Chain Reactions: Quantify Risk and Improve Capacity Planning

Capacity planning is risky business for so many different industries. Supply chain risk management is one way to take some of the uncertainty out of your decision-making process.

Capacity investment decisions have traditionally been guesswork. Educated guesswork, to be sure — but still guesswork. Fickle consumers, competitive threats, economic factors and a host of unforeseeable factors often combine to create highly uncertain levels of demand for new products, components and capacity. A new approach to capacity decision-making uses financial engineering techniques, similar to those used on Wall Street, to help organizations reduce their tooling investments, direct material spends, inventory write-offs and total sourcing costs.

Both Ends of the Bullwhip

Capacity decisions are high-stakes decisions, particularly in volatile industries such as semiconductors. Here, competitive advantage is won or lost on unique feature sets and within extremely compressed time-to-market windows. Chip manufacturers typically make billion-dollar investments in foundries, employing the newest technologies, and they must optimize capacity to remain competitive — despite downturns, thin margins and fierce competition. Yet, while these capital investments approach 15 percent of sales, only rudimentary processes and solutions currently exist to aid in capacity investment decisions, with costly consequences.

During upturns, demand spikes quickly deplete safety stocks, resulting in ballooning component lead times, product shortages and lost revenue. With pressure to maintain positive forecasts and service levels, purchase orders often continue to flow and inventories accumulate even after demand softens. Downturns can be just as costly when companies are often left with millions of dollars of custom inventories and orders that can't be returned or cancelled. Ultimately they must be written off.

For example, a large Silicon Valley semiconductor testing equipment manufacturer was chronically caught short when the market would begin to heat up. Without forward-looking capabilities, they picked up the demand indicators late, began ordering components and when demand began to fall off, they could not shut the pipeline down quickly enough to avoid being caught with excess inventory. They're not alone ¬— for the U.S. semiconductor industry as a whole, these kinds of capacity decisions amounts to a $10 billion gamble.

What Will Consumers Want Next?

Similarly, the auto industry represents a hyper-competitive market with many new models and even more options introduced each year. Unlike semiconductor time-to-market measured in weeks or months, bringing new auto models and features to market requires enormous tooling and capacity investments with long lead and ramp-up times. Attempting to discern fickle consumer tastes two to three years in advance exposes automakers to considerable risk — diminished market share if they cannot meet demand for a hot vehicle or evaporating profits if they must deep-discount vehicles.

Rising prices on key commodities such as steel and alloys further squeeze margins. Buy now at a high price to guarantee a quantity and forfeit possible future savings. Or buy only the minimum quantities now, and risk higher prices or shortages. Either way, risk plays a leading role in the equation.

Consumer packaged goods companies also face uncertain consumer adoption rates. Growth in this industry relies on a healthy stream of new products, and leading firms introduce hundreds of new products per year. Forecasts — while notoriously inaccurate — form the basis for key planning and sourcing decisions. With many new products in the pipeline, companies often adopt a rule-of-thumb approach for determining finished goods and raw materials inventory levels. To hedge the possibility that demand will exceed available supply, companies often stockpile, so as not to be caught short. Compounded by significant forecast error for new products, this approach frequently results in substantial inventory write-offs.

Suppliers Risk Even More

With original equipment manufacturers (OEMs) as customers, suppliers face perhaps even greater challenges. They are expected to continuously innovate and improve responsiveness at shrinking costs. Because most OEMs rely on point forecasts, suppliers are forced to shoulder a disproportionate share of the risk with no visibility from the OEM to help them manage it. And while the OEMs can use just-in-time (JIT) techniques to wish inventory away, suppliers must build stock levels to meet expected demand, considerably increasing their liabilities.

OEMs cannot simply demand more flexibility from suppliers. Traditional one-size-fits-all flexibility programs can't ensure supply continuity in allocated markets. They can't provide adequate liability protection if demand collapses. Even products in the same business cycle may have very different flexibility requirements based on configurations or target market segments. In addition, how does a company define flexibility? How do they size it? How much should they be willing to pay for it?

These are questions that cannot be answered by standard point forecasts and educated best guesses. Regardless of industry, product, or component, forecasting is a treacherous business. Everyone charged with making capacity decision has a point forecast. Everyone knows that his or her forecast will never be exactly right. But almost nobody knows just how wrong it can be — in either direction. With so much at stake, companies need more precise tools for measuring risk, performing trade-off analyses and improving the profitability of their capacity decisions. Supply chain risk management addresses these very issues.

Supply Chain Risk Management — Financial Engineering for Manufacturers

Supply chain risk management (SCRM) is a new approach that applies standard Wall Street financial engineering techniques — hedging, stochastic modeling and real options theory — to capacity planning. For manufacturers, the ability to hedge, that is, capture and quantify the possible demand scenarios, provides far more flexibility in preparing for changing demand conditions. Stochastic modeling incorporates the effects of variable market factors and reflects the impact of these factors on capacity decisions. Real options theory then combines the range of possible demand scenarios with their impacts on the variable market factors to provide companies with a resilient portfolio of tooling, capacity and sourcing alternatives.

SCRM can help organizations to systematically quantify demand and supply uncertainty and make strategic supply chain decisions based on forward-looking, risk-aware metrics. Companies that have experimented with SCRM have moved from old rule-of-thumb, heuristic or best-guess strategies to quantifying the risks they face and the costs associated with managing a range of demand and supply scenarios. From this perspective, they have been able to position their supply chains and supplier relationships to avoid errors and unprofitable bet-the-business decisions.

Understanding Uncertainty — How Wrong Can A Forecast Be?

The first step in reducing supply chain risk is to understand and quantify uncertainty. People always underestimate uncertainty. We rarely anticipate how bad or how good things can go. In manufacturing, this has meant focusing on a single number — the point forecast — instead of looking at the true range of possible demand.

A SCRM approach quantifies the range of possible demand. With SCRM, a company can analyze demand data from analogous situations to accurately predict demand ranges for a new product, component or capacity decision. For example, using SCRM techniques, one company found that its rule-of-thumb forecast of 20,000 to 30,000 units really could range from 10,000 to 80,000 units. SCRM techniques quantify this range in a range forecast using advanced stochastic modeling algorithms, giving companies a precise picture of how high and how low demand can actually go.

Understanding the Costs of Alternatives — Assigning Costs to the Range Forecast

Once the range of uncertainty is quantified, the company can next identify its alternatives for managing the possibilities and what each alternative will cost. There are numerous variables associated with each step up — or down — in capacity; and complex, often sequential dependencies complicate matters. For example:

  • How does a company know exactly when a capacity increase is needed?

  • How exactly should the company expand capacity? In what amounts and at what locations?

  • Should the company build a new plant or add a production line in an existing one? Building a new facility takes time and significant investment — will market demand last? Will it be subject to low capacity utilization or high shortages when the capacity comes on-line? How does this affect revenue and market presence?

  • Should the company outsource production? To whom, for how long and at what price?

While many companies turn to outsourcing, it can only provide flexibility based on properly structured supplier agreements, which historically are based on point forecasts. For example, a larger volume commitment to the external supplier often yields a better price but exacerbates the risk of excess capacity. Alternately, the company may reserve flexible capacity with an option to release if it is not required, at a price that can wipe out flexibility gains. With SCRM techniques, the decision-making process captures these complex dimensions, allowing the company to not only quantify the range of possibilities it may encounter, but also to know precisely its alternatives at each step.

A Portfolio of Options

Each alternative for a new product launch, component or capacity decision constitutes an option. An option consists of the ability to set up capacity expansion ahead of time, knowing how much it will cost, how long it will take to put into effect, and what its additional output will be. All sources of capacity — in-house or outsourced — should be viewed as options in a portfolio.

The company with a capacity range of 10,000 to 80,000 units would find it cost-prohibitive to spend on 80,000 units' worth of capacity. Instead, a SCRM approach would provide a portfolio of options that defined the costs of an alternative that would take the company from 30,000 to 40,000 units; another option that take them from 40,000 to 50,000 and so on. By breaking uncertainty into options, SCRM enables companies to cost-effectively structure their flexibility ahead of the event — and be prepared.

While many organizations rely on visibility solutions to alert them to shortages or excess inventory, by the time these solutions come into play, it's too late to cost-effectively avoid the problem. In contrast, SCRM techniques employ forward-looking capabilities — predictive analytics — enabling notification when an anticipated situation is pending — with time to exercise the appropriate option. They will also guide the company when to use its options, how much to use them and when to take them off the table. In effect, an SCRM approach will tell the company It looks as though a risk is pending; this is the appropriate option for avoiding the risk; this is what it will cost and this is the benefit.

Structuring Supplier Agreements That Win

With range forecasts in hand, OEMs can work with their suppliers to structure agreements that provide a fairer sharing of risks and rewards. After using a SCRM approach, the semiconductor testing equipment manufacturer communicated with its suppliers that demand would fall between X and Y levels, and provided them with trade-off analyses that granted flexibility through different lead times and location considerations. The resulting supply agreements reflected low-, medium- and high-demand scenarios, enabling suppliers to make their own decisions about how they could best meet each scenario. Eventually, the OEM was able to prioritize its supplier relationships based on suppliers' abilities to perform under the new approach.

Minimizing Under- or Over-Investment

An SCRM approach enables companies to minimize under- or over-investment in inventory or capacity and improve profitability. Through it, the semiconductor testing equipment manufacturer was able to quantify low- and high-demand ranges and easily determine trade-offs and potential overlooked opportunities. It also determined the actual flexibility that would be required — down to the component level. As a result, the company reduced inventory costs by $1.3 million; it realized a $10.8 million improvement in its net cash position; and it reclaimed $0.4 million in gross margin otherwise forgone through lost sales.

Reducing Fixed Capacity Costs

A global automaker used SCRM techniques to help it plan capacity and tooling for new car options. Predicting consumer uptake is inexact at best, further complicated by consumer budget inelasticity. Deciding to purchase one option, such as an improved stability system, increased air-bag protection or a navigation system, usually decreases the customer's ability to purchase others. This can cannibalize current best-selling options and throw off revenue forecasts. Taking an SCRM approach, the automaker created a range forecast and evaluated supply quotes, measured a variety of capacity plans and measured suppliers' response times against it. As a result, the automaker achieved a 12 percent reduction in fixed-capacity costs in the supply base, which translated into $40 million in savings.

Reducing Total Sourcing Costs

With a range forecast, a European automaker created a portfolio of options for introduction of a new navigation system. The company used SCRM techniques to quantify the price reduction risk it would face if demand were only medium or low — 3.7 percent. At the same time, it could estimate what its shortage risk would be if demand exceeded expectations — 4.2 percent. By adjusting supplier contracts and taking the best and worst-case ranges into account, the company saved 18 percent in total sourcing costs — or 0.7 million Euros — on just this one decision, on one model of their line.

Complementary to Traditional Solutions

SCRM complements traditional supply chain, procurement change, collaboration, and supplier consolidation solutions by delivering actionable data that can be used by these solutions. However, SCRM provides an objective, repeatable way to approach capacity and launch decisions. It is not subject to defensive decisions based on turf, an individual's personal experience or best-efforts guessing. In addition, organizations that employ SCRM as an approach can scale it to virtually any supply chain or capacity problem as opposed to having individuals create ad-hoc, non-repeatable processes that are abandoned when the employee transitions to another role.

A Positive Chain Reaction

Organizations can apply science, in the form of financial engineering techniques, to supply chain and capacity decisions — without having to be financial experts themselves. Supply chain risk management techniques can provide results to companies that are seeking repeatable methods for reducing total sourcing costs and improving the quality of their capacity decisions.

About the Author: Heiko Pieper is director of Business Consulting at Vivecon Corp., a company that applies financial engineering techniques to design and manage supply chains.

Loading