CASE 15-3 IBM Analysis of Exchange Rate Effects: Multiple Currencies

Transcription

1 CASE 15-3 IBM Analysis of Exchange Rate Effects: Multiple Currencies INTRODUCTION CASE OBJECTIVES IBM is one of the world s largest multinational corporations, and changes in currency rates have pervasive effects on the firm s financial statements. As IBM provided supplementary data regarding its foreign operations for many years, we can use the company to illustrate the analysis of multinational corporations. The objectives of this case are to use IBM to: IBM DISCLOSURES RELATED TO FOREIGN OPERATIONS ESTIMATION OF COMPOSITE EXCHANGE RATES 1. Show the effects of exchange rate changes in levels and trends of revenue, income, cash flow, and financial position. 2. Show the effect of exchange rate changes on financial ratios. 3. Calculate the effect of exchange rate changes on assets and liabilities. 4. Calculate translation gains and losses resulting from exchange rate changes. Exhibit 15C3-1 contains IBM s balance sheet at December 31, 1989, and Within the stockholders equity section, we see translation adjustments of $1,698 and $3,266 billion (4.4 and 7.6% of net assets), respectively. These entries tell us that the company has significant non- U.S. operations and its uses foreign functional currencies. If the company used the U.S. dollar as the functional currency for all foreign operations, all gains and losses would have been included in income. Exhibit 15C3-2 contains IBM s consolidated statement of cash flows for the three years ended December 31, The only reference to translation in the cash flow statement is the effect of exchange rate changes on cash and cash equivalents near the bottom. Exhibit 15C3-3, which provides the primary raw material for our analysis, is supplementary data on IBM s non-u.s. operations. Although much of this disclosure is not required (and, unfortunately, rarely provided), it enables us to obtain an understanding of the effect of changing exchange rates on the company s financial condition and operating performance. The first part of Exhibit 15C3-3 contains summarized balance sheets and income statements for IBM s non-u.s. operations. These data suggest steady growth in foreign revenue, net earnings, and net assets over the period 1988 to Comparison of these data with IBM s consolidated balance sheet and income statement indicates that foreign operations accounted for 60% of revenue for We return to the analysis of these data shortly. Before starting our analysis, we need data on the exchange rates that affect IBM s financial statements. For a company operating in a single currency (such as Foreign Subsidiary, in the chapter, or AFLAC in Case 15C1-1), we can obtain year-end and average exchange rates covering the period being analyzed. For a multinational such as IBM, we need data on many currencies and a breakdown of IBM s operations by functional currency. The latter is unavailable (to an external user), and the analysis of many currencies is very time-consuming. 1 We need a shortcut. 1 In some cases, annual reports for foreign subsidiaries of multinational companies are available, either because of local filing requirements or subsidiary financing. These reports can shed light on significant foreign operations. However, these reports are generally prepared in local currencies according to local accounting standards, not in U.S. dollars under U.S. GAAP. In some cases, reports are available only in the local language, further hampering use. Nonetheless, when a company has one or a few highly significant foreign subsidiaries, the subsidiary annual report may provide insights not available from the parent s consolidated financial statements. W108

2 ESTIMATION OF COMPOSITE EXCHANGE RATES W109 EXHIBIT 15C3-1. IBM Balance Sheet At December 31: (Dollars in millions) Assets Current Assets Cash $ 1,189 $ 741 Cash equivalents 2,664 2,959 Marketable securities, at cost, which approximates market 698 1,261 Notes and accounts receivable trade, net of allowances 20,988 18,866 Other accounts receivable 1,656 1,298 Inventories 10,108 9,463 Prepaid expenses and other current assets 1,617 1,287 38,920 35,875 Plant, Rental Machines, and Other Property 53,659 48,410 Less: Accumulated depreciation 26,418 23,467 27,241 24,943 Investments and Other Assets: Software, less accumulated amortization (1990, $5,873; 1989, $4,824) 4,099 3,293 Investments and sundry assets 17,308 13,623 21,407 16,916 $87,568 $77,734 Liabilities and Stockholders Equity Current Liabilities: Taxes $ 3,159 $ 2,699 Short-term debt 7,602 5,892 Accounts payable 3,367 3,167 Compensation and benefits 3,014 2,797 Deferred income 2,506 1,365 Other accrued expenses and liabilities 5,628 5,780 25,276 21,700 Long-Term Debt 11,943 10,825 Other Liabilities 3,656 3,420 Deferred Income Taxes 3,861 3,280 Stockholders Equity: Capital stock, par value $1.25 per share 6,357 6,341 Shares authorized: 750,000,000 Issues: ,618,795; ,775,560 Retained earnings 33,234 30,477 Translation adjustments 3,266 1,698 42,857 38,516 Less: Treasury stock, at cost (Shares: ,604; ,723) 25 7 Source: IBM Corporation, 1990 Annual Report. 42,832 38,509 $87,568 $77,734

3 W110 CASE 15-3 IBM ANALYSIS OF EXCHANGE RATE EFFECTS: MULTIPLE CURRENCIES EXHIBIT 15C3-2. IBM Statement of Cash Flows For the Year Ended December 31: (Dollars in millions) Cash Flow from Operating Activities: Net earnings $ 6,020 $ 3,758 $ 5,806 Adjustments to reconcile net earnings to cash provided from operating activities: Depreciation 4,217 4,240 3,871 Amortization of software 1,086 1, Loss (gain) on disposition of investment assets 32 (74) (133) (Increase) in accounts receivable (2,077) (2,647) (2,322) Decrease (increase) in inventory 17 (29) (1,232) (Increase) in other assets (3,136) (1,674) (1,587) Increase in accounts payable Increase in other liabilities 1,020 1, Net cash provided from operating activities 7,472 7,372 6,080 Cash Flow from Investing Activities: Payments for plant, rental machines, and other property (6,509) (6,414) (5,390) Proceeds from disposition of plant, rental machines, and other property Investment in software (1,892) (1,679) (1,318) Purchases of marketable securities and other investments (1,234) (1,391) (2,555) Proceeds from marketable securities and other investments 1,687 1,860 4,734 Net cash used in investing activities (7,144) (7,080) (4,120) Cash Flow from Financing Activities: Proceeds from new debt 4,676 6,471 4,540 Payments to settle debt (3,683) (2,768) (3,007) Short-term borrowings less than 90 days net 1, ,028 Payments to employee stock plans net (76) (29) (11) Payments to purchase and retire capital stock (415) (1,759) (992) Cash dividends paid (2,774) (2,752) (2,609) Net cash used in financing activities (306) (609) (1,051) Effects of Exchange Rate Changes on Cash and Cash Equivalents 131 (158) (201) Net Change in Cash and Cash Equivalents 153 (475) 708 Cash and Cash Equivalents at January 1 3,700 4,175 3,467 Cash and Cash Equivalents at December 31 $ 3,853 $ 3,700 $ 4,175 Supplemental Data: Cash paid during the year for: Income taxes $ 3,315 $ 3,071 $ 3,405 Interest $ 2,165 $ 1,605 $ 1,440 Source: IBM Corporation, 1990 Annual Report. Fortunately, there are indices of the value of the U.S. dollar against a basket of foreign currencies, normally computed on a trade-weighted basis. Using such a series for IBM requires us to make the assumption that IBM s business has the same currency mix (distribution over various currencies) as U.S. trade flows. Although that assumption might be untenable for a smaller company with more limited foreign operations, it appears reasonable for a giant multinational such as IBM. Exhibit 15C3-4 shows average and year-end exchange rates for the period covered by our analysis.

4 BALANCE SHEET EFFECTS W111 EXHIBIT 15C3-3. IBM Data on Non-U.S. Operations Non-U.S. Operations (Dollars in millions) At end of year: Net assets employed: Current assets $24,337 $20,361 $20,005 Current liabilities 15,917 12,124 11,481 Working capital 8,420 8,237 8,524 Plant, rental machines, and other property, net 11,628 9,879 9,354 Investments and other assets 9,077 6,822 5,251 29,125 24,938 23,129 Long-term debt 5,060 3,358 2,340 Other liabilities 2,699 2,607 2,505 Deferred income taxes 2,381 1,814 1,580 10,140 7,779 6,425 Net assets employed $18,985 $17,159 $16,704 Number of employees 168, , ,904 For the year: Revenue $41,886 $36,965 $34,361 Earnings before income taxes $ 7,844 $ 7,496 $ 7,088 Provision for income taxes 3,270 3,388 3,009 Net earnings $ 4,574 $ 4,108 $ 4,079 Investment in plant, rental machines, and other property $ 3,020 $ 2,514 $ 2, net earnings before cumulative effect of accounting change for income taxes. Non-U.S. subsidiaries which operate in a local currency environment account for approximately 90% of the company s non-u.s. revenue. The remaining 10% of the company s non-u.s. revenue is from subsidiaries and branches which operate in U.S. dollars or whose economic environment is highly inflationary. As the value of the dollar weakens, net assets recorded in local currencies translate into more U.S. dollars than they would have at the previous year s rates. Conversely, as the dollar becomes stronger, net assets recorded in local currencies translate into fewer U.S. dollars than they would have at the previous year s rates. The translation adjustments, resulting from the translation of net assets, amounted to $3,266 million at December 31, 1990, $1,698 million at December 31, 1989, and $1,917 million at December 31, The changes in translation adjustments since the end of 1988 are a reflection of the strengthening of the dollar in 1989 and the weakening of the dollar in Source: IBM Corporation, 1990 Annual Report. BALANCE SHEET EFFECTS Exhibit 15C3-3 states that IBM had non-u.s. net assets of approximately $19 billion. What functional currencies did the company use to account for its foreign operations? The exhibit reports that non-u.s. subsidiaries which operate in a local currency environment account for approximately 90% of the company s non-u.s. revenue. The remaining 10%... is from subsidiaries and branches which operate in U.S. dollars or whose economic environment is highly inflationary.

5 W112 CASE 15-3 IBM ANALYSIS OF EXCHANGE RATE EFFECTS: MULTIPLE CURRENCIES EXHIBIT 15C3-4 Dollar s Trade-Weighted Exchange Index, 1988 to 1990 ( ) December 31 Index Year Average Rates for Year, 1980 to 1990 Index Sources: Economic Report of the President, February 1991 (annual data) and Federal Reserve Bank of St. Louis (December 31 data). In other words, the local currency is the functional currency for 90% of IBM s foreign operations. The U.S. dollar is the functional currency for the remainder, including subsidiaries operating in hyperinflationary economies. Assuming that the 90% figure applies equally to the balance sheet, we conclude that IBM had net assets in nondollar functional currencies of $ billion (90% of total nondollar net assets of $ billion) at December 31, The corresponding figures for year-end 1989 and 1988 were $ billion (0.90 $ billion) and $ billion (0.90 $ billion), respectively. These amounts represent IBM s exposure to changes in exchange rates under SFAS 52. Translation gains and losses resulting from exchange rate fluctuation have been accumulated as a component of stockholders equity, in accordance with SFAS 52. The text of Exhibit 15C3-3 gives us the cumulative translation adjustments at each year-end: December 31 Cumulative Translation Adjustments 1988 $1.917 billion These calculations enable us to compute the actual increase in IBM s foreign net assets in functional currencies. By taking the reported change and subtracting the effects of translation (change in accumulated adjustment), we get the real change ($ in millions): Year Reported Translation Real 1989 $ 455 $ (219) $ ,826 1, From the reported change, it appears that IBM s foreign net assets increased more rapidly in 1990 than The reality is that the large 1990 increase was mostly due to the appreciation of foreign currencies against the dollar; before translation (in real terms), the 1989 increase was larger.

6 BALANCE SHEET EFFECTS W113 The year-to-year change in the cumulative translation adjustment account is the effect of translation for each year. Compare those changes with IBM s exposure: 1989: 1990: $1.698 billion $1.917 billion 1.46% $ billion $3.266 billion $1.698 billion 10.15% $ billion These calculations reveal that the IBM-weighted functional currency composite declined by 1.46% against the dollar in 1989 and rose by 10.15% against the dollar in Turning to our trade-weighted index in Exhibit 15C3-4, we see that the percentage changes are 1989: 1.0% 1990: 11.9% These changes approximate the IBM-weighted changes, reassuring us that our index is a good proxy. But when possible, we use the IBM-weighted index that we have now derived. First, consider the company s inventories. Exhibit 15C3-3 does not break out non-u.s. inventory, so we must assume that inventories are a constant percentage of current assets. 3 At December 31, 1989, consolidated inventories were 26.4% of consolidated current assets (Exhibit 15C3-1). We assume that non-u.s. inventories also were 26.4% of non-u.s. current assets of $ billion or $5.375 billion, of which $4.838 billion (90%) were in nondollar functional currencies. Applying the IBM-weighted exchange rate change of 10.15% results in an estimated increase in non-u.s. inventories of $491 million due to changing exchange rates. This accounts for most of the $645 million ($ billion $9.463 billion) increase in IBM s consolidated inventories during 1990 (data from Exhibit 15C3-1). These calculations suggest that most of the 1990 inventory increase was due to the impact of changing exchange rates rather than to operating changes. We can confirm this result from the company s cash flow statement. In Exhibit 15C3-2, we find that IBM s inventory change, excluding the effect of translation, was a decrease of $17 million, suggesting that the true effect of exchange rate changes was $662 million [$645 million actual change less ( $17 million) real change]. 4 Although our estimated effect of $491 million is not equal to the true effect of $662 million for 1990, they are not unreasonably far apart. Clearly, our assumptions did not precisely hold. But even if we did not have the true figure, our estimate would still have told us that IBM s inventory increase in 1990 was mostly due to currency effects rather than operating causes. It is this conclusion that makes the analysis worthwhile. This technique, although superfluous when the cash flow statement excludes the impact of exchange rate changes, is useful when cash flow statements (such as those for non-u.s. firms) are not adjusted to exclude that impact. We can perform this same analysis for IBM s fixed assets. Exhibit 15C3-3 shows that non- U.S. fixed assets were $9.879 billion; we estimate that $8.891 billion (90% of $9.879 billion) was in nondollar functional currencies. The estimated effect of currency changes is $902 million (10.15% of $8.891 billion). The actual impact of currency changes on fixed assets was disclosed in IBM s 10-K report in Schedules V and VI. These reconciliations of fixed assets (gross) and accumulated depreciation reveal that translation increased fixed assets by $963 million ($2,143 million for gross fixed assets less $1,180 million for accumulated depreciation). Again, our estimate is approximately correct, despite the assumptions required. Consolidated net fixed assets rose by $2.298 billion in 1990 (Exhibit 15C3-1), or 9.2%. Nearly half the gain resulted from exchange rate changes rather than new investment. Even if the 10-K data had not been available (Schedules V and VI are no longer required), we would have the same knowledge. 2 Perceptive readers will note that we have omitted the effect of changing exchange rates on the increase in IBM s net assets in functional currencies. Given the small change in those assets (in functional currency terms) over the period 1988 to 1990, we have opted for simplification. 3 IBM uses the FIFO inventory method worldwide. For companies with significant LIFO inventories, this calculation should be made on a FIFO basis by adding back the LIFO reserve (see Chapter 6). 4 This computation, and similar computations in this case, are possible only because IBM made no purchase method acquisition during Chapter 14 discusses the impact of purchase method acquisitions on the statement of cash flows.

7 W114 CASE 15-3 IBM ANALYSIS OF EXCHANGE RATE EFFECTS: MULTIPLE CURRENCIES INCOME STATEMENT EFFECTS Turning to the income data (Exhibit 15C3-3), we note that IBM had revenues of $ billion in currencies other than the dollar, an increase of 13.3% from the 1989 level of $ billion. On the surface, it appears that the 1990 gain in foreign sales was much larger than the 1989 increase (up 7.6% from the 1988 level of $ billion). However, analysis reveals that exchange rate effects distort the data. In 1990, non-u.s. sales of $ billion (90% of $41.886) were in operations with nondollar functional currencies (FC) (with the remainder in operations with nondollar local currencies but the dollar as functional currency). These revenues (and all expenses) were translated into dollars at the average rate for Using the data in Exhibits 15C3-3 and 15C3-4, we can compute the effect of rate changes for each year: ($ in millions) Non-U.S. revenues ($, Exhibit 15C3-3) $34,361 $36,965 $41,886 Non-$ FC revenues ($, 90%) 30,925 33,268 37,697 % Increase +7.6% +13.3% Dollar index (Exhibit 15C3-4) FC revenues FC 28,667 FC 32,802 FC 33,588 % Increase +14.4% +2.4% The last entry, FC revenues, is an artificial index, derived by multiplying estimated non-$ FC revenues by the dollar index. 5 The result is a measure of revenue from which the impact of changes in the value of the dollar has been removed. As a result, we can estimate the real change in foreign revenues. We find that the decline in the value of the dollar accounted for most of the gain in foreign revenues in 1990; the increase is only 2.4% when that factor is removed. Conversely (since the dollar rose in value in 1989), the real (FC) gain is 14.4% as compared with a gain of 7.6% in dollars. The rise in the dollar in 1989 resulted in a smaller percentage sales gain in dollars than local currencies. (These calculations assume that local currency prices were unaffected by exchange rate changes.) This exercise, therefore, approximates the impact of changing exchange rates on IBM s nondollar revenues. A similar calculation approximates the effect on net income. IBM s annual report to shareholders provides virtually no disclosure of this impact. Exhibit 15C3-5 contains the result of this analysis for the 11-year period 1980 to Comparison of the reported data with the adjusted data reveals differences that are quite significant. The year-to-year percentage changes in both revenues and pretax income are, in most years, quite different after adjustment for changes in the value of the dollar. We have already 5 We must use the index because we do not have average IBM weights, only year-end to year-end data. As we have shown that the index tracked the IBM weights well, we can use it to examine the trend of revenues and pretax income. 6 The analysis in Exhibit 15C3-5 uses total non-u.s. sales rather than the proportion for which IBM uses nondollar functional currencies. This proportion has declined over the 1980 to 1990 period, but the disclosure on this point is vague. For simplicity and because we believe the analysis would not be significantly affected, we omit that step in our analysis. In principle, it is preferable to use only sales in nondollar functional currencies, as in the 1988 to 1990 computations above. Although other foreign sales are also affected by exchange rate changes, there is an important difference. Foreign sales for which the dollar is the functional currency are likely to be in hyperinflationary countries or where local selling prices are the local currency equivalent of dollar prices. In these cases, changes in exchange rates may affect volume but do not affect dollar prices; they do not create income statement distortion as discussed in this section. In addition, the index derived from changes in the cumulative translation adjustment is not applicable to these situations. In practice, however, the proportion of sales for which the dollar is the functional currency is rarely available and, therefore, the analyst must use total foreign sales for analytic purposes.

8 INCOME STATEMENT EFFECTS W115 EXHIBIT 15C3-5 Analysis of IBM s Foreign Operations, 1980 to 1990 Year Revenues % Change Pretax Income % Change Reported Data ($U.S. in millions) 1980 $U.S. 13,787 $U.S. 2, , % 2, % , , , , , , , , , , , , , , , , , , Adjusted Data (FC Units in Millions) 1980 FC 12,050 FC 2, , % 2, % , , , , , , , , , , , , , , , , , , discussed the impact on the period 1988 to For a broader perspective, we have summarized the data for the entire period: Percentage Changes in IBM Foreign Results, 1980 to 1990 Revenues Pretax Income Period Reported Adjusted Reported Adjusted % 155.7% 100.1% 227.3% Source: Data in Exhibit 15C3-5. Over the entire ten years, the reported and adjusted trends are quite similar. As the dollar showed a very small increase in value over the period, we conclude that local currency revenue growth was only slightly greater than revenue growth reported in dollars. But for the two subperiods, the adjusted data tell a completely different story from the reported data. During the period 1980 to 1985, the value of the dollar rose sharply; the data in Exhibit 15C3-4 show that the average value of the dollar in 1985 was 63.6% higher in 1985 than 1980 (143.0/ ). Thus, revenues and earnings in foreign currencies were continuously devalued when translated into dollars. The growth in revenues during this period was 155.7% in local currencies, but only 56.3% after translation into dollars. Pretax income was similarly devalued; the local currency growth was 227.3%, whereas the dollar growth was only 100.1%. The individual year-to-year changes also reflect the impact of the strengthening dollar. In 1981, for example, reported pretax income declined by 3.9%; after adjustment, there was a gain

9 W116 CASE 15-3 IBM ANALYSIS OF EXCHANGE RATE EFFECTS: MULTIPLE CURRENCIES of 13.7%. In every year during the period 1980 to 1985, the performance of IBM s foreign operations was better in local currencies than U.S. dollars. During the second half of the decade, 1985 to 1990, the impact of exchange rates reversed. The value of the dollar declined in most years, and by 1990 it had returned to a level very close to The declining value of the dollar inflated foreign currency revenues and income when translated into dollars. Over the 1985 to 1990 period, IBM s foreign revenues (in dollars) increased by 94.4%, higher growth than in the 1980 to 1985 period. The adjusted data suggest that the reverse was true; IBM s local currency revenues grew by only 21.1% over the second half of the decade, a marked slowing from the 155.7% growth during the first half. Although the decline of the dollar was not consistent, some of the individual year data echo this conclusion. In both 1986 and 1987, foreign revenues (in dollars) rose sharply, suggesting favorable performance trends. The adjusted data show that, for both years, foreign currency revenues declined. The pretax income data also appear significantly different after adjustment for changes in the value of the dollar. Over the period 1985 to 1990, foreign pretax earnings rose by 41.4% in dollars, but declined by 11.9% in local currencies. The years 1986 and 1990 are the clearest examples of this effect in individual years: In both cases, pretax income rose in dollars but declined in local currencies. It is important to caution, however, that this analysis makes a crucial assumption that IBM s foreign operations were unaffected by exchange rate changes. For some firms, selling prices (and, therefore, revenues and earnings) are affected by variations in exchange rates, which impact the cost of imported components, and the prices of competitive products. We cannot assume that local currency results are always independent of exchange rates. Nonetheless, it is apparent that the rising value of the dollar during the 1980 to 1985 period disguised the excellent performance of IBM s foreign operations. It is equally clear that the dollar decline during the second half of the decade masked the deterioration of the operating performance of the company s foreign subsidiaries. These conclusions show that analysis of a multinational enterprise is seriously deficient unless the impact of changing exchange rates is taken into account. Despite the approximations and assumptions required, the analyst gains important insights into operating trends and can use these to question management more perceptively about its real operating performance. RATIO EFFECTS The impact of foreign currency changes on IBM s financial ratios is hard to determine because of inadequate data. Since IBM uses functional currencies other than the U.S. dollar for 90% of its non-u.s. operations, we can conclude that income statement ratios in dollars largely replicate the local currency data. This would also be true of ratios using only balance sheet data, such as the current or debt-to-equity ratios. The increased importance of foreign operations in 1990, resulting from the weakness of the dollar, gave foreign operations more weight in the consolidated total in 1990 than Without details of the income statement and balance sheet for foreign operations, we cannot easily tell which ratios are improved (or worsened) by this effect. 7 CONCLUDING COMMENTS As stated at the outset, the analysis of IBM was made possible by the voluntary disclosures (the first part of Exhibit 15C3-3) regarding its non-u.s. operations. Few companies provide similar data; IBM stopped providing extensive disclosures after its 1991 Annual Report. Why, then, have we devoted a case to this analysis? Our major objective is to illustrate how changing currency rates distort financial statements in the context of a real company. The analysis issues exist for all companies with significant foreign operations. Our goal is to enable analysts and other readers of this text to apply portions of this analysis of IBM to other companies. 7 By using cash flow data and the technique previously employed to estimate the effect of exchange rate changes on various balance sheet and income accounts, we can approximate ratios for IBM s foreign operations.

10 CONCLUDING COMMENTS W117 Required: Note: Make the simplifying assumption that IBM uses local currencies as the functional currency for all foreign subsidiaries. 1. Using Exhibit 15C3-3, the balance sheet and income statement for IBM s non-u.s. operations after translation to U.S. dollars: (a) Convert the 1989 and 1990 balance sheets to FC units. (b) Convert the 1990 income statement to FC units. (c) Using only FC net income, try to reconcile the change in FC equity (net assets) during Provide one possible reason for the discrepancy. 2. Exhibit 15C3-3 states that IBM invested $3,020 million in plant, rental machine, and other properties during Calculate the amount in FC units. Using this result, estimate depreciation expense (in FC units) for IBM s non-u.s. operations. 3. [Cash flow analysis of IBM foreign operations] (a) Assume that cash is 5% of the current assets shown in Exhibit 15C3-3. Prepare a 1990 cash flow statement in FC units for IBM s non-u.s. operations. (b) Convert the FC unit cash flow statement prepared in part (a) to a U.S. dollar cash flow statement. (c) (i) Compute the percentage of IBM s 1990 consolidated cash from operations that came from its non-u.s. operations. (ii) Compute the percentage of IBM s 1990 consolidated borrowings made by its non-u.s. operations. (iii) Compute the percentage of IBM s 1990 investment in fixed assets that took place in its non-u.s. operations. (iv) Discuss how your answers to parts (i) through (iii) contribute to your understanding of the importance of IBM s non-u.s. operations to the company. (v) Discuss the limitations of your answers to parts (i) through (iii). (d) Using the cash flow data calculated in part (c) estimate the effect of exchange rate changes on cash and cash equivalents. Compare your results to the amount shown in IBM s statement of cash flows (Exhibit 15C3-2).