Supply Chain Resilience

It’s been about a month since the UK’s covid hospitalization rate began to bend up again (see chart below). Over the last several months the US has typically lagged UK by four to six weeks. Hospitalization rates have been especially troublesome (e.g., delta or omicron) when similar demand curves coincided for the UK, Denmark, Israel, and the United States. Right now the UK’s demand curve may have plateaued. Israel and Denmark have continued to show declining demand for hospitalization (the difference seems to be mostly a matter of proportional vaccination rates and, especially, booster rates). It is likely that US covid hospitalizations will increase during April and probably into early May. Meanwhile another new variant (well, more than one) seems to be emerging, with potential demand implications for late Spring and early Summer.

Goldman Sachs has conducted a textual analysis of earnings call transcripts since 2010 for Russell 3000 companies. This focused on supply chain references overtime. The researchers found that these companies have recently focused on three strategic options: “Reshoring appears to be limited so far, as construction of new manufacturing facilities has mostly gone sideways and imports are still growing faster than domestic manufacturing output. Diversification of supply chains appears to be further along, and inventory overstocking is the strategy that’s most clearly underway.”

Many want to conceive and craft more resilient supply chains. Rana Foroohar calls out the choices — and balances? — between localization, decentralization, and redundancy.  Gillian Tett focuses on mapping concentrations to work through potential disruption scenarios and/or stockpiling, and/or lateralization (diversification of sources and operational functions). Geoffrey Gertz suggests mapping, stress testing, and diversification.  

It says more about how human cognition works than the essential structures of demand and supply networks, but some form of three dimensional analysis often recurs. Usual suspects include:

1. How Much Stuff: Also-Known-As overstocking, buffering, safety stock, just-in-case capacity… Perhaps mass?

2. From How Far Away: AKA reshoring, back-shoring, near-shearing, localization, close-to-customer… Perhaps distance? If distance, then potential velocity must be part of the resilience package.

3. From How Many Sources: AKA de-concentration, decentralization, diversification, redundancy… Perhaps centrality? (More below)

Is “de-globalization” an issue of how far away or how many sources?  Some of each, it seems to me, depending on demand for what is needed (or wanted) where.

Mass, distance, and velocity are familiar topics; centrality less so. I suspect that centrality is often the key resilience multiplier (or divider) for demand and supply networks. In a network of many nodes, some nodes and connected channels have the capacity for much more flow. Some nodes feature many short paths to other nodes, many nodes have only a few paths. The more central nodes and channels can facilitate high volume, high velocity flows. Loss of these more central nodes can seriously constrain flow capacity. Below is a less than six-minute video with further explanation.

More mass moving faster over less distance from more places is usually less vulnerable to disruption and more likely to prove resilient in case of serious disruption — unless sources are placed so close as to subvert the benefits of more places [supply chain “clusters” are a thing]. Distance of source capacity from disruption can be helpful, if sufficient channels remain open or can be reopened. High proportion dependence on concentrated sources very close by will increase risk of catastrophic failure. So, perhaps: More mass moving more-even proportions at a comparatively constant speed from more disparate places far and near toward well-defined demand targets, better ensures demand fulfillment. Distributing capacity far and wide minimizes the risk of losing network capacity.

Demand is often implied in these trinitarian approaches, but treated as an independent, even non-constant variable. Yet supply chain professionals and processes depend on demand mostly remaining within a predictable range. Supply capacity organizes around demand capacity. Highly concentrated demand can create vulnerabilities as extreme as concentrated sources or channels or modes. Excess demand, demand destruction, or extreme demand volatility will quickly increase various frictions in flow. To achieve enhanced supply chain resilience, we need to more fully engage how pull shapes push to create flow… another example of the rule of three.

+++

Yesterday’s release by the Bureau of Economic Analysis on February Personal Income and Outlays includes good news for supply chains. Overall personal expenditures increased slightly, from $16,678 billion in January to $16,713 billion in February. Considering inflation, this implies a potential reduction in product flow. This potential seems to be confirmed by a 2.5 percent decline in durable goods expenditures compared to (an admittedly high) January pull signal. Meanwhile expenditures on services continued to increase and are now above pre-pandemic levels. Almost two years of intense demand for durable goods is dissipating as other consumption channels become more available (and, perhaps, as inflation-influenced prices encourage alternatives). In late December and early January we saw a decline in flow that was precipitous enough to cause concern. But even before yesterday’s numbers, we had seen February freight volumes recover a more sustainable level (see chart below, March 14 release for February). The demand patterns in yesterday’s report reinforce this sense of improving equilibrium between pull and push. This was the flow context just before the war in Ukraine spiked fuel prices and introduced significant uncertainty into flows of key agricultural products and strategic minerals. The lock-downs in Hong Kong, Shenzhen, Shanghai, and elsewhere have since contributed their own ebbs and flows. We continue to ride some rapids. Class III or even more exciting?

Thursday BlackRock’s chairman, Larry Fink, included the following concise assessment in his annual letter:

Russia’s aggression in Ukraine and its subsequent decoupling from the global economy is going to prompt companies and governments worldwide to re-evaluate their dependencies and re-analyze their manufacturing and assembly footprints – something that Covid had already spurred many to start doing. And while dependence on Russian energy is in the spotlight, companies and governments will also be looking more broadly at their dependencies on other nations. This may lead companies to onshore or nearshore more of their operations, resulting in a faster pull back from some countries. Others – like Mexico, Brazil, the United States, or manufacturing hubs in Southeast Asia – could stand to benefit. This decoupling will inevitably create challenges for companies, including higher costs and margin pressures. While companies’ and consumers’ balance sheets are strong today, giving them more of a cushion to weather these difficulties, a large-scale reorientation of supply chains will inherently be inflationary.

Dependencies will persist, but diversification and decentralization can reduce network-wide vulnerability. Trading near-term cost reduction and flow optimization for longer-term strategic resilience will be more acceptable now that existential risks associated with over-concentration have become more viscerally experienced.

How, where, and especially how-much and how-quickly this de-concentration happens will depend on density of demand, velocity of demand, distances between demand and supply, perceived geopolitical risk, access to capital, comparative friction of flows between nations involved, and perceived host-nation regulatory/ confiscatory risk. Price competition, related opportunism, and innovation will continue to stir the pot.

Where Larry Fink and his peers decide to invest — and even more pointedly: not invest — in these demand and supply networks will do a great deal to determine pace and places.

+++

A few hours after writing what’s above I read online at the Wall Street Journal a piece about supply chains become more web-like. Very complementary to what Larry Fink and I are saying. [March 27 Update] More weekend catch-up reading: Exposing Capitalism’s Great Illusion at Bloomberg. This is a helpful, important exploration of a century-plus adventure in economic integration and disintegration. Which way are we leaning today? Tomorrow?

There are real risks of insufficient sources. These include:

Lengthening duration, deepening scale, and expanding scope of drought. For example, this year has started very dry in California’s San Joaquin Valley, Sacramento Valley and Central Coast. According to the National Oceanic and Atmospheric Administration, “Many of these areas have experienced record dryness since the start of 2022, which has driven some reservoirs to record low levels and resulted in widespread stream flows and soil moisture ranking below the 2nd percentile. Water availability is a real concern as allocation from the Central Valley Project is likely to be either much reduced or non-existent for many farmers in California’s Central Valley, according to the U.S. Bureau of Reclamation.” (More) Increased demand for water will not make more water, and even prompt the opposite.

Reduced flows of agricultural fertilizers and grain. Drought has reduced crop yields in Southern Brazil and Northern Argentina, the Horn of Africa, and California’s Central Valley. War and sanctions related to Ukraine and Russia threaten to significantly reduce availability of significant proportions of wheat, corn, and sunflowers. Remaining flows will experience– are experiencing — significant price volatility until future global flows can be better anticipated (more). If anything, demand for food is increasing.

Fewer passenger vehicles are being manufactured. Production is down from one-tenth to one-third depending on the market and vehicle category. Errors predicting demand — and ensuring related semiconductor flows — have played an important role in this shortage. So have exogenous factors such as covid-related production cut-backs and earthquakes and trade conflicts and maritime port congestion. For the United States this means that supply of new vehicles has been one-fifth to one-quarter below likely demand for much of the last year (see chart below). At some point car manufacturers will reclaim previous capacity. Then the disequilibrium of demand and supply will transition from today’s “scarce” to “sporadic” as pent-up demand is gradually fulfilled and downstream flows even out.

Differentiating scarce (rare, deficient) from sporadic (scattered, dispersed) flows is a crucial diagnostic step on the way to better balancing demand with supply. Actions based on inaccurate diagnosis almost always result in further disruption of whatever flows are available. Wealth will often pull what it wants even from scarce flows. Effectively managing demand targets and supply speed (velocity) can often mitigate human need while dealing with reduced volumes. Poverty tends to be excluded from flows, experiencing drought even when nearby there is flooding.

Because I work mostly with high-income places and people, my problems more often involve satisfying wealthy wants than existential need. But whether flows are scarce or sporadic or abundant, calibrating push (supply) with pull (demand) is fundamental.

The three examples above share a concentration characteristic. When drought is concentrated in a place with many people or high proportional production, the consequences of scarcity are amplified. When capacity — production or distribution or consumption — is highly concentrated, any disruption — even modest and/or temporary — will often have amplified wide-area consequences. The more dependent on any particular place or piece or process or person or whatever, the more vulnerable the entire network.

Increasing concentration can produce economies of scale and related competitive advantages. But at some point, concentration is always a Faustian bargain (here and here). Systemic diversification generates long-term strategic advantages.

Many late-Twentieth Century supply chains have become demand and supply networks where much more volume is moved much more accurately, quickly, and confidently than ever before. This enhanced fluidity is the result of policy, technology, and processes that reduce friction in discerning effectual demand and completing physical and financial transactions to fulfill what is discerned. More assured pull has prompted significantly more — and better — push.

Over the last thirty years concentration and consolidation have been helpful tools in reducing friction and cultivating flows. Covid, climate change, and war now suggest that too much concentration can self-generate congestion and, with just a little complication, friction-shedding, target-enhancing bottlenecks can too easily become convulsive chokepoints.