Caltex Ltd v Federal Commissioner of Taxation

106 CLR 205

---

(Judgment by: Taylor J)

Between: Caltex Ltd
And: Federal Commissioner of Taxation

Court:

Judges: Dixon CJ
Fullagar J
Kitto J

Taylor JMenzies J

Subject References:
Taxation and revenue
Income tax
Debt payable in foreign currency
Exchange loss
Allowable deduction

Legislative References:
Income Tax Assessment Act 1936 (Cth) - s 51(1); s 80; s 136

Hearing date: 7 September 1959; 8 September 1959; 9 September 1959; 10 September 1959; 1960 September 1959; 5 April 1959
Judgment date: 5 April 1960

SYDNEY

---

Judgment by:
Taylor J

According to its financial records as they stood at the end of June 1936 the appellant, a company incorporated pursuant to the laws of New South Wales but controlled principally by non-residents, owed to a company, described as The Texas Company -New York, the sum of 10,186,408 United States dollars payable in New York. The latter company is referred to in the case as the old supplier. The business of the appellant was that of importing into and selling in Australia petrol and petroleum products and its indebtedness was for stocks of these products and for some minor capital items which had been supplied to it by the old supplier over a number of years prior to 30th June 1936. As and from 1st July 1936 no further orders for stock were placed with the old supplier though supplies continued to come forward in fulfilment of past orders until September 1936. On this account the appellant, during this latter period, incurred a further dollar liability to the old supplier payable in New York. The additional amount was, apparently, 643,605 dollars. The appellant's aggregate indebtedness to the old supplier was discharged by two payments made during that year, namely, a payment of 7,000,000 dollars made in New York on 1st July 1936 and a further payment of 3,830,013 dollars in the same place on 27th November 1936. The circumstances in which these payments were made is of prime importance in the case and it will be necessary shortly to discuss the material facts as they have been agreed upon by the parties.  

Before doing so, however, it should be mentioned that on and after 1st July 1936 the appellant placed its orders for supplies of petrol and petroleum products exclusively with The California Texas Oil Company Limited, a company which was formed in New York at the beginning of that year. This company, which was formed by the old supplier and the Standard Oil Co of California, is referred to in the case as the new supplier. The course of the appellant's trading with this company is of no consequence in the case; it is mentioned only because of the part which it played in connection with the discharge by the appellant of its liability to the old supplier.  

As already appears the appellant had for many years obtained its supplies of stock from the old supplier. Its purchases were frequent and extensive and for a number of years during which it incurred liabilities for purchases there were marked variations in the exchange rate between Australia and the United States of America. In general, the movement was against Australia though there were periods of partial recovery. In these circumstances, the appellant adopted the practice, as detailed in the case stated, of making monthly adjustments to the old supplier's account in its books- which were kept in dual currency-in order to reflect in the credit entries in Australian currency the fluctuations of the exchange rate from time to time. Of necessity a system of averaging was employed with respect to balances carried forward in this account from month to month but it is unnecessary for the purposes of the appeal to discuss, in detail, the method employed. What is of importance to observe is that when on 1st July 1936 and 27th November 1936 the appellant made the two payments mentioned in discharge of its liability to the old supplier the Australian equivalent of the United States dollars then paid was substantially higher than, either, the cost at which its supplies had, in Australian currency, originally been taken into account, or, the amount at which, by reason of adjustments of the nature mentioned, that cost then stood in the appellant's books. In these circumstances, the appellant claims that in 1936 it incurred expenditure which pursuant to s. 51 of the Income Tax Assessment Act 1936-1940, was an allowable deduction from its assessable income. It should be mentioned that by arrangement with the respondent the appellant had adopted calendar years for the purposes of its accounting and, therefore, it was said, a deduction on this account should have been allowed in respect of the year of income which ended on 31st December 1936.  

According to the method of accounting adopted by the appellant the so-called "exchange loss" which it suffered amounted to nearly PD400,000 but, after excluding so much of this as was referable to items chargeable to capital account, it is claimed that a deduction of PD380,000 should have been allowed. The respondent denies that any such "exchange loss" was incurred; he claims that the nature and effect of the relevant transactions were not such as involved the appellant in any loss of that character. But he does concede that, if, in the circumstances of the case, it is proper to conclude that such a loss was incurred, it was substantial in amount and that if it had been taken into account in the year ended 31st December 1936 it would have resulted in a trading loss which, apart from one matter which requires our special consideration, might have been carried forward to the year ended 31st December 1939 pursuant to the provisions of s. 80 of the Act. The particular matter to which I refer is that it was contended by the respondent that for the year ended 31st December 1939 the appellant was assessed pursuant to the provisions of s. 136 of the Act and that this circumstance precluded the appellant from taking into account in that year any unabsorbed loss or losses from previous years. We are, therefore, concerned to inquire first of all whether in the circumstances the appellant did, as claimed, incur revenue "losses" or "outgoings" of the nature suggested in gaining or producing its assessable income or in carrying on its business for the purpose of gaining or producing such income and, if so, whether the appellant was, pursuant to s. 80, entitled to a deduction in the income year ended 31st December 1939 of any part of the loss incurred in the earlier year.  

In approaching the first question it is, of course, necessary to examine the circumstances in which the two payments in question were made. What had happened was that either because the appellant company needed working capital for the conduct of its business, or, because it became uneconomic to convert depreciated Australian currency to United States dollars, or, what is more likely, because of a combination of these factors, the appellant's liabilities to the old supplier had, to a large extent, been allowed to remain outstanding. Remittances were, however, made from time to time and "exchange losses" were incurred when this was done. These "losses" were the subject of discussion in an earlier case before the appellant assumed its present name (Texas Co (Australasia) Ltd v Federal Commissioner of Taxation ^[F25] ) when it was held that they constituted allowable deductions pursuant to s. 51. It does not appear precisely how these remittances were made but, presumably, each one was made through the appellant's bankers who, at the cost of the appellant in Australian currency, made available to its use in New York an equivalent amount in United States dollars which was then used to discharge, by payment in New York, part of its outstanding liability. Whether or not the cost to the appellant was contemporaneously discharged by it or merely charged to its account does not appear. But this is of little moment for it either expended the necessary amount in Australian currency or incurred an obligation to pay it at some future time and, in either case, it was in a position to assert that in making these payments to the old supplier it had incurred revenue "losses" or "outgoings" of the character specified in s. 51.  

This, it seems to me, is of the essence of the decision in the earlier case and it is not out of place to examine why the members of the Court came to this conclusion. No member of the Court experienced any difficulty in holding that a trader incurs an additional "loss" or "outgoing" when, in purchasing foreign currency to discharge an overseas debt, he is compelled to lay out a larger sum in Australian money than it would have been necessary for him to find if he had paid for his goods at the time of their purchase. Nor, does the contrary appear to have been seriously contested by the respondent. What was contended in answer to the taxpayer's claim was that any additional expenditure so incurred in any one year in respect of purchases made in an earlier year was not incurred in the course of gaining or producing the assessable income of the year of remittance and, alternatively, that, in the special circumstances of the case, the additional expenditure in question was of a capital nature. The latter argument was founded on the assertion that the creditor had allowed the indebtedness to remain outstanding over a lengthy period in order to provide the debtor with working capital. But the members of the Court directed their attention to the initial point as well as to the problems involved in these contentions and, in the result, it was held that the taxpayer was entitled to deductions as claimed. Latham C.J. took the view that the fact that a taxpayer

"has made a preliminary estimate of the amount required to discharge his foreign debts does not ... preclude him from claiming later a deduction of any increased amount which in fact he has to pay". ^[F26]

Dixon J. (as he then was) took a somewhat different view. He was disinclined to accept the proposition that any additional expenditure so incurred should be regarded as "a falsification of a prior estimate". ^[F27] Rather, he thought, it should be regarded as part of the cost of defraying the original liability and he added

"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 borne as a business expense". ^[F28]

There was, however, common agreement that the additional expense or cost could be taken into account in the year of remittance. I should, perhaps, add that Starke J. understood the Commissioner to concede that "such losses or outgoings might, in circumstances other than exist in the present case, be legitimately claimed as deductions under and by virtue of the Income Tax Assessment Act 1922-1934". ^[F29]  

At first sight it might appear that the present case is no more than a repetition of its predecessor. But, although the decision in the earlier case solves some of the problems which might, otherwise, have arisen in the present case, it is said that there is a significant difference between the manner in which the earlier payments were made and that in which the two final payments were made. There was, of course, no such difference. Every payment, whether preceded by a remittance, or not, was a payment made in dollars by the appellant to the old supplier in New York. But the dollars required for the two final payments were obtained by the appellant in a manner quite different from the foreign currency which was the subject of the earlier "remittances". When the appellant was about to close its account with the old supplier it was decided in consultation with the new supplier that its outstanding indebtedness to the former should be discharged. In pursuance of this decision the new supplier, on 1st July 1936, advanced to the appellant in New York a sum sufficient to enable the latter to make a payment of 7,000,000 United States dollars to the old supplier. This advance was made by the old supplier paying the necessary sum in dollars to the credit of the appellant's bank account in New York. Thereupon, on the same day, the appellant drew a cheque on that account for 7,000,000 dollars and paid it into the New York account of the old supplier. The old supplier's account in the appellant's books was on the same day debited with the amount paid in dollars and with the Australian pounds equivalent obtained by converting the amount so paid at the rate of exchange then prevailing which was approximately four dollars to the pound. Dollars for the final payment on 27th November 1936 were provided by an advance made in precisely the same manner so that it will be seen that the provision of dollars to meet these two payments did not involve the appellant in the immediate expenditure of any Australian currency. It did, no doubt, incur an obligation to repay the loans which were made to it but this obligation was, it seems to me, on capital account and any subsequent "exchange loss" incurred in discharging that liability was for that account also.  

It is this last-mentioned circumstance that is said to differentiate the present case from the earlier case. Whilst it is conceded that if the appellant had laid out Australian currency to purchase the United States dollars necessary to make the two final payments it might properly have claimed as a deduction so much of the amount expended as exceeded the cost at which its supplies had originally been taken into account in Australian currency, the assertion is made that, since there was no expenditure of Australian currency, there was no transaction involving exchange and, therefore, no so-called "exchange loss". It should, perhaps, be mentioned that it is suggested that the decision in the earlier case supports the contention involved in this proposition and some members of the Board of Review seem to have thought that it did. But the decision in that case was concerned only with "remittances" made in the manner shown by the facts before the Court and it is no authority for the proposition that a so-called "exchange loss" can occur only when that procedure is followed in discharging an obligation to pay foreign currency in another country. It is, of course, true that if an Australian company has dollar credits of its own readily available in the United States or, if it obtains dollar credits in some other way without laying out Australian currency, it does not incur an "exchange loss" in the sense that it purchases dollars at an unfavourable rate of exchange. But a trader does not incur an "exchange loss" merely by purchasing foreign currency in those circumstances; the result of his purchase is that he obtains money's worth for his Australian currency. Indeed, immediately after his purchase he may reconvert his foreign currency to Australian currency without loss. His "exchange loss" occurs when he expends his foreign currency so obtained in payment for goods purchased at a time when the rate of exchange was less unfavourable. In the result he incurs an additional expenditure either as part of the cost of his goods or as part of the cost of paying for them. But if he has readily available dollar credits of his own can it be said that he does not incur any additional cost if, instead of purchasing a draft, he resorts to those credits for the purpose of discharging his liability? It is, I think, incontestable that an Australian trader who, by payment, discharges a liability to pay a million dollars in New York, expends in terms of Australian currency precisely the same amount whether he has his own dollar funds available or whether he purchases a draft or adopts some other means of securing the necessary United States funds. In either case his expenditure will, for the purposes of his business, be taken into account in Australian currency. Indeed, since s. 20 of the Income Tax Assessment Act 1936-1939 required that, for all the purposes of the Act, income wherever derived and any expenses wherever incurred should be expressed in terms of Australian currency, this is precisely what the appellant was required to do.  

These latter observations serve to indicate that a too literal understanding of the expression "exchange loss" may tend to obscure the problem in the case by suggesting that the fundamental question is whether the taxpayer has actually exchanged an amount in one currency for its equivalent in another. But the fact that, in such a case, the trader obtains an equivalent in foreign currency denies that the transaction, at that stage, results in a loss. His loss, as already pointed out, occurs when he discharges an overseas liability contracted at a time when the rate of exchange was less unfavourable to him. Likewise, he may incur an "exchange loss" as a payee. If an Australian trader were to sell goods on credit to a United States buyer for a specified number of dollars payable in Australia he would incur such a loss, if before payment, the rate of exchange were to move in favour of Australia. Initially, the sale price would be taken into his accounts in Australian currency at the rate of exchange prevailing at the date of the sale. But, although upon payment he would be paid the equivalent of the specified number of dollars, he would be entitled to receive and would actually receive a lesser sum in Australian currency than that originally taken into account. (Anderson v Equitable Assurance Society of the United States, ^[F30] Ottoman Bank of Nicosia v Chakarian, ^[F31] Cummings v London Bullion Co Ltd, ^[F32] In re Russian Commercial and Industrial Bank, ^[F33] and in Re United Railways of the Havana and Regla Warehouses Ltd. ^[F34] ) To my mind the defect in the respondent's argument is that it selects as its basis an expression-"exchange loss"-which is no more than a commercial term conveniently used to denote some of the effects which fluctuating rates of exchange may produce in trading transactions and then, after giving a too literal meaning to the expression, it denies that any such loss can occur except in the case of a trader who has exchanged a sum in one currency for its equivalent in another. The problem is, however, considerably wider than the literal meaning of that expression would appear to suggest to some minds. But it cannot be solved in the manner initially suggested on behalf of the appellant. At the outset of the appeal it was suggested that the mere fact that in 1936 the appellant's dollar liability to the old supplier, expressed in terms of Australian currency, had by adjustment in its books assumed a magnitude in excess of the cost at which its supplies had originally been taken into account establishes that an "exchange loss" was made. But mere entries in books of account do not create "exchange losses" although they may, in appropriate circumstances, reveal them. What we are concerned with are not merely entries in books of account; the critical question is whether, in making the two final payments, the appellant incurred losses or outgoings within the meaning of s. 51. To my mind the answer to this question must, in the circumstances of the case, be answered in the affirmative, for the difference between the amount at which its cost for stock was originally taken into account and the cost of discharging its liability, both being expressed in terms of Australian currency, truly represented an additional cost incurred in carrying on its business. The use of a trader's own dollar funds to discharge a dollar indebtedness overseas means, of course, that the trader is no longer in a position to convert the dollars used to Australian currency. And, expressed in terms of that currency, he incurs an additional cost which is just as real and significant as if it had been necessary for him to expend Australian currency to secure dollar funds for the purpose of making the payment in question. The truth of this proposition may, I think, be demonstrated by a simple illustration. An Australian trader sells goods in New York on 30th June and contemporaneously receives there the sum of a million dollars as the purchase price. Hypothetically, the rate of exchange that day is four dollars to the PDA. For the purposes of Australian income tax he will, therefore, be required to bring into account the sum of PDA.250,000 as assessable income. But if, on 30th June, he owes in New York the sum of 1,000,000 dollars for goods purchased six months previously-when the rate of exchange was five dollars to the PDA-he may use his 1,000,000 dollars credit to discharge this liability. If he does he will be left without any dollars to remit to Australia. But, unless the view which I have expressed be correct he would be required, for income tax purposes, to bring into account as assessable income the sum of PDA.250,000 and would obtain no deduction in respect of his purchases over and above the sum of PDA.200,000 which was his cost, or estimated cost, at the time when he made them. This would mean, of course, that although he had received 1,000,000 dollars in New York and had expended the whole of that amount there on the same day to pay for his earlier purchases, he would be treated as if he had retained a surplus, expressed in Australian currency, of PD50,000.  

The respondent, however, maintains that even if the views already expressed be correct they do not conclude the answer to the problem in favour of the appellant. As was pointed out it is, of course, true that the new supplier might have become substituted for the old supplier as the appellant's creditor without the latter incurring any so-called exchange loss. Such a result might have been achieved by an assignment of the old supplier's debt to the new supplier or, with the concurrence of all three parties, by the process of novation. If either of these courses had been adopted it would not have been possible for the appellant to say that it had expended dollar currency to discharge its debt. The fact is, however, that there was no assignment of the debt and there was no agreement by way of novation; the old supplier's debt was discharged by payments made by the appellant out of its own moneys. Yet, since the moneys lent to the appellant by the new supplier were lent for the purpose of enabling it to make these payments the respondent asserts that we should deal with the case as if there had been an assignment or an agreement by way of novation. To my mind the answer to this contention is that the appellant discharged its dollar indebtedness for stock-in-trade out of its own dollar funds and it is of no consequence that it borrowed dollar currency to enable it to make the payments. The substance of the argument for the respondent on this point is that, even if the appellant would have been entitled to succeed if the payments to the old supplier had been made out of dollar funds available to it in New York for its general purposes, the fact that the loans made by the new supplier were made to it for the express purpose of making those payments precludes it from asserting that it suffered any so-called exchange loss. This is an argument which I find difficult to understand for the payments were in fact and in law made out of its own moneys and, to my mind, it is of no consequence whether it borrowed the moneys which enabled it to make the payments in question for that specific purpose or not. Again, it was said, that the loan from the new supplier was on capital account and that, therefore, any "exchange loss" suffered was of the same character. But this contention should be rejected. It may be-though at this stage it is impossible to say-that an "exchange loss" will be incurred when, and if, the loan is repaid and if this should occur the loss will, of course, be on capital account. But the mere fact that the loan represented a capital liability does not mean that when the moneys advanced were used for the purpose of discharging the appellant's income liability any additional cost thereby incurred and expressed in terms of Australian currency was not chargeable to income and properly deductible under the provisions of s. 51.  

The result of the foregoing observations is that I think that the first question raised by the case stated should be answered in favour of the appellant and it is necessary, therefore, to consider the problems raised by the second question. For the purposes of this question it must be assumed that the appellant's trading activities during 1936 resulted in a loss though the facts before us do not enable us to specify its magnitude. This, by agreement between the parties, will be a matter for further consideration in the event of the second question being answered in the affirmative. In the circumstances it is sufficient to say that the answer to the first question means that the appellant made a trading loss during 1936 and that it is apparent from the case stated that no part of this loss had, pursuant to s. 80 of the Act, been claimed or allowed as a deduction in the income years which intervened between 1936 and 1939. In the ordinary course of events, therefore, the appellant would have been entitled in 1939 to deduct the loss from its assessable income or so much of it as was required to reduce its taxable income to nil. But as already indicated, events did not follow the ordinary course. First of all, the appellant was assessed to income tax in respect of the income year which ended on 31st December 1936 pursuant to s. 136 of the Act. This section at the relevant time was in the following form:

"Where any business carried on in Australia-

(a)

is controlled principally by non-residents;

(b)

is carried on by a company a majority of the shares in which is held by or on behalf of non-residents; or

(c)

is carried on by a company which holds or on behalf of which other persons hold a majority of the shares in a non-resident company,

and it appears to the Commissioner that the business produces either no taxable income or less than the amount of taxable income which might be expected to arise from that business, the person carrying on the business in Australia shall, notwithstanding any other provision of this Act, be liable to pay income tax on a taxable income of such amount of the total receipts (whether cash or credit) of the business as the Commissioner determines".

For the purposes of this section the respondent determined that PD513,914 was the amount of the total receipts on which the appellant was liable to pay income tax and he assessed the tax payable at PD25,695 14s. 0d. (See Notice of Assessment-annexure A). The next thing to be mentioned is that this assessment was the subject of a number of objections and, that thereafter, several conferences took place between the respondent and representatives of the appellant. These are referred to in pars. 8 to 14 of the case stated and the relevant annexures. It is, however, necessary that some particular reference should be made to the various matters which were discussed and to the decision which was ultimately reached. The conferences were concerned with the appellant's taxation liability for the years 1929 to 1938 inclusive and not for the year 1936 alone. Furthermore, grounds exist for thinking that the conditions which the parties ultimately accepted as fixing their respective rights and liabilities were determined more by a comprehensive view of the appellant's trading results over this lengthy period rather than by independent reviews of the results of each particular year. Indeed, so much may be thought to appear from the respondent's letter of 26th March 1941 (annexure B). The proposal made by this letter was the basis of the subsequent "settlement" but following this letter there were further discussions because the appellant was not prepared to accept the conditions contained in pars. (iii) and (iv). Subsequently, on 30th April 1941 it was proposed by the appellant's advisors that for these paragraphs there should be substituted the provision set out in their letter of that date (annexure D). This amendment was acceptable to the respondent and on 5th May 1941, he purported to determine the matters in dispute in the manner appearing in the memorandum (annexure E). The relevant portion of this memorandum is in the following terms:

"(a)

I assess the taxable income for the year ended-

31.12.1929 at PD65,592

31.12.1930 at PD55,620

31.12.1931 at PD37,613

31.12.1933 at NIL

(b)

I accept the decision of the Board of Review for the year ended 31.12.1932 and accordingly the taxable income for that year is reduced to PD232,434.

(c)

In respect of each of the income years ended 31.12.1934 to 31.12.1937 (for which years the company has been assessed on an amount equal to 20% of its total receipts) I consider that the company is properly assessed under the provisions of Section 28 (Section 136 for the income year ended 31.12.1935 and subsequent years) but in the circumstances do not think it proper to assess and charge tax on any percentage or amount of the total receipts of the business.

(d)

In respect of the income year ended 31.12.1938 for which year no assessment has yet issued, I consider that the company is properly assessable under the provisions of Section 136, but I do not think it proper to assess and charge tax on any amount of the total receipts of the business."

This decision was communicated to the appellant by letter on 26th May 1941, which intimated:

"With reference to the communication of the 26th March last, addressed to Messrs. Minter Simpson & Co , of Sydney from the Commissioner's Office, I desire to inform you that your Assessments for the years ended the 31st December, 1929, to the 31st December, 1938, inclusive, have been adjusted in the manner indicated therein. The resultant credit of PD251 14s. 0d. is being applied in part payment of the Assessments in respect of Non-Resident Insurers, Notices of which are enclosed and the balance of tax now payable is PD349 15s. 10d. In view of the action now taken to reduce to `Nil' your Assessments on income derived during each of the years ended the 31st December, 1934, to the 31st December, 1937 inclusive, the Objections lodged in connection with those Assessments are being regarded as having lapsed."

It remains only to be said that no amended assessment or assessments issued and that the parties were, apparently, satisfied to treat this letter as a final and effective determination concerning the appellant's liability to income tax for the years under review. This, of course, meant that it was accepted that the appellant's taxable income for 1936 was "Nil" and that, therefore, there was no existing liability to tax in respect of that year.  

For the respondent it is contended that where liability to tax in respect of any one year arises because of a determination of the commissioner pursuant to s. 136 the actual trading results of that year, as determined by ordinary accounting methods, are of no consequence as far as s. 80 is concerned. This proposition is, of course, denied by the appellant but it seems apparent that the fundamental and only purpose of s. 80 is to permit a taxpayer, within specified limits, to carry forward from one year to another revenue expenditure which has not been absorbed by income of the earlier year. This contemplates the setting off of expenditure against income in the earlier year and, in the event of a deficiency of income, the allowance of a deduction, calculated by reference to the deficiency, in the later year. In other words it provides a means whereby, in the deduction of revenue expenditure incurred over a period of years, a taxpayer may disregard the existence of separate and successive accounting periods when seeking to ascertain the extent of his taxable income in some later year of that period. The section, of course, operates only when a taxpayer's allowable deductions for an earlier year exceed his assessable income for that year and, therefore, where it appears that he had no taxable income. But if he is assessed pursuant to s. 136 in any year he becomes liable to tax upon a conventional sum which, for the purposes of the Act, is "taxable income". And this is so even if, on ordinary accounting principles, his business has produced no taxable income. That being so it seems reasonably clear that if the machinery provided by the lastmentioned section be employed in assessing a taxpayer in one year he cannot hope to invoke the aid of s. 80 to carry forward from that year, as a deduction in a later year, any business loss disclosed by ordinary accounting practice. To conclude otherwise would be to ignore entirely the fundamental purpose and effect of s. 80. That section, in effect, is intended to authorize the Commissioner to disregard ordinary accounting principles where they do not provide a true reflection of the income producing capacity of a business and it would be strange indeed if a taxpayer, precluded by its provisions from asserting for income tax purposes that he had suffered a loss in one year and having been assessed to tax upon some portion of his total receipts, should then be permitted to carry the asserted loss forward to the next income year as a deduction in that year.  

But, as already appears, the original assessment pursuant to s. 136 did not, in the events which happened, finally determine the extent of the appellant's income tax liability in respect of the income year 1936. There were objections to this assessment and, thereafter, a considerable amount of discussion took place in an attempt to resolve the differences between the parties. As a result of these discussions, it was said, the matter was "settled" but there was no formal amendment of the assessments under review. Yet the parties regarded what was done as effective and final and it is of some importance in the case to examine exactly what happened in relation to the 1936 assessment. First of all, it may be noticed that on 26th March 1941 the respondent intimated that he was "prepared to finalize all outstanding matters in connection with the assessments of the Company for all years up to and including the income year ended 31.12.38" on the basis thereinafter specified. As far as 1936 was concerned it was proposed that the appellant should be dealt with on the basis that its taxable income for that year was "Nil" and, therefore, that no tax was payable. But one of the conditions specified was that liability to income tax for each year subsequent to 31st December 1938 should "be determined without taking any account of the trading results of the business of the Company prior to that date". This condition was not acceptable to the appellant and after further discussion and correspondence it was, in effect, agreed that it should not be taken from any agreement of the parties that the respondent had "allowed a deduction in respect of Exchange losses and in respect of `losses of previous years' either wholly or in part" but that the appellant should "be at liberty to contend in respect of any income year subsequent to 31.12.1938 that the Commissioner should allow deductions for Exchange losses and `losses of previous years' respectively". Thereafter the respondent, though expressing the view "that the company is properly assessed under the provisions of Section 28 (Section 136 for the income year ended 31.12.1935 and subsequent years)", formed the opinion that it was not proper "to assess and charge tax on any percentage or amount of the total receipts of the business". This decision was not conveyed to the appellant in these words; it was notified that its assessments for the years ended 31st December, 1929 to 31st December 1938 inclusive had "been adjusted in the manner indicated" in the earlier letter of 26th March 1941. The respondent's notification then went on to say that "In view of the action now taken to reduce to `Nil' your Assessments on income derived during each of the years ended 31st December, 1934, to 31st December, 1937 inclusive, the Objections lodged in connection with those Assessments are being regarded as having lapsed".  

Of course, as both parties agreed, the assessments already made could not have been reduced to nil merely by an informal "settlement". Yet the appellant asserts, and the respondent agrees, that as a result of what happened, the liability of the former for income tax originally assessed in respect of the income year 1936 was discharged. According to the respondent what happened resulted in an amendment of the s. 136 assessment. But it is asserted that the amended assessment was itself an assessment pursuant to s. 136. In my view, however, there are insuperable difficulties in accepting this last proposition for the only manner in which the authority given by s. 136 may be exercised is by the Commissioner "determining" some amount of the total receipts of a business as taxable income. In spite of the respondent's expressed view "that the company is properly assessed under the provisions of Section 28 (Section 136 for the income year 31.12.1935 and subsequent years)" his opinion that it would not, in the circumstances, be proper "to assess and charge tax on any percentage or amount of the total receipts of the business", conclusively indicated a failure at that stage to exercise any authority committed to him by s. 136. Furthermore, it is difficult to see how it could have appeared to the respondent, whilst he held that opinion, that the company's taxable income was "less than the taxable income which might be expected to arise from" its business and this, of course, was a condition precedent to the operation of the section (Texas Co (Australasia) Ltd v Federal Commissioner of Taxation ^[F35] ).  

In its objections which it lodged at an earlier stage the appellant had claimed that the assessment for the income year 1936 was, for a number of reasons, excessive and that the provisions of s. 136 had been applied in error. It is unnecessary to go into the substance of the objections but it is clear that the subsequent discussions between the parties were, in a substantial measure, concerned with matters which they raised. It is, perhaps, not too much to say that discussion and consideration of some of the matters induced the Commissioner to reconsider the assessment and to make a fresh decision. In these circumstances, it seems to me, the better solution of the problem now under consideration is to hold that, in effect, the respondent must be taken to have accepted sufficient of the substance of these objections to enable him to exercise his powers under s. 170 (7). Upon that view there was, pursuant to the objections originally raised, an amendment of the original assessment though it was not made in the usual formal manner and since, for the reasons already given, it was not made in the exercise of any authority given by s. 136, the appellant was entitled in the income year 1939 to take into account the loss made during the income year 1936.  

For these reasons each of the questions raised by the case stated should, in my view, be answered in the affirmative.