide but it need not be exact for the process of estimation involves a process of approximation….

    The task of estimation of a liability in a case such as the present is a commercial one….

    In my view the making of an estimate by reference to a case by case analysis of files by  a person expert in the field is a reasonable way of making such an estimate … That was the estimate adopted in the Bank’s accounts which were audited and which, so far as appears, was subject to no reservation by the auditors in the certification of those accounts.  In my view the present is not a case where it can be said that there can be no process of estimation or that no process of estimation was in fact made because the figure adopted was not reasonably arrived at.  Rather, the establishment of provisions in the present case was an exercise capable of approximate calculation based on probabilities.”

1. It would appear that in RACV, Commercial Union or ANZ there was not any attempt  made to discount the amount of the estimate by the fact that claims would be paid in the future.   The estimates in each case allowed as a deduction produced a figure representing the face value of claims as they were expected to be paid out in the future. In none of the cases was there any argument addressed to the need for any such discount to be made.  Each case proceeded on the basis that the amount claimed as a deduction equated to the amount required to be provided for in the taxpayer’s accounts in accordance with accepted accounting principles so as to ensure that the accounts gave a full and true view of the profit or loss of the insurer.   However, in Commercial Union Newton J commented at 50-51:

   “I wish expressly to state that it was not suggested by counsel for the commissioner that the calculation of the sum of $5,864,866 was incorrect, because no allowance was made for a discount of future payment: compare Southern Railway of Peru Ltd v Owen [1957] AC 334. Indeed no such allowance was made in the estimate of $54,155,259 in respect of communicated claims against the Commercial Union pool which were unpaid as at 30th June 1973, although some of those claims could reasonably be expected not to be settled for a considerable time. However, since, as earlier stated, there was an overall underestimate of outstanding claims against the pool as at 30^th June 1973 (including both communicated claims and claims incurred but not reported) by an amount of more than two million dollars, it appears to me that this question is not of practical importance in the present case.”

   The Accounting Treatment – How the problem arises

2. In each of the present appeals the amount claimed by Mercantile Mutual as a deduction under s 51(1) for claims incurred but not reported and claims reported but not settled in the year of income differed from the provision in its accounts for some items, which provision was calculated in accordance with accounting standards and, in accordance with the expert accounting evidence, correctly so calculated.

3. In simpler times it may have been possible for a company carrying on an insurance business to examine all reported claims to arrive at an estimate of the face value of each claim by reference to the facts reported.  That was the process adopted by the ANZ Banking Group Ltd as a self-insurer.  Clearly such a process could not be adopted by a commercial insurer.  By definition it could never be the case with claims incurred but not reported.  The alternative, adopted by actuaries, is to base the calculation on claims experience over time. This was the method adopted by Mercantile Mutual.  As the starting point of the calculation it was not disputed by the Commissioner.

4. The evidence established that the first step adopted by Mercantile Mutual in calculating the provision which it was required to make in its accounts was the calculation on an actuarial basis of the actual dollar amount likely to be paid out by the insurer in future years of income in settlement of claims.  This figure was referred to in evidence as the “inflated (undiscounted)” amount.   It is referred to as “inflated” because the figure arrived at took into account the changing value of money where claims were to be paid in the future.  Thus the calculation took into account expected rates of inflation.  It is referred to as “undiscounted” because it made no attempt to calculate the present value of the future liability to pay out claims.

5. Because the calculation was statistical and because the actuarial models used were obtained from variable past experiences there was a degree of uncertainty associated with them.  The figure initially derived as a result of the statistical process was referred to in the evidence as being the “central estimate”.  An expert actuary giving evidence discussing the concept of central estimate said:

   “Statistically, the Estimates were intended to have an equal (ie 50%) chance of ultimately proving to be too high or too low.  In other words, the probability that any central estimate would prove to be sufficient (ie too high) was intended to be 50% and (by definition) the probability that the estimate would prove to be insufficient (ie too low) was also intended to be 50% …”

6. Because the probability of the central estimate being too low was 50% it involved a higher than desirable risk of underestimation.  As the same actuary pointed out:

   “No allowance was made in the Estimates for an unpredictable deterioration in experience.  This may occur because of increased court awards, changing social attitudes and so on.”

7. Hence it was desirable that there be an adjustment to the central estimate in the nature of a “safety or prudential margin” to take account of such contingencies.  That adjustment was a matter upon which managerial judgment had to be brought to bear.  The extent of such an adjustment was likely to be influenced by a number of factors including the overall level of solvency margin thought desirable for the company as a whole, and assessments as to how future experience might be though to be likely to differ from past experience.  There was no suggestion that the adoption of a safety or prudential margin was inappropriate as such.  One actuary, a Mr Buchanan, referred to it as involving an estimate of the value of uncertainty.  It was the normal practice of insurers to add a prudential margin to the actuarial central estimate. It operated to increase the probability that the central estimate was neither too low nor too high, and thus ensure that the provision in the accounts was more likely to be sufficient.

8. As will be seen there is a dispute between the parties in part as to the concept of a prudential margin adopted, and in part as to whether the percentage applied in the accounts should be applied to the central estimate directly.

9. The second step adopted was the calculation of the inflated and discounted figure.  This was no more than the calculation of the present value in the year of income of the figure derived at in the calculation of the inflated, undiscounted figure.  This was in accordance with the Accounting Standard AASB 1023 “Financial Reporting of General Insurance Activities’ (paras 5.1 and 5.1.1) that:

   “a liability for outstanding claims must be recognised in respect of both direct business and inwards reinsurance business and must be measured as the present value of the expected future payments.

   … This measurement approach is based on the view that the liability for outstanding claims ought to reflect the amount which, if set aside as at the reporting date, would accumulate so as to enable the insurer to pay the amounts of claims as they fall due.  The approach requires estimation of the ultimate costs of settling claims and the discounting of those amounts to a present value”

10. It is also in accordance with Accounting Guidance Release AAG13, published by the Public Sector Accounting Standards Board of the Australian Accounting Research Foundation and the Australian Accounting Standards Board (“Determination of Discount Rates for Measuring Certain Liabilities at Present Value”) which requires  liabilities to be settled in the future to be brought into the profit and loss account by reference to present value, whether the liability is that of an insurer or of some other similar accounting entity. 

11. The third step in deriving the provision taken up in the accounts of Mercantile Mutual is the application of the percentage prudential margin to the inflated and discounted figure.  In the years of income in question this was 10% or 11%.

12. By way of illustration of the difference between the inflated and discounted figure (adopted in the accounting provision of MMIWC in the 1989 year) on the one hand and the inflated (undiscounted) figure adopted in the tax return on the other:  the inflated (undiscounted) amount was $96,115,000;  the inflated and discounted figure was $61,435,058.   

13. In the return the computation commenced with the central estimate.  This was then increased by applying the same percentage prudential margin factor as had been applied in arriving at the final inflated and discounted figure used in calculating the provision in the accounts.  The difference is that the prudential margin percentage was, in the return, applied to the central estimate, whereas it was applied, in calculating the accounting provision, to the present value of the inflated (undiscounted) figure to arrive at the computation of the inflated and discounted figure.

The Judgment Appealed from

1. The learned primary judge concluded that the correct approach required by the Act was to allow a deduction for the face value of estimated claims in the year the claim arose. His Honour accepted that it was appropriate for that face value to be adjusted by way of a prudential margin. He pointed out in this context that actual claims experience after the years of income in question had shown that the estimates arrived at fell short of the amounts actually required to settle claims. He made no comment as to the application of the prudential margin to the central estimate.

2. In arriving at his decision the primary judge was influenced by passages appearing in the judgment of the High Court in Commissioner of Taxation v The Myer Emporium Ltd (1986-7) 163 CLR 199 at 216-7 and in Commissioner of Taxation v Orica (1998) 154 ALR 1 at 12-13 and 24 to the effect that historical cost was the basis of calculating profits and losses for the purposes of the Australian income tax law. He referred also to Burrill v Commissioner of Taxation (1969) 67 FCR 519, a decision of a full court of this Court which had applied these passages in the context of s 70B of the Act, regarding that case as governing the outcome of the present appeals unless distinguishable. His Honour refused to distinguish it.

3. In particular his Honour accepted that while it was necessary for accounting purposes to provide in the accounts as a liability on an annual basis the amount required to be set aside and invested to provided for the payment in the future of claims (ie the inflated and discounted amount) this did not govern the taxation outcome.  His Honour noted that the accounting treatment was said to depend upon “the matching principle”.  He quoted one witness, Professor Peirson as having said in evidence that to provide as a liability in the accounts the inflated and undiscounted figure would give:

   “an incorrect measure of the liability for outstanding claims and, therefore, an incorrect measure of the claim’s expense.  In turn, this will give an incorrect measure of profit by overstating the expenses and understating the profit”.

4. Commenting on this passage and the matching principle his Honour said:

   “The concept of ‘the matching principle’ is to be found in the passage from RACV cited above, in which Menhennitt J referred to s 51(1) as being its ‘statutory recognition’.  The approach of Menhennitt J was approved by the High Court in Coles Myer Finance Ltd v Federal Commissioner of Taxation (1993) 176 CLR 640 at p 665-6. … I have great difficulty in understanding how it can be asserted that the applicants’ ascertainment of the deductions to be made for outstanding claims liabilities in a fiscal year can transgress the matching principle enunciated by Menhennitt J when the procedures which were the subject of explanation and approval in that case appear to have been followed. In my opinion, there has been no violation of any matching principle when that principle is understood as a jurisprudential rather than a commercial postulate. So understood, what must be matched is the premium component of the assessable income of the year in question and the amounts which must be paid in the future to satisfy claims the risk of which is covered by the insurance purchased by those premiums. The payments required to discharge the liabilities arising from those claims will necessarily be made in the future. They will be made in the money value of the day of payment and will accord with the levels of compensation appropriate to that time. The discounted amounts contemplated by Professor Peirson and Mr Buchanan are not moneys paid out to indemnify policy holders in respect of claims made against them, against the risk of which they have obtained cover by the payment of the premiums. They are amounts invested for the purpose of the later meeting of those claims. The income from such investments will itself form part of the applicants’ assessable income and, as such, be liable to tax. They do not answer the description of ‘outgoings incurred’ in s 51(1).”

   The Submissions on behalf of the Commissioner

5. It was submitted for the Commissioner that the learned trial judge had erred in three ways.  First, it was submitted that his Honour should have held, consistently with what was said by Menhennitt J in Commercial Union, that there was a distinction between money payable in the future which the Commissioner conceded was deductible at face value rather than at its present value at end of the year of income and insurance claims.  It was submitted that the liability to pay money was not encountered in the case of insurance claims when those claims arose and thus not incurred.  What was encountered in the year of income was the happening of the contingency insured against which gave rise to a liability in the future.  A distinction had to be drawn between the happening of the contingency and the liability to pay. Since the liability to pay only arose in the future what was deductible was the provision required to be made for the liability, not the liability itself.

6. Secondly, and this submission was not necessarily related to the correctness of the first submission, the decision of the High Court in Coles Myer Finance required that the amount incurred and thus deductible under s 51(1) had to be referrable to the year of income. This led to the conclusion, it was submitted, that what was deductible was the provision required to be set aside in the year of income to meet the payment required to be made in the future. To hold otherwise would, it was submitted, lead to distortion and was contrary to accepted accounting practice, in particular the matching principle.

7. Finally, it was submitted, if the first two propositions were not accepted, his Honour erred in accepting the calculation of the prudential margin, since this was applied in the accounts to the inflated and discounted figure, whereas in the taxation computation the percentage was applied to the central estimate.

A distinction between the happening of an event and a liability to pay money?

1. In my opinion this submission is, with respect, misguided.  Moreover, it involves placing more emphasis than is warranted to the passage earlier cited from the judgment of Newton J in RACV.

2. Section 51(1) provides for a deduction to be allowed for a “loss” or an “outgoing”.  In some case, banks or finance companies are examples (although the circumstances where losses are deductible are not limited to such examples), losses which are made in the course of an investment activity will be allowable deductions just as net gains rather than gross income are included in assessable income. The case of Coles Myer already referred to provides an illustration. In the more general case the available deduction is granted for an outgoing. The most obvious examples are rent and interest, although again I use these by way of illustration without suggesting that s 51(1) is limited to them. This is consistent with the operation of the Act which requires the computation of taxable income to start with assessable income, usually gross income subject to statutory definition, and to subtract therefrom all allowable deductions. In the present case what falls for deduction if incurred is an outgoing.

3. The concept of when an outgoing is incurred has been the subject of considerable judicial discussion.  As the High Court in Coles Myer observed at 662 the Australian authorities have adopted a legal or jurisprudential analysis in determining when a loss or outgoing is incurred in preference to a commercial test, cf the criticism of the Privy Council in Commissioner of Inland Revenue v Mitsubishi Motors New Zealand Ltd [1996] AC 315 at 324 that the Australian approach is “an unusual approach to a taxing statute”. 

4. In New Zealand Flax Investments Ltd v Federal Commissioner of Taxation (1938) 61 CLR 179 at 207 Dixon J in a passage often repeated, said:

   “ ‘Incurred’ does not mean only defrayed, discharged, or borne, but rather it includes encountered, run into, or fallen upon.  It is unsafe to attempt exhaustive definitions of a conception intended to have such a various or multifarious application.  But it does not include a loss or expenditure which is no more than impending, threatened, or expected.”

5. It was pointed out in James Flood at 506 that it was not a necessary precondition to a deduction that the taxpayer have an immediate obligation enforceable at law whether payable presently or at a future time.  That comment must, as the High Court in Coles Myer observed, be read against the question which fell for decision in that case; see at 661-2.  It is, however, necessary that a liability must be “presently existing” before, in the relevant sense, it is incurred.  (Subject to Coles Myer, discussed later, the discussion of the meaning of “incurred” in Ogilvy & Mather Pty Ltd v Commissioner of Taxation (1990) 95 ALR 663 in my view correctly summarises the principles which have been adopted in determining when a loss or outgoing is incurred.)

1. There is no doubt that once an event insured against has happened (and I put to one side matters of notice and the like which may affect the generality of what is here said) the insurer comes under a legal liability to pay money to the policy holder.  That legal liability is encountered the moment the event insured against happens.  It is in all respects a presently existing liability, and for this purpose it matters not that the liability falls to be discharged in the future.  There is no difference for this purpose between the liability of an insurer to pay money in the future in settlement of a claim and the obligation of any other taxpayer who falls under a legal liability to pay money in the future.  Contrary to the Commissioner’s submission, there is no gap in time between the event insured against and the liability to pay.  There is a presently existing liability to pay monies in the future, which liability, like the event which gives rise to it, occurs in the year of income.

2. As the Commissioner’s submission accepts, there has been no case reported in Australia where it has been held that a non insurer under a presently existing liability to pay in the future was held not to have incurred that liability.  (The actual decisions in Flood and Nilsen Laboratories turn not on this point but rather on the view that no liability to pay money arose until the employees either took the leave entitlements or ceased employment;  see Coles Myer at 662.) Nor for that matter has it ever been suggested that, where there is a presently existing liability to pay money in the future, and where the outgoing which is deductible is constituted by that liability, what is available for deduction under s 51(1) in the year in which that liability arose is the present value then of the obligation to pay in the future. On the contrary, cases such as Commissioner of Taxation v Raymor (1990) 24 FCR 90 proceed on the basis that the presently existing liability to pay in the future is deductible. It is not submitted that such cases were wrongly decided.

3. It may be said that where inflation and interest rates are negligible such an issue does not assume great importance.  However, the fact remains that present value discounting of outgoings payable in the future has never been suggested in Australian tax jurisprudence.  And this is so notwithstanding that for commercial accounting purposes such discounting would, whether required or not is not the subject of evidence before us, be at least consistent with the accounting concepts contained in “Definition and Recognition of the Elements of Financial Statements” – SAC4, which defines ‘liabilities’ to be “the future sacrifices of economic benefits that the entity is presently obliged to make to other entities as a result of past transactions or other past events.”

1. In this context it is well to recall that Dixon J in New Zealand Flax, in response to a submission by Mr Bowen as he then was that the ascertainment of taxable income, both assessable income and allowable deductions, should follow the principles of commercial trading, commented at 208 in respect of allowable deductions that matters affecting the business propriety of making provisions must give way to the fact that the “Income taxAssessment Act is not framed to give effect to such considerations. I shall return to this case shortly.

2. If all entities, insurers or otherwise were required to bring in the future value of liabilities payable in the future, it would be necessary to consider whether additional amounts were incurred in each year of income as the time for payment drew closer and to make adjustments.  In the alternative it might be necessary to await the year of payment before a deduction for the difference between the original present value of the liability and the amount required to settle it arose.  No doubt the problems which arise in cases not involving estimation explain the fact that the Commissioner sought to differentiate insurers from other taxpayers.

3. If it were appropriate to discount liabilities to present value there would then arise the question whether the same process should be applied to revenue.  A trader who sells stock on terms is required to bring the income derived from the sale into assessable income in the year the sale is made: J Rowe & Son Pty Ltd v Federal Commissioner of Taxation (1971) 124 CLR 421. It has never been suggested that some part of the amount payable should be deferred to a later year on the basis that a present value calculation should be made.

4. As the learned primary judge notes, the High Court has on two occasions emphasised that the Australian income tax system proceeds upon the basis of historical cost and not discounted value. That discussion is, however, in the context of the computation of profit and loss.  It is not necessarily determinative in the present context.  Thus in Myer at 216-7, Mason ACJ, Wilson, Brennan, Deane and Dawson JJ said:

   “The accounting basis which has been employed in calculating profits and losses for the purposes of the Act is historical cost … not economic equivalence. And so a taxpayer who lends money for a stipulated period at interest is treated as exchanging the money lent for a debt of the same amount, unless the loan is made at a discount or premium, in which case there may be a gain or loss on capital account. … In the ordinary case, the debt is brought into account in the same amount as the money lent. The amount of the debt is not reduced because the lender is kept out of the use and enjoyment of the money lent for the period of the loan.

   If economic equivalence were the appropriate accounting basis, the debt would be brought to account at the beginning of the period in an amount less than the amount of the money lent and would increase day by day until it equalled the amount of the money lent when the period expired.  On that basis the right to interest on the money lent would be brought to account at the beginning of the period at a  maximum figure reducing to nil when the period expired.”

5. This passage was endorsed by Gaudron, McHugh, Kirby and Hayne JJ in Orica at 23-24 but again in the same context.

6. Myer was applied by a Full Court of this Court in Burrill at 524-5. Certainly the discussion in that case proceeds to give to what the High Court had said in Myer a broader application than the context in which the remarks were made.  It is not necessary in my view to consider either the correctness of the conclusion reached in Burrill or the application in that case of the discussion on historical cost accounting. It suffices here to say that unless the concept of the incurring of a liability in s 51(1) requires the computation of the present value of the liability when the outgoing is payable in the future, there is no reason to conclude other than that s 51(1) allows as a deduction the outgoing to be made in the future in the year of income in which that outgoing is incurred.

7. Before proceeding to the more substantial issue for decision two propositions of law were advanced on behalf of the Commissioner to support this first submission.  Neither may be accepted, at least without qualification, in the form they are advanced. First it is said that the ordinary rule in tax law is that the provision for a future payment will not be allowed as a deduction.  For this proposition Flood and Nilsen Development Laboratories are cited.  It is no doubt correct to say on the authority of those cases that a mere provision which does not reflect a presently existing liability does not give rise to a deduction.   If that is all that is meant it is correct.  But the second submission is then quite inaccurate.  It is said to be that insurance companies are an exception to this first proposition.  The deduction allowed to insurance companies is more than a provision in the sense that a provision was involved in Flood and Nilsen.  It is a deduction allowed for a presently existing liability.  What distinguishes the insurance cases from the ordinary class of case is that the deduction involves the process of estimation.  There is no reason to suppose that in appropriate cases the allowance of a loss or outgoing may involve the process of estimation even where the taxpayer is not an insurance company.  In truth the deduction in Texas Co to which reference has already been made involves estimation, with the consequent need to adjust as a result of exchange fluctuation. 

Coles Myer – referability to the year of income

1. The substantial argument of the Commissioner relies upon Coles Myer.  The submission is simply that to allow a deduction for the whole of the central estimate with or without the prudential margin adjustment is to allow a deduction for some amount which is referable to future years.  This, on the authority of Coles Myer is not permissible.  It is thus necessary to consider that case more carefully.

2. The issue in Coles Myer involved the discounting at or around the end of the year of income of promissory notes and bills by a finance company.  The company claimed that it had incurred the discounting loss when the bills or notes were discounted. The Commissioner claimed that the loss was incurred only when the time for payment of the bills or notes arose.  That was in a subsequent year of income.  Both the taxpayer and the Commissioner relied upon a fall back position that the loss should be apportioned in some way over the two income years.

3. A full court of this Court, Sweeney, Northrop and Wilcox JJ (reported in 28 FCR 7) held that the taxpayer had not incurred a loss within the meaning of s 51(1) until the year of income in which the promissory notes or bills fell due for payment. The High Court allowed unanimously the appeal in part. The leading judgment is the joint judgment of Mason CJ, Brennan, Dawson, Toohey and Gaudron JJ. Deane J who agreed with the majority in the result delivered a separate judgment.

4. The first part of the judgment discusses the question of whether the drawer of an accommodation bill comes under a liability to the acceptor until payment.  It rejects the view taken by this Court that the drawer does not.  It considers then the question whether the maker of a promissory note comes under a present liability when the promissory note is made notwithstanding that the note is payable in the future, and decides that this is the case.  In so doing it discusses its earlier decision in W Nevill & Co Ltd v Commissioner of Taxation (1937) 56 CLR 290 upon which the judgment of this Court relied. It then proceeds to discuss s 51(1) and the question of the meaning of the word “incurred” by reference to the cases of Flood and Nilsen  to which reference has already been made and then says at 663:

   “But it is not enough to establish the existence of a loss or outgoing actually incurred.  It must be a loss or outgoing of a revenue character and it must be properly referable to the year of income in question.”

5. In support of the last part of this comment the Court cites from the judgment of Dixon J in New Zealand Flax  at 207.  Under a separate heading “What is the amount of taxpayer’s loss or outgoing referable to the year of income?”  the judgment says at 665-6:

   “The acceptance by this Court of the jurisprudential analysis of s 51 does not compel the conclusion that, once a taxpayer subjects itself in the year of income on revenue account to a present legal liability to pay in a future year of income an amount which generates, or gives rise to, a net loss or outgoing, the net loss or outgoing is deductible in full in the year of income.  The relevance of the present existence of a legal liability on the part of the taxpayer to meet the bills and notes at a future date is that it establishes that the taxpayer has ‘incurred’ in the year of income an obligation to pay an amount which gives rise to a net loss or outgoing, being the recurrent cost of acquiring working or circulating capital.  But there remains the question: how much of that net loss or outgoing is referable to the year of income?

   Although the legal liability to pay is incurred in the year of income, the amount in question is not payable until the subsequent year of income and, more importantly, the net loss or outgoing represents the cost of acquiring funds which the taxpayer puts to profitable advantage in both years of income. The cost is incurred by the taxpayer with a view to acquiring funds with which to engage in its profit-making activities during the currency of the respective bills and notes. As between the drawer and the holder of a note or bill, the burden of the liability incurred by the drawer increases with the passage of time between the discounting of the note or bill and its maturity. In ascertaining what is the taxpayer’s net income or profit for a particular year of income, it is proper to set against the taxpayer’s gross income or profit for that period the net losses or outgoings referable to that period. Under s 51(1) a loss or outgoing is a deduction only to the extent to which it is incurred in gaining or producing the assessable income. That provision has been described as: ‘a statutory recognition and application of the accountancy principle which all the accountants who gave evidence referred to as the matching principle’, to use the words of Menhennitt J. in R.A.C.V. Insurance Pty. Ltd. v. Federal Commissioner of Taxation.  Apportionment of the cost over the two years of income therefore accords with both accounting principle and practice and the statutory prescription.

   The correctness of this approach may be illustrated by example.  Let us suppose that the taxpayer raises finance by long term rather than short term bills, drawing bills which mature ten years after the date on which they are drawn and discounting them immediately.  The amount of the discount would be very substantial having regard to the very long life of the bills so that the deduction of the difference between the face value of the bills in the year in which they are drawn and the amount realized by discounting the bills, if permitted, would lead to a distortion of the taxpayer’s operations on revenue account in the year of income in which the bills are drawn and would open the way to inflating very considerably the amount of allowable deductions under s 51 for that year.

   Once the Court rejects the approach adopted by the Full Court of the Federal Court and as well the primary argument of the taxpayer that the entire cost is an allowable deduction in the year of income, it follows that the total cost should be apportioned and, having regard to the relatively short life of the bills and notes, the apportionment should be on an accounting straight line basis over the term of the relevant note or bill.  On this aspect of the case, the parties were not in dispute.”

6. On its face the judgment may be taken as concluding that the taxpayer had incurred the whole of the loss or outgoing when it came under the present liability to pay the notes or bills as the case may be and that there was then a second question, which like the question of the revenue character of the loss, fell for separate consideration, namely whether the loss was in whole or in part “referable to the year of income”.  However, it is difficult to accept that this was the case.  The question whether the loss or outgoing is on revenue account is a question raised by the express language of the statute as a criterion for deductibility.  The statute makes no reference to any test of referability.  It proceeds literally on the basis that the deductibility of a loss or outgoing depends upon whether it is incurred in the year of income.  If it is, and subject to the statutory tests of inclusion being satisfied and those of exclusion being absent, the loss or outgoing is deductible.  It seems more likely, therefore, that their Honours were distinguishing between the incurring of the present liability, an event which occurred before the year of income concluded and the incurring of the loss which their Honours took as occurring over time and thus incurred over time.  It was in this latter context that the test of referability was raised.

7. The High Court in Commissioner of Taxation v Energy Resources of Australia Ltd (1996) 185 CLR 66 at 75 discussed briefly what had been said in Coles Myer.  Again the case was one where what was incurred was a loss on the maturity of promissory notes.  Dawson, Toohey Gaudron, McHugh and Kirby JJ said:

   “The taxpayer contended that it incurred a loss or outgoing on the maturity date of each promissory note.  However, that contention is contrary to the holding in Coles Myer Finance Ltd v Federal Commissioner of Taxation where this Court held that, for the purpose of s 51 of the Act, a taxpayer incurs a liability in respect of a promissory note upon its issue and not upon its payment. Section 51(1) ‘has been interpreted to cover outgoings to which the taxpayer is definitively committed in the year of income although there has been no actual disbursement’. If there is a presently existing liability to pay a pecuniary sum of a revenue nature, the taxpayer incurs an outgoing for the purpose of s 51(1). Similarly, when the difference between a receipt and a later payment constitutes a loss that is an allowable deduction under the sub-section, the loss will usually, but not always, be incurred for the purpose of s 51(1) when the liability to pay becomes binding. So, in Coles Myer, the Court held that the difference between the discounted proceeds of an issue of promissory notes and the amount payable on their maturity was a loss or outgoing that was incurred when the note was issued.  However, the majority in that case held that the cost of the discount was deductible in the year in which it was incurred only ‘to the extent to which it is incurred in gaining or producing the assessable income’ of that year.”  (underlining added)

8. Senior Counsel for Mercantile Mutual submitted that the test of referability as discussed in Coles Myer should be confined to situations where what was to be deducted was a loss, rather than an outgoing.  Certainly the facts of the case were concerned with the deductibility of a loss and not the deductibility of an outgoing, although in the passage from ERA set out above the Court seems not to have distinguished between loss and outgoing in explaining Coles Myer.  Nor is it, with respect, clear to me why the test of referability (treated as being part of the concept of “incurred”) should as a matter of principle, apply to the deductibility of losses, but not to the deductibility of outgoings.It is true, however,  that it is easier to conceive of losses being incurred over time than it is to conceive of an outgoing (that being the entire quantum of the liability  to be expended in the future) being incurred over time.  But interest, which accrues from day to day is apportionable, and has been held to be in this Court in Federal Commissioner of Taxation v Australian Guarantee Corporation Ltd (1984) 2 FCR 483.

9. The passage cited above from ERA does, except in the part underlined, however, support a differential approach as between present liabilities which are incurred at the moment the present liability to pay arises when the outgoing is incurred, and losses which may be deductible over time.  It supports also, however, read as a whole, the view that there may be some losses or outgoings which may be wholly incurred in the year of income, but whose deductibility in the year is limited to the extent that the loss or outgoing is incurred in gaining or producing the assessable income of that year.

10. Before turning to cases subsequent to Coles Myer it is useful to say something about New Zealand Flax on which the decision in Coles Myer was based, for that may illustrate better what it was that the High Court meant.

11. As I pointed out in Ogilvy & Mather Pty Ltd v Commissioner of Taxation (supra) the facts in New Zealand Flax were complex, as was the scheme which the taxpayer had adopted.  The taxpayer agreed to issue bonds.  Payment at the option of the purchasers of bonds could be made cash down with the balance over two and a half years.  Within five years the taxpayer was to purchase land, plant flax and erect a mill.  In the first four years holders of fully paid bonds were to receive interest at 7% per annum.  The company paid commissions to salesmen for bonds as and when they were paid or instalments were paid.  The company’s accounts treated the face value of bonds sold, whether fully or partly paid for as income and made a provision for all its future expenditure including the purchase of land and mill, the planting of flax, interest and commissions  The Court held that subscription monies for bonds not received in the year of income were not derived in that year as assessable income.  It appears to have treated this amount as having the character of remuneration for services, rather than as capital, a matter not raised in the objection (see at 205).  Clearly the full provision for outgoings could not be allowed in the year of income when the bonds monies had not yet been derived.  Only a proportion could be allowed referable or attributable to the proceeds of the bonds.  The future interest might or might not have properly been deductible up front – that required further inquiry – likewise commission.  It was for that reason that his Honour was of the view that the assessments should be remitted to the Commissioner to allow whatever part of the deductions claimed for interest and deferred commissions as appeared referable to the accounting periods under assessment.  In the context referability related to the connexion between the interest or commission on the one hand and the income to which it related, namely the bond proceeds on the other.

1. Coles Myer was the subject of some discussion by a full court of this Court in Commissioner of Taxation v Woolcombers (WA) Pty Ltd (1993) 47 FCR 561. In that case the taxpayer had entered into forward contracts to purchase wool. The purchase price of the majority of contracts was fixed as at the date of purchase, although some were variable. The obligation to pay arose in the future. The taxpayer sought a deduction for its liability under the contracts. Expert accounting evidence was divided as to the correct accounting treatment. However, the taxpayer did not treat the liabilities under the forward contracts as trading expenses in the profit and loss account. Counsel for the taxpayer sought to restrict the application of Coles Myer to the situation where a deduction was sought for a net loss, and as not applying where a deduction was sought for a liability.  This submission was not expressly accepted by the Court.  Beaumont, French and Foster JJ, after referring to Coles Myer and New Zealand Flax in both of which cases apportionment had been held appropriate, said at 573:

   “In our view, Lee J was correct in his analysis of the effect of the contractual provisions, that is to say, as has been noted, that ‘[t]here was [here] an accrued obligation or present liability imposed on the taxpayer by a definite contractual commitment’.  It follows, in out view, that an ‘outgoing’ was ‘incurred’ within the meaning of s 51(1) in the 1988 year; and that no apportionment was appropriate.”

1. Their Honours then continued,  after determining that there was a present liability payable in the future:

   “As has been said, the complexity of the scheme in New Zealand Flax called for an inquiry of the kind there ordered.  In our  opinion, there is no such complexity here.  In Coles Myer, because of the special nature of the financing transaction, it was held, by the majority, that apportionment was appropriate.  Likewise, in the financial arrangements considered in Australian Guarantee, apportionment of the total sum of the interest was proper.  But there are no similar features in the present matter, which concerns a relatively simple forward contract for sale without any financing aspect;  no question arises here of a liability accruing daily, as interest does, or otherwise accruing periodically.”

2. Woolcombers has recently been considered by Nicholson J in Merchant v Commissioner of Taxation [1999] FCA 49, 5 February 1999 (unreported), although the deduction there in question was disallowed in full by his Honour for another reason not presently relevant.

3. It is difficult to discern the reasoning in Woolcombers which led to the conclusion that apportionment was not ‘appropriate’ in that case, unless it was the general proposition that outgoings which are presently existing liabilities are incurred for the purposes of s 51 (1) in the year those liabilities are incurred.  But such a principle would be inconsistent with New Zealand Flax and Australian Guarantee Corporation discussed above.

4. No doubt it is preferable to consider each case as it arises rather than to lay down some general principle.  However, I am prepared in the present case to accept the general principle that to be incurred an outgoing has to be referable to the year in which the liability is incurred applies both to losses and outgoings.  But that principle appears to mean, as New Zealand Flax makes clear, that the outgoing or loss must be referable to the income of the year to which it relates, which in New Zealand Flax had not been derived. In the case of insurance the outgoing which becomes payable as a result of the contingency insured against is referable to the insurance premium income which is derived year by year, including the year in which that event occurs. That has been the basis up to the present for allowing a deduction in the year in which the claim arises, although it may be conceded that the matter was never argued. The accounting evidence shows that what is required in preparing accounts is the calculation of the amount which it would be necessary in the year of claim to set aside and which with interest as it is earned will when the claim becomes payable produce the amount to be paid out. That is not the outgoing incurred at all, it is a provision directed at a different question. To relate the outgoing required to be paid to investment income which subsequently might be earned on a capital sum set aside in the accounts is not what s 51(1) is concerned with. Accordingly, in my view, there is no room in the present case for reducing the amount of the liability which has accrued by the making of a present value calculation.

   The Prudential Margin

5. This final matter can be shortly dealt with.  It is true that in calculating the provision in the accounts the prudential margin is applied not to the central estimate but to the discounted value of that estimate.  It is true also that for tax purposes the prudential margin percentage has been applied to the central estimate.  But mathematically the two are the same thing, save that the one figure is a discounted version of the other. Once it is accepted that the application of the prudential margin leads to an estimate which is more likely to be correct than the figure initially arrived at, there is no reason why the adjustment should not be made.  As has already been noted, claims experience still shows that the provision in the accounts, and therefore the central estimate as adjusted, understates the value of claims.

6. I would accordingly dismiss the appeal with costs.

I certify that the preceding seventy-one (71) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Hill

Associate:

Dated:             1 April 1999

IN THE FEDERAL COURT OF AUSTRALIA
NEW SOUTH WALES DISTRICT REGISTRY	NG 963 of 1998 NG 964 of 1998

ON APPEAL FROM A SINGLE JUDGE OF THE FEDERAL COURT

BETWEEN:	COMMISSIONER OF TAXATION Appellant
AND:	MERCANTILE MUTUAL INSURANCE (WORKERS COMPENSATION) LIMITED Respondent NG 963 OF 1998
BETWEEN:	COMMISSIONER OF TAXATION Appellant
AND:	MERCANTILE MUTUAL INSURANCE AUSTRALIA LIMITED Respondent NG 964 of 1998

JUDGES:	HILL, SACKVILLE & HELY JJ
DATE:	1 APRIL 1999
PLACE:	SYDNEY

REASONS FOR JUDGMENT

SACKVILLE J

1. Hill J’s judgment sets out the facts and competing contentions and also analyses the relevant authorities.  I do not need to repeat what his Honour has said.

2. It is important to identify the issue presented for determination to the learned primary Judge and the arguments relied on by the appellant (“the Commissioner”) in this Court.  The Commissioner has chosen to take what might fairly be described as a limited approach to the present case, doubtless reflecting a particular view of the authorities.  The outcome of the appeal must rest with the fate of the arguments relied on by the Commissioner.   However, since the question of the taxation treatment of outstanding claims incurred by liability insurers is a very significant one, it is appropriate, in my view, to identify the limits of the Commissioner’s arguments.

3. Each of the respondents to the present appeals (“the taxpayers”) initially appealed to the Court, pursuant to s 14ZZ(c) of the Taxation Administration Act 1953 (Cth) (“Administration Act”), against an appealable objection decision of the Commissioner. The Commissioner had disallowed each taxpayer’s objection against an amended assessment, in the one case for the year ended 30 September 1989 and, in the other, for the year ended 30 September 1992. Each of the amended assessments quantified the deductions allowed in respect of outstanding claims, pursuant to s 51(1) of the Income Tax Assessment Act 1936 (Cth) (“Assessment Act”), by reference to what the primary Judge described as their “present value” or their “inflated and discounted” value. 

4. As the primary Judge explained, the Commissioner’s approach found expression in Income Tax Ruling IT 2663 (“IT 2663”).  IT 2663 identified, without pretending to be exhaustive, four possible methods of estimating the amount allowable as an income tax deduction for an outstanding claims provision at the end of a year of income.  These were stated as follows (par 110):

   “(a)the additional amount set aside or ‘earmarked’ each year to pay year end outstanding claims in the future;

   (b)a figure which represents the amount which would be required to settle all outstanding claims if they were to be fully paid out at the end of the year of income;

   (c)a figure determined by calculating the present value of expected claim payments using expected inflation rates only; or

   (d)the amount estimated as the quantum to be finally paid out.”

5. IT 2663 then explained why the Commissioner had chosen method (a):

   “111.The method in sub paragraph 110(a) would produce, we believe, a substantially correct reflex of a general insurer’s true (net) income in respect of general insurance business it conducts in each year of income, particularly in relation to its outstanding claims for the year. 

   112.An example is attached to this Ruling to illustrate how the method in sub paragraph 110(a) applies.  In essence, it entails two broad steps namely:         

   (a)An estimate is made at the end of a year of income of the total amount a general insurer expects to pay in future years to finalise outstanding claims ($30,000 in the example being the gross liability of $60,000 less reinsurance proceeds of $30,000);

   (b)A calculation is then made of funds that need to be set aside or earmarked at the end of the year which, when invested, will provide the amount in subparagraph (a) needed to pay out the claims ($18,000 in the example).

   113.The amount deductible for the outstanding claims provision in the income year is the amount arrived at in subparagraph 112(b)($18,000 in the example).

   114.In our view, this method is a proper and reasonable one to use to estimate a general insurer’s outstanding claims provision for income tax purposes.  We consider that it is the most appropriate for a general insurer to use to produce a true and fair view of the insurer’s taxable income for an income year.

   Why other methods have not been adopted

   115.A present pay-out cost method (subparagraph 110(b)) would ignore the realities of a continuous business being conducted by a general insurer.  A present value method (sub paragraph 110(c)) is not one calculated by all insurers and it is not used for accounting purposes.  An estimated final pay-out method (sub paragraph 110(d)) would entail a deduction in the income year of the original estimate for the fully inflated cost of the claims even though any inflation element would be incurred in a later year....”.

6. By virtue of s 14ZZO(b) of the Administration Act, the taxpayers in the present case had the burden of proving that the Commissioner’s assessment was excessive.  They sought to discharge this burden by contending that the Commissioner’s approach was incorrect in law, because the authorities required that the “inflated” or “face value” figure in respect of outstanding claims, and not the “inflated and discounted” or “present value” figure, be accepted as the appropriate amount to be deducted.  To use the primary Judge’s words, the taxpayers argued that

   “the authorities ... establish, as a matter of principle, that the ‘inflated’ or ‘nominal value’ or ‘face value’ figure is the appropriate one for deduction.”

   It was this argument which the primary Judge accepted and which provided the basis for the orders he made.

7. No issue arose in the proceedings at first instance as to whether, if the taxpayer’s contention were rejected, the Commissioner’s decision to employ the present value approach was erroneous on other grounds.   Thus, for the Commissioner to have succeeded, it was not necessary for him to demonstrate that the present value approach was the correct approach.  It was enough for the Commissioner to establish that he was not bound, as a matter of law, to use the face value approach to the calculation of outstanding liabilities adopted by the taxpayers.

8. At some points in the appeal, the Commissioner seemed to accept a heavier burden than was necessary for his case to succeed.  The first ground of appeal, for example, contended that

   “His Honour erred in holding that the appropriate method of determining outstanding claims provisions was an inflate-only method rather than an inflate and discount method or, alternatively, a no-inflate and no-discount method.”

   A ground of appeal in this form assumes a choice between alternatives (albeit that the second alternative has two branches), rather than making the less burdensome contention that his Honour erred in holding that only the face value method could be used to value outstanding claims incurred by the insurer in the relevant tax year.

9. The Commissioner’s submissions in this Court seemed to accept that the face value approach had been authoritatively established as the only method of valuing an existing liability to pay a fixed amount in the future.   The Commissioner may well have taken the view that this proposition flows from the rejection of the concept of “economic equivalence” in Australian taxation law in cases such as Federal Commissioner of Taxation v The Myer Emporium Ltd (1987) 163 CLR 199, at 216-217 and Federal Commissioner of Taxation v Orica Ltd (1998) 154 ALR 1 (HC), at 23-24. Certainly the language used by a Full Court of this Court in Burrill v Commissioner of Taxation (1996) 67 FCR 519 was unequivocal. The Court in Burrill said this (at 523-525):

   “It is a fundamental principle of Australian income tax law that rights to receive money and obligations to pay money are taken into account in calculating income and outgoings, gains and losses, at their nominal value... .

   ... An accruals taxpayer brings to account a loss or outgoing incurred (s 51) even though it may not be defrayed, discharged or borne for some time.  The amount brought to account is the nominal value of the loss or outgoing at the time it is incurred, and not its actual value at that time.  In other words, the taxpayer does not suffer a reduction in the amount of a deduction otherwise allowable by reason that the present value of the loss or outgoing at the moment it is incurred is less than the nominal amount of the money required to discharge it sometime in the future.  An obligation to pay $100 in six months time is less onerous to the taxpayer than an obligation to pay that amount forthwith.  Yet in both cases the deduction allowed is for the full $100.”

10. In the face of this unequivocal language, the Commissioner sought to distinguish cases such as Burrill, on the ground that they involved an existing liability to pay a fixed amount in the future, rather than an existing liability to pay amounts the quantum of which can only be the subject of estimate.   (I leave to one side the fact that in Burrill the liability was not that of the taxpayer but of the Victorian Government.)   The Commissioner argued that, although a liability to pay a fixed amount in the future must be measured by the face value of the liability, it was open to him to value a liability to pay a future amount capable only of estimation by reference to the discounted value of the estimate.   I agree with Hill J that the distinction the Commissioner seeks to draw is not sound.  As a matter of principle, it can make no difference whether the taxpayer’s liability to make future payments involves fixed dollar amounts or amounts that cannot be estimated until they actually fall due for payment.

11. Similarly, in my view, the Commissioner cannot succeed in the present case on the basis of the “referability” principle applied by the High Court in Coles Myer Finance Ltd v Federal Commissioner of Taxation (1993) 176 CLR 640. As Hely J as pointed out, in the High Court’s view, there was a link between portion of the cost incurred by the taxpayer and the derivation of income by the taxpayer in a future tax year. It is difficult to see how the taxpayer’s liability in the present case in respect of outstanding claims can be treated in the same way.

12. I therefore agree with Hill and Hely JJ that the arguments relied upon by the Commissioner do not establish that the learned primary Judge was in error in setting aside the Commissioner’s decision.   Yet I must confess that I reach this conclusion somewhat reluctantly.  The evidence in this case demonstrates what seems in any event to be obvious, that for the purposes of preparing an insurer’s profit and loss account, its liability to meet outstanding claims must be measured by reference to the present value of expected future payments (although there may be a question as to how the precise calculations are to be made).  More importantly, having regard to the issue in the present case, the evidence shows that to measure the value of an insurer’s liability to meet outstanding claims by reference to the face value of those claims is capable of producing serious distortions in the taxpayer’s operations.  So much appears from Accounting Standard AASB 1023, referred to in the judgment of Hill J.

13. The distorting effect of reliance upon the face value of outstanding claims also appears from the evidence of Professor Peirson, Professor of Accounting at Monash University.  Professor Peirson, whose evidence was not challenged in cross-examination, said this in his second report:

    “[T]he liability for outstanding claims must be measured as the present value of the expected future payments.   As a result, to estimate the expected future payments without discounting those expected future payments to a present value will give an incorrect measure of the liability for outstanding claims and, therefore, an incorrect measure of the claims expense.  In turn, this will give an incorrect measure of profit by overstating the expenses and understating the profit.”

    Professor Peirson’s first report set out in some detail the “principle in financial economics .... that money has a time value” and its application to cases where moneys are to be received or paid in the future.  

14. The evidence was primarily directed at establishing, from an accounting viewpoint, the accepted and acceptable approach to the measurement of the value of outstanding liabilities in a particular accounting period.  However, it supports the proposition that the face value method of measuring an insurer’s outstanding liabilities gives an incorrect measure of those liabilities.  The evidence, at least arguably, justifies the inference that, if an insurer’s outstanding long term liabilities are to be measured by their face value, a deduction for outstanding liabilities calculated on that basis will be distorted.  The distortion flows from the fact that the taxpayer will be entitled to a deduction from income derived in year one (largely, if not entirely, measured in year one dollars), the face value of payments not due to be made until, say, year six or year eight.  

15. Neither party to the appeal sought to analyse the accounting and actuarial evidence in detail, in order to shed light on the length of the “tail” constituted by the outstanding liabilities of each of the taxpayers in the relevant years of income.  However, the evidence shows that the “run-off” for outstanding claims for the taxpayers was expected to continue over a very substantial period.   For example, an actuarial report which analysed the outstanding employers liability claims of Mercantile Mutual Insurance (Australia) Limited (“MMIA”), as at 30 September 1990 showed that the run-off was expected to take place over a twenty year period (that is, until the year 2009).  While greatest cash outflows in respect of 1990 outstanding claims were expected to occur in the five years following 30 September 1990, substantial cash outlays were also expected to occur in subsequent years (that is, after September 1995).

16. The very great differences between the face value and the present value of outstanding claims for the second taxpayer, Mercantile Mutual (Workers Compensation) Limited (“MMIWC”) are shown by the following table:

Financial Year	Face Value $m	Present Value $m
1987	149.3	102.0
1988	126.6	80.6
1989	96.1	61.4
1990	68.6	44.5

(These figures include the “prudential margin”.   The declining amounts reflect the winding down of MMIWC’s portfolio during the relevant period.) 

1. I have said that the decision in Coles Myer does not determine the outcome of the present case.  However, the High Court in Coles Myer recognised the difficulty that can arise when applying s 51(1) of the Assessment Act in circumstances where a taxpayer subjects itself to a present liability to pay in a future year of income an amount which generates a net loss or outgoing.  In Coles Myer, the joint judgment (Mason CJ, Brennan, Dawson, Toohey and Gaudron JJ) held that the so called “matching principle”, formulated in RACV Insurance Pty Ltd v Federal Commissioner of Taxation [1975] VR 1, at 14, required the apportionment of the “cost” of bills drawn and sold by the taxpayer over two years of income. In that context, their Honours made the following observation (at 666):

   “The correctness of this approach may be illustrated by example.  Let us suppose that the taxpayer raises finance by long term rather than short term bills, drawing bills which mature ten years after the date on which they are drawn and discounting them immediately.  The amount of the discount would be very substantial having regard to the very long life of the bills so that the deduction of the difference between the face value of the bills in the year in which they are drawn and the amount realized by discounting the bills, if permitted, would lead to a distortion of the taxpayer’s operations on revenue account in the year of income in which the bills are drawn and would open the way to inflating very considerably the amount of allowable deductions under s 51 for that year”.  

1. Deane J in Coles Myer went further.  He expressed the view that, where a theoretically contingent liability to make a future payment constitutes an outgoing or loss for tax purposes, the amount of the loss or outgoing which has been “incurred” at the end of the relevant year will not necessarily be the nominal amount which the taxpayer will ultimately have to pay.  His Honour continued (at 672-673):

   “regardless of whether it be absolute or contingent, the amount of loss or outgoing which has been incurred at the end of the tax year will depend, in the case of an obligation to pay a future amount, upon the circumstances of the particular case .... [t]he circumstances may well be such that the amount or ‘value’ of the loss or outgoing ‘incurred’ at the end of the relevant tax year is clearly less than the full amount of the ultimate liability.  Thus, in the example given above, the liability to pay $1 million on 1 July 2005 could not properly be treated, for the purposes of the tax year ending 30 June 1993, as a loss or outgoing, of $1 million.   For the purposes of that tax year, the amount of the liability would be the amount which represented the appropriate valuation of the liability as at 30 June 1993.  Necessarily, that amount would reflect the fact that no interest was payable on the $1 million until the time when it would actually have to be paid.”

2. The reasoning underlying Coles Myer perhaps might be used to support a valuation of a liability insurer’s outstanding claims by reference to something other than this face value. The joint judgments justified the so called “referability” principle by observing (at 665) that

   “The acceptance by this Court of the jurisprudential analysis of s 51 does not compel the conclusion that, once a taxpayer subjects itself in the year of income on revenue account to a present legal liability to pay in a future year of income an amount which generates, or gives rise to, a net loss or outgoing, the net loss or outgoing is deductible in full in the year of income.”

   The observations already quoted from the joint judgment indicate that, whether a net loss or outgoing will be deductible in full, may depend, inter alia, on whether to do so “would lead to a distortion of the taxpayer’s operations ... in the year of income”. 

3. Of course, the general comments in Coles Myer do not provide a licence to rewrite s 51(1) of the Assessment Act. Nor do they provide a warrant for overturning an established line of authority obliging the Court to hold that an insurer’s deduction for outstanding claims must be measured by the face value of those liabilities, regardless of whether the evidence suggests that this course will distort the taxpayer’s operations for the year of income.

4. Because the Commissioner conceded that a liability to pay a fixed amount in the future was to be measured by the face value of the liability, neither party analysed in detail whether the authorities in truth compel acceptance of the proposition that outstanding claims must always be measured by this face value, regardless of the accounting evidence.  While it is not necessary to attempt to resolve that question, it is perhaps not inappropriate to point out that the precise issue posed by the present case (bearing in mind the accounting evidence) does not appear to have been addressed.

5. The authorities dealing with the insurer’s long-term “tail” have recognised that it is one thing to establish the existence of a present liability to pay an amount in the future, but another to determine the quantum of the deduction that should be allowed in respect of that liability.   In RACV Insurance, Menhennitt J addressed the two issues separately, as did Newton J in Commercial Union Assurance Company of Australia Ltd v Federal Commissioner of Taxation (1977) 32 FLR 32.   In each of these cases, as Hill J has pointed out, the Court accepted estimates, said to be in accordance with accounting principles, based on the face value of the sums payable in the future.  In neither case was it argued that the estimate should be made by reference to discounted values.

6. Similarly, in Federal Commissioner of Taxation v Australian Guarantee Corporation Ltd (1984) 2 FCR 483 (not an insurer case) the taxpayer was held to be entitled to deduct the face value of future interest payments due under deferred interest debentures. The deduction was held to be available in the year the debentures were issued, since that was the year the taxpayer incurred the liability to pay interest. As McGregor J observed (at 501) the grounds of appeal did not raise any contention that the failure to call evidence as to the present value of the liability made the deduction impermissible. The Court was asked to approach the case on a “face value or nothing” basis.

7. In Australia and New Zealand Banking Group Ltd v Commissioner of Taxation (1994) 48 FCR 268, a Full Court of this Court held that the principle established in RACV Insurance applied to a self-insurer.   Hill J, with whose judgment Northrop and Lockhart JJ agreed, accepted that a self-insured employee was entitled to deduct the value of its existing statutory liability to make future compensation payments to injured workers (including periodic compensation in respect of lost earnings) if the quantum of the liability was capable of reasonable estimation.  His Honour said (at 280) that the “task of estimation in a case such as the present is a commercial one”.   On the facts, Hill J held that the estimate made by the taxpayer were bona fide and not unreasonable.   No issue arose as to whether the deduction available to the taxpayer should have been limited to the present value of the future payments.

8. The question remains as to whether the cases rejecting the principle of economic equivalence and the decision in Burrill compel the conclusion that only the face value method of assessing an insurer’s outstanding liabilities is permissible, regardless of the accounting evidence.   As I have already said, I do not think that Burrill can be distinguished simply on the ground that in that case the sum payable in the future was a fixed amount.   Even so, there may be a question as to whether Burrill is distinguishable on other grounds, for example, because it involved the construction of s 70B of the Assessment Act, the terms and purpose of which are different from s 51(1). Similar questions may arise in relation to the economic equivalence cases. However, as they were not explored in argument, I take them no further.

9. I agree with the orders proposed by Hill J. 

I certify that the preceding twenty-six (26) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Sackville.

Associate:

Dated:             1 April 1999

IN THE FEDERAL COURT OF AUSTRALIA
NEW SOUTH WALES DISTRICT REGISTRY	NG 963 OF 1998

ON APPEAL FROM A SINGLE JUDGE OF THE FEDERAL COURT

BETWEEN:	COMMISSIONER OF TAXATION Appellant
AND:	MERCANTILE MUTUAL INSURANCE (WORKERS COMPENSATION) LIMITED Respondent
NG 964 OF 1998
BETWEEN:	COMMISSIONER OF TAXATION Appellant
AND:	MERCANTILE MUTUAL INSURANCE AUSTRALIA LIMITED Respondent

JUDGES:	HILL, SACKVILLE AND HELY JJ
DATE:	1 APRIL 1999
PLACE:	SYDNEY

REASONS FOR JUDGMENT

HELY J:

1. I have had the advantage of reading a draft of the judgment of Hill J which sets out the background to the present appeals, and which contains a comprehensive review of the authorities which touch and concern the issues arising for decision.

2. It is common ground that, in the substituted accounting periods in question, the taxpayers had incurred a liability to make payments in relation to claims, both reported and unreported, which were outstanding at the end of those periods. The contest is as to the correct approach to the quantification of the allowable deduction in respect of such claims pursuant to s 51(1) of the Income Tax Assessment Act 1936 (Cth) (as amended) (“the Act”).

3. By way of illustration, Mercantile Mutual Insurance (Workers Compensation) Ltd (“MMIWC”) claimed to be entitled to a deduction of $96,115,000 in relation to the 1989 financial period.  In establishing that claim, an actuarial calculation was made of the expected future payments in terms of current values using the development pattern for claims derived from past experience.  The current value projections were then inflated, so as to assess the payments which will actually emerge in the future, taking account of the likely impact of inflation generally, and claims inflation in particular.  That produced a figure which is referred to in the evidence as “inflated (undiscounted)” or “the central estimate”.  A prudential margin of 10 percent was added to this figure, and other, but uncontroversial adjustments made, resulting in a claimed deduction of $96,115,000.

4. The corresponding outstanding claims provision included in the MMIWC’s accounts for the financial period was $61,435,058 (the “inflated and discounted” sum).  That sum represented the present value of the inflated, undiscounted figure, plus a prudential margin of 10 percent and other, but uncontroversial adjustments.  That accounting treatment of the provision is consistent with AASB 1023 “Financial Reporting of General Insurance Activities” par 5.1.1:

   “The liability for outstanding claims will need to be measured as the present value of expected future payments.  This measurement approach is based on the view that the liability for outstanding claims ought to reflect the amount which, if set aside as at the reporting date, would accumulate so as to enable the insurer to pay the amounts of claims as they fall due.  The approach requires estimation of the ultimate cost of settling claims and the discounting of those amounts to a present value.”

   An allowable deduction limited to the provision in the accounts

5. The Commissioner’s first submission is that the appropriate deduction is the provision made in the taxpayer’s financial accounts, ie the inflated and discounted figure.  I agree with Hill J that this submission should be rejected.  The foundation for the submission is that insurance companies are treated as an exception to the ordinary rule that a provision for a future payment will not be allowed as a deduction.  But, to put the matter in that way is to confuse the estimation of the amount of a liability which has been incurred, with the making of a provision for a liability which has not, but which is merely “impending, threatened or expected”.

6. A provision of the latter type is not an estimate of the actual amount of the liability which has been incurred; it is an estimate of the amount which must now be set aside so that when invested, it will in the future produce a fund sufficient to meet the liability when it emerges.  A provision of that type is, of course, not an allowable deduction.  Thus the submission is based upon an insecure foundation.

7. The Commissioner accepted (subject to questions of “referability” to be addressed later) that fixed amounts which are receivable or payable in the future are not discounted back to their net present value for tax purposes.  The amount brought to account is the nominal amount, or face value, of the item in question.

8. The authorities referred to by Hill J, particularly and most recently Burrill v Commissioner of Taxation (1996) 67 FCR 519, support that approach. The Commissioner seeks to distinguish those authorities upon the basis that, in the case of insurance companies, the deduction is allowed, as counsel put it, “in anticipation of” the ultimate liability. This misconceives the nature of the deduction; the deduction is allowed for a presently existing liability, the quantum of which is established by a process of estimation.

Coles Myer – Referability to the year of income

1. The Commissioner’s second submission is that insofar as there is included in the central estimate an allowance for the effects of inflation expected to be encountered in future years, there is an impermissible attempt to obtain a deduction in the year of income for matters which are not “referable” to it.  A deduction for the effects of inflation on the quantum of the claim is allowable in the years in which the inflationary effects are encountered, rather than in the financial period in which the premium is received and the outstanding claims liability first assessed.

2. The Commissioner’s submissions on this alternative argument assumed that the consequence of their acceptance was that the appropriate deduction is the provision made in the taxpayer’s financial accounts, ie the inflated and discounted sum.  It is by no means obvious to me that this is so, as acceptance of the submission would involve the excision from the estimate of the effects of future inflation, which would take one back to the assessment of expected future payments in current values.  But it is not necessary to pursue the issue of quantification, as the submission fails for other reasons.

3. The authorities establish that where the amount of an allowable deduction is an estimate, later revision of the estimate as more becomes known, will give rise to a further deduction, or to assessable income, depending upon whether the original estimate was too low, or too high: Texas Co (Australasia) Ltd v Federal Commissioner of Taxation (1940) 63 CLR 382; Commercial Union Assurance Co of Australia Ltd v Federal Commissioner of Taxation (1977) 32 FLR 32 at 48.

4. But where the object of the exercise is to assess, whether on a case by case basis, or by actuarial means, the quantum of the liability based on the probabilities (see Australia & New Zealand Banking Group Ltd v Commissioner of Taxation (1994) 48 FCR 268 at 281B) there is no reason for leaving out of account the effect of the passage of time before the claim is likely to be paid and of inflation during that period, when the probabilities are that the amount to be paid will be affected by those factors.

5. What is incurred in the year of income is the liability to meet the claim when it matures.  What is to be estimated is the amount payable on maturity of the claim.  The estimated amount is an allowable deduction, and it is beside the point that a component of the estimate reflects an expectation that by reason of future inflationary factors the amount to be paid in the future will be higher than might otherwise be the case.

6. Insofar as the matching principle is relevant, losses or outgoings are to be matched against the revenue to which they relate.  The estimate established for outstanding claims at the end of the financial year is to be matched against the premium income derived in or with respect to that period: Commercial Union Assurance Co case at 48-9.  In the present case, unlike Coles Myer (1993) 176 CLR 640, there is no link between the increased cost reflected by the expected impact of inflation in subsequent periods and the income to be derived in those periods, so as to warrant the apportionment of the cost over the period of the tail of the claim. There is or may be some nexus between the provision established in the accounts, and the investment income expected to be derived in later years. But it is not the provision which is claimed as the allowable deduction. It was critical to the decision in Coles Myer that the loss arising from the discounting of the bills represented the cost of acquiring funds which were put to profitable advantage in both the years of income over which the cost was apportioned.

7. In my respectful opinion, whatever the rationale for the difference between the view of the majority in Coles Myer and that of McHugh J, it does not bear upon the outcome of the present case on any view, because, the whole of the outgoing constituted by the estimated liability is referable to the year of income in which the estimate was established, there being no nexus between that outgoing, and income expected to be derived in later years.

8. The accounting evidence which was called in this case was directed towards establishing that, from an accounting viewpoint, the correct measurement of profit for the periods in question required that the provision for outstanding claims be established by a process of inflating and discounting, ie the inflated and discounted sum.  Our attention was not drawn to any accounting evidence to the effect that the application of the “matching principle” required that the effects of inflation in years following the substituted accounting periods in question should be deferred until those effects are encountered.

9. In the view which I take it is not necessary to form an opinion as to whether the referability principle referred to in Coles Myer, when considered in the light of Commissioner of Taxation v Energy Resources of Australia (1996) 185 CLR 66 at 75, applies to both losses and to outgoings.

The Prudential Margin

1. The prudential margin was applied so as to increase what would otherwise have been the outstanding claims provision in the company’s financial accounts.  It was also applied so as to increase the inflated undiscounted sum.  As a consequence of a concession made by the Commissioner, the propriety of including a prudential margin does not arise if the Commissioner’s primary submission that the appropriate deduction is the provision made in the taxpayer’s financial accounts, ie the inflated and discounted figure, were upheld.  As this submission fails it is necessary to consider the propriety of including the prudential margin in the amount for which a deduction is claimed.

2. The Commissioner’s contention is that the best estimate of the outstanding claims is the central estimate.  Having estimated the amount of the liability, the amount of the allowable deduction cannot be increased by adding a margin for prudence or for error.  The addition of such a margin does not make the estimate more accurate.  A margin set aside for prudence is, by definition, an amount which is in excess of that estimated as the amount of the liability.  Whilst one may add a prudential margin to a provision with a view to increasing the chances of it proving sufficient to meet the claims experience that does not alter the amount of the estimated liability.

3. As to this, the trial judge made the following finding:

   “In my opinion, the evidence amply demonstrates that its (ie the prudential margin’s)  inclusion was a proper commercial decision and one that rendered the estimate of the amount of the future liability more reliable.  Indeed, according to the evidence, the relevant estimates have fallen short of the amounts actually required to provide the indemnity contracted for.  Once it is accepted that the figure for deduction can be arrived at by estimation, then the inclusion of this allowance can be seen as an integral and proper step in that process.”

4. The question is one of fact.  No error has been shown in the trial judge’s conclusion that the inclusion of the prudential margin was part of the process by which the amount of the liability incurred was estimated, nor in his Honour’s finding that inclusion of the prudential margin improved the reliability of the estimate.

5. I agree that the appeal should be dismissed with costs.

I certify that the preceding twenty-two (22) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Hely.

Associate:

Dated:             1 April 1999

Counsel for the Appellant:	D H Bloom QC;  G T Pagone QC;  K M Connor
Solicitor for the Appellant:	Australian Government Solicitor
Counsel for the Respondent:	R F Edmonds SC;  B J Sullivan
Solicitor for the Respondent:	Ernst & Young Legal Services (NSW)
Date of Hearing:	25 February 1999
Date of Judgment:	1 April 1999