Eduzan

Learning Hub

Eduzan

Eduzan / 04 Financial Statement Analysis

FSA 3 – Multinational Operations

Worked examples are fully visible. Check-yourself items are study aids you can reveal one at a time.

Cross-border trade is not a specialist activity carried on by a handful of giant firms. World Trade Organization data put global merchandise exports at close to US$15 trillion in 2010, over double the US$7.4 trillion recorded in 2003 and over four times the US$3.7 trillion recorded in 1993. China ranked first among exporting countries in 2010, ahead of the United States, Germany, Japan and, in fifth place, the Netherlands. Within the United States, 293,131 firms were counted as exporters that year, of which just 2.2 percent were large, meaning more than 500 employees. The overwhelming majority of exporting entities were small or medium sized.

Two parties on opposite sides of a border must agree on a settlement currency, usually the currency of the buyer or the currency of the seller. That single decision determines which of the two carries an accounting problem. It also determines which of the two carries a market risk. The purpose of this reading is to trace both consequences through the financial statements, and then to read a real set of disclosures well enough to say what currency movement actually did to reported results.

Companies with foreign subsidiaries face a second, larger version of the same problem. Nestlé SA reports factories spread across 83 countries and a footprint in nearly every country on earth. Procter and Gamble discloses more than 400 subsidiaries located in more than 80 countries. Foreign subsidiaries generally keep their accounting records in the currency of the country where they sit, so a parent preparing consolidated statements must first restate every one of those sets of books into a single currency.

Three currency labels that are easy to confuse

Level 2 questions on this reading very often turn on keeping three labels apart. They are frequently the same currency, which is exactly why candidates stop distinguishing them and then answer the one question where they differ.

The three currency concepts

Concept	Definition	Who chooses it
Local currency	The national currency of the country where the entity is located	Nobody: it is a fact of geography
Functional currency	The currency of the primary economic environment in which the entity operates, normally the currency in which it generates and expends cash	Management, applying the indicators, with judgement where they conflict
Presentation currency	The currency in which the financial statements are presented	Management, usually the currency of the country where the parent is located

For accounting purposes, a foreign currency is defined relative to the functional currency, not relative to geography. Once the functional currency of an entity is fixed, every other currency counts as foreign to that entity, and a foreign currency transaction is simply one denominated in something other than the functional currency. That definition has a consequence worth stating plainly: a company can transact abroad without having a foreign currency transaction at all.

Consider a Mexican components manufacturer selling to a customer in Finland with settlement in Mexican pesos. The Finnish buyer has a foreign currency transaction, because it must convert peso amounts into euro to record the inventory and the account payable. The Mexican seller has an export sale but no foreign currency transaction whatsoever. It records revenue and a receivable in the currency it already keeps its books in, and nothing needs translating.

Because most entities generate and spend cash in the currency of the country where they sit, the local currency is usually the functional currency, and the functional currency is usually the presentation currency. A multinational with subsidiaries in many countries therefore usually has many functional currencies. The Thai subsidiary of a Japanese parent will normally treat the baht as functional while the parent works in yen.

The exception is common enough to matter. A foreign subsidiary can have the parent’s functional currency as its own. Ahead of the McAfee acquisition in 2011, Intel Corporation had concluded that every one of its significant foreign subsidiaries used the US dollar as its functional currency. Intel then noted in its 2011 annual report that some of the operations acquired from McAfee ran on a functional currency other than the dollar. One acquisition changed the answer for part of the group.

How the functional currency is identified

IFRS list nine factors. The first five look at the economics of the entity itself; the last four look at the relationship between the entity and its parent.

Factors considered in determining the functional currency under IFRS

No.	Factor
1	The currency having the dominant influence on the prices charged for goods and services
2	The currency of the country whose regulations and competitive pressure set those selling prices
3	The currency exerting the strongest pull on wages, materials and the other input costs
4	The currency raised through financing activity
5	The currency in which operating receipts are normally held rather than converted
6	Whether the operation runs with meaningful independence, or simply extends what the parent already does
7	Whether dealings with the parent form a large or a small share of what the entity does
8	Whether the cash the operation produces feeds straight into parent cash flow and can be remitted
9	Whether the operation earns enough operating cash to cover its own debt service, or has to draw on the parent

Where the indicators point in different directions, IFRS require management judgement, but factors 1 and 2 take priority over factors 3 through 9. US GAAP provide similar, though not identical, indicators.

The three-step functional currency approach

Accounting standards do not ask you to translate straight from bookkeeping currency to presentation currency. They ask for three steps.

Identify the functional currency of the foreign entity.
Translate any foreign currency balances the entity holds into that functional currency.
Translate the functional currency balances into the parent’s presentation currency at the current exchange rate, if the two differ.

Work this through with a US parent and a Mexican subsidiary that keeps its records in Mexican pesos and happens to hold an account payable in Guatemalan quetzals. If the peso is the functional currency, step 2 converts only the quetzal payable into pesos at the current rate, and step 3 then converts the whole peso balance sheet into US dollars using the current rate method.

Now change one assumption. Suppose the subsidiary operates primarily in US dollars, so the dollar is its functional currency. Every peso balance is now a foreign currency balance from the subsidiary’s own point of view, because pesos are not its functional currency. Each peso account, along with the quetzal payable, must be restated into dollars as though the books had been kept in dollars from the start: current values at the current rate, historical costs at historical rates. After step 2 the statements are already in dollars, which is both the functional currency and the presentation currency, so step 3 is unnecessary.

That second path is the temporal method, and the first path is the current rate method. Everything else in this reading is a consequence of which of those two paths the functional currency decision puts a subsidiary on.

A foreign currency transaction arises in two situations. The first is an import purchase or an export sale denominated in a foreign currency. The second is borrowing or lending where repayment or receipt is fixed in a foreign currency. In both, the entity ends up holding an asset or a liability denominated in a currency that is not its own.

The governing principle is short: every transaction is recorded at the spot rate on the transaction date. Risk therefore arises only from the gap between the transaction date and the payment date. If the two coincide, there is nothing to measure.

Which side is exposed, and to what

Import purchase. The importer owes foreign currency and is allowed to defer payment. It is exposed to the foreign currency strengthening, because more functional currency would then be needed to buy the foreign currency required to settle. This is a liability exposure.
Export sale. The exporter is owed foreign currency and has allowed the customer time to pay. It is exposed to the foreign currency weakening, because the receipt would then convert into fewer units of functional currency. This is an asset exposure.

Both IFRS and US GAAP require the movement in the functional currency value of that asset or liability to be reported as a gain or loss in the income statement. The size of that gain or loss is mechanical.

Where S is quoted as units of functional currency per unit of foreign currency. A positive result is a gain on a receivable and a loss on a payable.

Example 1 · Worked

On 1 November 20X1 FinnCo buys goods from a Mexican supplier for 100,000 Mexican pesos. The credit terms give it 45 days, and it settles the 100,000 pesos on 15 December 20X1. The functional and presentation currency of FinnCo is the euro, and its fiscal year ends on 31 December.

Spot rates, euro per Mexican peso

Date	Rate
1 November 20X1	MXN1 = EUR0.0684
15 December 20X1	MXN1 = EUR0.0703

How does FinnCo account for the transaction, and what is the effect on the 20X1 financial statements?

Solution. The payable and the inventory are both recorded on 1 November at the spot rate that day: MXN100,000 × EUR0.0684 = EUR6,840. That is the amount FinnCo could have settled for if it had paid immediately, and it is the cost that stays in the inventory account.

By 15 December the peso has strengthened to EUR0.0703, so acquiring MXN100,000 costs 100,000 × 0.0703 = EUR7,030. The cash outflow is EUR7,030 while inventory carries EUR6,840, and the difference of 7,030 − 6,840 = EUR190 is a foreign exchange loss reported in 20X1 net income.

This loss is realized, because FinnCo genuinely spent an extra 190 euro. On the balance sheet, cash falls by 7,030, inventory rises by 6,840, and retained earnings fall by 190, so the balance sheet still balances. Note what the loss is not: it is not part of the cost of inventory. Deferring payment is a financing decision, and its cost is reported separately.

When a balance sheet date falls in between

If the settlement date falls in a later accounting period, both IFRS and US GAAP require the foreign currency balance to be retranslated at the balance sheet date and the change taken to income. This creates one of the very few situations in which accounting rules require a company to recognise a gain in income before it has been realized.

The gains and losses for the two periods add up. The amount recognised from the transaction date to the balance sheet date, plus the amount recognised from the balance sheet date to settlement, equals the total realized gain or loss on the transaction.

Figure 1: Splitting one transaction across two reporting periods

The period 1 amount is recognised in income while still unrealized. The two period amounts always sum to the realized total, however far apart the rates move in between.

Example 2 · Worked

On 15 November 20X1 FinnCo makes a sale to a United Kingdom customer for GBP10,000, with the pounds due to be received on 15 January 20X2. The functional and presentation currency of FinnCo is the euro and its fiscal year ends 31 December.

Euro per British pound, and the euro value of a GBP10,000 receivable

Date	Rate	Euro value	Change in euro value
15 Nov 20X1	1.460	14,600	Not applicable
31 Dec 20X1	1.480	14,800	+ 200
15 Jan 20X2	1.475	14,750	− 50

1. What is recognised in 20X1?

Solution. The receivable is recorded at 10,000 × 1.460 = EUR14,600. At the balance sheet date it is retranslated to 10,000 × 1.480 = EUR14,800. The pound has strengthened against a euro-functional exporter holding an asset exposure, so 20X1 income includes a transaction gain of EUR200, computed as 10,000 × (1.480 − 1.460). The gain is unrealized: no cash has arrived.

2. What is recognised in 20X2, and what is the total?

Solution. By settlement the pound has slipped to 1.475, so the receivable is worth 10,000 × 1.475 = EUR14,750, which is EUR50 below the 31 December carrying amount. FinnCo recognises a transaction loss of EUR50 in the first quarter of 20X2.

Over the whole life of the receivable the pound rose by 1.475 − 1.460 = EUR0.015, giving a realized gain of 10,000 × 0.015 = EUR150. The two recognised amounts reconcile: 200 − 50 = 150.

The four-cell rule

Two things determine the sign: whether the exposure is an asset or a liability, and whether the foreign currency strengthens or weakens. Learn the grid rather than re-deriving it under time pressure.

Direction of the foreign currency transaction gain or loss

Transaction	Type of exposure	Foreign currency strengthens	Foreign currency weakens
Export sale on credit	Asset: a foreign currency receivable	Gain	Loss
Import purchase on credit	Liability: a foreign currency payable	Loss	Gain

Why the interim number is a weak forecast of the final one

Recognising an unrealized amount at the balance sheet date carries an implicit assumption: that the position at that date is a reasonable estimate of where the transaction will finally settle. Real currency paths do not cooperate. Consider a French company that purchased goods from a Canadian supplier on 1 December 20X1 with C$100,000 payable on 15 May 20X2.

Euro per Canadian dollar and the euro value of a C$100,000 payable

Date	Euro per C$	Euro value	Change in euro value
1 Dec X1	0.7285	72,850	Not applicable
31 Dec X1	0.7571	75,710	2,860 loss
31 Mar X2	0.7517	75,170	540 gain
15 May X2	0.7753	77,530	2,360 loss

The Canadian dollar strengthened into the year end, so the French company booked a loss of EUR2,860 in the fourth quarter of 20X1. It then reversed course over the first quarter of 20X2, producing a gain of EUR540, before strengthening again to give a further loss of EUR2,360 in the second quarter. The realized loss on settlement is 77,530 − 72,850 = EUR4,680, which is well above the EUR2,860 an analyst reading only the 20X1 statements would have seen. The interim figure was neither the direction nor the magnitude of the eventual outcome.

Standard setters tell companies that transaction gains and losses belong in net income. They do not say where in net income. The two treatments seen in practice are to fold the amount into other operating income or expense, or to report it below the operating line as part of non-operating income or expense, sometimes inside net financing cost. Gross profit and net profit are identical either way. Operating profit is not.

Example 3 · Worked

FinnCo reports the following amounts in both 20X1 and 20X2, excluding the transaction gain of EUR200 in 20X1 and the transaction loss of EUR50 in 20X2 from Example 2.

FinnCo income statement components before the currency item

Item	20X1	20X2
Revenue	20,000	20,000
Cost of sales	12,000	12,000
Other operating costs, net	5,000	5,000
Non-operating costs, net	1,200	1,200

Under Alternative 1 the currency item goes into other operating costs. Under Alternative 2 it goes below the operating line into non-operating costs.

1. Compute the gross, operating and net profit margins for 20X1 under each alternative.

Solution. A gain reduces expenses and a loss increases them.

Alternative 1. Other operating expenses become 5,000 − 200 = 4,800. Gross profit is 20,000 − 12,000 = 8,000. Operating profit is 8,000 − 4,800 = 3,200. Net profit is 3,200 − 1,200 = 2,000.
Margins: gross 8,000 ÷ 20,000 = 40.0%; operating 3,200 ÷ 20,000 = 16.0%; net 2,000 ÷ 20,000 = 10.0%.

Alternative 2. Other operating expenses stay at 5,000 and non-operating expenses become 1,200 − 200 = 1,000. Operating profit is 8,000 − 5,000 = 3,000 and net profit is 3,000 − 1,000 = 2,000.
Margins: gross 40.0%; operating 3,000 ÷ 20,000 = 15.0%; net 10.0%.

2. Repeat for 20X2, when the currency item is a loss of EUR50.

Solution. Alternative 1. Other operating expenses become 5,000 + 50 = 5,050, so operating profit is 8,000 − 5,050 = 2,950 and net profit is 2,950 − 1,200 = 1,750.
Margins: gross 40.0%; operating 2,950 ÷ 20,000 = 14.75%; net 1,750 ÷ 20,000 = 8.75%.

Alternative 2. Non-operating expenses become 1,200 + 50 = 1,250, so operating profit stays at 3,000 and net profit is 3,000 − 1,250 = 1,750.
Margins: gross 40.0%; operating 15.0%; net 8.75%.

3. What does the comparison show about the volatility of reported operating profit?

Solution. Under Alternative 2 the operating profit margin is 15.0 percent in both years, because the currency item never touches the operating line. Under Alternative 1 it moves from 16.0 percent to 14.75 percent purely because the currency item flipped from a gain to a loss. Because exchange rates never move by an identical amount, or even in a consistent direction, from one period to the next, Alternative 1 injects currency volatility straight into operating profit and the operating margin. Gross profit and net profit are unaffected by the choice in both years.

The comparability problem follows immediately. Two companies in the same industry are free to choose different alternatives, so a direct comparison of their operating profit or operating margin can be distorted by a presentation choice that has nothing to do with operations.

What the standards require to be disclosed

IFRS call for the amount of exchange differences taken to profit or loss to be disclosed. US GAAP call for the aggregate transaction gain or loss folded into net income for the period. Neither framework asks the company to say which line item holds it, which is precisely the piece an analyst needs.

BASF AG illustrates both the usefulness and the limits of the disclosure. Its 2011 income statement carries no separate currency line, but Note 6 shows income from foreign currency and hedging transactions of EUR170 million inside other operating income of EUR2,008 million, and Note 7 shows an expense line covering foreign currency, hedging and market valuation of EUR399 million inside other operating expenses of EUR2,690 million. The expense item is close to 15 percent of other operating expenses, at 399 out of 2,690. Netting the two gives a loss of 399 − 170 = EUR229 million, which is 229 ÷ 8,970, or about 2.55 percent of income before taxes and minority interests of EUR8,970 million. What the statements do not reveal is whether any of it was realized in 2011.

BASF AG 2011: locating the currency effect

Item, EUR million	2011	2010
Income on foreign currency and hedging	170	136
Total other operating income	2,008	1,140
Expenses on foreign currency, hedging and market valuation	399	601
Total other operating expenses	2,690	2,612
Income before taxes and minority interests	8,970	7,373

Heineken NV places the same item somewhere else entirely. Its Note 2 states that the euro is the functional currency of the company, and Note 3 discloses that currency gains and losses are shown net inside other net finance income and expenses. Note 12 shows a net foreign exchange loss of EUR107 million in 2011 and a net gain of EUR61 million in 2010. Against profit before income tax of EUR1,982 million in 2010 and EUR2,025 million in 2011, the 2010 gain is 3.1 percent of pretax profit and the 2011 loss is 5.3 percent. A reader comparing BASF with Heineken on operating profit would be comparing two different definitions.

Yahoo! Inc. shows the third pattern, an explicit narrative disclosure. In its 2011 filing the company explained that its exposure comes from assets and liabilities, intercompany balances included, that sit in a currency other than the functional currency of the entity holding them, that it may use forward contracts or other instruments to reduce short-term fluctuations, and that gains and losses on those contracts might offset none, part, or more than the underlying transaction results. It reported net realized and unrealized transaction gains of $9 million in 2011 and $13 million in 2010 and a transaction loss of $1 million in 2009, recognised in other income, net, which is a non-operating caption. The 2011 gain of $9 million was only 1.1 percent of pretax income of $827.5 million.

When the amounts are immaterial, and why

Some companies disclose neither the amount nor the location, usually because the numbers are small. Four reasons recur.

The company undertakes few foreign currency transactions, and the amounts involved are small.
The exchange rates between the functional currency and the currencies it transacts in are relatively stable.
Gains on some transactions are naturally offset by losses on others. A US company selling to a Canadian customer for C$ receipt in 90 days while simultaneously buying from a Canadian supplier for C$ payment in 90 days has an exactly offsetting pair.
The company hedges. Forward contracts and currency options are the two instruments most commonly used against transaction risk.

Nokia Corporation described its policy in its 2011 Form 20-F: material transaction exposures are hedged unless hedging would be uneconomic because of market liquidity or hedging cost, exposures are defined using nominal values, most hedging instruments run less than a year, and forecast flows beyond two years are not hedged at all. Nokia also quantified the effect of a specific move: the 4.2 percent appreciation of the US dollar in 2011 lifted net sales expressed in euro, since 40 percent of net sales were in US dollars or currencies closely tracking it, but also raised product cost, since 60 percent of components were sourced in US dollars. The net effect on operating profit was slightly negative. That is the useful shape of a currency disclosure: exposure on both sides of the income statement, with the net direction stated.

Why an unrealized transaction gain is not a forecast

Recognising the balance sheet date position in income embeds an assumption that the rate on that date is the best available estimate of the settlement rate. That assumption is defensible for a single period but it is not a prediction. The French importer earlier lost EUR2,860 in one quarter, recovered EUR540 in the next, and then lost EUR2,360 more, ending with a realized loss of EUR4,680, roughly 64 percent larger than the first interim figure. When you model a company with large unhedged exposures, treat the recognised amount as a mark at a point in time, not as a settled outcome, and check the maturity profile of the underlying receivables and payables before extrapolating.

Consolidation requires the assets, liabilities, revenues and expenses of foreign subsidiaries to be added to those of the parent, which means they must first be expressed in one currency. BMW AG has to convert both the US dollar accounts of its American subsidiary and the rand accounts of its South African subsidiary into euro before either set can be added in. Two questions have to be answered before any arithmetic can start.

Which exchange rate applies to each financial statement item?
How is the resulting translation adjustment reflected, given that using more than one rate will knock the translated balance sheet out of balance?

The second question is not a technicality. If every item were translated at one rate, a balanced statement would stay balanced and there would be no adjustment at all. The moment two different rates are used, a plug is required, and the whole debate is about what that plug means and where it belongs.

A minimal example that isolates the issue

Spanco is a Spain-based company presenting in euro. On 31 December 20X1 it establishes a wholly owned US subsidiary, Amerco, by investing EUR10,000 when EUR1 = US$1, physically converting the investment into US$10,000. Amerco also borrows US$5,000 from local banks, buys inventory costing US$12,000 and keeps US$3,000 in cash.

Amerco balance sheet, 31 December 20X1, in US dollars

Assets	US$	Liabilities and equity	US$
Cash	3,000	Notes payable	5,000
Inventory	12,000	Common stock	10,000
Total	15,000	Total	15,000

On the day of formation every balance translates at EUR1.00, so the euro balance sheet is a straight copy and balances at EUR15,000. During the first quarter of 20X2 Amerco does nothing at all: no transactions, no trading. All that happens is that the US dollar weakens, so that on 31 March 20X2 the rate is EUR0.80 = US$1. Because the EUR10,000 originally invested is a historical fact that Spanco wants its accounts to keep showing, common stock is translated at the historical rate of EUR1.00 in every scenario below.

Example 4 · Worked

Translate the 31 March 20X2 balance sheet of Amerco into euro under two approaches: first translating all assets and liabilities at the current rate of EUR0.80, then translating only monetary assets and liabilities at the current rate and leaving non-monetary items at historical rates. Monetary items are cash together with amounts receivable or payable that are fixed in currency units. Inventory, fixed assets and intangibles are non-monetary assets, and deferred revenue is a non-monetary liability.

1. All assets and liabilities at the current rate.

Solution. Cash 3,000 × 0.80 = EUR2,400, a fall of EUR600. Inventory 12,000 × 0.80 = EUR9,600, a fall of EUR2,400. Total assets are EUR12,000, down EUR3,000. Notes payable 5,000 × 0.80 = EUR4,000, down EUR1,000. Common stock stays at EUR10,000.

Liabilities plus common stock now total 4,000 + 10,000 = EUR14,000 against assets of EUR12,000, so a negative translation adjustment of EUR2,000 is required in stockholders equity.

Read as a net gain or loss, this is a EUR600 loss on cash plus a EUR2,400 loss on inventory less a EUR1,000 gain on notes payable, giving a net translation loss of EUR2,000. It is unrealized, but it could be realized: if Spanco sold Amerco at its book value of US$10,000, the proceeds convert at 0.80 into EUR8,000 against an original investment of EUR10,000, a realized loss of exactly EUR2,000.

2. Only monetary assets and liabilities at the current rate.

Solution. Cash 3,000 × 0.80 = EUR2,400, a fall of EUR600. Inventory is non-monetary, so it stays at its euro historical cost of 12,000 × 1.00 = EUR12,000. Total assets are EUR14,400, down only EUR600. Notes payable 5,000 × 0.80 = EUR4,000, down EUR1,000. Common stock stays at EUR10,000, so liabilities plus common stock total EUR14,000.

Assets of EUR14,400 exceed EUR14,000, so a positive translation adjustment of EUR400 is required. As a net figure that is a EUR600 loss on cash against a EUR1,000 gain on notes payable.

This gain is also unrealized, and again there is a path to realizing it. Suppose Amerco used its US$3,000 of cash to pay down liabilities, leaving US$2,000 outstanding. At the 31 December 20X1 rate Spanco would have had to send EUR2,000 to settle US$2,000. At the 31 March 20X2 rate it needs only 2,000 × 0.80 = EUR1,600. The saving of EUR400 is the gain.

3. Why do the two approaches disagree by EUR2,400?

Solution. The only item treated differently is inventory. Under the first approach it is revalued at the current rate and loses 12,000 × (1.00 − 0.80) = EUR2,400 of euro carrying value. Under the second it is frozen at historical cost and loses nothing. The adjustment therefore moves from −2,000 to +400, a swing of exactly EUR2,400. Nothing economic happened to the inventory in either case: the entire difference is a measurement convention.

Balance sheet exposure

Items translated at the current rate are revalued from one balance sheet to the next in parent currency terms, so they are exposed to translation adjustment. Items translated at historical rates never change in parent currency terms and are not exposed. This exposure is called balance sheet translation exposure, or accounting exposure, and it is a completely separate idea from the transaction exposure of the previous section.

A net asset balance sheet exposure arises where the exposed assets, meaning those carried at the current rate, outweigh the exposed liabilities.
A net liability balance sheet exposure is the mirror image: exposed liabilities outweigh exposed assets.

Sign of the current period translation adjustment

Balance sheet exposure	Foreign currency strengthens	Foreign currency weakens
Net asset	Adjustment is positive	Adjustment is negative
Net liability	Adjustment is negative	Adjustment is positive

This grid has the same shape as the transaction gain and loss grid, which is not a coincidence: both describe what happens when a position denominated in one currency is remeasured into another. In the monetary approach of Example 4, exposed assets of US$3,000 were less than exposed liabilities of US$5,000, so Amerco had a net liability exposure, and the foreign currency weakened, which produced the positive adjustment of EUR400.

All terms measured in the presentation currency after applying each item’s own rate. The adjustment is always the balancing figure, never an independently computed number.

After the first period, the figure carried on the balance sheet is a cumulative translation adjustment: the running sum of each period’s adjustment. Suppose Spanco puts every asset and liability of Amerco on the current rate, so a net asset exposure exists, and the dollar softening over the first quarter of 20X2 produces the negative EUR2,000 already computed. Suppose the dollar then strengthens in the second quarter, with the net asset exposure still in place, producing a positive adjustment of EUR500 for that quarter. The current period adjustment is positive, but the cumulative adjustment at 30 June 20X2 is still negative, now at EUR1,500. A single quarter of movement rarely reverses a cumulative balance.

The two approaches demonstrated with Amerco have names. Translating all assets and liabilities at the current rate is the current rate method. Translating only monetary assets and liabilities at the current rate is the monetary and non-monetary method. A variant of that second method also translates non-monetary items that are carried at current value on the balance sheet date at the current rate, and that variant is the temporal method.

The organising idea of the temporal method is measurement basis preservation. If an item is carried in the foreign currency accounts at a current value, it should still be carried at a current value after translation, so it takes the current rate. If it is carried at historical cost, it should still reflect historical cost after translation, so it takes the historical rate. Neither the IASB nor the FASB uses these names, but the procedures each requires amount to one or the other.

Which method applies

The choice is not free. It follows from the functional currency of the foreign entity, which is why the functional currency decision of Section 1 carries so much weight.

Figure 2: From functional currency to method to reporting location

The left branch applies when the subsidiary operates in its own local currency. The right branch applies when the subsidiary effectively operates in the currency of the parent, and US GAAP call that path remeasurement.

Foreign currency is the functional currency: the current rate method

This is the common case. A Japanese subsidiary of a French parent will usually have the yen as its functional currency while the parent presents in euro. The procedures are:

Every asset and every liability takes the current rate at the balance sheet date.
Equity accounts apart from retained earnings take historical rates.
Revenues and expenses take the rate in force when each transaction occurred, which in practice means a period average.

The cumulative translation adjustment is reported as a separate component of stockholders equity. The concept behind the method is that the entire net investment in the foreign entity is exposed, so everything is revalued at each balance sheet date. The resulting gain or loss is unrealized and will be realized only on disposal, at which point the cumulative amount related to that entity is transferred to net income as a realized gain or loss.

Because every asset and every liability is translated at the same current rate, and total assets exceed total liabilities for any entity with positive equity, the current rate method produces a net asset balance sheet exposure in all but the rare case of negative equity. A strengthening foreign currency therefore increases a positive cumulative adjustment or shrinks a negative one, and a weakening foreign currency does the reverse.

The parent presentation currency is the functional currency: the temporal method

A German manufacturer might own a Swiss distribution subsidiary that operates day to day in euro. Swiss law still requires that subsidiary to keep its books in Swiss francs, so the franc statements have to be restated into euro as though the transactions had been recorded in euro all along. US GAAP call this remeasurement. IFRS describe it as reporting foreign currency transactions in the functional currency. The procedures are:

Monetary assets and monetary liabilities take the current rate. Non-monetary items held at historical cost take historical rates. Non-monetary items held at current value take the rate on the day that current value was struck.
Equity accounts apart from retained earnings take historical rates.
Revenues and most expenses take the rate when the transaction occurred, with period averages used in practice. Expenses tied to non-monetary assets, meaning cost of goods sold, depreciation and amortisation, take whatever rates were applied to the assets they relate to.

Both frameworks then push the balancing amount through net income as a gain or loss. The justification is one of timing: if the entity effectively operates in the parent currency, then its foreign currency monetary items will produce effects realized in the near future, so they belong in current income rather than parked in equity.

The temporal method can produce either a net asset or a net liability exposure, because only some items sit in the exposed group. In practice most liabilities are monetary, whereas only cash and receivables are monetary assets and other assets are usually carried at historical cost. Exposed liabilities therefore often exceed exposed assets, which is why the temporal method typically produces a net liability balance sheet exposure.

Figure 3: Which balance sheet items are exposed under each method

Under the current rate method the exposed group is the whole balance sheet, so exposure equals net assets. Under the temporal method the exposed group excludes inventory carried at cost and fixed assets, which usually leaves exposed liabilities larger than exposed assets.

Retained earnings and inventory cost flow

Equity accounts translate at historical rates under both methods, which creates a small problem for retained earnings, since retained earnings are an accumulation of past income less past dividends. The convention is a roll-forward rather than a single rate.

Net income is translated by whichever method applies to the income statement, and dividends are translated at the rate on the date they were declared. In the first year of operations the beginning balance is zero.

Under the temporal method the company must keep a record of the exchange rates in force when each non-monetary asset was acquired, since inventory, prepaid expenses, fixed assets and intangibles are all translated at historical rates. No such record is needed under the current rate method, and this bookkeeping burden is one practical reason the temporal method is harder to apply.

The historical rate used for inventory and cost of goods sold depends on the cost flow assumption. Under first in, first out, ending inventory consists of the most recently acquired items and is translated at relatively recent rates, while cost of goods sold is translated at older rates. Under last in, first out the pattern reverses: ending inventory is translated at older rates. Under weighted average cost, both ending inventory and cost of goods sold are translated at the weighted average rate for the year.

Translation rules under IFRS and US GAAP

Item	Functional currency is the foreign currency: current rate method	Functional currency is the parent presentation currency: temporal method
Monetary assets, such as cash and receivables	Current rate	Current rate
Non-monetary assets held at current value	Current rate	Current rate
Non-monetary assets held at historical cost	Current rate	Historical rates
Monetary liabilities, such as payables, accruals, long-term debt, deferred taxes	Current rate	Current rate
Non-monetary liabilities held at current value	Current rate	Current rate
Non-monetary liabilities not held at current value, such as deferred revenue	Current rate	Historical rates
Equity accounts apart from retained earnings	Historical rates	Historical rates
Retained earnings	Roll-forward: opening balance plus translated net income less dividends at historical rate	Roll-forward: opening balance plus translated net income less dividends at historical rate
Revenues	Average rate	Average rate
Most expenses	Average rate	Average rate
Expenses tied to assets translated at historical rates, such as cost of goods sold, depreciation, amortisation	Average rate	Historical rates
Treatment of the translation adjustment	Accumulated as a separate component of equity	Included as a gain or loss in net income

Differences between IFRS and US GAAP on translation are minimal, with the single significant exception of entities located in highly inflationary economies.

The hyperinflation override

Once a foreign entity sits inside a highly inflationary economy, its functional currency stops mattering for the choice of method. IFRS require the statements to be restated for local inflation under IAS 29 first and then translated at the current rate. US GAAP do not permit restatement for inflation and instead require remeasurement as if the functional currency were the reporting currency, which is the temporal method.

US GAAP set the threshold at a cumulative three-year inflation rate above 100 percent, though the test is meant to be applied with judgement, since the trend of inflation can matter as much as its level. A cumulative three-year rate of 100 percent corresponds to an average of roughly 26 percent per year, since 1.26 compounded over three years is very close to 2.0. IAS 21 offers no numerical definition, but IAS 29 indicates that a cumulative rate approaching or exceeding 100 percent over three years is an indicator of hyperinflation. If a country stops being classified as highly inflationary, the functional currency of the entity has to be identified again to pick the appropriate method.

The rules are only half the skill. The other half is the order of operations, which differs between the two methods and is where most calculation errors occur. Under the current rate method you translate the income statement first, carry the resulting retained earnings into the balance sheet, and let the translation adjustment be the plug. Under the temporal method you translate the balance sheet first, let retained earnings be the plug, carry that figure back into the income statement, and let the translation gain or loss be the amount that makes the income statement arrive at it.

Interco is a Europe-based company presenting in euro. On 1 January 20X1 it establishes a wholly owned Canadian subsidiary, Canadaco, funding it with an equity investment and a long-term note payable to a Canadian bank used to purchase property and equipment.

Canadaco opening balance sheet, 1 January 20X1, Canadian dollars

Assets	C$	Liabilities and equity	C$
Cash	1,500,000	Long-term note payable	3,000,000
Property and equipment	3,000,000	Capital stock	1,500,000
Total	4,500,000	Total	4,500,000

Canadaco results for 20X1 and closing balance sheet, Canadian dollars

Income statement and retained earnings	C$	Balance sheet at 31 Dec 20X1	C$
Sales	12,000,000	Cash	980,000
Cost of sales	(9,000,000)	Accounts receivable	900,000
Selling expenses	(750,000)	Inventory	1,200,000
Depreciation expense	(300,000)	Total current assets	3,080,000
Interest expense	(270,000)	Property and equipment	3,000,000
Income tax	(500,000)	Less accumulated depreciation	(300,000)
Net income	1,180,000	Total assets	5,780,000
Less dividends, 1 Dec 20X1	(350,000)	Accounts payable	450,000
Retained earnings, 31 Dec 20X1	830,000	Long-term notes payable	3,000,000
		Capital stock	1,500,000
		Retained earnings	830,000

Inventory is measured at historical cost on a first in, first out basis.

Exchange rates, euro per Canadian dollar

Date or basis	Rate
1 January 20X1	0.70
Average, 20X1	0.75
Weighted average over the inventory purchase period	0.74
1 December 20X1, the dividend declaration date	0.78
31 December 20X1	0.80

The Canadian dollar strengthened steadily against the euro through 20X1, from 0.70 at the start of the year to 0.80 at the end.

Example 5 · Worked

Translate the 20X1 financial statements of Canadaco into euro, first assuming the functional currency is the Canadian dollar, then assuming it is the euro.

1. Current rate method: translate the income statement and statement of retained earnings.

Solution. Every income statement line takes the average rate of 0.75, and dividends take the 1 December rate of 0.78.

Sales 12,000,000 × 0.75 = 9,000,000. Cost of goods sold 9,000,000 × 0.75 = 6,750,000. Selling expenses 750,000 × 0.75 = 562,500. Depreciation 300,000 × 0.75 = 225,000. Interest 270,000 × 0.75 = 202,500. Income tax 500,000 × 0.75 = 375,000.

Net income = 9,000,000 − 6,750,000 − 562,500 − 225,000 − 202,500 − 375,000 = EUR885,000.
Dividends 350,000 × 0.78 = 273,000, so retained earnings = 885,000 − 273,000 = EUR612,000, which is carried into the balance sheet.

2. Current rate method: translate the balance sheet and find the translation adjustment.

Solution. Assets and liabilities take 0.80; capital stock takes the historical 0.70.

Cash 980,000 × 0.80 = 784,000. Receivables 900,000 × 0.80 = 720,000. Inventory 1,200,000 × 0.80 = 960,000, so current assets are 2,464,000. Property and equipment 3,000,000 × 0.80 = 2,400,000, accumulated depreciation 300,000 × 0.80 = 240,000, giving total assets of EUR4,624,000.

Accounts payable 450,000 × 0.80 = 360,000 and notes payable 3,000,000 × 0.80 = 2,400,000, so total liabilities are 2,760,000. Capital stock 1,500,000 × 0.70 = 1,050,000. Retained earnings are 612,000 from the income statement.

Translation adjustment = 4,624,000 − 2,760,000 − 1,050,000 − 612,000 = +EUR202,000, reported in equity. Total equity is 1,050,000 + 612,000 + 202,000 = EUR1,864,000.

3. Temporal method: translate the balance sheet first.

Solution. Monetary items take 0.80. Inventory takes the 0.74 rate at which it was acquired. Property and equipment and the related accumulated depreciation take the 0.70 rate at acquisition.

Cash 784,000, receivables 720,000, inventory 1,200,000 × 0.74 = 888,000, so current assets are 2,392,000. Property and equipment 3,000,000 × 0.70 = 2,100,000, accumulated depreciation 300,000 × 0.70 = 210,000. Total assets are EUR4,282,000.

Liabilities are unchanged at 2,760,000 because both are monetary, and capital stock is unchanged at 1,050,000. Retained earnings must therefore be the plug: 4,282,000 − 2,760,000 − 1,050,000 = EUR472,000.

4. Temporal method: translate the income statement and find the translation loss.

Solution. Sales, selling expenses, interest and tax take the average 0.75. Cost of goods sold takes 0.74 because it follows inventory, and depreciation takes 0.70 because it follows the fixed assets.

Sales 9,000,000. Cost of goods sold 9,000,000 × 0.74 = 6,660,000. Selling expenses 562,500. Depreciation 300,000 × 0.70 = 210,000. Interest 202,500. Tax 375,000.
Income before the translation item = 9,000,000 − 6,660,000 − 562,500 − 210,000 − 202,500 − 375,000 = EUR990,000.

Net income has to reconcile to the retained earnings figure from the balance sheet: 472,000 + 273,000 of dividends = EUR745,000. The plug is therefore a translation loss of 745,000 − 990,000 = EUR245,000, reported inside net income. Total equity is 1,050,000 + 472,000 = EUR1,522,000.

5. Explain the sign of each adjustment.

Solution. Under the current rate method Canadaco has a net asset exposure, since total assets of C$5,780,000 exceed total liabilities of C$3,450,000, and the Canadian dollar strengthened. Net asset exposure plus a strengthening foreign currency gives a positive adjustment, and the result is +EUR202,000.

Under the temporal method the exposed assets are cash plus receivables of C$1,880,000 while the exposed liabilities are payables plus notes of C$3,450,000, so there is a net liability exposure. Net liability exposure plus a strengthening foreign currency gives a negative adjustment, and the result is the EUR245,000 loss. The two adjustments have opposite signs on the same subsidiary in the same year, purely because the two methods define the exposed group differently.

Canadaco translated statements, euro, side by side

Item	C$	Current rate, rate applied	Current rate, EUR	Temporal, rate applied	Temporal, EUR
Sales	12,000,000	0.75 average	9,000,000	0.75 average	9,000,000
Cost of goods sold	(9,000,000)	0.75 average	(6,750,000)	0.74 historical	(6,660,000)
Selling expenses	(750,000)	0.75 average	(562,500)	0.75 average	(562,500)
Depreciation expense	(300,000)	0.75 average	(225,000)	0.70 historical	(210,000)
Interest expense	(270,000)	0.75 average	(202,500)	0.75 average	(202,500)
Income tax	(500,000)	0.75 average	(375,000)	0.75 average	(375,000)
Income before translation gain or loss	1,180,000		885,000		990,000
Translation gain or loss	Not applicable		Not applicable	to balance	(245,000)
Net income	1,180,000		885,000		745,000
Less dividends, 1 Dec 20X1	(350,000)	0.78 historical	(273,000)	0.78 historical	(273,000)
Retained earnings, 31 Dec 20X1	830,000		612,000	from balance sheet	472,000
Cash	980,000	0.80 current	784,000	0.80 current	784,000
Accounts receivable	900,000	0.80 current	720,000	0.80 current	720,000
Inventory	1,200,000	0.80 current	960,000	0.74 historical	888,000
Total current assets	3,080,000		2,464,000		2,392,000
Property and equipment	3,000,000	0.80 current	2,400,000	0.70 historical	2,100,000
Less accumulated depreciation	(300,000)	0.80 current	(240,000)	0.70 historical	(210,000)
Total assets	5,780,000		4,624,000		4,282,000
Accounts payable	450,000	0.80 current	360,000	0.80 current	360,000
Long-term notes payable	3,000,000	0.80 current	2,400,000	0.80 current	2,400,000
Total liabilities	3,450,000		2,760,000		2,760,000
Capital stock	1,500,000	0.70 historical	1,050,000	0.70 historical	1,050,000
Retained earnings	830,000	from income statement	612,000	to balance	472,000
Translation adjustment	Not applicable	to balance	202,000		Not applicable
Total liabilities and equity	5,780,000		4,624,000		4,282,000

The same Canadian subsidiary, the same underlying year, and two sets of consolidated numbers that differ materially. This is the point of the reading for an analyst: the reported figures are a function of a functional currency judgement, and comparing companies without knowing which judgement each made is comparing different things.

Canadaco under the two methods, euro

Item	Current rate	Temporal	Difference, percent
Sales	9,000,000	9,000,000	0.0
Net income	885,000	745,000	+18.8
Income before translation gain or loss	885,000	990,000	−10.6
Total assets	4,624,000	4,282,000	+8.0
Total equity	1,864,000	1,522,000	+22.5

Differences are expressed relative to the temporal method figure.

Read the first two income rows together, because they say opposite things. Net income is 18.8 percent higher under the current rate method, computed as (885,000 − 745,000) ÷ 745,000. Income before the translation item is 10.6 percent lower under the current rate method, computed as (885,000 − 990,000) ÷ 990,000. The reversal happens entirely because the current rate method keeps the translation adjustment out of income while the temporal method forces it in. Strip the translation loss out and the temporal method is the more profitable presentation, because cost of goods sold and depreciation are translated at the lower historical rates of 0.74 and 0.70 rather than the average of 0.75.

Total assets are 8.0 percent higher under the current rate method because every asset is translated at 0.80 rather than at the historical rates of 0.74 and 0.70 for inventory and fixed assets. Total equity is 22.5 percent higher for two compounding reasons: a positive EUR202,000 adjustment is added to equity, and no EUR245,000 loss is deducted through retained earnings.

Ratios computed three ways

The cleanest way to see the distortion is to compute the same ratios from the original Canadian dollar statements and then from each translated set.

Canadaco ratios from local currency and from each translated set

Ratio	C$	Current rate, EUR	Temporal, EUR
Current ratio	6.84	6.84	6.64
Debt to assets	0.52	0.52	0.56
Debt to equity	1.29	1.29	1.58
Interest coverage	7.22	7.22	7.74
Gross profit margin	0.25	0.25	0.26
Operating profit margin	0.16	0.16	0.17
Net profit margin	0.10	0.10	0.08
Receivables turnover	13.33	12.50	12.50
Inventory turnover	7.50	7.03	7.50
Fixed asset turnover	4.44	4.17	4.76
Return on assets	0.20	0.19	0.17
Return on equity	0.51	0.47	0.49

Twelve ratios. Debt is taken as the long-term note payable of C$3,000,000; earnings before interest and taxes in Canadian dollars is 1,180,000 + 500,000 + 270,000 = 1,950,000; equity in Canadian dollars is 1,500,000 + 830,000 = 2,330,000.

Three patterns are worth extracting from that table.

Only one ratio survives both methods unchanged. Of the twelve, receivables turnover alone is identical at 12.50 under both, because it is the only ratio in which numerator and denominator take the same type of rate under both methods: sales at the average rate and receivables at the current rate. Every other ratio has at least one component whose rate type changes between methods. The current ratio moves from 6.84 to 6.64 solely because inventory goes from 0.80 to 0.74, and because 0.80 is higher than 0.74 the current rate method produces the larger figure.

The current rate method preserves the local currency relationships that live inside a single statement. Compare the C$ column with the current rate column: the current ratio, both leverage ratios, interest coverage and all three margins are identical. Each of those is built purely from the balance sheet or purely from the income statement, so the single rate applied within that statement cancels between numerator and denominator. The ratios that break are exactly those that mix the two statements, meaning the turnover and return ratios, because the balance sheet uses the current rate of 0.80 while the income statement uses the average of 0.75. In this case every turnover and return ratio is larger in Canadian dollars than in the current rate euro column. Those distortions would vanish if revenues and expenses were also translated at the current rate.

The temporal method distorts almost everything. Comparing the C$ column with the temporal column, only inventory turnover is preserved, at 7.50, because cost of goods sold and inventory are both translated at 0.74. The direction of the distortion cannot be generalised. Under the temporal method, gross profit margin and operating profit margin come out higher than the Canadian dollar figures, at 0.26 and 0.17 against 0.25 and 0.16, while net profit margin comes out lower, at 0.08 against 0.10. Receivables turnover falls from 13.33 to 12.50, inventory turnover is unchanged, and fixed asset turnover rises from 4.44 to 4.76. Every one of these movements comes from the mix of rate types, not from anything the business did.

Why receivables turnover is the exception

A ratio is unaffected by the choice of method when both of its components change by the same factor. Sales are translated at the average rate under both methods, and receivables are translated at the current rate under both methods, so the pair 9,000,000 over 720,000 appears in both columns. Nothing deeper is going on. The same logic tells you which other ratios would be immune under a different fact pattern: any ratio whose numerator and denominator both consist entirely of monetary items, or entirely of items translated at the average rate, will match across the two methods. This is a faster test than recomputing the ratio twice.

One more caution on net income. The temporal method produced a smaller net income here only because the standards require the translation loss to run through income. The loss arose because the Canadian dollar strengthened while Canadaco held more monetary liabilities than monetary assets. Reverse the monetary position and the sign reverses with it, which is what the next section demonstrates.

Two variables drive every translation outcome: the sign of the balance sheet exposure and the direction of the currency move. The following two examples hold everything else constant and change one variable at a time, which is the fastest way to build the intuition an exam item set will test.

Example 6 · Worked

Take Canadaco at its 1 January 20X1 starting point: C$1,500,000 of cash and C$3,000,000 of property and equipment. Under Case A the equipment is funded by a long-term note payable, which leaves net monetary liabilities of C$1,500,000, being C$3,000,000 of debt against C$1,500,000 of cash. Under Case B the equipment is funded by capital stock instead, which leaves net monetary assets of C$1,500,000.

So that nothing but the monetary position differs, hold interest expense at C$270,000 in Case B as well, despite the absence of borrowing. The assumption is unrealistic but it strips out every other source of difference. The 20X1 results, dividends of C$350,000 and total assets of C$5,780,000 at year end are the same in both cases, the functional currency is the euro, so the temporal method applies, and the exchange rates are those already used for Canadaco.

1. Translate the 31 December 20X1 balance sheet under the temporal method in each case.

Solution. The asset side is identical in both cases: cash 784,000, receivables 720,000, inventory at 0.74 giving 888,000, current assets 2,392,000, property and equipment at 0.70 giving 2,100,000, accumulated depreciation at 0.70 giving 210,000, and total assets of EUR4,282,000.

Case A. Accounts payable 360,000 plus notes payable 2,400,000 gives liabilities of 2,760,000. Capital stock 1,500,000 × 0.70 = 1,050,000. Retained earnings must be 4,282,000 − 2,760,000 − 1,050,000 = EUR472,000.

Case B. Liabilities are only the payable of 360,000. Capital stock 4,500,000 × 0.70 = 3,150,000. Retained earnings must be 4,282,000 − 360,000 − 3,150,000 = EUR772,000.

2. Explain the EUR300,000 difference in retained earnings.

Solution. The difference of 772,000 − 472,000 = EUR300,000 is exactly the translation loss caused by holding a Canadian dollar note payable while the Canadian dollar strengthened: C$3,000,000 × (0.80 − 0.70) = EUR300,000. Notes payable are exposed under the temporal method; capital stock is not, because it is translated at a historical rate. Financing the equipment with equity rather than debt would have avoided the loss entirely.

3. Translate the income statement in each case and identify the translation item.

Solution. Income before the translation item is the same EUR990,000 in both cases, since every revenue and expense line is identical.

Case A. Net income must be 472,000 + 273,000 = EUR745,000, so a translation loss of EUR245,000 is subtracted.
Case B. Net income must be 772,000 + 273,000 = EUR1,045,000, so a translation gain of EUR55,000 is added.

The EUR300,000 gap in net income equals the loss on the note payable, exactly as it did on the balance sheet. Note the practical lesson: under the temporal method a company can neutralise its translation exposure by matching monetary assets to monetary liabilities. Under the current rate method the equivalent would require total assets to equal total liabilities, which means zero stockholders equity, so exposure is effectively unavoidable.

Example 7 · Worked

Return to the original Canadaco fact pattern and translate it under three exchange rate scenarios: the Canadian dollar stable, strengthening, and weakening.

Euro per Canadian dollar under three scenarios

Date or basis	Stable	Strengthens	Weakens
1 January 20X1	0.70	0.70	0.70
Average, 20X1	0.70	0.75	0.65
Weighted average over the inventory purchase period	0.70	0.74	0.66
Dividend declaration date	0.70	0.78	0.62
31 December 20X1	0.70	0.80	0.60

1. Current rate method: what happens to sales, net income and the translation adjustment?

Solution. Under a stable rate every line takes 0.70. Sales are 12,000,000 × 0.70 = 8,400,000 and net income is (12,000,000 − 9,000,000 − 750,000 − 300,000 − 270,000 − 500,000) × 0.70 = 1,180,000 × 0.70 = EUR826,000. Dividends are 245,000 and retained earnings are 581,000. Total assets are 5,780,000 × 0.70 = 4,046,000, liabilities are 2,415,000, capital stock is 1,050,000, so the translation adjustment is 4,046,000 − 2,415,000 − 1,050,000 − 581,000 = zero.

Under the strengthening scenario, already computed: sales EUR9,000,000, net income EUR885,000, adjustment +EUR202,000, total assets EUR4,624,000, total equity EUR1,864,000.

Under the weakening scenario every line takes the lower rates. Sales 12,000,000 × 0.65 = 7,800,000 and net income 1,180,000 × 0.65 = EUR767,000. Dividends 350,000 × 0.62 = 217,000, so retained earnings are 550,000. Total assets 5,780,000 × 0.60 = 3,468,000, liabilities 3,450,000 × 0.60 = 2,070,000, capital stock still 1,050,000, so the adjustment is 3,468,000 − 2,070,000 − 1,050,000 − 550,000 = −EUR202,000 and total equity is EUR1,398,000 against EUR1,631,000 in the stable case.

The pattern is uniform: under the current rate method a stronger foreign currency raises revenues, income, assets, liabilities and equity, and a weaker foreign currency lowers all five.

2. Temporal method: what happens under the same three scenarios?

Solution. Under a stable rate the temporal method gives the same answers as the current rate method, since every rate is 0.70: total assets EUR4,046,000, income before the translation item EUR826,000, retained earnings EUR581,000, and a translation item of zero.

Under the strengthening scenario: total assets EUR4,282,000, income before the translation item EUR990,000, translation loss of EUR245,000, net income EUR745,000, retained earnings EUR472,000 and total equity EUR1,522,000.

Under the weakening scenario, inventory takes 0.66 giving 792,000, so current assets are 588,000 + 540,000 + 792,000 = 1,920,000. Property and equipment stay at the historical 0.70, so net fixed assets are 1,890,000 and total assets are EUR3,810,000. Liabilities are 2,070,000 and capital stock 1,050,000, so retained earnings must be EUR690,000 and total equity is EUR1,740,000.
On the income statement, cost of goods sold takes 0.66 giving 5,940,000 and depreciation stays at 210,000, so income before the translation item is 7,800,000 − 5,940,000 − 487,500 − 210,000 − 175,500 − 325,000 = EUR662,000. Net income must be 690,000 + 217,000 = EUR907,000, so a translation gain of EUR245,000 is added.

Because Canadaco carries a net monetary liability position, a stronger Canadian dollar produces a loss and a weaker one produces a gain. Note the counterintuitive result: a weaker Canadian dollar gives smaller sales, at EUR7,800,000 against EUR8,400,000, but larger net income, at EUR907,000 against EUR826,000, and larger equity, at EUR1,740,000 against EUR1,631,000.

Figure 4: The temporal method outcome grid

Under the temporal method the row you are in decides the sign. Under the current rate method there is effectively only the net asset row, so a stronger currency always gives a positive adjustment.

Effect of a currency move on the consolidated statements, against a stable currency baseline

Scenario	Temporal, net monetary liability	Temporal, net monetary asset	Current rate
Foreign currency strengthens	Revenues up, assets up, liabilities up, net income down, equity down, translation loss	Revenues up, assets up, liabilities up, net income up, equity up, translation gain	Revenues up, assets up, liabilities up, net income up, equity up, positive translation adjustment
Foreign currency weakens	Revenues down, assets down, liabilities down, net income up, equity up, translation gain	Revenues down, assets down, liabilities down, net income down, equity down, translation loss	Revenues down, assets down, liabilities down, net income down, equity down, negative translation adjustment

Two features of that table deserve emphasis. Revenues, assets and liabilities always move in the same direction as the foreign currency, in every column. Net income and equity are the only lines whose direction depends on the method and the exposure, and they are precisely the lines analysts anchor on.

This is the one area where IFRS and US GAAP genuinely diverge, and the divergence is not cosmetic: the two frameworks can report very different amounts for the same subsidiary. US GAAP require the temporal method with the translation gain or loss in net income, and no restatement for inflation. IFRS require the statements to be restated for local inflation under IAS 29 first, and only then translated at the current rate.

The disappearing plant problem

The FASB originally proposed the restate-then-translate approach and met firm resistance from US multinationals. Requiring the temporal method solves a specific pathology. In a high inflation country the local currency loses purchasing power internally and usually loses value externally at the same time. Translating the unchanged historical cost of land or buildings at successively weaker exchange rates makes those assets shrink year after year in the consolidated accounts until they nearly vanish, even though the physical asset is untouched and its local currency value is probably rising.

Example 8 · Worked

At the start of this century Turkey still counted among the handful of highly inflationary countries. A US-based company set up a Turkish subsidiary on 1 January 2000 and sent it US$1,000, which bought land costing TL542,700,000, calculated as TL542,700 per US dollar multiplied by US$1,000. Assume no other assets or liabilities. Turkey no longer carries that classification, and in 2010 the International Practices Task Force, part of the SEC Regulations Committee of the Center for Audit Quality, indicated that Venezuela had crossed the thresholds.

Turkish lira per US dollar and annual inflation

Date	Exchange rate	Year	Inflation rate
01 Jan 2000	TL542,700 = US$1
31 Dec 2000	TL670,800 = US$1	2000	38%
31 Dec 2001	TL1,474,525 = US$1	2001	69%
31 Dec 2002	TL1,669,000 = US$1	2002	45%

1. Translate the land at the current rate each year and compute the annual and cumulative translation gains or losses.

Solution. Divide the unchanged carrying value of TL542,700,000 by each year end rate.

31 Dec 2000: 542,700,000 ÷ 670,800 = US$809, an annual loss of 809 − 1,000 = US$191.
31 Dec 2001: 542,700,000 ÷ 1,474,525 = US$368, an annual loss of 368 − 809 = US$441.
31 Dec 2002: 542,700,000 ÷ 1,669,000 = US$325, an annual loss of 368 − 325 = US$43.

Cumulative losses run US$191, then US$632, then US$675. Land bought with US$1,000 sits on the consolidated balance sheet at US$325 after three years, and land is not even depreciable. This is why the approach is unacceptable under both IFRS and US GAAP for a highly inflationary economy: it captures the exchange rate move but ignores the local currency appreciation that inflation is producing at the same time.

2. Repeat under the temporal method as required by US GAAP.

Solution. Land is a non-monetary asset carried at historical cost, so it is translated at the historical rate of TL542,700 every year: 542,700,000 ÷ 542,700 = US$1,000 at each date. There is no annual or cumulative translation gain or loss on it at all, because items translated at historical rates are not exposed. The asset cannot disappear.

3. Repeat under the IFRS approach: restate for inflation, then translate at the current rate.

Solution. Restate the carrying value by each year of inflation, then divide by the year end rate.

31 Dec 2000: 542,700,000 × 1.38 = TL748,926,000; divided by 670,800 gives US$1,116, an annual gain of US$116.
31 Dec 2001: 748,926,000 × 1.69 = TL1,265,684,940; divided by 1,474,525 gives US$858, an annual loss of US$258.
31 Dec 2002: 1,265,684,940 × 1.45 = TL1,835,243,163; divided by 1,669,000 gives US$1,100, an annual gain of US$242.

Cumulative amounts run US$116, then a negative US$142, then a positive US$100. The land ends up at US$1,100 with a cumulative unrealized gain of US$100. That gain could be realized if the land appreciated in lira terms at the rate of local inflation, the subsidiary sold it for TL1,835,243,163, and the proceeds converted at the 31 December 2002 rate into US$1,100. Short of a formal appraisal, this approach probably comes closest to economic reality, because it picks up the lira value the land is likely to have gained as well as the currency move itself.

Figure 5: Translated carrying value of the same piece of land

The lower line is the disappearing plant: US$1,000 of land shrinks to US$325 in three years. The flat line is the US GAAP answer. The upper dashed line is the IFRS answer, which tracks both inflation and the currency.

What IAS 29 restatement actually does

The IFRS route has its own set of procedures, and they are close in spirit to the temporal method: split the balance sheet into items already expressed in current purchasing power and items that are not.

Monetary assets and liabilities, meaning cash, receivables and payables, are not restated, because they are already expressed in the monetary unit current at the balance sheet date.
Non-monetary assets and liabilities are restated for the change in general purchasing power. An item carried at historical cost is restated by the index movement between acquisition and the balance sheet date. An item carried at a revalued amount, such as property, plant and equipment revalued under IAS 16, is restated only from the revaluation date onward.
Every component of stockholders equity is restated using the price level movement from the start of the period, or from the contribution date if that falls later.
Every income statement line is restated using the index movement between the date the item was first recorded and the balance sheet date.
Net income also picks up the purchasing power gain or loss produced by holding monetary items through a period of inflation.

GPI is the general price index. Monetary items take a factor of one, since the index at the balance sheet date appears in both numerator and denominator.

Purchasing power gains and losses

Suppose the general price index is 100 on 1 January 20X1, so a representative basket of goods and services costs $100 that day. By the end of 20X1 the same basket costs $120, an inflation rate of ($120 − $100) ÷ $100 = 20 percent. Cash of $100 bought one basket in January. A year later, with the index at 120, the same $100 buys only 100 ÷ 120 = 83.3 percent of a basket. Since $120 is now needed to buy what $100 once bought, holding $100 of cash through the year produced a purchasing power loss of $20.

Borrowing works the other way. A company that expects $120 of cash at the end of 20X1 can wait and buy exactly 1.0 basket when the index is 120. If instead it borrows $120 on 1 January when the index is 100, it can buy 120 ÷ 100 = 1.2 baskets immediately, a purchasing power gain of $20, less whatever interest the borrowing costs.

A net monetary asset position produces a loss instead. This is directly analogous to a translation gain under the temporal method with a net monetary liability position when the currency is weakening.

Only monetary items are exposed to inflation risk, because everything else is restated. That is the same architecture the temporal method uses for currency risk, which is why the two approaches can converge.

Example 9 · Worked

On 1 January 20X1 ABC Company set up a subsidiary abroad, funded with a mix of debt and equity. That subsidiary bought land the same day and rents it to a local farmer. Its first year statements, expressed in foreign currency units (FC), appear below.

Foreign subsidiary statements in FC

Item	1 Jan 20X1	31 Dec 20X1
Rent revenue		1,000
Interest expense		(250)
Net income		750
Cash	1,000	1,750
Land	9,000	9,000
Total assets	10,000	10,750
Note payable at 5 percent	5,000	5,000
Capital stock	5,000	5,000
Retained earnings	0	750
Total liabilities and equity	10,000	10,750

The general price index stood at 100 on 1 January 20X1, averaged 125 over the year and reached 200 on 31 December 20X1, so 20X1 inflation was 100 percent and the country plainly qualifies as highly inflationary. The FC weakened accordingly: US$1.00 per FC on 1 January, an average of US$0.80, and US$0.50 on 31 December.

1. Apply the IFRS approach: restate for inflation, then translate at the current rate.

Solution. Restatement factors are the year end index of 200 over the index at the relevant date, and every restated amount is then translated at US$0.50.

Cash 1,750 × 200/200 = FC1,750, translated to US$875. Land 9,000 × 200/100 = FC18,000, translated to US$9,000. Total assets FC19,750, or US$9,875.
Note payable 5,000 × 200/200 = FC5,000, or US$2,500. Capital stock 5,000 × 200/100 = FC10,000, or US$5,000. Retained earnings are the remainder: 19,750 − 5,000 − 10,000 = FC4,750, or US$2,375.

On the income statement, revenue 1,000 × 200/125 = FC1,600, or US$800. Interest 250 × 200/125 = FC400, or US$200. The subtotal is FC1,200, or US$600. Net income has to reach FC4,750, so the purchasing power gain is FC3,550, or US$1,775.

Because every restated amount is translated at the same current rate, no translation adjustment arises at all.

2. Verify the purchasing power gain of FC3,550 directly.

Solution. Take each monetary position and apply the change in the index over the period it was held.

Gain from holding the note payable: 5,000 × (200 − 100) ÷ 100 = +FC5,000.
Loss from holding the opening cash balance: −1,000 × (200 − 100) ÷ 100 = −FC1,000.
Loss on the cash accumulated during the year, measured from the average index: −750 × (200 − 125) ÷ 125 = −FC450.
Net purchasing power gain: 5,000 − 1,000 − 450 = FC3,550.

3. Apply the US GAAP approach: the temporal method with no restatement.

Solution. Cash 1,750 × 0.50 = US$875. Land is non-monetary at historical cost, so 9,000 × 1.00 = US$9,000. Total assets US$9,875. Note payable 5,000 × 0.50 = US$2,500 and capital stock 5,000 × 1.00 = US$5,000, so retained earnings must be 9,875 − 2,500 − 5,000 = US$2,375.

Revenue 1,000 × 0.80 = US$800 and interest 250 × 0.80 = US$200, giving a subtotal of US$600. Since no dividend was paid, net income must equal the increase in retained earnings of US$2,375, so the translation gain is US$1,775.

The two frameworks produce identical US dollar amounts on every single line.

4. Why do they agree here, and what breaks the agreement?

Solution. They agree because the exchange rate moved by exactly the inverse of the price index. The index doubled from 100 to 200, so the FC lost half its purchasing power, and the FC also lost exactly half its dollar value, from US$1.00 to US$0.50. That is purchasing power parity holding precisely, which almost never happens.

Change one number to see the divergence. If the 31 December rate had been US$0.60 per FC rather than US$0.50, US GAAP would give net income of 1,750 × 0.60 + 9,000 − 5,000 × 0.60 − 5,000 = US$2,050, while IFRS would give 19,750 × 0.60 − 5,000 × 0.60 − 10,000 × 0.60 = US$2,850. A difference of US$800 appears from a single rate assumption, which is why this is the one area of the reading where the framework a company reports under genuinely changes the answer.

A group with subsidiaries in many countries will usually apply both methods at the same balance sheet date, because some subsidiaries operate in local currency and others operate in the currency of the parent. The consolidated statements then carry both artefacts simultaneously: a translation gain or loss inside net income from the temporal method subsidiaries, and a separate cumulative translation adjustment in equity from the current rate method subsidiaries.

Exxon Mobil Corporation described its policy in its 2011 annual report. It selects the functional currency for each international subsidiary based on the primary economic environment. Downstream and chemical operations mainly use local currency. In countries with a record of high inflation, mostly in Latin America, and in Singapore, whose output goes largely into the dollar export market, the dollar is used instead. Upstream operations that are largely self-contained inside one country also use local currency, and the report names Canada, Norway, the United Kingdom and continental Europe. Certain upstream operations in Africa and Asia use the dollar, since their crude and natural gas is sold into dollar-priced markets.

Chevron Corporation, in the same industry and the same year, reached almost the opposite conclusion. Its 2011 annual report states that the US dollar is the functional currency for substantially all of its consolidated operations and those of its equity affiliates, with remeasurement gains and losses in current period income and a cumulative translation adjustment used only for the few entities on other functional currencies. Two petroleum majors, the same standards, and a judgement applied differently enough that Chevron runs primarily on the temporal method while Exxon Mobil runs largely on the current rate method.

What must be disclosed

Two disclosures are required under both frameworks. The first is how much exchange difference reached net income. The second is the cumulative translation adjustment carried as a separate component of equity, with its opening and closing balances reconciled. US GAAP add a third: the amount moved out of stockholders equity into current income when a foreign entity is disposed of.

The exchange differences recognised in net income are a sum of two very different things: foreign currency transaction gains and losses from Section 2, and translation gains and losses produced by the temporal method. Neither framework requires the two to be split out, and most companies do not split them. BASF AG is an exception, showing EUR170 million of income on foreign currency and hedging separately from EUR42 million earned on translating foreign currency financial statements, both sitting inside other operating income, with matching detail on the expense side. That distinction matters because the two components respond to different exposures and are managed in different ways.

Example 10 · Worked

Yahoo! Inc. is a US-based company reporting in US dollars under US GAAP. Selected figures from its 2011 filings, in thousands of dollars, are set out below.

Yahoo! accumulated other comprehensive income and net income

Item	2009	2010	2011
Accumulated other comprehensive income, opening	120,276	369,236	504,254
Change in unrealized gains and losses on securities available for sale, after tax	(1,936)	3,813	(16,272)
Currency translation adjustments, after tax	250,896	131,205	209,887
Accumulated other comprehensive income, closing	369,236	504,254	697,869
Net income		1,244,628	1,062,699

Yahoo! revenue excluding total acquisition costs, by segment

Segment	2011 as reported	Currency effect stated in the MD and A
Americas	3,142,879	6,000 lower without the currency move
EMEA	407,467	16,000 lower without the currency move
Asia Pacific	830,482	59,000 lower without the currency move
Total revenue excluding total acquisition costs	4,380,828

1. By how much did accumulated other comprehensive income change in 2011, and where is that visible?

Solution. 697,869 − 504,254 = an increase of $193,615 thousand. Two documents carry it: the equity section on the face of the balance sheet, and the consolidated statement of changes in stockholders equity.

2. What amount of currency translation adjustment reached other comprehensive income in 2011, and where does it come from?

Solution. $209,887 thousand. It arises from applying the current rate method to subsidiaries whose functional currency is a foreign currency. Assuming those subsidiaries hold positive net assets, a positive adjustment in 2011 means foreign currencies strengthened, or equivalently that the US dollar weakened.

3. If the translation adjustment had run through net income instead, how would the change in income from 2010 to 2011 look?

Solution. As reported, net income fell from 1,244,628 to 1,062,699, a change of (1,062,699 − 1,244,628) ÷ 1,244,628 = −14.6 percent.

Adding the translation adjustments gives 1,244,628 + 131,205 = 1,375,833 for 2010 and 1,062,699 + 209,887 = 1,272,586 for 2011. The change becomes (1,272,586 − 1,375,833) ÷ 1,375,833 = −7.5 percent. The decline is roughly halved, because the translation adjustment was larger in the weaker year.

4. What share of revenue excluding total acquisition costs came from Asia Pacific in 2011, and what would it have been at unchanged exchange rates?

Solution. As reported, 830,482 ÷ 4,380,828 = 19.0 percent.

Removing the currency effect from each segment gives Americas 3,142,879 − 6,000 = 3,136,879, EMEA 407,467 − 16,000 = 391,467, and Asia Pacific 830,482 − 59,000 = 771,482, for a total of 4,299,828. Asia Pacific would then be 771,482 ÷ 4,299,828 = 17.9 percent, with the Americas at 73.0 percent against 71.7 percent as reported and EMEA at 9.1 percent against 9.3 percent. Currency movement flattered the fastest-growing segment.

5. From whose perspective does Yahoo! frame its currency risk?

Solution. From the perspective of movements in the US dollar, because the dollar is its reporting currency. A stronger dollar necessarily means weaker foreign currencies, which reduces the translated revenues, expenses and income of the foreign operations. Framing matters when reading these disclosures: a company that says a currency strengthened may be describing its own currency or the foreign one, and the direction of the reported effect flips accordingly.

Adjusting reported income for comparability

Comparing net income across two companies that use different predominant methods is comparing a figure that includes translation effects with one that excludes them. A partial fix is to add the change in the cumulative translation adjustment reported in equity back into income for both companies, which puts all currency effects on the same footing.

Example 11 · Worked

Exxon Mobil uses the current rate method for a substantial number of subsidiaries and reports the resulting adjustments within accumulated other non-owner changes in equity. Chevron uses the temporal method for substantially all subsidiaries and reports the adjustments of the few remaining entities in accumulated other comprehensive loss. Amounts are in millions of US dollars.

1. Compute adjusted net income for both companies for 2009, 2010 and 2011.

Solution. Add the translation adjustment for the year to reported net income.

Exxon Mobil. 2011: 42,206 + (−867) = 41,339. 2010: 31,398 + 1,034 = 32,432. 2009: 19,658 + 3,629 = 23,287.
Chevron. 2011: 27,008 + 17 = 27,025. 2010: 19,136 + 6 = 19,142. 2009: 10,563 + 60 = 10,623.

2. How does the relative size of the two companies change once the adjustment is made?

Solution. On reported figures, Exxon Mobil earned 42,206 ÷ 27,008 = 1.6 times Chevron in 2011, 31,398 ÷ 19,136 = 1.6 times in 2010, and 19,658 ÷ 10,563 = 1.9 times in 2009.

On adjusted figures the ratios become 41,339 ÷ 27,025 = 1.5 in 2011, 32,432 ÷ 19,142 = 1.7 in 2010, and 23,287 ÷ 10,623 = 2.2 in 2009. The 2009 gap widens from 1.9 to 2.2 times and the 2011 gap narrows from 1.6 to 1.5 times.

The direction of the translation adjustment agrees between the two companies in 2009 and 2010 but not in 2011. Exxon Mobil consistently has much larger adjustments, which is exactly what you would expect from a company that designates the local currency as functional for a substantially larger share of its operations.

3. Does the adjustment fully solve the comparability problem?

Solution. No. It puts the equity-deferred amounts and the income-recognised amounts on a common basis, but it does not undo the deeper effect: the two methods translate cost of goods sold, depreciation and non-monetary assets at different rates, so reported operating results were never comparable in the first place. Adjusted net income is a better comparison than reported net income, not a clean one.

The broader idea behind that adjustment is clean-surplus accounting, in which all non-owner changes in stockholders equity are included in the determination of income. The opposite, dirty-surplus accounting, routes some income items straight to equity, and the translation adjustment arising when a foreign currency is the functional currency is one of the classic dirty-surplus items in both IFRS and US GAAP. The disclosures required by both frameworks give an analyst enough detail to rebuild income on a clean-surplus basis, and both now require a statement of comprehensive income in which these deferred amounts are shown.

Two further consequences of operating across borders show up in the statements and are regularly examined. The first is the effect on the effective tax rate. The second is the effect on the quality of reported sales growth.

Multinational operations and the effective tax rate

In general a multinational incurs income tax in the country where the profit is earned. Transfer prices, meaning the prices related companies charge each other on intercompany transactions, decide how profit is allocated between those companies, so a group operating across jurisdictions with different tax rates has an incentive to set transfer prices that push profit toward lower-rate jurisdictions. Countries counter this with transfer pricing regimes, described as the body of law and practice by which a country makes sure that goods, services and intellectual property moving between related companies carry market-based prices, so that profit is reported correctly in each jurisdiction. Most countries are also party to tax treaties that prevent the same corporate profit from being taxed twice, by granting a credit for tax paid elsewhere.

Whether tax is also owed at home depends on the domestic regime. In the United States, for example, a multinational has historically owed only a residual amount on foreign income once a credit for foreign tax paid on that income is applied, so tax falls due at home only to the extent the US corporate rate sits above the foreign rate. Much foreign income earned by US multinationals was also not taxed until repatriated.

The disclosure an analyst uses is the reconciliation between the effective tax rate and the statutory rate, which standards require so that users can judge whether the current relationship between tax expense and accounting profit is unusual and what could change it.

The reconciliation bridges this rate to the statutory rate line by line, and the foreign jurisdiction line isolates the multinational effect.

Example 12 · Worked

Below are extracts from the effective tax rate reconciliations of Heineken NV, a Dutch brewer, and Colgate-Palmolive, a US consumer products company.

Heineken NV 2011 reconciliation of the effective tax rate, EUR million

Item	2011 percent	2011 amount	2010 percent	2010 amount
Tax at the domestic rate of the company	25.0	446	25.5	456
Foreign jurisdiction rate effect	3.5	62	1.9	34
Non-deductible expense effect	3.2	58	4.0	72
Tax incentives and exempt income	(6.0)	−107	(8.2)	−146
Temporary differences recognised for the first time	(0.5)	−9	(0.1)	−2
Earlier tax losses now used or recognised	(0.3)	−5	(1.2)	−21
Current year tax losses left unrecognised	1.0	18	0.8	15
Tax rate changes	0.1	1	0.2	3
Withholding tax	1.5	26	1.4	25
Prior year over and under provisions	(1.5)	−27	(2.3)	−42
Other items	0.1	2	0.5	9
Total	26.1	465	22.5	403

Profit before income tax was EUR2,025 million in 2011 and EUR1,982 million in 2010. After removing the share of net profit of associates and joint ventures of EUR240 million and EUR193 million respectively, the base for the reconciliation is EUR1,785 million in 2011 and EUR1,789 million in 2010.

Colgate-Palmolive reconciliation, percentage of income before income taxes

Item	2011	2010	2009
Tax at the US statutory rate	35.0%	35.0%	35.0%
State income tax after federal benefit	0.4	1.1	0.5
Earnings taxed at rates other than the US statutory rate	(1.7)	(4.6)	(2.5)
Transition charge on Venezuelan hyperinflation	Nil	2.8	Nil
Other items, net	(1.1)	(1.7)	(0.8)
Effective rate	32.6%	32.6%	32.2%

1. Which company has the lower home country statutory rate?

Solution. Heineken, at 25.0 percent in 2011, against 35.0 percent for Colgate-Palmolive.

2. What was the effect of multinational operations on each 2011 effective tax rate?

Solution. For Heineken the foreign jurisdiction rate line shows that multinational operations raised the effective rate by 3.5 percentage points, worth EUR62 million. For Colgate-Palmolive the line for earnings taxed at rates other than the US statutory rate shows that multinational operations lowered the effective rate by 1.7 percentage points. The sign depends entirely on whether the home rate is high or low relative to the jurisdictions where profit is earned.

As a check on the Heineken figures, the domestic rate line of EUR446 million is 25.0 percent of the EUR1,785 million base, and the total tax of EUR465 million divided by that same base gives 26.1 percent, which is the disclosed effective rate.

3. What do the Heineken figures suggest about the geographic mix of profit in 2011?

Solution. The foreign jurisdiction effect rose from 1.9 percentage points in 2010 to 3.5 percentage points in 2011. That is consistent with the profit mix shifting toward countries with higher marginal tax rates. It is not the only explanation: marginal rates in the countries where Heineken already earns profit could simply have risen. A change in this line is a prompt to check the segment disclosures, not a conclusion on its own.

Currency and the quality of sales growth

For a multinational, reported sales growth has three drivers, not two.

Only the first two are within management control, and only the first two are plausibly repeatable.

Growth from volume or price is more sustainable than growth from a currency move, and management has far more influence over it. An analyst therefore separates the components both to forecast forward and to judge a management team backwards. General Mills disclosed the split for its international segment in its 2011 annual report.

General Mills components of international net sales growth

Component	Fiscal 2011 versus 2010	Fiscal 2010 versus 2009
Volume contribution	6 pts	Flat
Price realization and mix, net	1 pt	3 pts
Currency translation	Flat	1 pt
Total reported sales growth	7 pts	4 pts

General Mills international sales growth by region, fiscal 2011

Region	Reported change in net sales	Currency contribution	Change at constant currency
Europe	5%	−2%	7%
Canada	8	5	3
Asia and Pacific	14	5	9
Latin America	−5	−16	11
Total international segment	7%	Flat	7%

The segment total conceals the regional detail entirely. Latin America reported a 5 percent decline while growing 11 percent in constant currency, and Canada reported 8 percent growth on 3 percent constant currency growth.

That contrast is the analytical point. An aggregate currency effect described as flat is the net of a positive 5 percentage points in Canada and the Asia and Pacific region against a negative 16 percentage points in Latin America. Two of the four regions had currency contributions larger than their underlying growth.

Example 13 · Worked

Procter and Gamble reconciles reported sales growth to organic sales growth in its investor materials. Organic sales growth is a non-GAAP measure that strips out acquisitions, divestitures and foreign exchange from year on year comparisons. Selected period averages from the 2012 CAGNY conference slides are below.

Procter and Gamble reported and organic sales growth, period averages

Period	Net sales growth	Foreign exchange impact	Acquisition and divestiture impact	Organic sales growth
Average, JAS 06 to JAS 08	11%	−4%	−2%	5%
Average, OND 08 to JAS 09	−7%	8%	0%	1%
Average, OND 09 to OND 11	5%	−1%	0%	4%

The impact columns are the adjustments added back to reported growth, so reported growth plus the two impact columns equals organic growth.

1. Why does the company present organic sales growth at all?

Solution. On its own account, because the measure gives investors a clearer view of underlying sales trends by putting growth on a consistent basis, and because organic sales is among the yardsticks applied to senior management and feeds into at-risk compensation. The second reason is worth noting: a non-GAAP measure tied to compensation deserves scrutiny of its definition, not just of its level.

2. On average, across the four quarters from October 2008 to September 2009, how did exchange rates affect reported sales growth?

Solution. The effect was strongly negative. Reported net sales growth averaged −7 percent while organic growth averaged +1 percent, and the reconciliation shows −7 + 8 + 0 = 1. Currency movement alone accounted for 8 percentage points of drag. Without any exchange rate effect, reported growth and organic growth would both have been 1 percent.

Compare that with the earlier average period, where reported growth of 11 percent contained a negative 4 point currency effect and a negative 2 point portfolio effect, leaving 5 percent organic. Reported growth swung by 18 percentage points between the two periods while organic growth swung by only 4.

Sensitivity disclosures

The most useful currency disclosures name the exposures and then quantify the profit effect of a move. The 2011 BMW AG annual report does both. Its management report explains that sales outside the euro zone create exchange risk, that roughly two-thirds of group currency exposure in 2011 sat in only three pairs, against the Chinese renminbi, the US dollar and the British pound, and that the group uses cash-flow-at-risk models and scenario analysis to measure it. Risk is managed strategically through natural hedging, meaning more purchasing denominated in foreign currency and more local production, with the Spartanburg plant in the United States and the new Tiexi plant at the Shenyang site in China cited as examples, and operationally through financial market hedges with counterparties of good credit standing.

BMW Group principal currency exposures and modelled downside, EUR million

Currency pair	Exposure 31.12.2011	Exposure 31.12.2010	Potential negative impact 31.12.2011	Potential negative impact 31.12.2010
Euro and Chinese renminbi	7,114	6,256	180	265
Euro and US dollar	4,281	3,888	121	103
Euro and British pound	3,266	3,056	182	184
Euro and Japanese yen	1,334	1,086	23	30

The downside figures are computed at 95 percent confidence over a horizon of as much as one year per currency, using current market prices and exposures. Aggregating them produces a risk reduction effect because the portfolios are correlated.

Read the two halves of that table together rather than separately. The renminbi exposure is the largest at EUR7,114 million but carries a modelled downside of EUR180 million, while the sterling exposure is less than half the size at EUR3,266 million yet carries a slightly larger downside of EUR182 million. Exposure size alone does not rank risk; volatility, correlation and the hedges already in place do. The renminbi downside also fell from EUR265 million to EUR180 million while the exposure grew, which is the signature of increased hedging or lower modelled volatility rather than reduced business.

Companies differ widely in how much of this they publish. Where the disclosure is detailed, an analyst can combine it with an independent exchange rate forecast and build the currency effect explicitly into projected profit and cash flow. Where it is not, the sensitivity figures still serve a purpose: they calibrate how far the downside case should sit below the base case.

Check yourself

1. A Brazilian subsidiary of a Japanese parent keeps its books in reais, sells almost entirely into the domestic Brazilian market at prices set by Brazilian competition, and finances itself with local bank debt. Which translation method applies, and where does the adjustment go?

The functional currency indicators all point to the real: it drives sales prices, the competitive environment determines those prices, and financing is generated locally. The functional currency is therefore a foreign currency from the parent perspective, so the current rate method applies. All assets and liabilities translate at the current rate, equity other than retained earnings at historical rates, and revenues and expenses at the average rate. The balancing amount accumulates as a separate component of equity and only reaches net income if the subsidiary is sold.

2. A euro-functional exporter has a US dollar receivable of USD500,000 booked at EUR0.90 per dollar. At the balance sheet date the rate is EUR0.86 and at settlement it is EUR0.92. State the amounts recognised in each period and the realized total.

Period 1: 500,000 × (0.86 − 0.90) = a loss of EUR20,000, recognised in income while unrealized. Period 2: 500,000 × (0.92 − 0.86) = a gain of EUR30,000. The realized total is 500,000 × (0.92 − 0.90) = a gain of EUR10,000, which equals −20,000 + 30,000. The interim figure had the wrong sign, which is the general warning about reading one period of currency effects as a forecast.

3. Under the temporal method a subsidiary reports income before the translation item of EUR990,000 and a translation loss of EUR245,000. Under the current rate method the same subsidiary reports net income of EUR885,000. Which method shows the more profitable operations, and why is the question ambiguous?

On net income the current rate method looks better, at EUR885,000 against EUR745,000, a difference of 18.8 percent. On income before the translation item the temporal method looks better, at EUR990,000 against EUR885,000, so the current rate figure is 10.6 percent lower. The reversal is entirely a reporting artefact: the current rate method parks its adjustment in equity while the temporal method forces its adjustment through income. Compare the pre-translation lines if you want to see operations, and remember that even those differ because cost of goods sold and depreciation carry different rates.

4. Why is receivables turnover the only ratio in the Canadaco table that is identical under both translation methods, and what general rule does that illustrate?

Sales are translated at the average rate under both methods and receivables at the current rate under both methods, so the ratio 9,000,000 over 720,000 = 12.50 appears in both columns. The general rule is that a ratio is unaffected by the choice of method whenever every component of the numerator and the denominator takes the same type of rate under both. Inventory turnover fails this test between the two euro columns, at 7.03 against 7.50, because inventory and cost of goods sold move from the average and current rates to the 0.74 historical rate.

5. A subsidiary translated under the temporal method holds monetary assets of C$1,880,000 and monetary liabilities of C$3,450,000. The foreign currency weakens over the year. What happens to net income relative to a stable currency, and what would management have to do to remove the exposure?

Exposed liabilities exceed exposed assets, so there is a net monetary liability exposure. A weakening foreign currency shrinks the parent currency value of those liabilities more than that of the exposed assets, producing a translation gain that raises net income relative to the stable case, even though revenues and assets both translate lower. To remove the exposure, management would have to bring monetary assets into line with monetary liabilities. Under the current rate method the equivalent would require total assets to equal total liabilities, meaning zero equity, which is why exposure cannot practically be eliminated there.

6. A subsidiary in a highly inflationary economy holds land at historical cost. Under which of the three possible approaches does the land shrink toward nothing in the consolidated accounts, and why do the standard setters reject that approach?

Translating unrestated historical cost at the current rate, which is the plain current rate method. In the Turkish illustration, US$1,000 of land fell to US$809, then US$368, then US$325 across three years, a cumulative loss of US$675, while the land itself was unchanged and its local currency value was rising with inflation. The approach captures the currency move but ignores the offsetting local price change, so both IFRS and US GAAP block it: US GAAP by requiring the temporal method, which holds the land at US$1,000, and IFRS by requiring restatement for inflation before translation, which produced US$1,100.

7. Two companies in the same industry report net income of 1,000 and 900. The first recognised a translation loss of 60 in income; the second deferred a positive translation adjustment of 80 in equity. How would you make the two comparable, and what does the adjustment still fail to fix?

Put both on a clean-surplus basis by adding the change in the cumulative translation adjustment to reported income: the first stays at 1,000, since its currency effect is already inside income, and the second becomes 980. The gap narrows from 100 to 20. What remains unfixed is that the temporal method translates cost of goods sold, depreciation and non-monetary assets at historical rates while the current rate method uses average and current rates, so the underlying revenue and expense lines were never on the same basis. The adjustment improves comparability, it does not deliver it.

8. A company reports international net sales growth of 7 percent with a currency effect described as flat, and a segment reporting a 5 percent sales decline. Why is the aggregate figure potentially misleading, and what should you look at instead?

A flat aggregate currency effect is a net of offsetting regional effects, not an absence of them. In the General Mills disclosure, the flat total nets a positive 5 percentage points in Canada and in the Asia and Pacific region against a negative 16 percentage points in Latin America, and Latin America grew 11 percent in constant currency while reporting a 5 percent decline. Work from the constant currency figures region by region, because the aggregate can hide a business that is growing strongly behind a currency that is falling fast, and the reverse.

End of lesson.

Back to 04 Financial Statement Analysis

FSA 3 – Multinational Operations

Overview

Presentation currency, functional currency and local currency