Minister of National Revenue vs Anaconda American Brass Ltd.

Rewritten Version Notice: This is a rewritten version of the original judgment.

Court: supreme-court

Case Number: Not extracted

Decision Date: 13 December 1955

Coram: VISCOUNT SIMONDS

On 13 December 1955 the Supreme Court of India delivered a judgment in the case titled Minister of National Revenue versus Anaconda American Brass Ltd. The judgment was read by Viscount Simonds. The appeal arose from a decision of the Supreme Court of Canada, which had dismissed an appeal taken from a judgment of the President of the Exchequer Court of Canada. The matter presented a question of novelty and importance. The appellant was the Minister of National Revenue and the respondent was the limited company Anaconda American Brass Ltd. The Minister had, on the basis of the company’s assessment for the year 1947, increased the amount of taxable income declared by the company by the sum of $1,611,756.43. As a result the excess profits tax for that year was raised by approximately $241,000 and the income tax by about $483,000. The company challenged this assessment before the Exchequer Court; the President of that court allowed the company’s appeal, and subsequently the Supreme Court of Canada dismissed the Minister’s appeal against that judgment. The present appeal before the Indian Supreme Court concerned only the excess profits tax component of the assessment. The excess profits tax was imposed under the Excess Profits Tax Act, 1940, which in section 3 provided that, in addition to any other tax or duty payable under any Act, a tax would be levied at the rate specified in the Second Schedule upon the excess profits of every corporation or joint-stock company residing or ordinarily resident in Canada or carrying on business there. The rate set out in the Second Schedule was fifteen per cent of the excess profits for corporations and joint-stock companies. Under the same Act the term “profits” was defined to mean, for any taxation period, the amount of net taxable income for that period as determined under the provisions of the Income War Tax Act. That Act, for the purposes relevant here, defined income as the annual net profit or gain that could be ascertained and computed as the profits from a trade, commercial, financial or other business.

The precise issue before the Court was whether the Minister had correctly assessed the annual net profit or gain of the company for the year 1947, or, more accurately, whether the company had established that the Minister’s assessment was incorrect. The material facts of the case were not in dispute. The company was engaged in the business of purchasing metals, converting them into sheets, rods and tubes, and selling the finished metal products. Approximately ninety-eight per cent of the metals purchased by the company consisted of copper—over eighty per cent—and zinc—about fifteen per cent. The remaining principal metals included lead, tin and nickel. The prices charged by the company for its products closely reflected the prices it paid to replace the metals used in manufacturing, and the company adjusted its product prices promptly whenever market prices of the purchased metals rose. The company maintained a continual practice of purchasing metals to replace those being used. At the beginning of each month it estimated the quantity of metals it would need during the following month and ordered that quantity from its suppliers. It kept an inventory of metals equal to roughly one-third to one-quarter of its annual requirements, turning over the inventory three to four times each year. About two-thirds of the copper and zinc inventory remained continuously in process within the plant, indicating an average processing time of approximately eight weeks. The company did not keep detailed records from which the exact quantities of metal actually used during the year could be identified or the amounts paid for those metals could be determined, although it retained other records pertaining to its operations.

The Court observed that the company’s selling prices were directly linked to the prices it paid for the metals used in manufacturing its products, and that whenever market prices for the purchased metals rose, the company promptly raised the prices of its finished goods. The company continually bought metals to replace those that were consumed in production. Its routine was to estimate at the start of each month the quantity of metals that would be required for the following month and to place orders with suppliers for that same quantity. The company maintained a standing inventory of metals amounting to roughly one-third to one-quarter of its annual needs, allowing it to turn over this inventory three to four times a year. Approximately two-thirds of the copper and zinc inventory was constantly in the manufacturing process, indicating an average processing period of about eight weeks. The Court noted that the company did not keep detailed records that would enable it to identify the specific metals actually used during the year nor the exact amounts paid for those metals. Nevertheless, the company did keep records showing the quantities of metals that were in inventory at the beginning of the year, the quantities purchased during the year, and the quantities remaining in inventory at the end of the year. It also retained records of the purchase prices of the metals at the times they were bought. Consequently, for the fiscal year 1947 the company was aware of the opening stock, the purchases made during the year, and the closing stock, as well as the prices paid for the metals purchased. However, the company could not determine, for all the metals it had used during the year, the specific price paid for each, nor could it ascertain, for all the metals remaining in inventory at year-end, the price at which each had been acquired. The Court explained that in order to compute the annual profit or gain for 1947, the company needed to assign the correct cost to the metals consumed and the correct cost or value to the metals still in stock. In the absence of precise knowledge, the company was compelled to make assumptions or estimates, and it was at this point that the dispute between the parties arose. The Court then found it useful to outline how the company prepared its return for 1947. The premise was that, while raw-material prices had been relatively stable for some time before that year, 1947 saw substantial price increases because wartime controls were relaxed and eventually removed. For example, the price of copper rose on 22 January from 11.5 cents per pound to 16.625 cents per pound and again on 10 June to 21.5 cents per pound. It was under these circumstances that, for the first time in 1947, the company adopted a particular method of valuation for its income-tax return.

In this case the appellant prepared its income-tax return for the fiscal year 1947 by employing a valuation method that is commonly referred to as Lifo, an abbreviation for last-in-first-out. The company had previously used this method for its own internal accounting of copper, but neither the company nor any other taxpayer in Canada had ever applied Lifo for the purpose of computing taxable income. The Court first set out what Lifo actually entails. It explained that Lifo does not require the metal that was most recently received into inventory to be the first metal physically processed and sold. Instead, the Court emphasized that the actual physical flow of the raw material is considered immaterial for the method. Copper that was bought in earlier years and remained in stock at the beginning of the financial year, as well as copper purchased during the year, could both have been processed and the resulting products sold in that year, and that fact does not affect the Lifo computation. Lifo is concerned solely with the assignment of cost to inventory. In plain terms the method assigns to the next sale of processed metal the unit cost of the most recent purchase of copper that was added to the inventory. Consequently, the inventory that remains on hand at the end of the accounting period must be assigned the cost of copper that has not yet been exhausted by the cost already allocated to the metal that was consumed. This portion of cost is described in the judgment as the unabsorbed residue of cost.

Applying the Lifo approach to the appellant’s situation, the Court noted that the closing inventory of copper on 31 December 1947 comprised fourteen million two hundred ninety-one thousand seven pounds. Under the Lifo system the Company divided those pounds into several layers, each layer representing copper purchased in a particular year and valued at the corresponding average unit cost. The first layer consisted of six million five hundred thousand pounds valued at seven and a half cents per pound, reflecting the average cost of copper when the Lifo method was first adopted by the Company in 1936; this layer was assigned a monetary value of four hundred eighty-seven thousand five hundred dollars. The second layer comprised eight hundred two thousand six hundred ninety-seven pounds valued at nine point four six six cents per pound, the average price paid in 1936, resulting in a value of seventy-five thousand nine hundred eighty-three dollars and thirty cents. The third layer included seventeen thousand five hundred seventy-seven pounds valued at eleven point one nine one cents per pound, the average price paid in 1937, giving a value of one thousand nine hundred sixty-seven dollars and four cents. The fourth layer consisted of six hundred thirty-nine thousand eight hundred seven pounds valued at ten point four four three cents per pound, representing the average price paid in 1938, which amounted to sixty-six thousand eight hundred forty-seven dollars and four cents. The fifth layer comprised nine hundred seventy-three thousand four hundred seventy-seven pounds valued at eleven point zero three six cents per pound, the average price paid in 1939, resulting in a value of one hundred seven thousand four hundred thirty-two dollars and ninety-two cents. The sixth layer consisted of three million one hundred fifty-one thousand six hundred eighty-four pounds valued at eleven point five cents per pound, the price paid in 1945, giving a value of three hundred sixty-two thousand four hundred forty-three dollars and sixty-six cents. The seventh and final layer comprised two million two hundred five thousand seven hundred sixty-five pounds also valued at eleven point five cents per pound, the price paid in 1946, amounting to two hundred fifty-three thousand six hundred sixty-two dollars and ninety-seven cents. Adding together the values of all seven layers produced a total assumed cost of one million three hundred fifty-five thousand eight hundred thirty-six dollars and ninety-three cents for the copper on hand at the end of the year. The Court recorded that this total figure was the amount that the Company entered on its income-tax return as the value of its closing inventory.

In the matter of income tax, the Court observed that the company had assigned a price of seven and a half cents per Pound to nearly half of its inventory, a price that was lower than any price that had been in effect since 1935. By allocating the higher cost to the metals that had been processed and the lower cost to the metals that remained in stock, the company was able to report profits that were considerably lower than they would have been if it had employed the usual and traditional method of accounting for inventory. The central issue, therefore, was whether this new method of valuation could be permitted for income-tax purposes. The Court emphasized that the phrase “for income-tax purposes” was crucial. A substantial portion of the argument presented on appeal and throughout the proceedings focused on demonstrating that, in Canada, the last-in-first-out (Lifo) method was, under certain circumstances, a proper and widely accepted accounting practice, and that those circumstances were clearly present in the appellant’s case. This view had been accepted by the President of the Exchequer Court of Canada, upheld by the Supreme Court, and was fully endorsed by the members of the Court.

The Minister, acting as the appellant, maintained that although the Lifo method might be appropriate for the company’s internal corporate purposes, it did not faithfully represent the company’s profit for income-tax calculations. Consequently, the Minister increased the assessable income by the sum of one million six hundred eleven thousand seven hundred fifty-six dollars and forty-three cents. In doing so, the Minister applied the first-in-first-out (Fifo) method, which presumes that the metal purchased first is the metal used first. Under this assumption, the inventory held at the close of 1947 was treated as the most recently purchased metal. The Minister noted that during 1947 the company had purchased at least sixty-three million two hundred sixty-eight thousand five hundred fifty-five Pounds of copper, and that more than fourteen million two hundred ninety-one thousand seven Pounds—the amount recorded as closing inventory—had been bought at a price of two hundred fifteen cents per Pound in the final three months of the year. Accordingly, the Minister valued the copper in the closing inventory at three million seventy-two thousand five hundred sixty-six dollars and fifty cents, a figure that contrasted sharply with the one million three hundred fifty-five thousand eight hundred thirty-six dollars and ninety-three cents reported by the company.

At this stage the positions of the parties could be delineated more precisely. The Minister acknowledged that the amount of copper actually used during the financial year needed to be determined and did not dispute that some of the metal remaining in inventory might have been purchased in earlier years. Nevertheless, the Minister argued that the method he had adopted more closely achieved the result envisioned by Lord Loreburn L.C. in Sun Insurance Office v. Clark, namely that true gains should be ascertained as accurately as possible. He contended that income-tax law, drawing from commercial accounting practice, required that the values of stock-in-trade at the beginning and end of the accounting period be entered at either cost or market value, whichever was lower, and that for this purpose the actual stock on hand and its actual cost should be considered. In his view, the Fifo method better reflected the factual circumstances than the Lifo method.

In the Minister’s view, the quantities of copper that could be ascertained should be determined as far as possible, and any assumption or estimate should be used only when exact ascertainment is impossible. On that basis, the Minister argued that the FIFO method more closely reflects the actual facts than the LIFO method. He supported this argument by pointing first to the large purchases of metal made in the final months of 1947, purchases that could scarcely have been processed within that same year, and second to the implication that the LIFO method would require assuming that £6,500,000 worth of copper bought in or before 1936 remained in stock at the close of 1947. Accordingly, the Minister maintained that no evidence had been shown to demonstrate that the LIFO method gave a truer picture of the company’s income than the FIFO method. He did not claim that the company, or any other taxpayer, could never establish that a method other than FIFO might more accurately represent income for tax purposes when the raw material is homogeneous and no distinct portion can be identified. He further observed that it was unnecessary for the Lords to decide whether other methods, such as an average-cost approach, might in some circumstances be properly applied for tax purposes. The essential question the Minister posed for the Lords’ determination was whether, in the present case, the LIFO method produced a less accurate representation of true income than the FIFO method. The company, in contrast, contended that the Income War Tax Act contains no definition of “annual net profit or gain” and offers no direction on how profits or cost of sales should be ascertained. Consequently, the company argued that such matters must be resolved by applying ordinary commercial principles, except where the Act expressly excludes those principles. The company further asserted that the identification of ordinary commercial principles is a matter of fact, and that the President of the Exchequer Court had found, a finding affirmed by the majority of the Supreme Court, that the company’s return had been prepared in conformity with those principles. In short, the company maintained that annual income for tax purposes should be determined by accepted accounting practice unless the statute provides otherwise, and that the President had concluded not only that LIFO was an acceptable accounting method but that it was the most appropriate method, whereas the Minister’s chosen method was improper. The company’s arguments, in the view of the Lords, had been rejected by the minority of the Supreme Court, namely Kerwin C.J. and Estey J. They noted that the income-tax law of Canada, as applied in the United Kingdom, rests on the foundations outlined by Lord Clyde in Whimster & Co. v. Inland Revenue Commissioner, a passage quoted by the Chief Justice and repeated here: “In the first place, the profits of any particular year or accounting period must be taken to consist of the difference between the receipts.”

The Court explained that profit for any accounting year is calculated as the difference between the receipts earned from the trade or business during that year and the expenditures incurred to obtain those receipts. It further stated that the profit-and-loss account prepared to determine that difference must be drawn up in accordance with the ordinary principles of commercial accounting, insofar as they are applicable, and must also comply with the rules set out in the Income-Tax Act, including any modifications made by the provisions and schedules of the Acts that regulate Excess-Profit Duty, where relevant. As an illustration, the Court noted that ordinary commercial accounting requires that, in the profit-and-loss account of a merchant or manufacturer, the values of stock-in-trade at the beginning and at the end of the period be recorded at either cost or market price, whichever is lower, even though the statutes contain no explicit provision to that effect. For many years before the decision under discussion and ever since, the tax authorities had valued the opening and closing stock for tax purposes on the basis of the actual physical stock, to the extent that it could be ascertained. The Court observed that, from the standpoint of tax law, it was a novel and even revolutionary suggestion to disregard those physical facts, even when they could be wholly or partly known, for the purpose of determining opening and closing inventories, and to replace them with a theoretical assumption based on a supposed “flow of cost” and an “unabsorbed residue of cost.” An expert witness called for the company testified that he could not imagine any of the company’s own witnesses claiming that a quantity of metal held on hand, acquired in 1936, was still valued at its original cost. Yet the expert pointed out that the 1947 closing inventory contained no less than six and a half million pounds of copper, to which the 1936 cost had been applied. He added that, year after year, the same situation would arise as long as the business continued and existing stocks were not substantially reduced; consequently, in 1987, just as in 1947, the closing inventory would still reflect stock valued at 1936 costs. This, the Court said, clearly illustrated the operation of the LIFO method and demonstrated how far it had drifted from the traditional approach that had previously governed the assessment of income for tax purposes. The respondent had strongly urged that, when dealing with homogeneous material, the “actual user” test—if it could be applied by identifying parcels purchased at different prices—would produce capricious and illogical results. Assuming that this argument were correct and that the actual user test should not, in some circumstances, be regarded as the decisive test, the Court concluded that such reasoning did not, in its view, establish a sufficient basis for adopting LIFO. It further observed that, when properly applied, a method such as LIFO, like the FIFO method, would bring the cost of every purchase into the accounts.

In the order that was issued, the Court observed that the Lifo method, when a business persists and retains stock into future periods, can result in large purchases never being reflected in the profit account. The Court did not dispute that the Lifo system, or a variation of it, might be suitable for the commercial objectives of a trading corporation. Nevertheless, the Court stressed that businessmen and their accounting advisers must look beyond the single fiscal year that concerns the Minister of Revenue. It was noted that it could be prudent for them to record inventories at a figure that does not represent the current market value or the actual acquisition cost, but rather the lower historic cost at which comparable stock was obtained long ago. By doing so a hidden reserve is created, which may prove useful in later years. The Court further pointed out that the Income-Tax Act of 1947 was not concerned with the years 1948 or 1949; by those later dates a company might have ceased to exist and its assets could have been distributed. The Court recalled the remark of Lord Herschell that surplus consists of receipts from trade exceeding the expenditure necessary for earning them, and observed that this principle implies that no assumption should be made unless a fact cannot be ascertained, and even then only to the extent that it remains unascertained. Accordingly, the Court held that there was no room for speculative theories about the flow of costs, nor was it legitimate to treat the closing inventory as an unabsorbed residue of costs rather than as a tangible stock of metal awaiting processing. The Court concluded that the failure to recognize, or the deliberate disregard of, facts that could be ascertained and that must be given proper weight, undermined the application of the Lifo method to the present case. The same reasoning rendered the testimony of the export witness—who asserted that the Lifo method was generally acceptable and, in this instance, the most appropriate accounting method—insufficient to decide the question before the Court. While those facts might be established by the Exchequer Court and affirmed by the Supreme Court, the remaining issue was whether the method complied with the prescription of the Income-Tax Act. In the Court’s view, it did not. The Court also provided a brief historical note: the Lifo method had originated in the United States and had been adopted by an American corporation that was the parent of the respondent company sometime before 1938. It was only through an amendment to the existing revenue law in that year that the method became permissible for tax purposes, and further amendments in 1939 allowed its use subject to statutory conditions, which the Court summarized.

In the judgment of Estey J. of the Supreme Court, the Court observed that the numerous differences that exist between the revenue laws of the United States and Canada prevent Canada from placing great reliance on the proposition that it seeks, without specific legislation, to accomplish in Canada what the United States permits only when statutory safeguards are in place. The Court further noted that this observation nevertheless supports the view that accounting theories, even when widely accepted and practiced by businessmen, do not ultimately determine a trading company’s assessable income for tax purposes. The Court then acknowledged that, although this appeal must be decided on the basis of Canadian law, it was appropriate to look at comparable discussion in United States jurisprudence, particularly a case that examined the so-called “base stock” method. In that United States case, Justice Brandeis expressed language that is also applicable to the LIFO method, stating that in periods of rising prices the “base stock” method understates income because it ignores the gains actually realized when low-price stock is liquidated at higher market prices; he further explained that such a method, like many reserves created by businessmen for their own purposes, may serve to smooth results of operations over several years. The Court cited Lucas v. Kansas City Structural Steel Co. to illustrate the clear distinction drawn in that passage between what is permissible for tax purposes and what prudent business practice might deem appropriate. The Court also referred to a recent United Kingdom decision, Patrick v. Broadstone Mills Ltd., where Justice Singleton declined to accept the base stock method for income-tax purposes, noting that while the method might be acceptable in accounting practice, it was not conformable to tax law. Accordingly, the Court concluded that the appeal should be allowed, that the respondent company’s judgment against the Minister’s assessment be set aside, and that Her Majesty should be advised of this result. The Court emphasized that this conclusion was not reached by disagreeing with any factual findings of the President of the Exchequer Court, whose detailed and clear judgment the Court could not criticize, but rather because those findings did not provide a complete solution to the legal question. Finally, the Court ordered that the respondent company must bear the costs of this appeal and of the related proceedings in Canadian courts.