In recent decades, value chains have grown in length and complexity as companies expanded around the world in pursuit of margin improvements. Since 2000, the value of intermediate goods traded globally has tripled to more than $10 trillion annually. Businesses that successfully implemented a lean, global model of manufacturing achieved improvements in indicators such as inventory levels, on-time-in-full deliveries, and shorter lead times.
However, these operating model choices sometimes led to unintended consequences if they were not calibrated to risk exposure. Intricate production networks were designed for efficiency, cost, and proximity to markets but not necessarily for transparency or resilience. Now they are operating in a world where disruptions are regular occurrences. Averaging across industries, companies can now expect supply chain disruptions lasting a month or longer to occur every 3.7 years, and the most severe events take a major financial toll.
The risk facing any particular industry value chain reflects its level of exposure to different types of shocks, plus the underlying vulnerabilities of a particular company or in the value chain as a whole. New research from the McKinsey Global Institute explores the rebalancing act facing many companies in goods-producing value chains as they seek to get a handle on risk—not ongoing business challenges but more profound shocks such as financial crises, terrorism, extreme weather, and, yes, pandemics.
Today technology is challenging old assumptions that resilience can be purchased only at the cost of efficiency. The latest advances offer new solutions for running scenarios, monitoring many layers of supplier networks, accelerating response times, and even changing the economics of production. Some manufacturing companies will no doubt use these tools and devise other strategies to come out on the other side of the pandemic as more agile and innovative organizations.
The COVID pandemic has delivered the biggest and broadest value chain shock in recent memory. But it is only the latest in a series of disruptions. In 2011, a major earthquake and tsunami in Japan shut down factories that produce electronic components for cars, halting assembly lines worldwide. The disaster also knocked out the world’s top producer of advanced silicon wafers, on which semiconductor companies rely. Just a few months later, flooding swamped factories in Thailand that produced roughly a quarter of the world’s hard drives, leaving the makers of personal computers scrambling. In 2017, Hurricane Harvey, a Category 4 storm, smashed into Texas and Louisiana. It disrupted some of the largest US oil refineries and petrochemical plants, creating shortages of key plastics and resins for a range of industries.
This is more than just a run of bad luck. Changes in the environment and in the global economy are increasing the frequency and magnitude of shocks. Forty weather disasters in 2019 caused damages exceeding $1 billion each—and in recent years, the economic toll caused by the most extreme events has been escalating.1 As a new multipolar world takes shape, we are seeing more trade disputes, higher tariffs, and broader geopolitical uncertainty. The share of global trade conducted with countries ranked in the bottom half of the world for political stability, as assessed by the World Bank, rose from 16 percent in 2000 to 29 percent in 2018. Just as telling, almost 80 percent of trade involves nations with declining political stability scores.2 Increased reliance on digital systems increases exposure to a wide variety of cyberattacks; the number of new ransomware variations alone doubled from 2018 to 2019.3 Interconnected supply chains and global flows of data, finance, and people offer more “surface area” for risk to penetrate, and ripple effects can travel across these network structures rapidly.
Exhibit 1 classifies different types of shocks based on their impact, lead time, and frequency of occurrence. In a few cases, we also show hypothetical shocks like a global military conflict or a systemic cyberattack that would dwarf the most severe shocks experienced to date. While these may be only remote possibilities, these scenarios are in fact studied and planned for by governments and security experts. The impact of a shock can be influenced by how long it lasts, the ripple effects it has across geographies and industries, and whether a shock hits the supply side alone or also hits demand.
This analysis reveals four broad categories of shocks. Catastrophes are historically remarkable events that cause trillions of dollars in losses. Some are foreseeable and have relatively long lead times, while others are unanticipated. Larger patterns and probabilities can guide general preparedness; hurricanes strike in the Gulf of Mexico every year, for example. But the manifestation of a specific event can strike with little to no warning. This includes some calamities that the world has avoided to date, such as a cyberattack on foundational global systems.
Disruptions are serious and costly events, although on a smaller scale than catastrophes. They, too, can be split into those that telegraph their arrival in advance (such as the recent US–China trade disputes and the United Kingdom’s exit from the European Union) and unanticipated events such as data breaches, product recalls, logistics disruptions, and industrial accidents. Disruptions do not cause the same cumulative economic toll as catastrophes.
Companies tend to focus much of their attention on managing the types of shocks they encounter most often, which we classify as “unanticipated disruptions.” Some other shocks such as trade disputes have made headlines in recent years and, as a result, companies have started to factor them into their planning. But other types of shocks that occur less frequently could inflict bigger losses and also need to be on companies’ radar. The COVID pandemic is a reminder that outliers may be rare—but they are real possibilities that companies need to consider in their decision making.
All four types of shocks can disrupt operations and supply chains, often for prolonged periods. We surveyed dozens of experts in four industries (automotive, pharmaceuticals, aerospace, and computers and electronics) to understand how often they occur. Respondents report that their industries have experienced material disruptions lasting a month or longer every 3.7 years on average. Shorter disruptions happen even more frequently.
We analyzed 23 industry value chains to assess their exposure to specific types of shocks. The resulting index (Exhibit 2) combines multiple factors, including how much of the industry’s current geographic footprint is found in areas prone to each type of event, the factors of production affected by those disruptions and their importance to that value chain, and other measures that increase or reduce susceptibility.
Exposure to different types of shocks varies sharply by value chain. Aerospace and semiconductors, for example, are susceptible to cyberattacks and trade disputes, because of their high level of digitization, R&D, capital intensity, and exposure to digital data flows. However, both value chains have relatively low exposure to the climate-related events we have assessed here (heat stress and flooding) because of the footprint of their production.
Specific types of shocks are more likely to touch certain industries. Pandemics, for example, have a major impact on labor-intensive value chains. In addition, this is the one type of shock for which we assess the effects on demand as well as supply. As we are seeing in the current crisis, demand has plummeted for nonessential goods and travel, hitting companies in apparel, petroleum products, and aerospace. By contrast, while production has been affected in value chains like agriculture and food and beverage, they have continued to see strong demand because of the essential nature of their products.
In general, heat stress is more likely to strike labor-intensive value chains (and some resource-intensive value chains) because of their relatively high reliance on manual labor or outdoor work. Perhaps surprisingly, these same value chains are relatively less susceptible to trade disputes, which are increasingly focused on value chains with a high degree of knowledge intensity and high-value industries.
Overall, value chains that are heavily traded relative to their output are more exposed than those with lower trade intensity. Some of these include value chains that are the most sought after by countries: communication equipment, computers and electronics, and semiconductors and components. These value chains have the further distinction of being high value and relatively concentrated, underscoring potential risks for the global economy. Heavily traded labor-intensive value chains, such as apparel, are highly exposed to pandemic risk, heat stress (because of their reliance on labor), and flood risk. In contrast, the value chains including glass and cement, food and beverage, rubber and plastics, and fabricated metals have much lower exposure to shocks; these are among the least traded and most regionally oriented value chains.
All in all, the five value chains most exposed to our assessed set of six shocks collectively represent $4.4 trillion in annual exports, or roughly a quarter of global goods trade (led by petroleum products, ranked third overall, with $2.4 trillion in exports). The five least exposed value chains account for $2.6 trillion in exports. Of the five most exposed value chains, apparel accounts for the largest share of employment, with at least 25 million jobs globally, according to the International Labor Organization.4
Even value chains with limited exposure to all types of shocks we assessed are not immune to them. Despite recent headlines, we find that pharmaceuticals are relatively less exposed than most other industries. But the industry has been disrupted by a hurricane that struck Puerto Rico, and cyberattacks are a growing concern. In the future, the industry may be subject to greater trade tensions as well as regulatory and policy shifts if governments take action with the intent of safeguarding public health. The food and beverage industry and agriculture similarly have relatively low exposure overall, as they are globally dispersed. Yet these value chains are subject to climate-related stresses that are likely to grow over time. In addition to disrupting the lives and livelihoods of millions, this could cause the industries to become more dependent on trade or force them to undertake expensive adaptations.
Shocks inevitably seem to exploit the weak spots within broader value chains and specific companies. An organization’s supply chain operations can be a source of vulnerability or resilience, depending on its effectiveness in monitoring risk, implementing mitigation strategies, and establishing business continuity plans.
Some of these vulnerabilities are inherent to a given industry; the perishability of food and agricultural products, for example, means that the associated value chains are highly vulnerable to delivery delays and spoilage. Industries with unpredictable, seasonal, and cyclical demand also face particular challenges. Makers of electronics must adapt to relatively short product life cycles, and they cannot afford to miss spikes in consumer spending during limited holiday windows.
Other vulnerabilities are the consequence of intentional decisions, such as how much inventory a company chooses to carry, the complexity of its product portfolio, the number of unique SKUs in its supply chain, and the amount of debt or insurance it carries.5 Changing these decisions can reduce—or increase—vulnerability to shocks.
Weaknesses often stem from the structure of supplier networks in a given value chain. Complexity itself is not necessarily a weakness to the extent that it provides companies with redundancies and flexibility. But sometimes the balance can tip. Complex networks may become opaque, obscuring vulnerabilities and interdependencies. A large multinational company can have hundreds of tier-one suppliers from which it directly purchases components. Each of those tier-one suppliers in turn can rely on hundreds of tier-two suppliers. The entire supplier ecosystem associated with a large company can encompass tens of thousands of companies around the world when the deepest tiers are included.
Exhibit 3 applies network analytics to illustrate the complexity of the first- and second-tier supply ecosystems for two Fortune 500 companies in the computer and electronics industry. This is based on publicly available data and may therefore not be exhaustive.6 These multitiered, multinational networks span thousands of companies and extend to deeper tiers that are not shown here. This illustration also underscores the fact that even within the same industry, companies may make materially different decisions about how to structure their supply ecosystems, with implications for risk.
Companies’ supplier networks vary in ways that can shape their vulnerability. Spending concentrated among just a few suppliers may make it easier to manage them, but it also heightens vulnerability should anything happen to them. Suppliers frequently supply each other; one form of structural vulnerability is a subtier supplier that accounts for relatively little in spending but is collectively important to all participants. The number of tiers of participating suppliers can hinder visibility and make it difficult to spot emergent risks. Suppliers that are dependent on a single customer can cause issues when demand shocks cascade through a value chain. The absence of substitute suppliers is another structural vulnerability.
In some cases, suppliers may be concentrated in a single geography due to that country’s specialization and economies of scale. A natural disaster or localized conflict in that part of the world can cause critical shortages that snarl the entire network. Some industries, such as mobile phones and communication equipment, have become more concentrated in recent years, while others, including medical devices and aerospace, have become less so (Exhibit 4). The aerospace value chain, for example, has diversified in part due to secure market access.
Even in value chains that are generally more geographically diversified, production of certain key products may be disproportionately concentrated. Many low-value or basic ingredients in pharmaceuticals are predominantly produced in China and India, for instance. In total, we find 180 products across value chains for which one country accounts for 70 percent or more of exports, creating the potential for bottlenecks. The chemicals value chain has a particularly large number of such highly concentrated products, but examples exist in multiple industries. Other products may be produced across diverse geographies but have severe capacity constraints, which can create bottlenecks if production is halted. Geographic diversification is not inherently positive, particularly if production and sourcing expands into areas that are more exposed to shocks.
When companies understand the magnitude of the losses they could face from supply chain disruptions, they can weigh how much to invest in mitigation. We built representative income statements and balance sheets for hypothetical companies in 13 different industries, using actual data from the 25 largest public companies in each. This enables us to see how they fare financially when under duress.
We explore two scenarios involving severe and prolonged shocks:
- Scenario 1. A complete manufacturing shutdown lasting 100 days that affects raw material delivery and key inputs but not distribution channels and logistics. In this scenario, companies can still deliver goods to market. But once their safety stock is depleted, their revenue is hit.
- Scenario 2. The same as above, but in this case, distribution channels are also affected, meaning that companies cannot sell their products even if they have inventory available.
Our choice to model a 100-day disruption is based on an extensive review of historical events. In 2018 alone, the five most disruptive supply chain events affected more than 2,000 sites worldwide, and factories took 22 to 29 weeks to recover.7
Our scenarios show that a single prolonged production-only shock would wipe out between 30 and 50 percent of one year’s EBITDA for companies in most industries. An event that disrupts distribution channels as well would push the losses sharply higher for some.
Industries in which companies typically hold larger inventories and have lower fixed costs tend to experience relatively smaller financial losses from shocks. If a natural disaster hits a supplier but distribution channels remain open, inventory levels become a key buffer. However, the downstream company will still face a cash drain after the fact when it is time to replenish its drawn-down safety stock. When a disruption outlasts the available safety stock, lower fixed costs become important to withstanding a decline in EBITDA.
Having calculated the damage associated with one particularly severe and prolonged disruption, we then estimated the bottom-line impact that companies can expect over the course of a decade, based on probabilities. We combined the expected frequency of value chain disruptions of different lengths with the financial impact experienced by companies in different industries. On average, companies can expect losses equal to almost 45 percent of one year’s profits over the course of a decade (Exhibit 5). This is equal to seven percentage points of decline on average. We make no assessment of the extent to which the cost of these disruptions has already been priced into valuations.
These are not distant future risks; they are current, ongoing patterns. On top of those losses, there is an additional risk of permanently losing market share to competitors that are able to sustain operations or recover faster, not to mention the cost of rebuilding damaged physical assets. However, these expected losses should be weighed in the context of the additional profits that companies are able to achieve with highly efficient and far-reaching supply chains.
Today much of the discussion about resilience in advanced economies revolves around the idea of increasing domestic production. But the highly interconnected nature of value chains limits the economic case for making large-scale changes in their physical location. Value chains often span thousands of interconnected companies, and their configurations reflect specialization, access to consumer markets around the world, long-standing relationships, and economies of scale.
We set out to estimate what share of global exports could move to different countries based on the business case and how much might move due to policy interventions. To determine whether industry economics alone support a future geographic shift, we considered a number of factors. One is whether some movement is already under way. Between 2015 and 2018, for instance, the share of trade produced by the three leading export countries in apparel dropped. In contrast, the top three countries in semiconductors and mobile communications increased their share of trade markedly.
Other considerations include whether the value chain is highly capital- or knowledge-intensive, or tied to geology and natural resources. All of these make relocation less feasible. Highly capital-intensive value chains are harder to move for the simple reason that they represent hundreds of billions of dollars in fixed investments. These industries have strong economies of scale, making them more costly to shift. Value chains with high knowledge intensity tend to have specialized ecosystems that have developed in specific locations, with unique suppliers and specialized talent. Deciding to move production outside of this ecosystem to a novel location is costly. Finally, value chains with comparatively high levels of extraregional trade have more scope to shorten than those that are already regionalized. We also consider overall growth, the location of major (and rising) consumer markets, trade intensity, and innovation dynamics.
With respect to noneconomic factors, we consider governments’ desire to bolster national security, national competitiveness, and self-sufficiency. Some nations are focusing on safeguarding technologies with dual-use (civilian and military) implications, which could affect value chains such as semiconductors and communication equipment, particularly as 5G networks are built out. In other cases, governments are pursuing industrial policies intended to capture leading shares of emerging technologies ranging from quantum computing and artificial intelligence to renewable energy and electric vehicles. This, too, has the potential to reroute value chains. Finally, self-sufficiency has always been a question surrounding energy. Now the COVID pandemic has driven home the importance of self-sufficiency in food, pharmaceuticals, and certain medical equipment as well.
We estimate that 16 to 26 percent of exports, worth $2.9 trillion to $4.6 trillion in 2018, could be in play—whether that involves reverting to domestic production, nearshoring, or new rounds of offshoring to new locations. It should be noted that this is not a forecast: it is a rough estimate of how much global trade could relocate in the next five years, not an assertion that it will actually move.
The value chains with the largest share of total exports potentially in play are pharmaceuticals, apparel, and communication equipment. In dollar terms, the value chains with the largest potential to move production to new geographies are petroleum, apparel, and pharmaceuticals.8 In all of these cases, more than half of their global exports could potentially move. With few exceptions, the economic and noneconomic feasibility of geographic shifts do not overlap. Thus, countries would have to be prepared to expend considerable sums to induce shifts from what are otherwise economically optimal production footprints.
In general, the economic case to move is most viable for labor-intensive value chains such as furniture, textiles, and apparel. These value chains were already experiencing shifts away from their current top producers, where the cost of labor has risen. The continuation of this trend could represent a real opportunity for some developing economies. By contrast, resource-intensive value chains, such as mining, agriculture, and energy, are generally constrained by the location of natural resources that provide crucial inputs. But policy considerations may encourage new exploration and development that can shift value chains at the margins.
The value chains in the global innovations category (semiconductors, automotive, aerospace, machinery, communication, and pharmaceuticals) are subject to the most scrutiny and possible intervention from governments, based on their high value, cutting-edge technologies as well as their perceived importance for national competitiveness. But the feasibility of moving these value chains based on the economics alone is low.
Production networks have begun to regionalize in recent years, and this trend may persist as growth in Asia continues to outpace global growth. But multinationals with production facilities in countries such as China, India, and other major emerging economies are typically there to serve local consumer markets, whether or not they also export from those places. As prosperity rises in these countries, they are key sources of global growth that companies will continue to pursue.
In a McKinsey survey of supply chain executives conducted in May 2020, an overwhelming 93 percent reported that they plan to take steps to make their supply chains more resilient, including building in redundancy across suppliers, nearshoring, reducing the number of unique parts, and regionalizing their supply chains.
Strengthen supply chain risk management and improve end-to-end transparency
Global manufacturing has only just begun to adopt a range of technologies such as analytics and artificial intelligence, the Internet of Things, advanced robotics, and digital platforms. Companies now have access to new solutions for running scenarios, assessing trade-offs, improving transparency, accelerating responses, and even changing the economics of production.
Most companies are still in the early stages of their efforts to connect the entire value chain with a seamless flow of data. Digital can deliver major benefits to efficiency and transparency that are yet to be fully realized. Consumer goods giant Procter & Gamble, for example, has a centralized control tower system that provides a company-wide view across geographies and products. It integrates real-time data, from inventory levels to road delays and weather forecasts, for its own plants as well as suppliers and distributors. When a problem occurs, the system can run scenarios to identify the most effective solution.9
Creating a comprehensive view of the supply chain through detailed subtier mapping is a critical step to identifying hidden relationships that invite vulnerability. Today most large firms have only a murky view beyond their tier-one and perhaps some large tier-two suppliers. Working with operations and production teams to review each product’s bill of materials can reveal whether critical inputs are sourced from high-risk areas and lack ready substitutes. Companies can also work with their tier-one suppliers to create transparency. But in cases where those suppliers lack visibility themselves or consider their own sourcing to be proprietary information, risk management teams may have to turn to other information sources to do detective work. After mapping upstream suppliers, downstream companies need to understand their production footprint, financial stability, and business continuity plans.
Minimize exposure to shocks
Targeted measures taken before an event occurs can mitigate the impact of a shock or speed time to recovery. As more physical assets are digitized, for example, companies will need to step up investment in cybersecurity tools and teams.
One of the most important steps is building more redundancy into supplier networks. Relying on a single source for critical components or raw materials can be a vulnerability. In fact, even if a company relies on multiple suppliers, they may be concentrated in the same place. Taking the time to identify, prequalify, and onboard backup vendors comes at a cost. But it can provide much-needed capacity if a crisis strikes. Auditing and diversifying the supply chain can have the added benefit of reducing carbon intensity, raising environmental and labor standards, and expanding opportunities for women- and minority-owned businesses.
One way to achieve supply chain resilience is to design products with common components, cutting down on the use of custom parts in different product offerings. Auto manufacturers are perhaps the most advanced in this regard, having implemented modular manufacturing platforms that share components across product lines and production sites.
Physical assets may need to be hardened to withstand natural disasters. In regions that are vulnerable to worsening hurricanes and storm surges, this may involve installing bulkheads, elevating critical machinery and utility equipment, adding more waterproof sealing, and reworking drainage and valves. Many factories that are not air-conditioned today will need cooling systems to prepare for rising temperatures and potential heat waves in some parts of the world. Plants located in earthquake-prone areas may need seismic retrofitting. Companies can also build more redundancies into transportation and logistics.
When a shock does hit, companies need the ability to respond quickly
The shift to just-in-time and lean production systems has helped companies improve efficiency and reduce their need for working capital. But now they may need to strike a different balance between just-in-time and “just in case.” Having sufficient backup inventory of key parts and safety stock is a critical buffer that can minimize the financial impact of disrupted supplies. It can also position companies to meet sudden spikes in demand.
The ability to reroute components and flex production dynamically across sites can keep production going in the wake of a shock. This requires robust digital systems as well as the analytics muscle to run scenarios based on different responses. When the COVID pandemic hit, Nike used predictive analytics to selectively mark down goods and reduce production early on to minimize impact. The company was also able to reroute products from brick-and-mortar stores to e-commerce sales, driven in part by direct-to-consumer online sales through its own training app. As a result, Nike sustained a smaller drop in sales than some of its competitors.
When disaster strikes, companies have to be laser focused on cash management. But those at the top of a value chain also have a vested interest in preserving the supplier networks on which they depend. In the aftermath of the global financial crisis, some companies accelerated payments or guaranteed bank loans to give key vendors a lifeline.
Coming on the heels of Brexit and a flare-up in US–China trade tensions, the COVID pandemic has forced businesses to focus on building resilience in their supply chains and operations. Not everything that can go wrong actually does go wrong, but businesses and governments cannot afford to be caught flat-footed when disaster strikes. Preparing for future hypotheticals has a present-day cost. But those investments can pay off over time—not only minimizing losses but also improving digital capabilities, boosting productivity, and strengthening entire industry ecosystems. Rather than a trade-off between resilience and efficiency, this rebalancing act might deliver a win-win.