iia MINN: npr x Bei BY NY: aS, g=s=! if. iGeey Snake ae! erate! Be Ba L4NG th rjiet UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN BOOKSTACKS Digitized by the Internet Archive in 2011 with funding from University of Illinois Urbana-Champaign http://www.archive.org/details/decisionriskanal871hert FACULTY WORKING PAPER NGC. 871 Decision and Risk Analysis of a New Product ana Facilities Planning Problem: CCA Company's Egg’N Foam Project David B. Hertz Howard Thomas College of Commerce and Susiness Administration Bureau of Economic and Business Research University of lilinois, Urbana-Cnampaig: The person charging this material is re- sponsible for its return to the library from which it was withdrawn on or before the Latest Date stamped below. Theft, mutilation, and underlining of books are reasons for disciplinary action and may result in dismissal from the University. To renew call Telephone Center, 333-8400 UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN L161—O-1096 BEBR FACULTY WORKING PAPER NO. 871 College of Commerce and Business Administration University of Illinois at Urbana-Champaign May 1982 Decision and Risk Analysis of a New Product and Facilities Planning Problem: CCA Company's Egg'N Foam Project David B. Hertz Chairman of the Board Prime Time' Magazine Howard Thomas, Professor Department of Business Administration Abstract The paper examines a decision and risk analysis approach applied to a new product and facilities planning problem.* The discussion describes the risk analysis process, and presents the step-by-step approach used by the project team. The initial results of the analysis are presented and are discussed by the team in a process of policy dialogue. It is shown that the project team questioned a number of project assumptions, parti- cularly in the marketing area, and expressed concern about the perceived vagueness of CCA's objectives. The output of the dialogue process finally led to a strategy recommendation for Egg'N Foam, following an examination of the relationship of the Egg'N Foam project to the firm's long-term strategy path. *This problem is somewhat disguised in regard to project chronology and magnitudes of input data but the problem structure and new product area is unaltered. ) heen Aanorque aereneoie me vais ba 1 iaaner4 viet oer bosnene7e, oe teetann dy te aytuaen taasted at ‘afoand eal ‘y | “veda et a -nagedii Sette ren Ye-aasoase 9 i a ried de ¥ bye: onl “hevtaoreg ats 0d dexmoies. boniornns bns yeeis potiecane od fot cal vitor wawDond euqatatty nat hol amg3v0 ott stawbdontie: efang’ ‘Yo Op onal santas hh gatwolle wat oe ro} polaebmenmosi3 Years . 4 he i% aft * 7 a : Y A A : oA y _Xantoootts bicdh joie od byage e. bap lgoguth deitewod 22 mit “Saphery weet) bir: eanFONI 8 meldong ois stint namh quant tev INTRODUCTION The paper discusses the role of risk analysis in relation to the Egg N'Foam project--a potential investment opportunity in the plastic egg carton market. The paper illustrates the step-by-step nature of the application of risk analysis [see Hertz and Thomas {4]] and also show the procedures adopted at each stage. In subsequent sections of this paper, we will develop the problem structure in flow diagram form, state the problem assumptions, perform an initial sensitivity analysis, develop investment alternatives, assign subjective probabilities, pre- sent the results of the risk analysis, demonstrate the process of policy dialogue about the strategy to be chosen, and summarize our conclusions. The decision alternatives under consideration ranged from abandon- ment (which would involve the loss of the pilot plant) to expansion strategies (involving larger plants--so called "superplants"-~-and a series of small plants). The risk analysis approach involved three sets of people. The management group at CCA, the CCA project team (charged with the respon- sibility for project evaluation), and the consultants (providing tech- nical expertise and guidance). The most interesting element in the process involved the dialogue about choice and implementation which developed between the project team and management group at CCA. THE CCA COMPANY'S EGG'N FOAM PROJECT The Egg'N Foam project at CCA represented a large investment, a totally new market area, and a completely new product for CCA. For several years CCA had been developing Egg'N Foam, a plastic package as a competitive substitute for paper egg cartons, refining its product -2- design and operating a pilot plant in Nesquehoning, Pennsylvania. The project involved substantial amounts of uncertainty. The project team selected to work on the Egg'N Foam decision con- sisted of representatives from the Research and Development Division, Plastics Division, and Corporate Controller's Office. Those involved had been associated with the Egg'N Foam project for a considerable period of time, and were quite familiar with the development, production, and marketing aspects of this project. The following section discusses the step-by-step approach used by the project team to study the attractiveness of investing more funds in Egg'N Foam. THE RISK ANALYSIS PROCESS DEVELOPING THE FLOW CHART AND PROBLEM STRUCTURE The first step in applying risk analysis to Egg'N Foam was to construct a flow chart of the investment analysis. The process started with an assessment of basic economics of the investment project, entering each element in a simple chart. The initial Egg'N Foam flow chart shown in Chart 1 consisted of six elements representing: (1) the total market size for egg cartons, (2) CCA's foam market share, (3) selling price, (4) manufacturing cost, (5) total overheads, and (6) investment base. The initial flow chart was then expanded backward to encompass a large set of input factors. This involved establishing a value for each factor, analyzing the determinants of that value, questioning the validity of the determining factors, and where possible, exploding the CHART 1 INITIAL FLOW CHART EGG’N FOAM CCA TOTAL OVERNEAD v PROFIT RETURN ON INVESTMENT INVESTRHERT -4- elements into further detail. The final product was a flow diagram of all the factors relevant to determining the profitability of Egg'N Foam. The final flow chart for Egg'N Foam is considerably more detailed than the initial chart (Exhibit 1). The manufacturing costs, overhead costs, and investment base, were expanded into a substantial number of inputs. However, the unavailability of sufficient market data limited the amount of analysis which could be undertaken on the marketing ele- ments (i.e., market size,* market share, and price). STATING THE ASSUMPTIONS When the flow chart was completed, assumptions underlying each of the factors were stated explicitly. This step was particularly impor- tant because it placed a limit on the amount of detail required to substantiate the data inputs. For example, if the project team assumed CCA would capture a 10 percent share of the carton market, little data would be needed to estimate CCA's sales volume. As this assumption demonstrates, however, the magnitude that assumptions can take reveals the need to verify their reliability. Invalid assumptions can result in misleading and inaccurate conclusions from investment analysis. Thus, verification of assumptions is an integral and important part of the phase of stating the assumptions. *For the purpose of this study, the total market for egg cartons has been divided into five regions. Each represents a market that could be covered by the output from one Egg'N Foam plant. The West Coast was excluded from the analysis because entering this market was considered to be a separate decision not related to building plants in the other four regions. Egg'N Foam cartons cannot be shipped to the West Coast as the freight costs are too high. -5- On the Egg'N Foam project, the team was able to verify the validity of most of the data. This included information such as equipment rates, supplied by Research and Development Division personnel, and forecasts of total egg production, supplied by the U.S. Department of Agriculture. With some data, however, it was felt that further study was necessary to determine the accuracy of the data and assumptions. In particular this included the percentage of eggs cartoned, and Egg'N Foam sales volume and price. Exhibit II lists all of the assumptions underlying the most important input factors of the flow chart. ANALYZING SENSITIVITY The next step in the analysis was to construct a nonprobabilistic computer model of the flow chart, using the basic data shown in Exhibit III. The model was built for a single plant, as each region represented a similar size market that would support a single plant operation. The model performed all the calculations indicated by the flow chart, and provided answers in terms of the return on investment and cash flows. The model was also used to determine the sensitivity of investment returns and cash flows to changes in input variables. The team was quickly able to answer questions such as: "What would happen to the discounted rate of return if the price dropped by $0.50 per thousand cartons?" or “What would happen if the thermoformer cycle time were improved to 4-1/2 seconds?". The results of the sensitivity analyses on Egg'N Foam are shown in Chart 2. Changes in two factors--caliper (or thickness) and demand--had a dramatic effect on the rate of return, whereas changes in the value of other factors had only minor impact on 262 CHART 2 SENSITIVITY OF INVESTMENT RETURN TO CHANGES IN MAJOR INPUT FACTORS Mitlions of Cartons Year} Yenr 2 Year 3 Extruder rate = 525 |b./hr. Demand= 21 54 90 One extruder Thermoformer cycle time = 5.3 sec Physical capacity = 76 million cartons One thermoformer Caliper = .080 in. Discounted raic of return = 12.96% BASE CASE: CHANGE IN ADJUSTED FACTOR CHANGE RATE OF RETURN RATE OF RETURN *** Extruder rate 525 to 625 !b./hr. ; | | 42.24% 15.20% Thermoformer cycle time 5.3 to 4.2 sec. VY 41.71% | 14.67% : jams 3808) | Raw material casts 15.0¢ to 13.5¢/Ib. pellets +1.40% | 14.36% : ! Equipment cosis $730,000 to $990,000 , 1579 11.39% | Caliper * 0€0 to..103 in | == 10.2555 | Millions of Cartons | Demand Yesur 1 Year 2 Year 3 — 4.17% 8.79% From _21 54. 90 : To. 21 54 66 ** Price : Decreased 50¢ — 1.20% 11.765 3 .20/hr. Direct labor €xtruder $11.00 to $13.20/hr ~ 1.83% 11.13% Thermoformer $26.94 to $32.34/hr. * Caliper increase reduces the physica! capacity of the plant considerably ** Sales volume drops from 76 to 66 in year 3 *** Adjusted to use one extruder to two thermoformers —J= profitability. The high sensitivity of the investment return to changes in caliper and demand indicated the extreme importance of the accuracy of the assumed sales volume of Egg'N Foam. The sensitivity analysis was also valuable because it permitted the team to experiment on the computer with different equipment combina- tions, and plant sizes, to identify the optimal plant operation. When applied to Egg'N Foam, the sensitivity analysis suggested that one extruder could support two thermoformers instead of one thermoformer, and also demonstrated the feasibility of building larger plants con- taining two extruders and four thermoformers. DEVELOPING INVESTMENT ALTERNATIVES After the sensitivity analysis was performed, the single plant model was used as a building block in constructing investment strate- gies for Egg'N Foam.* The team developed four alternative strategies for CCA in the egg carton market. Two strategies involved building superplants, each containing two extruders and four thermoformers; the other two strategies called for small plants, each containing one extruder and two thermoformers. The difference between the two superplant strategies** and the two small plant strategies is speed of market entry. One of the superplant strategies and one of the small plant strategies represent a slow, con- servative investment pace, requiring the addition of thermoformers only *The appendix lists the input data that were used for each strategy. **Superplant strategy assumes construction of two plants to serve the four regions, while the small plant strategy assumes construction of four plants. -8- after the demand for the output of existing equipment is firmly estab- lished. Conversely, the other strategies assume that CCA would act as quickly as possible to acquire equipment, hire and train the necessary manpower, and open the plants. The four strategies are shown in Chart 3) expressed in terms of the timing and sequence of equipment installation by plant location. ASSIGNING THE PROBABILITIES The team's next step was to obtain probability information on input factors. Subjective probabilities were developed for the five inputs of the model that had either a significant impact on profitability, or a considerable degree of uncertainty: extruder rate, thermoformer cycle time, caliper, price, and demand. The probability distributions in Exhibit IV represent the best judg- ment of the CCA personnel most knowledgeable about Egg'N Foam. To pro- vide a better understanding of the graphs shown in the exhibit, the following discussion highlights the important characteristics of each probability distribution. Extruder Rate Information from CCA's production specialists, as well as from out- side extruder manufacturers, indicated that extruders brought to the market after 1970 would have higher throughput rates due to advances in design. Therefore, two probability curves were developed; one for the period 1968 to 1970, and the other for the years 1970 to 1975. The width of both distributions is rather tight around the expected value for each distribution. This implies that CCA personnel were con- fident that the extruders would operate at the expected pounds of CHART 3 INVESTMENT STRATEGIES CGG'N FOAM Number of LEGEND Thermoformoars MW = Midwest NE = Northeast 9 SUPER PLANT 0 ee on 8 STRATEGY > ieee saat tad laa = LMA tw 7 = a 2 MW B iw SLOW 6 4 Ce e2 ms 3 ke all 1 MW Ew 5 peat it anf } ‘MW 7; MW 4 i F is bo TOTAL INVESTMENT = $8 MILLION z tT] om 2 m 2 eee ee me Oe See eRe Or eater Here a 1968 1969 1970 1971 1972 1973 1974 1975 1975 , LEGEND Number of : : MW = Midwest SE = Southeast Thermoformers SW = Southwest NE = Northeast SMALL PLANT } STRATEGY pow \ 6 Gs =a aoe sa God [Mw 5 Fo Prva r MW 4 SE 3 wll S E TOTAL INVESTMENT = S8 MILLION 1968 1969 1970 1971 1972 1973 1974 1975 1976 y LLLP ALCL LE COLE CC OE OE OE ST SS | ST RY St SE SS SSE SS SS -10- foam per hour. Furthermore, the shapes of the extruder probability curves are almost identical, indicating that the CCA representatives had the same amount of confidence in the performance of existing and future extruder designs. Thermoformer Cycle Time The same advances are expected for the thermoformers. Improved equipment should be available by 1971, so two different probability distributions were necessary. The first one covers the years 1968 to 1970, and ranges from 6.0 to 4.5 seconds with an expected value of 5.3. The distribution for the 1971 to 1975 thermoformer design ranges from a 6.0- to 3.0-second cycle time, with an expected value of 4.2 seconds. The shapes of the two curves differ, however, because CCA personnel had greater confidence in the performance of 1968 equipment compared to that expected of future thermoformer designs. Caliper Similarly, two probability distributions were developed for the thickness of foam sheets to be processed by thermoformers, as shown in Exhibit IV. In this case, however, CCA production personnel were less confident in their predictions. The probability distribution shows that the caliper could fall anywhere between .090 and .078 inches thick- ness with equal likelihood. The flatness and width of both probability curves indicate the lack of confidence in CCA's ability to produce foam of uniform caliper. Price The project team developed probability distributions for price in each of the years between 1968 and 1975, and predicted the expected -l1- value would drop $0.30 per thousand cartons each year. The probability distributions for three of the seven years are shown in Exhibit IV, and are representative of the remaining price curves. Demand Estimates of demand were developed for each strategy, for each plant, and for each year (the base data for means and standard devia- tions are shown in the appendix), based on the assumption that CCA can capture 10 percent of the egg carton market or 360 million cartons* by 1974 with the fast strategy, or, by 1978 with the slow strategy. For the fast strategy, there is a greater probability of deviation from the expected demand than for the slow strategy. RESULTS OF THE "FIRST PASS” RISK ANALYSIS Exhibit V shows the logical flow chart for the risk analysis simu- lation. This was used as the basis for the "first-pass" evaluation of each of the alternative strategies. The results of the analysis are shown in summary form in Table I below: Table I RESULTS IN TERMS OF NPV RISK-ANALYSIS NPV SUCH THAT POINT SIMULATION (SM.) 95% PROBABILITY ESTIMATE STANDARD PROBABILITY OF EXCEEDING STRATEGY (SM.) MEAN DEVIATION (NPV > 0) THAT VALUE (SM.) 1 4.12 3653 0.84 0.99 2.0 2 0.85 0.74 0.45 0.94 0 3 2.90 0.94 ALES 0.88 -2.0 4 -1.38 -1.70 0.68 0 -3.0 NOTES FOR a) Discount rate in simulation calculation is assumed at 5% TABLES net-of-tax, i.e., the riskless rate. b) Point estimate calculation made using most likely values for each input variable and CCA's appropriate net-of-tax risk-adjusted rate of return. c) Strategy 1 = super-fast; strategy 2 = super-slow; strategy 3 = small-fast; strategy 4 = small-slow. *The 360 million cartons represent a 10% share of the market in four regions: East, Southeast, Midwest, and Southwest. -l 2- Chart 4 shows the cumulative density functions (CDF's) for net present value for each of the strategies. The performance measures shown in Table I are calculated in rela- tion to the discounted return on investment measure, net present value. Columns 4 and 5 present two possible choice criteria for each strategy. These are the probability that the net present value is positive, and also that the value of NPV is such that there is a 95% probability of exceeding that value. Using these criteria, it would appear that strategy 4 could be eliminated immediately. In addition, using principles of stochastic dominance, it would also appear that Strategy 1 dominates all other strategies. Therefore, strategy 1 is recommended. However, as a contingency plan, strategy 2 is the next best alternative, if, in the first four years, the market growth projections are revised downwards considerably. Policy Dialogue Following the initial risk analysis, the dialogue about strategy choice between the project and management teams at CCA focussed around two main concerns. Although it was agreed that the sensitivity and risk analyses had aided the process of strategy formulation, it did not lead to a clear-cut strategy choice. Much controversy centered around the inadequacy of the processes of problem identification (including key assumptions), and the lack of specificity and clarity about CCA's objectives. The strategic issues raised are outlined in subsequent paragraphs. LUM ATE FPR} BILITY DY pe JS TL 04 o7g OF Ob =3= Chart 4 iS) re -] 4- Problem identification Typical critical comments from management included: "The team seems to have considered heavily techno- logically oriented strategies, i.e., concentration on plant size and speed of introduction for the product. Little emphasis was placed on interpreting the demand picture, which seems to suggest that marketing related strategies have either not been considered or assumed away.” or “Why should the first step...be to construct a flow chart?” or "The team's effectiveness might improve if its com- position were better balanced, with the addition of some marketing personnel.” In general, it was felt that some problem assumptions and inputs should be examined more closely, and questioned in order to proceed to a more effective “structuring” of the risk analysis model. It was agreed that the listing of assumptions in Exhibit 2, and the sensitivity analysis in Chart 2, provided a valuable starting point. In relation to the assumptions in Exhibit 2, some further observa- tions were made. First, why did the CCA team assume that they will produce their own sheets of foam when the option of purchasing from an outside source might be a worthwhile alternative? Second, what addi- tional correlation effects (with their attendant measurement complica- tions) should be included in the model in relation to price elasticity, demand, production rates, etc.? Third, why were market decisions taken more or less as given, thus reducing CCA's decision to an apparent choice between one of four production strategies? -15- After considerable discussion, it was agreed that the greatest weakness concerned the marketing assumptions adopted in the model, and the level of empirical or research effort available to support these assumptions. Listed below are some of the areas in marketing which were questioned. - Market share will be 10% by 1978. (Should there be a spread of possible values?) - No manufacturer other than Diamond National will come in before 1978. - Paper carton manufacturers will not react violently to their share being cut in ten years from 50% to 5%. (Why would they not start a price war? Could they afford to do so?) - Pricing must be at Pillopost level. Why not consider the adoption of penetration pricing policies? - Market forecasts are assumed correct. Has any attempt been made to check their forecast accuracy? - Selling will be done by the plastics marketing department. It would appear that the transfer price will become an important motivating factor, and might indicate that a separate department for marketing Egg'N Foam in the carton market should be established. Two other financial management assumptions were seen to require further justification. First, that working capital cannot be regarded as a certain need. The amount required will depend upon the accuracy of demand predictions (e.g., a stock pile-up might occur), the speed of payments by debtors and many other, as yet unknown, factors. Second, the influence of inflation cannot be ignored. It may hit foam harder -16- than pulp, or vice-versa, seriously changing the competitive balance within CCA. Specificity and Clarity of Objectives The management felt that the project team were uncertain about the criteria by which choice should be made in relation to the available strategies. It was suggested, therefore, that attention should be directed towards the criterion for strategy choice in relation to CCA's overall portfolio of investments, both current and projected. The main concern was to relate project risk to the firm's overall risk, and to handle strategic issues. For example, the issue of whether the CCA Egg'N Foam project would be too risky, not only in the sense of varia- bility of returns, but also in terms of its effect on the firm's port- folio of projects. Senior managers commented that the carton project might focus too great a proportion of the company's resources (finan- cial, managerial, production, and marketing) on a single project to the detriment of other possible alternatives. Conclusions of the Initial Policy Dialogue It was agreed that certain tasks could be better performed by management than by the project team. It was suggested that the project team, (perhaps better balanced with additional marketing and financial expertise), should generate appropriate alternative strategies in line with management policy, and subsequently carry out some form of strate- gic evaluation using risk measures agreed jointly with management. Management's task should be to define the range of strategies to be -17- considered, to direct the project group in the decision-making task, and to determine the final strategy choice, perhaps in a dialogue pro- cess involving the consideration of other factors and attributes. It should be emphasized that both groups stressed the importance of management's role in the problem finding and formulation process. It was felt that the current strategy set had an overly production/R & D character. Also, it was agreed that the range of strategies for con- sideration should focus on a close scan of the environment as well as a screen of potential competitive market and technological uncertainties. As a result, the following recommendations were made concerning further strategy formulation and evaluation through the analytic process. First, that greater attention should be paid to the influence of compe- tition, competitive reaction and marketing forces on the Egg'N Foam decision. In other words, marketing/production mix combination strategies should be developed. Management felt that the most sensible approach would be to develop a number of marketing scenarios (using a Delphi-type approach), and if appropriate, carry out a simulation of each scenario. For example, scenarios of the following type should be developed in association with each of the set of production alternatives. (i) steady price; (ii) failure of the market to grow meaningfully; (iii) tough competitive reaction, eag-, price war initiated by Diamond National and others; (iv) a price war started by CCA to discourage other market entrants--i.e., an attempt to build-up share quickly; (v) consumers preferring paper-based to plastic cartons. -18- Second, that attention should be given in model development to working capital and cash flow management. For example, inflation effects and their impact on potential cash flow generation needed to be examined thoroughly. Third, that the output of the analytic process should be presented in terms of a series of performance measures over the project time-horizon (15 years). This recommendation developed from the belief that the influence of contingencies on a series of measures such as cash flow and sales projections and NPV measures, would give valuable input to the process of strategy choice. The performance measures suggested were: - Total $ sales Cash flow profile - Gross profit as a percentage of sales - Net profit after taxes - Net profit as a return on investment - Net present value - Further sensitivity analysis of results to key changes in input factors. Subsequent Policy Dialogue Management reviewed the subsequent output of the risk analysis which showed that Strategy 1 (super-fast) was still dominant, except under adverse marketing and financial scenarios, when Strategy 2 would be preferred. Interestingly, it was also determined (from the marketing scenarios), that CCA had a competitive strength with the plastic carton. As a result, if it adopted an even more aggressive aos capacity—building strategy, it would be able to obtain both the domi- nant market share position, and hence the lowest relative cost position. This information provided an additional interesting discussion topic. The management group commented upon the usefulness of the marketing and financial (e.g., cash flow) probabilistic forecasts provided by the “second-pass”" risk simulation. Quite apart from better understanding the influence of uncertain events on their business activities, they felt that it gave them an insight into the "spin-off" from this project to the firm's entire portfolio. As the project's NPV (for Strategies I and II) was positive when discounted at the required risk adjusted rate (determined from modern capital-asset pricing theories), then the project should certainly merit acceptance in portfolio terms. This was reinforced by the stra- tegic aspects of the risk simulation, which enabled management to judge the project's viability under a range of alternative future scenarios. Indeed, managers commented that the confirmatory positive indications of the set of performance measures under the range of scenarios, enabled them to better understand the project. Whilst they were now generally in favor of project acceptance, they wanted to examine CCA's goals in further detail. First, they wanted to ensure that this project was compatible with the firm's long-term growth plans and objectives. Second, in order to develop this strategic thinking process, they decided to simulate the firm's portfolio over a five to ten years horizon in terms of profitability and cash flow objec tives. By so doing, they hoped to anticipate other business areas with -20- future potential for decline/growth, and assess cash flow and financial implications for the Egg'N Foam project. This subsequent portfolio analysis showed sound long-term viability, and gave strength to management's view that the growth into cartons was a useful related-business ie ereteicacioon It was felt that it would be complementary in skills and growth potential to the firm's recent diversifications into the poultry processing and broiler chicken manu- facturing areas. Summary, Conclusions, Strategic Implication The value of risk analysis in a facilities planning decision is examined here in relation to a disguised recent planning situation called Egg'N Foam. In essence, following Gluck et al [3], risk analy- sis is presented as Stage II/III of a sensible strategic thinking process. That is, as a vehicle for forecast/uncertainty based planning, in which an understanding of future scenarios, cash flow projections, and synergies between marketing and production activities is developed. This also enables managers to search more creatively in order to iden- tify the menu of strategic options. The strategic risk analysis presented here differs from that pre- sented by some other authors in the same area [e.g., Spetzler and Zamora [6]; Buffa [2]] in terms of its treatment of risk preference and criteria for strategy choice. Risk preference involves the decision- maker's attitude towards risk, and is commonly handled via utility function assessment and certainty equivalence concepts. In this instance, the management group felt uncomfortable with the utility - 21- concept, even though they regarded risk preference as a policy question. They preferred to treat risk preference through a number of “lenses,” i.e., by examining the risk simulation output in mean/variance terms, in terms of "risk of ruin" criteria (i.e., prob NPV > 0), and by looking at future cash flow profiles. In addition, they accepted that a project with a positive NPV, when discounted at a “risk-adjusted” rate, would increase the value of the firm in portfolio terms. Management also felt that for strategy choice, both the project and the firm's portfolio should be examined in terms of a number of perfor- Mance measures (cash flow, NPV, sales) specified over the fifteen-year project time horizon. This criterion of policy choice, specified in terms of a time stream of indicator measures (cash flow etc.) rather than in terms of a single criterion (such as expected utility), is con- sistent with the work on preferences over time (see Bell [1] and Meyer [5]). In terms of strategic implementation, this risk analysis approach was successful. This was because of the value of risk analysis in encouraging policy dialogue amongst the management group about future uncertainty impacts. The continued questioning of assumptions and problem formulation is seen as essential for the effective formulation and evaluation of alternative strategy positions. By such continued questioning and dialogue, a meaningful consensus about strategy choice emerged. This consensus process, which involved “three passes" of a risk analysis process, could not have been achieved without adaptive mechanisms and flexibility being built into the process during the course of dialogue and successive risk analysis passes. =99= REFERENCES Bell, D. E., (1977). “A Decision Analysis of Objectives for a Forest Pest Problem,” in Conflicting Objectives in Decisions (1977); Bell, D. E., Keeney, R. L., and Raiffa, H. (eds.). New York: John Wiley. Buffa, E. S. (1980). Modern Production/Operations Management, 6th ed., pe. 108. New York: John Wiley. Gluck, F. W., Kaufman, S. P., and Walleck, A. S. “Strategic Manage- ment for Competitive Advantage," Harvard Business Review, July/August 1980, pp. 154-161. Hertz, D. B., and Thomas, H. Risk Analysis and its Applications. Chichester: John Wiley (forthcoming). Meyer, R. F., (1976). “Preferences Over Time,” Chapter 9 in R. L. Keeney and H. Raiffa, (1976) Decisions with Multiple Objectives, New York: John Wiley. Spetzler, C. S., and Zamora, R. M., (1974). “Decision Analysis of a Facilities Investment and Expansion Problem," in Readings in Decision Analysis. Palo Alto: Stanford Research Institute. D/73 Exhibit 1 FINAL FLOW CHART EGG'N FOAM | ene on TOTAL US EGG FAOOUCTION WTITRREGIONAL EGG SHIPMENTS —ccs fF | ff A OEMARD Peet ».4 FOLM SHARE =| FOAM MARKET | lea SHARE = _ = & CAPACITY Now Carton Use of Ean Murder of Eatructer les Murterc! Cree low t~ Catraterr = Rate fe emote an Tree + Fawnet Punterty Gon == 1 { ! 1 | 1 1 | ! I I I ! 1 I | I 1 ! | I t 1 oe wre et es es ee ee ie ee ae eee ee ae ee ee ae ee eee ae ee —, — ee ee ee ee ee Teteg apn ond Telepree Espeeimental Cares Dues art Sutacnanes Cus rear Batarary —_—— parser —_ Tote! Soles Corp bites Expene Total Varaule Cameron Cost Trawet ast Ertenaaren: Sela Renayement Salas Salamen’s Sotanes Torat Cantemnng Proxtucton Cost Paeopng * Mocnene rae | Alin ince. ————— To: at Genera! ond Ad mrestrotive Exe le ae 2 ee eee Torat Sates Evpeme aan ane are eee — onl j cot CROSS € GN FOAM " anes R TOTAL on cca i 1. AX RATE SA. ES VOUUME »,4 CCA PRICE j = REVENUE —— f MANUFACTURING =- { Lee ING OTe ce: preraxenosir! xX aaeecerd =e | Tote Continuing | Productoon Cont | Toto Guect Cow | Total Varserte Crmwerton Cost The revi former Varwote Cour Mors Amor ranon 4 Wnts | -io- ee FINAL FLOW CHART EXHIBIT | EGGN FOAM CCA AFTER TAX PROFIT RETURN OM INVESTMENT INVESTMENT aS ee erence A — MAJOR ASSUMPTIONS UNDERLYING EGG’ N FOAM PROJECTIONS | ce 1. . Foam cartons will sell at the Pillopost price level 1 ‘ 2. Ww EXHIBIT i PRICE Pillopost prices are expected to decline over the years Market volume for foam cartons will drop sharply at prices above Pillopost level Diamond National will be the price leader; price competition will be moderate Price-volume discount schedule is unlikely to change significantly over the next decade No transfer price for foam is considered SELLING EXPENSES , att Egg N Foam distribution will be handied by the Plastics Division’s Marketing Department 2. Selling costs include: — Unrecovered charges (design and plates, speculative art work, customer machinery) — Supplies — Dues and subscriptions — Advertising and promotion 3. There will be one-time start-up cost | MARKET SHARE | . The market under consideration is the potential number of foam egg cartons sold . CCA will be able to produce a high-quality foam carton which is acceptable to customers and will close satisfactorily on existing egg packers machines 3. Product characteristics of Egg’N Foam will be competitive to Pillopost 4. CCA can successfully sell foam egg cartons through existing channels of distribution (to each type of customer in the market) 5. CCA's sales volume will be penalized by building plants in later years . CCA will sell out a plant’s capacity regardless of when the plant is built 7. CCA will have captured 10 percent of the assumed carton market by 1978 (360 million cartons). . No additional carton suopliers are expected to become a significant factor in the eag carton market in the next 10 years | through existing channels of distribution in four regions (West Coast is excluded) \ . Colored foam will not be a factor in the market before 1978 . CCA will equalize freight paid by customers when foam cartons ave shipped into other regions . Significant cost savings can be achieved by building super plants 1 2 3. 4 5 6 . No major changes will take place in the manufacturing concept In an Egg'N Foam plant no other products than foam egg cartons will oe produced . CCA will produce its own foam sheets . CCA foam will have uniform quality and density characteristics . Through-put rates of the extruder and thermoformer will improve over time due to advancements in design-and manufacturing knowledge Inflation in future years has been excluded since it should affect both revenues and costs SIZE OF CARTON MARKET 1. U.S.D.A. statistics on egg production by state are correct ( 2. Estimates from the editor of the Poultry Tribune are correct on eggs cartoned as percentage of total egg production 3. Statistics of U.S. Economics Corporation are correct 4. ivial market is divided into five regions determined by: — Probabie organization of CCA field sales force — Size of areas where a 19 percent share would be the equivalent of a small size plant oO Foam will have better merchandising appeal than pulp or paper and therefore ; f aa (i) 2 a an o = | 2 tAin rtons over tha nanyt ' © t inna an Mabry Ua eens Foun A =" ie) =, “< oD o bal a) 6. By 1978, share of market by carton type will have changed as follows: — Paperboa:d from 45%to 5% — Pulp from 53% to 50% i | Ca-—_— f--— AM en ACOFL | CONTINUING COSTS 1. Straight-line depreciation is used for calculating the return on investment. Double declining balances are used for calculating depreciation for tax pusposes and the discounted cash flows i: 2. Life of the equipment is equivalent to standards allowed by the Internal Revenue Service — Building 30 years — Production equipment 7 years — Trucks 5 years EXHIBIT tl FIXED INPUT DATA — EGG’N FOAM RISK ANALYSIS MODEL YIELD PERCENTAGES Extruder garbage loss: 1.0% Thermoformer mold loss: 1.0% Printer loss: 2.5% Extruder reclaimable scrap: 7.0% Thermoformer buttends loss: 1.0% Printer reclaimable scrap: Thermoformer reclaimable scrap: 13.0% EFFICIENCIES - EXTRUDER THERMOFORMER Mechanical efficiency £0% 90% Scheduling efficiency 95% 100% | MATERIAL PRICES | Foam nellets: $0,15/Ib. | Supplies: $1.01/thousand cartons MANUFACTURING COSTS* ONE EXTRUNER TWO EXTRUDERS t Evt der wairanle nacte os es SeolaG | Thermoformer variatie costs ZZ.2) 38.94 Continuing costs : $237,350 $393,600 * The cost for three thermoformers is equal to the cost for one plus the cost for two. The cost for four thermoformers is equal to twice the cost of two. ‘ | MANUFACTURING PROCESS Extruders will work ona four-shift basis (7-day week); thermoformers on a three-shift basis (5-day week) EQUIPMENT SPEEDS AND | | SPECIFICATIONS 1968 1969 1970 1971 1972 1973 1974 $26.00 5. $24.50 $24.00 Price/thousand cartons 2 $26.90 $26.90 $26.80 $26.70 $26.60 Extruder Rate (lb./hr.) 420-600 A 520-700 Caliper (inches) .080-.087 .078-.087 Thermoformer Cycle Time” econ WS Ax sce | 6.030 * Thermoformers will use a twelve-up mold EGG CARTON MARKET 1968 1978 3.5 billion cartons 4.7 billion cartons Continuing costs include: Direct costs include: Selling costs include: A small plant sales force includes: — Plant overhead — Material — — Sales force salaries -- 1/2 sales manager — Office expenses — Labor — Payroll taxes, employee — 3 salesmen a — Property tax — Supplies benefits — 2 packaging machinery service = izati — Utiliti — Travel and entertainment Mold Sree Stn Utilities (25% of salaries) A large plant sales force would — Depreciation — Maintenance 3 include: — Other production — Telephone and telegraph A enlesumariaaar expenses — Office rent nee — Unrecovered charges aS ae eee : 2 | — Supplies ey est y service — Dues and subscriptions CHART 1 EXTRUDER RATE* Probability i¢) 400 Pounds/hour CHART 3 CALIPER* Probability 100% aimee tl SR IS 5 mee A 75 1971-75 50 25 0.030 0.086 0.082 Inches PROBABILITIES ASSIGNED TO MAJOR INPUT FACTORS LEGEND CHART 2 THERMOFORMER CYCLE TIME* Probability oOo ro) ——— oO ies) 3 “nh N = Ses ee ee ee i 3 25 \: E i eS t \ % | Tonalin fe) Sched ! > 8 7 6 5 4 3 Seconds CHART 4 PRICE* fi a ea Dollars * The probability distributions for these input factors are the same for each of the four strategies EXtiBit tv G Represents expected value rn — SSN ee SYOLCIVO TLOoOW CUTTS! EX tia (7 gS Fok YEAR & —_—_—_—_ Nuwiber a) Type t& Sxtrsion Machuun Extrsien Rate 3 Machine 1. Extrusion Rate 4 Madune 2. Annval Extrusion Gapaaty pe ceaea Demitund Which is Greater ? Plant CapactyfP—] lubich 9 ? Dadual “Aero fre Goaaty Girone 4 Material Cost pee canton Fixed Costs Say GY" een fe} Ray Gok Pn DIScouNT " Is : [| Caleulahen J ev ee oT | Input 4 Vari able [] Inputa (strategy dependent) constant 7 Input- 4 ( strotery vrdependont) Cowrtamt Ablenrx: VARIABLE INPUT DATA — EGG'N FOAM RISK ANALYSIS MODEL STRATEGY: SUPER FAST PLANT LOCATION: Nesquehoning Working Selling Mean Standard Building Equipment Trucks Capital No. Expenses Demand Deviation No. Thermo- ($000) Extruders formers ($000) Carton Millions 1 20 1 70.0 22 30 2 0 120 0 1 121.7 41 30 3 720 701 0 2 3% | 189.0 86 30 4 O 0 0 2 4 204.1 134 30 5 0 0 0 0 2 4 204.1 180 20 6 0 ee 8 ) 0 2 4 204.1 | 180 30 7 (6) fe) 0 0 2 4 H 204.1 |} 180 ia) | 8 0 ) 0 0 2 4 204.1 180 | 30 | 9 0 0 0 0 2 4 204.1 180 20 i0 0 fr) ) 0 2 4 204.1 160 | 30 | mt | o | au | No 0 2 pat femmenr Hamme Ime 0S 2 j wh o WI 0 # | 0 Ora a ee: f 204.1 “180 | 2 { | 13 | o 0 0 Demers ee? | 4 204.1 180 30 | 0 0 0 @ : 2 4 204.1 120 | ca) | 0 ) 0 ) oe | 4 204.1 iso «6| la | aap HE = | | PLANT LOCATION: Cairo Working TT. | Seiling Mean | stencard | Canital Ree aaa Expenses Demand | Deviation Exiruders formers (S209) Carton Millicns rg 0 0 30 pat 0 30 % % 70.0 30 | ) 1% 3 180.0 54 30 0 0 ) 2 4 204.1 115 30 0 0 “9 ) 2 4 204.1 139 30 G 0 0 0 2 4 204.1 180 30 0 O° 0 0 2 4 204.1 180 30 0 0 0 0 2 4 204.1 180 30 0 oF J =.0 0 2 4 204.1 180 30 0 0 0 o 2 4 204.1 |. 180 20 - 0 0 0 0 “2 4 204.1 180 30 0 0 0 0 2 4 204.1 180 30 0 0 0 0 2 4 204.1 180 30 ) ) ) () 2 4 204.1 180 — 30 VARIABLE INPUT DATA — EGG'N FOAM RISK ANALYSIS MODEL STRATEGY: SUPER SLOW pant vocation: Aesquehaning Working s Selling Mean Standard Building Equipment Trucks Capiial No. Expenses Demand Deviation No. Thermo- Year 00) Extruders formers ($000) Carton Millions 1 2 0 10) 3 0 10} 4 720 0 5 0) rt) 4 H G 0 9) O2F oI ° 2 4 204.1 | fog at 10 7 0 (@) 0 0 2 4 204.1 180 10 8 (a) (@) 0 10] 2 4 204.1 | 180 10 S$ | Oo | 9) ) 0) | 2 4 204.1 | 1329 1G 19 : a ot =r! a | 2 : 2 I A to onay | son | any } | i | | H | | ji | | | $1 (a) 0 0 0 2 4 } 204.1 | 1280 | 10 12 Oo 0 Q 0 2 4 204.1 180 10 13 0 0 0 Cae eae | ee 204.1 180 10 | 14 ie} 0 0 0 | 2 4 204.1 189 10 } id i 6 | 0 cs) G | 2 4 204.1 isd id i PLANT LOCATION: Cairo Working Buildina Equioment Trucks Capital No Year (S000) Extruders Selling Mean Standard Exnenses Demend Deviation Carton Miliions 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 N oooo oO h oO cS) - ©Oo Oo oO _ ooooeoeoooeoe90o0d 9 1 2 3 4 5 6 7 8 9 _— ~ Oo (os © © © ee © ee © ee © ee =) oo) S3o..0 70 oo ee ee ee ob W N NO NN NY N NY NY NY DN — pb bh b> & bb bh BB ae ; VARIABLE INPUT DATA — EGG'N FOAM RISK ANALYSIS MODEL STRATEGY: SMALL FAST PLANT Location: Nesquehoning Working s Satiing Buiidiag Equloment Trucks Capital | No. Expanses No. Thorrro- } (€000) Moan Standard Denuind Deviation Extruders formers ($000) Caron M hoes 1 720 G71 20 250.0 1 70.0 21 21 2 0) 120 0) 201.2 1 1% 1a ee 21 3 fv) () ) () 1 2 121.7 57 21 4 ) ) fe) O. 1 2 121.7 30 21 5 f°) 0 fc) ft) 1 2 121.7 90 21 6 ) ) 0 0) 1 2 121.7 90 21 7 0 0 0 0 1 2 121.7 so 21 8 ) fc) ) 0 1 2 121.7 90 21 9 0) 0 ) 0. 1 2 121.7 90 21 10 0) fc) fc) 0) 1 2n° 121.7 90 21 11 fe) fe) fc) ) 1 2: 121.7 90 21 ip l2 (0) 0) fc) 6) 1 2 121.7 90 21 13 ) 0 Q ) 1 2 121.7 90 21 14 rc) QO 0) 0) 1 2 121.7 80 21 15 ) a) 0 0) 1 2 121.7 90 2 PLANT iOcation:, Southoast ——— RR ERT I RES RT PS SF EET BE OR ES RE 1 9) 0) 0) ) fe) 0) ) 0) 21 2 0 fe) 0) 0 0) 0 ) 0) 21 3 720 791 20 4512 1 % 100.0 27 21 4 | ) | Q a) Q | 1 H 2 124,72, 62 21 ‘ 5 H rv) : U I (6) ‘ O i 1 ! 2 fh Wei? 8 iste} | 21 ! ' G | Oo 0) | o | 0) | 1 2 J 121.7 | 90 | 21 H 7 o } C | ) 0, 1 .2 121.7 $0 21 | 8 0) ) 0 f°) 1 2 121.7 90 21 3 | a) | a) fe) fe) i 2 121.7 ole) 21 10 ) ) 0 ft) 1 2 121.7 co 21 4 ) fs) ) ) 1 ‘ 2 121.7 so 21 12 0 0) 0) ) 1 2 121.7 30 21 + 42 3 Q ql S 1 2 oth 53 ; % ; 14 0 0 0 0 | 1 2 | | | 15 0) ) 0) a) 1 2 i H ns ee ee neers 2 D PLANT LOCATION: Adidivest ~“ N cos) No * cooooaoo0o0oo°0ce000o oooo0oo0o0Ccoo0co°o ee ee ee NNNNNNNNNNN AN PLANT LOCATION: Southwest (6) - 0 (0) 0) 0) 0 0 (e) (e) 0 0) (6) (0) (0) (0) (0) 720 791 20 451.2 1 1% 0 10) (e) fe) 1 2 0 (0) fe) fe) 1 2 (0) 0) ) (a) 1 2 rf) (a) 0 0 q 2 ) (a) 0 0 1 2 ) (0) 0 0 1 2 0 0 (0) 0 1 2 ) (0) ry) () 1 2 (0) 0 0 (a) 1 2 te) 0 0 (a) 1 2 oO ro) fe) 0 4 ise) A-3 en 7 VARIABLE INPUT DATA — EGG'N FOAM RISK ANALYSIS MODEL A-4 STRATEGY: SMALL SLOW VeOrTkiNg = Seltin | Mean Stendard | Capital No. | Bunce + Demand Deviation No. Thermo- , Yeor ¢ Extruders formers ($090) | Carton hiillions I 3 1 720 ie} | 1 1 7 2 0 10} 0 50.0 1 1 7 3 ie) fe) 0 : 1 2 7 4 0) 0 ie) 0: 1 2 7 5 fe) fe) e) (0) 1 2 7 6 0 0 0 (e) 1 2 7 yi 0 ie) (0) ie) 1 2 7 8 0 0 te) 0 1 2 7, 9 0 ie) 10] oO. 1 2 7 10 ie) 0 0 0 1 2 7 H 11 QO 0 ce) ce) 1 on 7 12 fe) fe) ie) 0 1 2 a M3 ie) 0 0 QO 1 2 7 4 ie) fe) 0 0 1 2 Z 15 0) 10) 0 (0) 1 2 7 PLANT LOCATICN: a 1 Bs 0 0 0 0 { 0 0 0 7 j ‘2 0 0 0 ) 0 ) 0 7 3 5 ft) ) r) f) 9) 0 0 7 | H 4 720 671 20 300.0 % % | 70.0 7 ee t 5 ! o | 120 | 6 ! 131.2 ' 4 t z { 12107 H 3+ ' 7 t b 10) U U U 1 Zz iZi./ fore) ‘ 7 q o o a }, | i | "2 | i247 cowl ea | 8 rt) 0 re) te) 1 | 2 121.7 so 7 | | 9 9 0 Q ) 1 2 421.7 90 7 10 0 0 ) ) 1 2 121.7 90 7 | 11 0 A) 0 0) 1 2 121.7 90 7 12 i) ) ) 0 1 3 124.7 | 30 | 7 ne ! 0 0G ao 0 | 1 { 2 1 o4o42 | ary | Sy t 14 0 0 0 nt) 1 2 1207h a). sOOha ak ee. 15 r) r) 0 r) 1 2 121.7 | S000 ye "a7 L d ve i : PLANT LOCATION: Aticivest ie) QO ie) G ie) 9 te) 0 70.0 16 " 100.0 50 ocooooocoocooco0co0co[o0o000 = as SEE ee OS ee a — ee wee woe ot oO OO OC O NNNNNNNNN|-|OO00 NNW NNN NNN NNN NYS PLANT LOCATION: Southwest oooo°co BoeteRoRoRonokononon om ono kok a) POoOooocoocooeoceoceoeo0o0°o SNNNNNNN NIN SN PNNMNNNNN PTS | AGA Re IO a eh : WENO Be VPA AOS Sty - t : WOE OOM STAR | P ony thw Pern wy, vein te | ’ "eines fone) a ome Se eb) Sb : i at | ar, } ee ee Aina aye mm nt ee AHA MN MOA : Sac z , & rs & = wet aim ee ne ae > be Be etek ; 10 i 4 i 1 OG, { j '7 3 ne . ¢ rt 43 | — t | id | Oe. Y & 4 i c¢ 7 a a i] ent ne Sa ee fa ie ao PLANT TOCATY ‘ ) uae eee ' ad ’ 6 ' I | 8 t “es ~ 0 qe PR s : 5 rj : f fy yh. 5 - : . | | oJ? , ; 5 ' ts 4 on , , iY Y i ow be - t ‘Oe » ; va ' | wo. on i | : i] At. out p 7. (| .. a | 3 « en mn. ' re Pai TP lo x on ' ' , =. RARSRoRgeCEel vasasce= > —— Ci 3-4 b ty = y ri _ - em yy mee i terme any ie Tila yin 4 tis al , | a if a 1- ¢ o oe hat | : aioe ea , | 0 +] | Med oe vt { ny i : Ti : | 7) “¢ Y ¢ ' } | of f of a & } z, g° rt 12 + i ad ( 0 | ‘ 9 7 | ey i > ii a ¢g ¢ a i by ! * | i ° a { a, ae) ¢ . f a « ; } ‘ wy cv } '¢ of ‘ at i iy ote f i ¢ } LP JUN 95 VERSITY OF ILLINOIS-URBANA Do in 3 0112 060296 sy tere A(Sereh is Hevtite al WORE % § rier Ponca Cp aitee aR eM Rit anny Jetpdan ah ath