Stock Analysis Report ALCOA Inc. Analyst: Sok Gun Song (song.173@osu.edu) March 5, 2004 Summary This report is to estimate intrinsic value of Alcoa. First of all, to understand what makes this business tick and to know how Alcoa gets in its present shape, I tried to analyze the industry using Porter's Model. Then, as an attempt to gauge the company's operational efficiency and ability to generate profits, company financial statements were analyzed. Based on industry and company analysis, forecast and valuation were made through ReOI model and Monte Carlo simulation. Valuation result is 22 dollars for normal case and 36 dollars for alumina shortage case. My recommendation based on the company's intrinsic value is WAIT until the middle of April because uncertainty related to alumina is too big right now. Table of Contents Page 1 Company Profile 3 2 Basic Market Data 3 3 Industry Analysis 4 4 Company Analysis 11 5 Forecast 16 6 Valuation 20 7 Recommendation 21 8 Accounting Quality 21 Appendices 1 ~ 17 1. Company Profile 2 Alcoa, since founded in 1888, has long been the world's leading producer of aluminum, fabricated aluminum and alumina. It controls more than 70 % of US primary production and produces about 17% of world primary aluminum production. Its 2003 revenue of 21,504 million is 58.5% of total combined revenue of other aluminum companies in North American continent including Alcan and secondary production companies. The company is fully integrated through the whole value chain in the industry. Operating 41 countries, it is active in bauxite mining, refining, smelting, fabrication and recycling with significant investment in Brazil, Australia and China. In 2002 US market and Europe market accounted for 63% and 21% of its revenue respectively. By industry sector, major markets are packaging and consumer (24.7%), automotive (13.3%), building and construction (10.8%) and aerospace (7.4%). By product segment, revenue from engineered products and flat-rolled products accounted for about 47% of total revenue. Figure 11 Sales Change by Product Sales (in million) 25000 20000 Alumina Other 15000 Packaging Engineered 10000 Flat Rolled 5000 Primary Metal 0 1998 1999 2000 2001 2002 2. Basic Market Data ▪ Share Price 37.39 (Feb.27,2004) ▪ Trailing PE 30.93 ▪ Shares Outstanding 868.49 Million ▪ Forward PE 14.90 ▪ Market Capitalization 32.47 Billion ▪ PEG Ratio (5yr exp) 1.59 ▪ 52 week high/low 39.44 / 18.45 ▪ Price to Sales 1.5 ▪ Beta 0.878 ▪ Price to Book 2.69 ▪ Average Volume (3 M) 5.35 Million ▪ % Held by insiders 1% ▪ Ticker AA (NYSE) ▪ % Held by Institution 80.21% 3. Industry Analysis 1 source: company annual reports 3 3-1 Supplier Power Supplier power in aluminum industry is potentially huge. A recent dramatic example of supplier power was witnessed when US aluminum industry's output dropped by 28.1% in 2001 because of a hike in electricity price. Electricity: Since electricity cost is the largest element of aluminum manufacturing production (above 20%), securing sources of low cost electricity is critical to aluminum producers. For example, Alcan has been known as the lowest cost producer because of its high level of electricity self sufficiency (about 62%), which is far higher than the industry average of 28%.2 Bauxite: Even though bauxite could be found almost everywhere, large scale commercial production is highly concentrated in a handful of mines in a few countries. In 2001, Australia, Jamaica, Brazil and Guinea produced 70% of total world production. Alumina: Alumina refineries are built close to bauxite mines because refineries have to be tailored to particular ore characteristics of the mine and bauxite transportation is very costly ($22/ton). For that reason, alumina production is geographically concentrated as well. In addition, alumina from each refinery is different in particle characteristics and smelters prefer to use alumina from a single refinery. Possible shortage of alumina in 2004 and 2005 causes a lot of confusion for future aluminum prices. It will turn out that this report's final recommendation is up to future alumina supply demand balance. To address potentially high supplier power, Alcoa is highly vertically integrated. Alcoa is the world largest raw material producer as well as the largest aluminum manufacturer, operating 2 stand alone bauxite mines, 9 refineries and 28 smelters and various sources of electricity worldwide. Alcoa's recent distress was caused by rising electricity cost in US. Alcoa had to shut down smelters in Oregon and Texas in 2002, reduced production in North Carolina in 2002 and decided to idle some part of its facilities in the Pacific Northwest and New York State in 2003. Alcoa's long term strategy should be to move smelter capacity to countries with low cost sources of electricity. Alcoa already dismantled Tacoma and Troutdale and decided to invest in a new smelter in Iceland and other hydroelectricity projects in Brazil. This transition process, in my opinion, will take more than a dozen years and will result in transforming the world aluminum industry as well as Alcoa and US aluminum market. Higher labor cost and increasing environmental cost will add impetus to this trend. 2 Alcan 2002 Annual Report 4 3-2 Customer Power Customer power is disproportionately strong in aluminum industry. Major customers for aluminum producers are transportation (30%), packaging and containers (17%) and construction industry, which consist of relatively fewer numbers of major players. Figure 23 US Shipment by End Users 2003 24.9% 16.2% Construction Transportation Consumer Durables Electrical Machinery Packaging 6.2% 6.7% 37.8% 8.2% On the other hand, aluminum is basically a commodity and production differentiation is difficult. more significant is that some customers might choose a substitute such as plastic and steel when aluminum prices are too high. Since customers tend to be gigantic manufacturers, they usually want to secure long term large quantity contracts. No aluminum producer can afford to lose those big customers because refineries and smelters can not even temporarily shut down without incurring huge cost. This lack of manufacturing flexibility and extreme amount of fixed cost often forces companies to operate under their cost curve. Alcoa appears to have competitive advantage in dealing with customer power. Its size and financial stability has made itself one of favorites of big customers who need long term contracts. With increased practice of long term contracts Alcoa started using derivatives such as futures and options for hedging purpose. Strategic motives behind Alcoa's recent acquisitions of Alumax, Reynolds, Cordant and many others can be seen from various angles. However, one possible interpretation is to see them as an attempt to balance customer power through consolidation in US aluminum industry. For example, after the company acquired Reynolds, which dominated aluminum packaging market, in 2000, it 3 The Aluminum Association 5 acquired Ivex Packaging Corp. in 2002, allowing it to enter plastic packaging market to, in my view, maximize its market power in packaging industry. 3-3 Substitute Aluminum competes with other metals for various end users. Competition exists in packaging (glass, paper, plastics and steel), construction (composites, steel and wood), transportation (magnesium, titanium and steel) and electrical application (copper). New application of aluminum or a competitor, relative price change among competing materials, technological breakthrough in usage of a material, and introduction of new material can make significant effect on competing dynamics among metals. Alcoa has traditionally been addressing this issue by investing in R&D. Recently the company showed its unusual enthusiasm for entering substitutes market by acquiring Ivex Packaging Corp. in 2002, entering the thermoformed plastics and plastic extrusions markets. In the same year, Alcoa Fujikura Ltd became sole owner of Engineered Plastic Components, Inc., a supplier of automotive precision-molded components and assembles. Due to various causes including weakening dollar and increasing Chinese demand, metal prices hiked in 2003 while aluminum was one of the weakest performers. Surprisingly, price difference between aluminum and competing metals are at highest level during the scope of this report, 1987~2003 as depicted in Figure 3. I expect that aluminum will substitute for other metals in some instances. Further discussion will be made in Forecast Section. Aluminum, Copper & Nickel 3500 18000 16000 14000 2500 12000 2000 10000 1500 8000 6000 1000 4000 500 2000 0 0 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 Jan-04 3-4 Entry / Exit / Expansion Barrier 6 Nickel Aluminum/Copper 3000 Aluminum Copper Nickel It is very difficult for a new firm to enter aluminum industry. First, The industry is prohibitively capital intensive. Required capital to build a standard size smelter (250,000 tons) is estimated far above 1 billion dollars (Alcoa has 28 smelters). Second, existing producers preoccupy critical resources. A new producer, if it wants to be vertically integrated, needs to find, for example, commercially viable bauxite mines, almost all of which are already owned by existing producers. Third, mini-mills do not have competitive advantages in aluminum industry. In steel industry, better efficiency of mini-mills drove steel giants to a corner. Aluminum mini-mills lack the technology to make high quality can sheet, which is most lucrative market in the industry. However, outside North America, many aluminum producers are government owned. Pechiney, recently acquired by Alcan, in France was wholly owned by its government and Norwegian government holds 51% of Norsk Hydro. In case, as it has been, a government wants to produce aluminum for as part of industrial development program or for national security purpose, entry barriers mentioned above can be abated significantly. Government intervention has contributed to a chronic problem in aluminum industry in that higher cost producers could stay with government subsidy. Incumbent producers have not found an easy way to expand and grow in the aluminum industry. Besides tremendous required capital for expansion, possibility of overcapacity has hindered companies to expand fast. As a result, Alcoa has sought to expand through acquisition of existing companies and brown field projects. In fact, Alcoa has not invested in a new smelter for the past 20 years. It, in my view, resulted in higher maintenance cost and decrease in asset efficiency. This issue will be revisited in Company Analysis section. 3-5 Competition Competition in aluminum industry is potentially very high. It stems mainly from two factors. First, aluminum is a commodity, for which product differentiation is hard to come by. Second, aluminum producers are basically price takers as prices are set by supply-demand balance in London Metal Exchange. However, lack of flexibility in assets and output, long lead time to add new capacity and the huge required capital imply that a company can not encroach on others' market share without risking prohibitively costly punishment by prolonged overcapacity and plummeting prices. Therefore, competition gets in the form of cost control. The question is whether a company can be profitable at a given aluminum price level. Alcoa's cost structure aggravated in recent years due to rising smelter operating costs in US including electricity cost, aging of existing facilities and acquisition of 7 high cost assets, each of which will be discussed again in Company Analysis section. Alcoa's endeavor to deal with this pressing issue can be summarized as reduction of US operation and an internal productivity improvement campaign called ABS(Alcoa Business System). The company claims that it saved $1 billion after 2000 through ABS and aims to cut additional $1.2 billion cost by 2006. US aluminum producers include Century Aluminum Co., Commonwealth Industries Inc., and Kaiser Aluminum Corp. Alcoa's US competitors are small in size and financially unstable. Table 14 US Primary Aluminum Producers (Unit in Million Dollars) Sales (2003) Alcoa Century Aluminum Commonwealth Kaiser Net Available to Common 21,504 1,034 782 (21) 920 (31) In Court Receivership Alcan, a Canadian aluminum producer spun off from Alcoa after World War II, can be considered a rival to Alcoa in terms of size and profitability. It is the second largest aluminum producer in the world and owns abundant low-cost electricity sources and efficient management. Particularly, Alcan became the sole owner of the third biggest aluminum producer, Pechiney of France, in February 2004. Table 25 World Leading Aluminum Producers (Unit in Million Dollars) Alcoa Alcan (Canada) Pechiney (France) Sales (2003) 21,504 13,640 13,314 Net Available to Common 1,034 283 (546) As already mentioned either company could not influence the other's market share without investing hugely in new capacity, which would be prohibitively risky and costly attempt. Rather the two companies will focus on cost cutting to be more profitable at foreseeable market prices. Another possibility is tacit collusion between the two dominants to affect market price since the combined primary production capacity of the two is estimated to be about 35% of world primary capacity. 3-6 Market Size and Growth 4 5 source: company annual reports and websites source: company annual reports and websites 8 World capacity, production and consumption have grown more than 200% during 1973-2003 periods as shown Figure 1. Figure 36 In contrast, US production and consumption increased merely 6.6% and 23% respectively as described in Figure 4. Figure 47 It is believed that the growth of aluminum industry is driven by GDP growth8, implying that aluminum industry is cyclical in line with general business cycle. 6 source: London Metal Exchange, Bloomberg source: US Bureau of Mines 8 To verify the general claim, I performed regressions between GDP and aluminum consumptions. The result can be found in Appendix 1. In Forecast section, I used this relation whenever a projected consumption is not clear. 7 9 Aluminum prices have relatively been stable through the last 100 years. Using US physical spot price for the period of 1900 – 2003, I found standard deviation of yearly change is mere $0.088 per pound. If we exclude years affected by supply or demand shocks such as world wars and oil shocks, then standard deviation can be as low as $0.044 per pound. Using BestFit, distribution of price changes during the last 36 years is found as a triangle distribution (-0.22, 0.0027, 0.26) with exclusion of outliers caused by oil shocks. Figure 5 Triang(-0.26636, 0.0027000, 0.40625) 6 5 4 This distribution will be used to estimate future aluminum prices in Forecast section. BestFit Student Version 3 For Academic Use Only 2 1 1.2% -0.2200 90.9% 7.9% 0.2600 0.5 0.4 0.3 0.2 0.1 0.0 -0.1 -0.2 -0.3 0 Figure 6 describes price movement from 1987 to 2003. After fluctuation of late 1980s and early 1990s that was caused by CIS producers, prices stabilized since mid 1990s Figure 6 LME 3 Months/Metric ton Aluminum Price (1987-2003) 2,500 2,000 1,500 1,000 1,987 1,991 1,995 1,999 4. Company Analysis 4-1 Overview 10 2,003 The overall performance of Alcoa and important market factors for the period of 1987 to 2003 are shown in Figure 79 for profitability and 8 for growth. Figure 7 1.20 1.10 1.00 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 40.0% 20.0% ROCE LME Price Aluminum Price vs Alcoa ROCE 0.0% -20.0% 1987 1991 1996 2000 Aluminum Price ROCE Figure 7 describes that Alcoa's profitability fluctuated as aluminum prices moved as commonly expected. Through the 17 year period, Alcoa recorded an average of 11.3% of ROCE.10 Figure 8 Metric tone 10000 220 8000 200 180 6000 160 140 4000 120 100 2000 0 US GDP Growth GDP, US Demand & Aloca Shipment 80 1987 1991 US Consumption 1995 1999 Alcoa Shipment 2003 US GDP The company grew solidly during that period in spite of setbacks in early 1990s and recent years. Its shipment increased 1.28 times from 2.2 million metric tones to 5.05 million metric tones with average annual growth rate of 5.75%. 4-2 Financial Statement Analysis 4-2-1 Common Size Analysis / Trend Analysis / Ratio Analysis In calculating ROCE for 1992, I did not include accounting charge related to the restructuring of retirement benefits. Appendix 2 is to show the effect of the accounting charge of $1,162 million to Figure 1 and regression result. 10 9.1% if the accounting charge in 1992 is included. 9 11 Discussion of implications from patterns to be revealed through these analyses will be incorporated in profitability analysis and growth analysis section. As you can find in Appendix 3, 4 and 5, common size analysis and trend analysis reveals apparently deteriorating patterns in almost all aspects. Most conspicuous are the following four patterns. First, long-term debt is increasing too fast even though the company reduced it by large amount in 2003. Second, growth rate of share holders' equity far exceeds that of retained earnings. Third, growth rate of PPE is deteriorating while goodwill gets to account for more portions of total assets. Fourth, the company pays less proportion of tax out of operating income with years. Even though a few patterns related to income items are shown but they are not apparent. As for the tax issue, it turns out that the company managed to lower effective tax rates through either recognizing benefits from foreign net operating losses or obtaining more favorable tax treatment from foreign governments. Alcoa's effective tax rate was 33% in 1997 and it went down to 21% in 4th quarter in 2003. Ratio Analysis in Appendix 6 provides additional issues. Besides a disturbingly increasing debt to total assets ratios, asset turnover ratios and capital expenditure ratios raise reasonable concerns. Asset turnover did not improve and capital expenditure ratios aggravated in 2003 implying that these issues might stem from some other problems in addition to cyclical downturn in aluminum industry. Figure 9 Selected Ratios 40.00 35.00 30.00 25.00 CAPEX Ratios 20.00 15.00 10.00 5.00 0.00 2.00 1.50 1.00 1/ATO 0.50 0.00 1996 1997 1998 1999 Total Assets to Capital Expenditure 2000 2001 2002 2003 PPE to Capital Expenditure 1/ATO Table 3 Selected Ratios 1996 Debt to Total Assets 1/(Asset Turnover) Total Assets to CAPEX PPE to CAPEX Sales to CAPEX 0.13 1.94 6.75 3.55 13.11 1997 0.11 1.00 14.52 7.53 14.59 1998 0.16 1.00 16.40 8.49 16.48 12 1999 0.16 0.95 18.83 9.96 17.80 2000 0.16 0.93 22.12 9.97 20.56 2001 0.23 0.75 25.66 10.42 19.23 2002 2003e 0.28 0.21 0.70 23.03 9.36 16.04 0.70 35.48 14.23 24.80 4-2-2 Profitability Analysis Alcoa's financial statements from 1996 to 2003 are reformulated to facilitate profitability analysis, growth analysis and forecast as you can see Appendix 7 and 8. Appendix 9 and 10 shows common size analysis and trend analysis from reformulated statements. Insights from those analyses are incorporated in the following profitability analysis and growth analysis. Table 6 summarizes profitability analysis. Table 6 2003 Return on Common Equity (ROCE) Return on Net Operating Asset(RNOA) 2002 2001 2000 1999 1998 1997 15.90% 10.76% 0.79% 1.54% 5.80% 4.75% 15.49% 11.74% 13.20% 10.50% 16.49% 13.73% 13.43% 12.13% 2.32% 8.45% 0.608 2.74% -1.21% 0.621 2.89% 1.86% 0.565 4.48% 7.26% 0.516 3.63% 6.87% 0.393 5.13% 8.60% 0.321 6.27% 5.87% 0.221 8.17% 3.44% 4.72% 0.505 2.14% 3.39% -1.25% 0.504 3.61% 1.61% 2.00% 0.487 9.34% 4.26% 5.08% 0.494 7.70% 3.29% 4.41% 0.561 9.72% 3.57% 6.15% 0.631 8.46% 3.64% 4.82% 0.690 9.85% 7.83% 0.055 0.203 0.125 0.024 0.044 1.42% 7.42% 0.054 0.198 0.124 0.020 -0.040 4.09% 10.07% 0.073 0.220 0.120 0.028 -0.032 8.09% 13.54% 0.088 0.245 0.109 0.047 -0.007 6.81% 11.76% 0.080 0.232 0.114 0.037 -0.012 7.85% 10.78% 0.070 0.222 0.114 0.037 0.008 6.58% 11.14% 0.068 0.229 0.117 0.043 -0.003 0.915 0.128 0.115 0.574 0.300 0.084 0.150 0.146 0.178 0.921 0.130 0.119 0.583 0.295 0.078 0.164 0.120 0.155 0.861 0.127 0.113 0.542 0.258 0.076 0.156 0.091 0.144 0.689 0.107 0.095 0.485 0.162 0.068 0.131 0.079 0.119 0.648 0.117 0.107 0.560 0.084 0.069 0.144 0.094 0.087 0.572 0.117 0.104 0.515 0.062 0.060 0.149 0.093 0.077 0.542 0.118 0.104 0.516 0.018 0.060 0.164 0.093 0.103 Financial Leverage Drivers Net Borrowing Cost (NBC) SPREAD Financial Leverage (FLEV) Operating Liability Leverage Drivers Return on operating assets(ROOA) Short term borrowing rate (after tax) Operating Leverage SPREAD Operating Liability Leverage (OLLEV) Next Levels Profit Margin Core Sales PM Sales PM Gross Margin Ratio Operating Expense Ratio Operating Tax Ratio Other Operating Income Ratio 1/Asset Turnover Receivables Turnover Inventory Turnover PPE Turnover Goodwill Turnover Accounts Receivable Turnover R $ PR Cost Turnover Other noncurrent Liabilities Other Turnovers The analysis provides clear insights on the company's profitability. While the company appears to have come back to impressively high level of ROCE, it was largely due to increased financial leverage and further reduced net borrowing cost, not due to improved operation. 13 The company's core ability to generate profits from sales as reflected in core sales profit margin numbers have kept deteriorating over time. ATO seems to have troubles in recent years. Individual profit margins and turnovers generally have gotten worse except accounts receivable turnover. Figure 10 Figure 11 18.00% 2.500 55.00% 14.00% 2.000 10.00% 1.500 6.00% 1.000 2.00% 0.500 Core Sales PM 70.00% 40.00% 25.00% 10.00% -5.00% 1997 1998 1999 2000 2001 2002 -2.00% 2003 Net Borrowing Cost (NBC) 1997 1998 1999 2000 2001 2002 2003 Core Sales PM 0.000 Asset Turnover Financial Leverage (FLEV) Return on Common Equity (ROCE) In other words, whereas ROCE has moved in line with aluminum price and other market factors as discussed in overview section of company analysis, its profit generating ability has decreased over time. It is worth noting that profit margin in 2003 is a historical high of 9.85% while core sales profit margin is just 7.83%. It was caused by abnormally high gains from unusual items such as currency translation and available for sale securities, not from its core operations. It is not easy to detect what might have caused it because almost all indicators moved in one direction together. Fluctuation of aluminum prices can not fully explain the disturbing fact because prices in recent several years were very stable and prices in 2003 were higher than those in 1998 and 1999. However, the pattern of deterioration gives a clue to understand this issue. Two turning points in the pattern appear to be 1998 and 2000, in which the company acquired Alumax and Reynolds respectively. I conjecture that troubles caused by the acquisitions should be one part of explanation. 4-2-3 Growth Analysis To measure of return for shareholders and a company's ability to grow, I am going to look into the company's residual earnings during the years between 1996 and 200311. 11 Residual earnings as (ROCE-Cost of Equity Capital)*CSEt-1 14 1/ATO Core Sales PM & ATO ROCE, FLEV & NBC Cost of equity capital and beta for Alcoa for the period of 1996 to 2003 are shown in table 7.12 Table 4 Yearly Change of Risk-Free and Beta Year Risk-Free Beta 1996 1997 1998 1999 2000 2001 2002 0.064 0.064 0.053 0.056 0.060 0.050 0.046 0.040 0.464 0.912 0.671 0.388 0.418 1.390 1.298 1.282 2003 In Appendix 11, I assumed three scenarios. In the first scenario I used changing risk free rates and betas from Table 7. Constant beta of 0.878 was used for the second scenario. For the third scenario constant cost of equity capital of 10.69% was assumed with constant risk free rate of 5.42% and beta of 0.878. Risk premium of 6% was used throughout all of the three scenarios. Results show that the company made negative residual earnings in two scenarios and earned negligible positive residual earnings in a scenario. In sum, the company does not appear to have earned residual earnings over the past 7 year period. Given the low interest rates in the period, it is quite disappointing. Fluctuation of growth rates of residual earnings is surprisingly large and big negative growth occurred, in all three scenarios, in 2000 and 2001. This analysis provides evidence that whereas Alcoa achieved solid growth of CSE and other items through the acquisitions, the acquired assets might not be properly functioning. 4-2-4 Interpretation My hypothesis is that the company's operating efficiency and ability to generate profit were damaged by a series of acquisitions, idling capacity and aging facilities. Patterns observed above indicate that the company had to borrow hugely to finance the acquisitions affecting debt ratio and FLEV. However, unfortunately, the acquired assets had higher cost structure resulting in lower core sales pm. On top of it, packaging market, which Reynolds serves, was struggling since 2001, aggravating the situation. Perhaps organizational turmoil in Reynolds and Alumax worsened their performance, losing core customers, suppliers, distributors or important employees. Idling capacity, which still consumes a certain level of cost, definitely must have driven down efficiency ratios and core sales pm. A large scale of layoffs since 12 Risk free rates are from 10 year US Treasury bond rates. Betas come from regressions with S&P as benchmark. For how the data was used, see Appendix 10. 15 2000 might have discouraged employees, lowering productivity. Rising electricity cost and employee benefit cost must have added some negative influence. Alcoa wanted to start investing heavily in long term assets to maintain long term capacity and efficiency because its assets started showing aging effects13. However, the company had been in financial constraints due to a series of acquisitions and therefore capital expenditure was restrained, pushing CAPEX ratios high to the ceiling. Fortunately, net borrowing costs had stayed low and FLEV was high and aluminum prices started to recover from 2002 low. Furthermore, the economy started picking up with weakening US dollar and soaring stock market. As a result, the company could show a good ROCE and PM numbers in 2003. My next question is whether the damage to efficiency and margin is permanent or temporary. To answer to the question, I will need interviews with Alcoa officials and additional information. For now, I will assume that the company will gradually recover from the damage for the coming 5 years. 5. Forecast As in many cases in valuation, it is inevitable to make a lot of assumptions. Particularly I face two serious obstacles. First, I will assume that Alcoa does not acquire other companies. This is problematic because companies in this industry grow through acquisitions. Increasing capacity with green field projects is extremely risky and costly in aluminum industry. However, it is more difficult to project when and which company Alcoa will acquire. Second problem stems from future aluminum price. Predicting a mid to long term aluminum price is as tricky as predicting a stock price. As a result I will use industry consensus projections for 2004 and 2005 and try random simulation for the period of 2006 to 2008 basing on historical data. 5-1. Sales Forecast 5-1-1 Price Forecast The O'Carroll Aluminum Bulletin estimate will be used for 2004 and 2005 with a moderate modification to incorporate substitute effect. The source is said to be one of a few to estimate supply and demand through bottom up approach. Its projection of supply-demand balance and prices is summarized as below. Table 5 Price Forecast 13 Average age of Alcoa's smelters is over 25 years in my calculation including acquired assets. 16 Projection for 2004 and 2005 Changes in Smelter Capacity (Western World) 2003 2004 2005 Projected World Consumption % Change 27,300 7.1% 29,234 7.1% 31,077 6.3% Projected World Production % Change 27,868 7.3% 28,948 3.9% 30,253 4.5% Projected World Capacity % Change 31,135 5.4% 32,073 3.0% 33,666 5.0% Capacity Utilization % 91.9% 91.6% 92.0% 568 (286) (824) 0.65 0.65 0.73 0.76 0.77 0.88 US Canada Latin America Western Europe Africa Asia Oceania Western World 2004 0 2 55 130 117 199 35 538 2005 0 148 52 50 347 31 628 Source: The O'Carroll Aluminum Bulletin Implied Inventory Change Price Price (Alumina Shortage) Source: The O'Carroll Aluminum Bulletin However, it appears that they did not incorporate possibility of substitute effect. As discussed in Industry Analysis section, prices of metals such as copper, steel and nickel have spiked while aluminum prices have relatively been stable. Current difference between aluminum and other metals is at historical highest within the scope of my study, 1987 ~ 2003. On top of it, prices of other metals are expected to go up further in, at least, coming a few years14. I expect other metals will be replaced with aluminum in a wide range of industries centering on small businesses, low value product users and, especially, easy switching sectors. Consequently I conjecture for 2004 and 2005 that approximately additional 4% demand increase for aluminum will be generated from electric equipment market (for wiring / replacing copper), packaging industry (for steel cans and closures / replacing ordinary steel), automotive (for ordinary steel components) and construction (for corrosion resistance / replacing nickel and galvanized steel), expecting 0.77 (0.82 in case of alumina shortage) for 2004 and 0.80 (0.90 in case of alumina shortage) for 2005. It is very controversial if there will be an actual alumina shortage. Some market observers claim that alumina shortage has begun in reference to alumina prices at spot market. However, considering spot market buyers are mostly Chinese manufacturers, who in many cases buy large amount of raw materials to stock for future usage, others believe that the shortage, if any, will be temporal. I attached alumina supply demand forecast prepared by The O'Carroll Aluminum Bulletin as Appendix 14. For prices after 2005 my projection is based on historical data. I used BestFit program to find distribution pattern of price movements during the past 36 years as shown in Industry Analysis section. The result is used as input data for Monte Carlo simulation. 14 See Appendix 12 for Forecasted Supply Demand Balance for copper, nickel and zinc. 17 5-1-2 Shipment and Unit Price Forecast Market estimate per product line is attached as Appendix 1315. Increasing shipment is largely due to Boeing's increased orders, recovery of autos and trucks, and electrical equipment. While it is generally estimated that electrical industry is likely to struggle due to lack of new investment, I believe some copper users, especially in wiring and appliance, will switch to aluminum in 2004 and 2005. Packaging market is forecast to be stagnant and modest growth is predicted for consumer durables. From the market forecast mentioned above, I prepared Alcoa's shipment and unit price forecast per product line as shown Table 6. I assume that, as prices spike in 2004 and 2005, Alcoa restarts some of its idling capacity of 595 thousand metric tones. I considered the scheduled addition of a new smelter capacity of 322 kmt in 2007 and expansion of Quebec smelter. For unit price for packaging line, possibility of introduction of new consumer products with higher margin was incorporated. 5-2. Core Sales PM and Asset Turnover My main idea was that, even though absolute numbers are likely to be very good in 2004 and 2005 thanks to high aluminum prices, in relative measure the company's core sales pm will be equivalent to 2003 level, improving a little in 2007 and 2008 as new low cost capacity comes on stream, high cost capacity in US is dismantled and other troubles are dissolved. A problem was caused by the fact that I could not obtain information about Alcoa's variable cost structure. I used industry data from mid 1990s16. For simplicity purpose, I assumed only two point estimations of core sales pm in every year forecasted. However, in reality, it should be continuous, not discrete. Electricity price, which is rising almost everywhere over the world, was assumed constant mainly because of lack of data. The company's asset efficiency to generate sales is expected to be a little higher in 2004 and 2005 with return of some of idle capacity. It is likely to show moderate improvement in 2007 and 2008 with addition of new efficient capacity and dismantlement and divestiture of less efficient assets. Source: The O'Carroll Aluminum Bulletin and various trade associations such as The Aluminum Association, Wards Automotive, United Brewers Association and others. 15 16 Some professional organizations sell updated industry cost data. 18 Table 6 Forecast of Shipment and Unit Price per Product Line Aluminum Price Capacity Idle Capacity Add on New Capacity Shipment Unit Price Unit Added Value Sales Primary Metal Shipment Unit Price Unit Added Value Sales Flat Rolled Shipment Unit Price Unit Added Value Sales Engineered Product Shipment Unit Price Unit Added Value Sales Packaging Shipment Unit Price Unit Added Value Sales Others Shipment Unit Price Unit Added Value Sales Alumina Shipment Unit Price Unit Added Value Sales World Real GDP Growth Developed Country Real GDP Growth US Real GDP Growth 2003 2004 2005 2006 2007 2008 0.7 0.82 0.8 0.8072 0.8130 0.8175 3345 595 3645 295 3645 295 3645 295 3967 0 5047 1.937 1.237 21504 5194 2.073 1.253 23691 5440 2.064 1.264 24703 5546 2.031 1.224 24782 3967 150 322 5915 1.855 1.043 24146 6005 1.823 1.005 24084 1952 0.752 0.052 3229 1850 0.820 0.000 3337 1900 0.800 0.000 3344 1900 0.807 0.000 3374 2181 0.813 0.000 3901 2181 0.818 0.000 3923 1819 1.203 0.503 4815 1946 1.350 0.530 5781 2044 1.330 0.530 5980 2095 1.337 0.530 6162 2147 1.343 0.530 6344 2201 1.348 0.530 6524 879 2.890 2.190 5589 984 3.320 2.500 5415 1063 3.300 2.500 5848 1106 3.307 2.500 6082 1133 2.813 2.000 4987 1162 2.618 1.800 4601 167 8.751 8.051 3215 165 8.870 8.050 2922 165 8.850 8.050 2922 170 8.857 8.050 3011 170 10.813 10.000 3740 170 10.818 10.000 3740 230 5.245 4.545 2654 248 8.820 8.000 4820 268 8.800 8.000 5194 276 7.807 7.000 4737 284 6.813 6.000 4250 292 6.818 6.000 4372 7671 0.119 7700 0.184 7700 0.184 7700 0.184 7700 0.120 7700 0.120 2002 1416 1416 1416 924 924 2.63 2.69 2.81 2.78 2.76 1.63 1.50 1.81 1.90 2.03 2.50 2.02 2.50 2.03 2.50 5-3. Cost of Operating Capital and ReOI Growth Rate The company's cost of operating capital in 2003 is estimated as 9.145% with risk free rate of 4.95% (10 year Treasury bond rate), beta of 0.878, market premium of 6%, and borrowing rate of 6.6%. Since it is very likely that risk free rates will rise in the near future, I set 10% as default for Alcoa's cost of operating capital. The effect from the company's repayment of debt will be offset by rising borrowing rates. Valuation incorporates sensitivity analysis for cases with different cost of operating capital from 9.15% to 14%. 19 Given the wild fluctuation of ReOI in the studied period, which might not be a typical cycle for aluminum industry, it is very hard to estimate long term ReOI for Alcoa. I assume Alcoa's long term ReOI to range from 2% to 3%. 6. Valuation (with Monte Carlo Simulation) Input list is attached as Appendix 15. Below is Monte Carlo simulation model for valuation. Table 7 Valuation Model (ReOI with Monte Carlo) 2003 A Sales 2004E 21,504 Growth in Sales Core Sales PM 1/ATO ReOI Growth in ReOI Cost of Operating Capital Discount Rate PV of ReOI Total PV Continuing Value PV of CV NOA 2003 Value of NOA NFO V 2003 Value per share (at 868.49 million) 468 0.0915 2005E 2006E 2007E 2008E 23,691 10.2% 0.125 0.75 1,185 153.1% 24,703 4.3% 0.115 0.78 914 -22.8% 24,782 0.3% 0.115 0.75 991 8.5% 24,146 -2.6% 0.13 0.7 1,449 46.1% 24,084 -0.3% 0.13 0.65 1,565 8.1% 0.1000 0.1000 118.5 0.1000 0.8264 755.4 0.1000 0.7513 744.8 0.1000 0.6830 989.5 0.1000 0.6209 972.0 3,580 24,927 15,478 20,050 39,108 6750 32,358 Cost of Operaqting Capital after 2008 0.1000 ReOI Growth Rate 0.035 37.26 Valuation Result and Sensitivity Analysis At 10% of cost of operating capital, valuation result is 22 dollars in normal case and 36 dollars in alumina shortage case. For summarized result of simulations, see Appendix 16 and 17. Distribution of output and input sensitivity for the default case are depicted as follows; 20 Regression Sensitivity for Value per share / 2003 A/J23 ... 4 Aluminum Price / 2008/G6 .817 3 Aluminum Price / 2005/D6 .144 @RISK Student Version 2 For Academic Use Only Aluminum Price / 2006/E6 .068 1 Unit Added Value / 2004/C33 0 -1 -0.75 -0.5 .066 -0.25 0 0.25 0.5 0.75 Std b Coefficients Valuation result varies depending on cost of operating capital as follows; Cost of Capital 10% (Default) 9.15% (present) 11% 12% 13% 14% Mean 30.47 35 26.37 23.22 20.71 18.68 Normal Case 22 25 19.5 17.8 15.6 14.2 Alumina Shortage 37.5 45 33 28.4 25.2 22.4 7. Recommendation17 Current price of 37.39 is appropriate if we assume alumina shortage will come to existence in 2004 and 2005, which is quite uncertain at present. Accordingly I recommend that you wait until alumina issue is clarified in a couple of months. I expect that the issue will be cleared when Alcoa and Alcan issue their first quarter result in the middle of April. 8. Accounting Quality Since the most of information I used for this report came from the company's financial statements, I checked accounting quality of Alcoa using Beneish 5 and 8 variables models. The result is as follows; 17 I made a lot of assumptions in estimating Alcoa's intrinsic value. Therefore, attention should be paid to any deviation from my assumptions such as interest rate change, electricity price, major end users' trend, other metals' price movement and many others. If any deviation happens, the intrinsic value should be reestimated. 21 1 The Full Beneish model for earnings manipulation detection (Based on Eight Variables) 2002 Year 2001 2000 1999 1998 1997 1996 INPUT VARIABLES Net Sales CGS Net Receivables Current Assets (CA) PPE (Net) Depreciation Total Assets SGA Expense Net Income (before Ext. Items) CFO (Cash flow from op.) Current Liabilities Long-term Debt 20,263 16,247 2,378 6,313 12,111 1,108 29,810 1,147 420 1,839 4,461 8,365 22,497 17,539 2,386 6,449 11,530 1,234 28,355 1,256 908 2,411 4,885 6,384 22,659 17,111 3,461 7,578 12,850 1,199 31,691 1,088 1,484 2,851 7,954 4,987 16,323 12,536 2,199 4,800 9,133 888 17,066 851 1,054 2,236 3,003 2,657 15,340 11,933 2,163 5,025 9,133 842 17,463 783 853 2,197 3,268 2,877 13,319 10,275 1,581 4,416 6,667 735 13,071 682 805 1,888 2,453 1,457 13,061 9,966 1,675 4,281 7,078 747 13,450 709 515 1,887 2,373 1,690 DERIVED VARIABLES Other L/T Assets [TA-(CA+PPE)] 11,386 10,376 11,263 3,133 3,305 1,988 2,091 1.1065 1.1120 1.0438 0.9007 1.1534 0.9863 -0.0476 1.0826 0.6944 1.1110 1.0296 0.9929 0.8828 0.8600 -0.0530 0.9732 1.1338 0.9475 1.9359 1.3882 1.0383 1.0858 -0.0431 1.2313 0.9554 0.9573 0.9701 1.0641 0.9526 0.9791 -0.0693 0.9424 1.1879 1.0290 1.2441 1.1517 1.1764 1.0032 -0.0770 1.1766 0.9258 1.0368 0.9782 1.0198 0.9614 1.0601 -0.0829 0.9901 M = -6.065+ .823 DSRI + .906 GMI + .593 AQI + .717 SGI + .107 DEPI M-score (5-variable model) -2.76 -3.07 -2.02 -2.97 -2.47 -2.95 #REF! M = -4.84 + .920 DSRI + .528 GMI + .404 AQI + .892 SGI + .115 DEPI -.172 SGAI + 4.679 Accrual to TA - .327 Leverage M-score (8-variable model) -2.62 -2.93 -1.95 -2.81 -2.46 -2.92 #REF! DSRI GMI AQI SGI DEPI SGAI Total Accruals/TA LVGI #REF! #REF! #REF! #REF! #REF! #REF! -0.1020 #REF! Note: if M > -2.22, firm is likely to be a manipulator The result indicates that the company might have manipulated earnings in 2000. Further investigation to identify the causes is need. However, M scores for other years are quite sound. Ending Page 22