Seagate Technologies: Operation Hedging
The Seagate Technologies as a group assignment in our Production Logistics course. The purpose and aim of this case it to learn the impact of each asset’s (location) capacity on the overall profitability of the processing network. In addition, investigate how the entire “capacity portfolio” can be designed to provide an optimal hedge against uncertainty.
We have been following six questions: i) what is Seagate’s corporate strategy? Describe and evaluate how its operational strategy and processes support the corporate strategy. Critically evaluate Seagate’s product and process development strategy, which calls for development in its respective product / process centre in U.S. and then exporting the developed process to site in the Far-East for high-volume production. ii) What are Seagate’s major risks? How does it manage those risks? iii) How would you describe the ‘capacity of the processing network’ if the current CAR capacity proposal were implemented? What is the expected profit and ROI under this investment? (Given the short product life, assume the firm is making its decision for a single time period of length one year, at the end of which manufacturing capacity will zero salvage value). iv) The case states that the true demand forecast contains uncertainty. Given this forecast, recommend a capacity portfolio that maximizes expected NPV. (Recall, capacity investment must be performed before you observe actual market demand). Verify financial attractiveness of your recommendation. What is the expected profit and ROI now? v) Interpret your recommended capacity portfolio in intuitive terms: in what sense does your capacity configuration prepare you to ‘hedging’ and why is your plan to be preferred? vi) In broad conceptual terms, what are the advantages of ‘sales-plan driven capacity planning?’ What is ‘wrong’ with that practice end how would you improve on it?
LITERATURE REVIEW
Operations Management:
In operations management, there are two streams of research originating from two separate, but conceptually similar, definitions of operational hedging. The first definition, as introduced by Huchzermeier (1991) and quoted in Ding and Kouvelis (2001, p.2), states that “Operational hedging strategies … can be viewed as real (compound) options that are exercised in response to demand, price and exchange rate contingencies faced by firms in a global supply chain context.”
Real options might have value-enhancing capabilities under uncertainty. The value-enhancing feature of real options under uncertainty is called “exploiting uncertainty.”
Huchzermeier and Cohen (1996) analyze operational flexibility, which they define as the ability to switch among different global manufacturing strategy options. Cohen and Huchzemeier (1999) illustrate how the deployment of excess capacity can be a source of operational flexibility in global supply chains. They argue that investing in capacity in excess of the aggregate demand forecast provides flexibility in coping with demand uncertainties. Additionally, excess capacity enables the firm to produce more in that location, providing a value-enhancing opportunity in addition to reducing its downside risks.
The second definition of operational hedging is found in Van Mieghem (2003). Without referring to real options, but making an analogy with its financial counterpart, financial hedging, Van Mieghem defines operational hedging as “mitigating risk by counterbalancing actions in a processing network that do not involve financial instruments.” He lists dual-sourcing, component commonality, having the option to run overtime, dynamic substitution, routing, transshipping, or shifting processing among different types of capital, locations or subcontractors, holding safety stocks and purchasing warranty guarantees as operational hedging strategies.
One of the main contributions of this definition is the observation that operational hedging can be employed in the absence of tradable risks, particularly exchange rate risk, as all the other academic fields mostly consider operational hedging in an exchange rate framework. Again departing from the literature, Van Mieghem does not consider any particular risk measure to formalize the effect of operational hedging in terms of risk mitigation. In addition, the term “counterbalancing actions” is not formalized: criteria to determine whether given actions are counterbalancing are not developed, this term corresponds to investing in more than one resource, or “betting on two horses” that is, investing in operational flexibility, similar to the former definition of operational hedging.
Finally, as with real options, counterbalancing actions described by Van Mieghem have a value-enhancing capability and increase expected profit in a risk-neutral setting. This is demonstrated on a two-product, two-stage production system where capacity imbalance is the operational hedging strategy (Harrison and Van Mieghem 1999, Van Mieghem 2003).
By purposely unbalancing the capacity vector, i.e. having safety capacity (in excess of the capacity that would be optimal in the deterministic case), firms can hedge against demand uncertainty and increase expected profit. Counterbalancing actions, taken in such a way as to maximize expected profit for a risk-neutral decision maker, are called operational hedges.
Finance
In the finance literature, operational hedging is the course of action that hedges the firm’s risk exposure by means of non-financial instruments, particularly through operational activities.
Similar to the operations management literature, operational flexibility is the major operational hedging strategy discussed in the finance literature.
In addition to operational flexibility, geographical diversification is another operational hedging strategy in a multinational context. Geographical diversification is aligning the costs and revenues of a firm so that they are exposed to the same risks. Domestic firms selling to foreign markets can ensure that their production costs and sales revenues are exposed to the same exchange rate uncertainties by opening a production facility in these markets. Therefore, geographical diversification reduces the total variability of cash flows.
Chowdry and Howe (1999) argue that the facility location decision is considered to be an operational hedging strategy only when firms are concerned with the variability of their operating profits.
Hommel (2003) argues that operational flexibility is employed as a hedging device when the exchange rate and demand volatility are sufficiently large (in that case the minimum profit constraint is violated); otherwise it serves as a value driver to enhance expected profits.
It is emphasized that because operational flexibility can be used for a purely value-enhancement motive, it is considered to be an operational hedging strategy only when there is a risk hedging motive for employing it. Generally speaking, operational actions are considered to be operational hedges if they are taken in order to reduce a risk measure of concern. In particular, if firms care about downside risk (e.g. having a minimum profit constraint), then operational hedges mitigate risk through a reduction in the downside exposure.
In summary, the finance literature defines operational hedging as mitigating firms’ risks by operational means. Operational flexibility achieved through various operational means (ability to shift production, transferring technologies, product differentiation etc.) and geographical diversification is the operational hedges of firms utilized in conjunction with financial hedges. Compared to their financial counterparts, operational hedges require higher levels of capital investment (opening a production facility), but create longer term hedges against risk exposures including risks that are not contingent on asset prices (e.g. demand risks, political risks).
Strategy and International Business
Diversification is defined as having different lines of business through mergers and joint ventures (Wang and Lim 2003), of which geographical diversification is one type.
Kogut (1985) analyzes diversification and operational flexibility as risk management tools of multinationals. He argues that an operational decision (the sourcing policy in this case) can create three different types of risk profile: speculative, hedged and flexible. The speculative profile is betting on one site mainly to benefit from economies of scale in operations. By matching the exchange rate exposure on the cost side with that on the profit side, the firm can create a hedged risk profile. This approach corresponds to the geographical diversification strategy. Finally, a flexible risk profile created through operational flexibility permits the firm to exploit uncertainties by creating real options.
In the international business literature, Pantzalis et al. (2001) define operational hedging as the firm’s operational decisions (related to marketing, production, sourcing, plant location, and treasury) that are best suited to managing the exchange rate exposure on the firm’s competitive position across markets.
In summary, the strategy literature focuses on operational flexibility and diversification as risk management tools without defining them as operational hedges. Operational flexibility achieved through several operational means (developing in-house capacity, product differentiation, keeping excess capacity etc.) creates both arbitrage and leverage opportunities for multinational firms. In addition to aligning costs and revenues, real option benefits of geographical diversification in the form of growth options are discussed. The international business research, similar to the finance literature, focuses on operational flexibility and geographical diversification as long-term operational hedges of multinationals against exchange rate exposures.
Analysis of Seagate Technologies:
Question 1:
What is Seagate’s corporate strategy?
Describe and evaluate how its operational strategy and processes support the corporate strategy.
Critically evaluate Seagate’s product and process development strategy, which calls for development in its respective product / process centre in U.S. and then exporting the developed process to site in the Far-East for high-volume production.
Answer:
Seagate Technologies Corporate Strategy:
‘The Barracuda’ & ‘The Cheetah’ are two new Seagate’s high-end disk-drive products families that are scheduled to go into volume production in the first calendar quarter of 1998. The capital appropriation request called for $ 103 million capital investment in two final assembly facilities, one for the Barracuda and one for the Cheetah and one for joint test facility. The company’s establishment and ongoing expansion of production facilities in Singapore, Thailand, Malaysia, China and Ireland are directed toward cost reduction.
Describe and evaluate how its operational strategy and processes support the corporate strategy.
Operational Strategy and Processes:
Manufacturing Strategy:
Process Choice – Establishment and maintenance of key vendor relationships. Produce and sell its disc drives in significant volume, continue to lower manufacturing costs and carefully monitor inventory levels. Transfer volume production of disc drives and related components between facilities, including transfer overseas to countries where labor costs and other manufacturing costs are significantly lower than in the U.S.
Infrastructure – The key element if the Seagate’s manufacturing strategy is high volume, low cost assembly and test; vertical integration in the manufacturing of selected components. Seagate continually evaluates its components and manufacturing processes. Seagate rapidly achieve high manufacturing yields in new production processes and obtain uninterrupted access to high quality components in required volumes at competitive prices.
Marketing Strategy:
Seagate’s ability to compete successfully depends on its ability to provide timely product introductions and to combine to reduce production costs. The company’s establishment and ongoing expansion of production facilities in Singapore, Thailand, Malaysia, China and Ireland are directed toward such cost reductions. The two new products were planned to be in volume production only for the four quarters of 1998. The capital investment to build production capacity was significant and had two components. First, there were significant fixed costs – estimated at about $ 40 million – associated with designing, commissioning, and starting up the three new facilities. The second component was that the capital expense of building new capacity increased with the amount of capacity: larger production capacity required larger space requirements and tooling costs, leading to an linear increase in the capital expense.
Seagate’ products include over 50 rigid disc drive models with from factors 2.5 to 5.25 inches and capacities from 1GB 10 23 GB. Seagate believes it offers the broadest range from of disc storage products available. It provides more than one product at some capacity points and differentiates products on a price / performance and form factor basis. Seagate typically devotes its resource to developing products with industries leading performance characteristics and to being among the first to introduce such products to market. The company continuously seeks to enhance its market presence in emerging segments of the rigid disc drive market by drawing on its established capabilities in high volume, low cost productions.
The Marathon and Medalist disk drive product lines are targeted for the personal mobile and desktop computing market, respectively, while the high end workstation and server/multi user systems market is served with the Barracuda, Cheetah, and Elite product families.
The Barracuda family of 3.5 inch drives was first introduced in 1992. At 7,200 rpm the Barracuda had the highest rotation speed of any drives produced at that time. In fiscal year 1997, Seagate introduced two new products in the Barracuda family, the Barracuda 4LP and the Barracuda 4XL, with 4GB and 4.5GB respectively. The Barracuda 4XL, which began volume production during the fourth quarter of fiscal 1997, was designed to provide a balance of price and performance for the workstation market as it matures.
In August 1996, the company announced the 3.5 inch Cheetah family – the world’s first drives to offer rotation speeds of 10,000 rpm for increased data throughout and lower latency times. The Cheetah drive is focused at the very high performance segment of the market. Volume production of the Cheetah 4LP and the Cheetah 9 began in the third and fourth quarters of fiscal 1997, respectively. Seagate is going to announce the fifth generation Barracuda 9LP and the second generation Cheetah 0LP in early fall 1997, with volume production schedule to begin in the first calendar quarter of 1998.
Finally, the Elite product line covers the high end 5.25 inch market. In the third quarter of fiscal year 1997, production commenced on the Elite 23, a high performance, 5.25 inch disc drive with 23 GB of formatted capacity, a rotation speed of 5,400 rpm and mean time between failures of 500,000 hours.
Critically evaluate Seagate’s product and process development strategy, which calls for development in its respective product / process centre in U.S. and then exporting the developed process to site in the Far-East for high-volume production.
Product and Process Development:
The content of the Seagate product/ process strategy:
Seagate has the superior strategy, i.e. business strategy or corporate strategy, requirements on the product portfolio.
It is described in the case the present state of the product portfolio.
It is also described in the case what would be the future state of the product portfolio.
A plan of action, i.e. how Seagate wanted product portfolio can be reached in practice.
The five steps/activities are described below:
Requirements on the product portfolio:
The most central activity in the process is the identification of the requirements on the product portfolio. The requirements should be found both in the superior strategy, i.e. business strategy or corporate strategy, and also in other functional strategies. Requirements put on the product portfolio consist of among other range, mix and volumes of products. Seagate has number of segments which is introduced in the market.
New product proposals:
Ideas for new products can arise in different ways; customer, market analysis etc. The new product proposal capture, visualize and preserve the ideas that are found within and outside the company. The aim of the new product proposals is to attain a more distinct product development funnel as shown in Figure 12, where several ideas are evaluated in parallel. Seagate’s strategy for new products emphasizes developing and introducing on timely and cost effective basis products that offer functionality an d performance equal to or better than competitive product offering.
Product development process:
The product development process should fit the actual company, its products and its manufacturing.
The product development process should also state which design method to use when and why. Seagate devotes significant resources to product engineering aimed at improving manufacturing processes, lowering manufacturing cost and increasing volume production. Seagate’s process engineering groups are located with the disc drive development groups and the reliability engineering groups in many cities of U.S. and also in Singapore.
Product portfolio:
When making decisions within the product development process it is important to have the product portfolio in mind and vice versa. Therefore it is emphasized that the same group of managers handles both the product development process and the product portfolio.
Reengineering or product deleting:
All products have a limited life span. Not unusual at companies aimed at in this research is some kind of facelifts of products during their lifetime. New requirements like new features, manufacturing processes, customer needs etc. on a product or product family require a reengineering or the product will be obsolete. Seagate’s product life cycles of disc drives are short (high volume products introduced and sold for about 6 to 7 quarters. Due to fast changing in technology in computer industries the product deleting is very short and re-engineering might cost extra money to Seagate due to rapid development.
Question 2:
What are Seagate’s major risks? How does it manage those risks?
Answer:
Competitive differentiation: (e.g., price, quality, time or customization)
Market: Fundamental change in supply and demand functions or global prices for commodities.
The rigid disc drive industry is intensely competitive, with manufactures competing for a limited number of major customers. In addition to the product performance dimension, the principal competitive factors in the rigid disc drive market include product quality and reliability, form factor, price per unit, price per megabyte, production volume capability and responsiveness to customers. The relative importance of these factors varies with different customers and for different products. Competitors offer new and existing products at prices necessary to gain or retain market share and customers. To remain competitive, Seagate believes it will be necessary to continue to reduce its price and aggressively enhance its product offering.
Technological capabilities (lead or follow in technology innovation)
With the proliferation of multimedia applications, the demand for increased drive capacities has and continues to increase at an accelerating rate since sound and moving pictures require many times the storage capacity of simple text.
Economic: Ability to attract and retain staff in the labour market; exchange rates affect costs of international transactions; effect of global economy.
Given the high demand uncertainty of the two product families, the current capital appropriation request moves Seagate towards financial risk in terms of expenditure.
Socio-cultural: Demographic change affects demand for services; stakeholder expectations change.
Operational: Relating to existing operations – both current delivery and building and maintaining.
A pessimistic scenario with likelihood estimated at 25%, would demand only 150,000 Cheetah’s and 350,000 Barracuda’s.
Mitigating Risk with Financial Hedging:
If the counterbalancing actions involve trading financial instruments, including short selling, futures, and options, this is financial hedging. Financial hedging yields an elegant approach to incorporating risk without having to resort to utility functions and price its present value using risk-neutral discounting. The basic idea is to construct a ‘perfect hedge’, which is a portfolio that provides a constant future value in any state of nature and therefore can be priced using risk-free discounting. Financial hedging requires writing an unambiguous contract that specifies capacity usages in a form that is divisible, trade able, and enforceable.
Mitigating Risk with Operational Hedging:
Processing flexibility such as dual or multi-sourcing, using component commonality, having the option to run overtime or to dynamically reroute or shift production (among different types of capital, locations, or subcontractors); holding safety stocks; having warranty guarantees, etc.
A variety of these actions can be grouped as ‘counterbalancing capacities’ to mitigate risk, often by inducing some form of resource pooling.
Question 3
How would you describe “capacity of processing network” if current CAR capacity proposal were implemented? What is the expected profit and ROI under this investment?
(Given the short product life, assume the firm is making the decisions for a single time period of length one year, at the end of which manufacturing capacity will have zero salvage value)
Answer:
Expected Capacity Scenario (Capacity both for Cheetah and Barracuda is 300000)
There are 2 different profit and cost structure
PS=Profit of solved product
C=Cost of unused capacity
Contribution Margins
Cheetah=$400
Barracuda=$300
Demands
Pessimistic (25%) Expected (50%) Optimistic (25%)
Cheetah 150 000 300 000 450 000
Barracuda 350 000 300 000 250 000
Demand for Cheetah in Pessimistic Scenario (0.25)
Profit= (PS-C)*(0.25)
PS= $400*150000 – ($30000*(150) +$80000*(150))
=$43.500.000
C= $30000*(150) + $80000(50) (The spare capacity cost is shared in Cheetah and Barracuda)
=$8.500.000
Profit= $35.000.000 * 0.25
Demand for Cheetah in Expected Scenario (0.50)
Profit= (PS)*(0.50)
PS= $400*300000 – ($30000*(300) +$80000*(300))
=$87.000.000
Demand for Cheetah in Optimistic Scenario (0.25)
Profit= (PS)*(0.25)
PS= $400*300000 – ($30000*(300) +$80000*(300))
=$87.000.000
Total profit for cheetah = (PS-C)*(0.25) + PS*(0.50) +PS*(0.25)
=$35.000.000 * 0.25+$87.000.000 * 0.50+$87.000.000 * 0.25
=$74.000.000
Demand for Barracuda in Optimistic Scenario (0.25)
Profit= (PS-C)*(0.25)
PS= $300*250000 – ($20000*(250) +$80000*(250))
=$50.000.000
C= $30000*(150)
=$4.500.000
Profit= $45.500.000 * 0.25
Demand for Barracuda in Expected Scenario (0.50)
Profit= (PS)*(0.50)
PS= $300*300000 – ($20000*(300) +$80000*(300))
=$60.000.000
Demand for Cheetah in Pessimistic Scenario (0.25)
Profit= (PS)*(0.25)
PS= $300*350000 – ($20000*(300) +$80000*(300))
=$60.000.000
Total profit for barracuda = (PS-C)*(0.25) + PS*(0.50) +PS*(0.25)
=$45.500.000 * 0.25+$60.000.000 * 0.50+$60.000.000 * 0.25
=$56.375.000
Total Profit for the System= Total profit for cheetah+ Total profit for barracuda -Fixed Cost
= $74.000.000+$56.375.000- $40.000.000
= $90.375.000
Question 4:
The case states that true demand forecast contains uncertainty. Given this forecast contains recommend a capacity portfolio that maximizes expected NPV. (Recall, capacity investment must be performed before you observe actual market demand).Verify financial attractiveness of your recommendation: What is the expected profit and ROI now?
Answer:
As mentioned above; we calculated the total profit according to the expected capacity scenario. In uncertainty situations we also calculate total profit pessimistic and optimistic scenarios as well
Pessimistic Capacity Scenario (Capacity for Cheetah 150000 and Barracuda is 350000)
Demand for Cheetah in Pessimistic Scenario (0.25)
Profit= (PS-C)*(0.25)
PS= $400*150000 – ($30000*(150) +$80000*(150))
=$43.500.000
C= $80000(50) (The spare capacity cost is shared in Cheetah and Barracuda)
=$4.000.000
Profit= $47.500.000 * 0.25
Demand for Cheetah in Expected Scenario (0.50)
Profit= (PS)*(0.50)
PS= $400*150000 – ($30000*(150) +$80000*(150))
=$43.500.000
Demand for Cheetah in Optimistic Scenario (0.25)
Profit= (PS)*(0.50)
PS= $400*150000 – ($20000*(150) +$80000*(150))
=$43.500.000
Total profit for cheetah = (PS-C)*(0.25) + PS*(0.50) +PS*(0.25)
=$47.500.000 * 0.25+$43.500.000 * 0.50+$43.500.000 * 0.25
=$44.500.000
Demand for Barracuda in Pessimistic Scenario (0.25)
Profit= (PS-C)*(0.25)
PS= $300*350000 – ($20000*(350) +$80000*(350))
=$70.000.000
C= $80000(50) (The spare capacity cost is shared in Cheetah and Barracuda)
=$4.000.000
Profit= $66.000.000 * 0.25
Demand for Barracuda in Expected Scenario (0.50)
Profit= (PS)*(0.50)
PS= $300*300000 – ($20000*(300) +$80000*(300))
=$60.000.000
Demand for Barracuda in Optimistic Scenario (0.25)
Profit= (PS)*(0.50)
PS= $300*300000 – ($20000*(300) +$80000*(300))
=$60.000.000
Total profit for barracuda = (PS-C)*(0.25) + PS*(0.50) +PS*(0.25)
=$66.000.000 * 0.25+$60.000.000 * 0.50+$60.000.000 * 0.25
=$61.500.000
Total Profit for the System= Total profit for cheetah+ Total profit for barracuda -Fixed Cost
= 44.500.000+$61.500.000- $40.000.000
= $66.000.000
Optimistic Capacity Scenario (Capacity for Cheetah 450000 and Barracuda is 250000)
Demand for Cheetah in Pessimistic Scenario (0.25)
Profit= (PS-C)*(0.25)
PS= $400*150000 – ($30000*(150) +$80000*(150))
=$43.500.000
C= $30000*(300) + $80000(50) (The spare capacity cost is shared in Cheetah and Barracuda)
=$13.000.000
Profit= $30.500.000 * 0.25
Demand for Cheetah in Expected Scenario (0.50)
Profit= (PS-C)*(0.50)
PS= $400*300000 – ($30000*(300) +$80000*(300))
=$87.000.000
C= $30000*(150) (The spare capacity cost 450-300=150)
=$4.500.000
Profit= $82.500.000 * 0.50
Demand for Cheetah in Optimistic Scenario (0.25)
Profit= (PS)*(0.50)
PS= $400*450000 – ($30000*(450) +$80000*(450))
=$142.500.000
Total profit for cheetah = (PS-C)*(0.25) + (PS-C)*(0.50) +PS*(0.25)
=$30.500.000 * 0.25+$82.500.000 * 0.50+$142.500.000 * 0.25
=$84.500.000
Demand for Barracuda in Pessimistic Scenario (0.25)
Profit= (PS-C)*(0.25)
PS= $300*250000 – ($20000*(250) +$80000*(250))
=$50.000.000
C= $80000(50) (The spare capacity cost is shared in Cheetah and Barracuda)
=$4.000.000
Profit= $46.000.000 * 0.25
Demand for Barracuda in Expected Scenario (0.50)
Profit= (PS)*(0.50)
PS= $300*250000 – ($20000*(250) +$80000*(250))
=$50.000.000
Demand for Barracuda in Optimistic Scenario (0.25)
Profit= (PS)*(0.50)
PS= $300*250000 – ($20000*(250) +$80000*(250))
=$50.000.000
Total profit for barracuda = (PS-C)*(0.25) + PS*(0.50) +PS*(0.25)
=$46.000.000 * 0.25+$50.000.000 * 0.50+$50.000.000 * 0.25
=$49.000.000
Total Profit For The System= Total profit for cheetah+ Total profit for barracuda -Fixed Cost
= 84.500.000+$49.000.000- $40.000.000
= $93.500.000
Therefore according to the capacity scenarios profits are formed;
Total Profit for Cheetah Total Profit for Barracuda
Pessimistic Capacity Scenario $44.500.000 $61.500.000
Expected Capacity Scenario $74.000.000 $56.375.000
Optimistic Capacity Scenario $84.500.000 $49.000.000
Question 5:
Interpret your recommended capacity portfolio in intuitive terms: in what sense does your capacity configuration prepare you to ‘hedging’ and why is your plan to be preferred?
Answer:
Newsvendor:
The newsvendor (or newsboy) model is a mathematical model in operations management and applied economics used to determine optimal inventory levels. It is (typically) characterized by fixed prices and uncertain demand. If the inventory level is q, each unit of demand above q is lost. This model is also known as the Newsvendor Problem or Newsboy Problem.
In the case of Seagate Technologies, let K1 is the capacity for ‘The Cheetah’ and K2 is the capacity for ‘The Barracuda’, K3 is the capacity of for ‘The Final Test’ and D is the demand for each product family. The sales plan is 300 thousand units of ‘The Barracuda’ and an equal amount of ‘The Cheetah’ i.e. the sales plan corresponds to a demand vector (in thousands) D = (300, 300). The associated capacity investment vector for the three resources that makes this sales plan feasible is Kb = (300, 300, 600). The capacity portfolio Kb is balanced in the sense that all three resources are fully utilized at the sales plan. Capacity balance means that K1 + K2 = K3.
Another capacity plan that may show up in practice is a plan that minimizes lost sales. In some settings, marketing managers may state that ‘a customer lost once is lost forever’ and advocate ample capacity to prevent that. We refer to such plan as a ‘total coverage capacity plan Kc. Obviously, a centralized, expected profit maximizing planner with knowledge of the probabilistic demand forecast can do better.
Maximization of expected profit leads to increasing the investment in resources with high marginal return compared to marginal investment costs. This generalized typical ‘newsvendor logic’ works in a coupled, multi-dimensional setting and show the risk-neutral or newsvendor network solution in Seagate case to be K* = (350, 350, 600).
The newsvendor network solution by definition yields the highest expected profit among all capacity portfolios; it has three properties that may impact the likelihood of its implementation in practice. First, the newsvendor capacity plan is patently unbalanced in the following sense: there does not exist a single demand realization under which all three resources would be fully utilized. The manager recommending a newsvendor investment plan therefore must explain to top management or the board why they should authorize cash to be invested in capacity that is known in advance to be never fully utilized. This counterbalancing investment is the best operational hedge: this intentionally unbalanced portfolio yields a network capacity configuration that is expected to work ‘best’ for any random draw of demand. This operational hedge through capacity imbalance is appropriate even for a risk-neutral firm.
The unbalanced configuration is related to the second property: given that the newsvendor solution is not optimal for any ex-ante known demand, it can never be the outcome of the ‘typical’ sales-plan driven capacity requirement planning. To put this in perspective, the capacity literature has typically focused on single resource investment for which a famous result was first shown in the seminal paper by Manne (1961): the optimal capacity facing stochastic demand forecasts can be found by an ‘equivalent deterministic problem’ that considers the same capacity investment problem but with a perhaps modified, but always deterministic demand and with some parameters (typically the discount rate) modified to incorporate the effect of uncertainty. Such result justifies the practice of sales-plan driven capacity planning or deterministic planning based on one demand scenario. In contrast, such ‘equivalent deterministic problem’ and hence related justification of practice does not exist for the capacity portfolio problem. The unbalanced optimal capacity portfolio can only be obtained by carefully weighing upside vs. downside for each capacity, which is expressed by intricate and coupled optimality conditions.
The third property is that the newsvendor network capacity vector defines the maximal risk extreme point of the efficient frontier of risk-return capacity configurations. Consider all possible capacity plans and see the below scatter diagram, where variance of returns was chosen as a measure of risk. Similar to financial portfolio theory, the efficient frontier (a line created from the risk-reward graph, comprised of optimal portfolios) is traced by those capacity plans that exhibit minimal risk for a given expected profit. By definition, the newsvendor has highest expected profit and is thus the right-end-point of the frontier. That immediately implies that a risk-sensitive decision maker may find it attractive to ‘optimally deviate’ from the newsvendor solution to trade off some risk for return. As shown in the figure, giving up only about 10% in returns typically cuts risk by an order of magnitude. Of a practical concern, one wants to know the direction along which one should adjust the newsvendor capacity vector so as to trace the frontier and ‘risk-optimally hedge’.
Newsvendor
K*
Total coverage
Kc
Balanced to mean and most likely demand
Kb
Risk-return scatter plot and efficient frontier for the capacity portfolio investment problem of the example. Each point corresponds to a specific capacity investment vector K.
X -Axis: Variance of network value σ 2 (K)
Y – Axis: Mean network value μ (K)
Question 6:
In broad conceptual terms, what are the advantages of ‘sales-plan driven capacity planning?’ What is ‘wrong’ with that practice end how would you improve on it?
Answer:
Advantages of ‘sales-plan driven capacity planning:
The most likely scenario of the demand forecast, sometimes called the ‘sales plan’ is the input to aggregate planning, MRP and CRP systems, such approach to capacity planning, which is called as deterministic or sales-plan driven capacity planning, ‘works’ under decentralized decision making and leads to ‘balanced’ investments if the planned scenario materializes. Managers know the academic dictum that ‘point forecasts’ in the form of a single, deterministic number are typically wrong as the exact number may not materialize. Yet, aggregating demand forecasts that feature not only means but, at a minimum, also variances and some measure of co-variances is not easy. Incorporating such uncertainty, which typically is first estimated at the plant level and consolidated and adjusted after some perturbations by top management, into the complex procedure of capacity planning is even harder. In addition, in contrast to the traditional assumption in operations research, demand is not completely exogenous; sales force incentives and compensation typically are set to enhance the likelihood of ‘meeting the numbers.’ Therefore, it is not uncommon for current commercial capacity planning software to only consider means or a most likely scenario.
What is ‘wrong’ with that practice end how would you improve on it?
Good capacity portfolio planning cannot be performed independently at the various plants with a corporate sales plan as input, but must be coordinated throughout the organization. Top management knows that traditional sales-plan driven capacity planning misses uncertainty and therefore heuristically incorporates uncertainty by perturbing its capacity proposals before implementing them. Adopting coordinated stochastic capacity portfolio planning, however, automates and optimizes these heuristic perturbations, thereby dramatically improving the fidelity of the capacity proposals and their usefulness for strategic planning.
The models could assess the ‘efficiency loss’ that results from adopting sales-plan driven capacity planning. The problem here would be to determine the ‘best’ deterministic sales plan assuming an optimal mechanism design for the sales force and manufacturing incentives. An agency problem is embedded: sales and marketing may agree on a plan, but it typically does not reflect an unbiased assessment of what will be sold. Porteus & Whang (1991) may provide a starting point for such research. In addition, sales faces a very non-symmetric penalty function with harsher punishment for falling short of the sales plan.
The newsvendor network capacity vector defines the maximal risk extreme point of the efficient frontier of risk-return capacity configurations. The newsvendor network solution by definition yields the highest expected profit among all capacity portfolios; it has three properties that may impact the likelihood of its implementation in practice. First, the newsvendor capacity plan is patently unbalanced.
Order Now