Using Volatility as a Driver for Supply Chain Success

There are two ways of meeting difficulties: you alter the difficulties, or you alter yourself to meet them. - Phyllis Bottome

An unrelenting barrage of disruptions continues to expose the fragility of global supply chains.

The Red Sea crisis and the Russia-Ukraine war have disrupted global trade, and the resulting cost increases further exacerbate inflation. The Baltimore bridge collapse had a significant impact on maritime logistics and trucking. Windows computers suffered the infamous ‘blue screen of death’ en masse when a CrowdStrike update went wrong, severely impacting global communications and hobbling multiple sectors.

The list goes on.

Fragility is a concept most understand. It refers to items that break easily under stress - including supply chains responsible for everyday products. Miss hunting for toilet paper during the pandemic? Yeah, me neither.

In his book Antifragile: Things That Gain from Disorder, Nassim Taleb argues the opposite of fragility isn’t resilience but antifragility.

What the heck is antifragility?

At first, I thought antifragility was a made-up word, like blamestorming or everywhen. Yes, those are all actual words.

Taleb coined the term to describe items that benefit from harm, which he proposes as the key to longevity.

The difference between resilience and antifragility is resilience allows you to survive adversity by developing the properties of malleability and elasticity, that is, bending without breaking and quickly returning to baseline form.?

Antifragility, on the other hand, thrives in adversity.

Antifragile entities welcome shocks that allow them to bend and break so they can regrow stronger. An example of antifragility is the Hydra, the many-headed serpent from Greek mythology, who grew two new heads every time one was cut off.

To better appreciate the concept of antifragility we need to adopt whole systems thinking, which helps us understand the relationship between structure and behavior.

Thinking in systems

A system is an interconnected set of elements working together to achieve something.?

There are large systems, like the US economy, which are made up of sub-systems, like public and private sectors. Going up or down levels you can find even more systems in play, with each system producing unique patterns of behavior. By understanding how each system works, we are better positioned to evaluate why certain behaviors lead to poor results and adjust these systems to achieve the desired behavior. Thinking in Systems: A Primer by Donella Meadows is a great book for anyone looking to learn more about systems thinking.

For a system to be antifragile, its subsystems must be fragile and subject to constant volatility. It is this combination of volatility and fragility that leads to the failure of individual elements, which act as information about what works and what doesn’t through established feedback loops. Antifragile systems use these feedback loops to overcompensate by building up extra capacity to deal with future stressors and shocks.

This can be seen in our body’s ability to repair and rebuild muscle that breaks down during resistance training, enabling us to build tolerance over time.

Antifragility is an inherent feature of all biological systems. Take evolution, for instance. The fragility of individual organisms is what makes evolution antifragile. Individual organisms die. All that matters is the required genetic code is passed on enabling subsequent generations to adapt to the changing environment.

Large complex man-made systems, like the economy, have also developed antifragility. The failure of Individual businesses within an industry enables the rest of the industry to learn, adapt, and grow.

To innovate and grow, failure is a prerequisite. The challenge is the limits of resilience in man-made systems are hard to see until they are exceeded. Often, these limits are exceeded too late and with catastrophic consequences.

The Great Recession was caused by sub-prime lending because the government failed to effectively regulate the financial industry; Shortages in medical supplies, staff, and capacity during the pandemic revealed the limits of global healthcare supply chains’ ability to handle large-scale health crises. In systems speak, these were broken feedback loops.

Identifying single points of failure within a system

A single point of failure (SPOF) will cause an entire system to fail, or significantly disrupt its functionality.

There can be more than one SPOF in a system. Everything from a person being solely responsible for a decision-making process, to the dependency on a single factory or warehouse for critical product flows.

The failure of a critical supplier disrupts the flow of materials supporting both dependent and independent demand, which then compounds into widespread supply chain disruptions, ultimately impacting product availability on a global scale.

The shortage of semiconductor chips during the pandemic exposed the Taiwan Semiconductor Manufacturing Company (TSMC) as a single point of failure for multiple industries. TSMC manufactures an estimated 90% of the world's leading-edge logic chips1, making telecom, automotive, AI, and smartphone companies among the many that rely on them.

It is impossible to predict risks.

By focusing on a SPOF and understanding the downstream impacts, we can begin to evaluate the fragility of the systems in which we operate. This enables us to transition from trying to predict risks, to assessing and controlling potential impacts.

Around 2009, when I was working in the MedTech industry, one of our suppliers went bankrupt and wasn't able to supply us with a key component, halting production and impacting sales. The reason? We extended our payment terms with the supplier to improve our cash conversion cycle, not fully appreciating the impact it would have on the supplier's cash flow as we accounted for a large percentage of their revenue.

Could we have avoided the incident? Perhaps. Hindsight is always 20/20. This does, however, highlight the consequences of changing various flows within a system without fully analyzing the potential impacts.

Let's look at an example of how we can better evaluate impacts. The Red Sea crisis highlights a SPOF in global logistics flows. Assessing how vessels going through the Red Sea vs. around Africa create variability in product flows is vital to measuring the downstream impacts.

One of these impacts is the resulting changes in safety stock at dependent downstream nodes in the supply chain. The basic safety stock optimization model depends on three factors - demand, lead time, and the safety factor for the desired service level as represented below2.

As highlighted in this post by Jason Miller, routing shipments via the Red Sea has a lower average lead time but greater lead time variability.

This means rerouting shipments around Africa has a smaller impact on safety stock for items with a coefficient of variation of demand (Stdev of Demand / Avg. Demand) significantly below 1, i.e., relatively stable demand. This is because the increase in the standard deviation of lead time is being multiplied by the average demand squared, which affects the standard deviation of lead time demand more significantly than an increase in average lead time.

Understanding how each SPOF impacts downstream operations is essential to building the necessary feedback loops to minimize disruptions.

Stress testing to minimize impacts

Stress testing systems by pushing them beyond their operating limits with small, controlled experiments are critical in identifying the limits of resilience in time to prevent large-scale disruptions.

The results of these tests help create, repair, and reinforce the required feedback loops. This can be as complex as introducing new legislation such as the 2010 Dodd-Frank Act, which was aimed at improving regulatory oversight in the wake of the Great Recession; or as simple as cross-training team members to build redundancy in case of promotions or early departures.

Stress testing to strengthen systems can include:

• Evaluating the dependency on single or sole-sourced products through network risk assessments, and using the results to develop a supplier diversification strategy including reducing the dependency on high-risk suppliers and establishing backup supplier capacity.
• Simulating disruption scenarios, like losing critical transportation routes, by using advanced modeling tools to simulate their impact on supply chain operations. Using these results to build network redundancy by developing flexible logistics strategies, including alternative routing options and strategic partnerships with multiple carriers.

Continuous testing is imperative in understanding how your system will react to unknowns. Developing a systematic way of taking small and controlled risks by intentionally injecting faults into each system through stress testing increases the likelihood of identifying potential failure points, aids in assessing the magnitude of downstream impacts, and enables system operators to take corrective action before a more significant system-wide failure.

Suppressing volatility does more harm than good

Both Taleb and Meadows warn against trying to artificially smooth out a system and make it tranquil without fully understanding the complexity.

Artificially tranquil environments result in fragile systems because they rob the system of the volatility needed for failure, adaptation, and growth. This means problems are not as apparent and lie dormant, potentially increasing in severity until they reach massive proportions.

We saw this in the challenges faced by companies adopting Just-In-Time (JIT) inventory management policies during the pandemic3. Creating a smooth and predictable flow of materials minimizes inventory levels and reduces carrying costs but makes the resulting supply chain highly sensitive to disruptions. Without buffer stock, even minor delays from suppliers or transportation issues can halt production lines, leading to significant operational disruptions. The lack of inventory buffers also means there is no cushion to absorb sudden demand spikes.

We also tend to suppress volatility by building too much excess capacity to manage future shocks.

While this overcompensation helps antifragile systems handle future stress by building up extra capacity, the long-term success of a system depends on the efficient use of resources, making redundancy wasteful. Moreover, unnecessary overcompensation, like adding too many buffers to a system, allows problems to lie dormant and worsen over time.

For instance, having too much inventory may ensure product availability, but it wastes capital that could be better utilized and prevents the identification and resolution of issues since there’s no immediate impact on operations. This masks inefficient processes, e.g., poor supplier product quality, excess scrap during manufacturing, increasing transportation lead time variability, etc. These issues accumulate and compound over time, leading to more significant problems down the road.

References:

1. Biden-Harris Administration Announces Preliminary Terms with TSMC

2. Understanding Safety Stock and Mastering Its Equations

3. Rethinking Your Just-in-Time Supply Chain