# Supply Chain Resilience Concept Card: [Company, network or course] > A concept card for the supply chain risk and resilience sessions of a second-year supply chain elective: how to find > the nodes whose failure would hurt most, using time to recover (TTR), time to survive (TTS) and performance impact > (Simchi-Levi, Schmidt and Wei 2014; Simchi-Levi et al. 2015), how to compare mitigation and contingency strategies > (Tomlin 2006), why redundancy and flexibility both buy resilience (Sheffi 2005), and why two suppliers are not always > two independent sources. It is for MBA students in an operations or supply chain elective and for supply chain, > procurement, risk and continuity managers. It builds on mba/mba-inventory-card (safety stock, the holding cost of a > buffer) and mba/mba-sourcing-procurement-card (Kraljic supply risk, dual sourcing), which it uses without > re-teaching, and on mba/mba-decision-trees-card for valuing options under uncertainty. Source basis: the works named > above plus Chopra and Sodhi (2004), Kleindorfer and Saad (2005) and Sheffi and Rice (2005), written up in original > words with fictional data (Northfield Outdoor Co., Corvane Robotics and Tallow Creek Medical Center are invented; > suppliers are generic labels). Replace every [bracketed] value and every row marked *Example*. This is a learning > aid, not legal, insurance or regulatory advice. ## Document control | Field | Value | |---|---| | Course or project | [Supply chain elective, risk sessions / resilience review] | | Learner or analyst | [Your name] | | Network scope | [Tier 1 suppliers, mapped tier 2 nodes, lanes, DCs, systems] | | TTR sources and dates | [Supplier survey YYYY-MM-DD, engineering estimate, past incidents] | | Inventory and margin data | [Stock report YYYY-MM-DD, margin by product family from finance] | | Probability estimates | [Incident history, insurer data, expert judgment; state which] | | Tool | [Excel / Google Sheets / Python with scipy] | | Reviewed by | [Study group / tutor / risk committee] | | Last updated | YYYY-MM-DD | | Version | [0.1] | ## Contents - [When to use it](#when-to-use-it) - [The idea in plain language](#the-idea-in-plain-language) - [Formulas and notation](#formulas-and-notation) - [Assumptions and limits](#assumptions-and-limits) - [How managers use it](#how-managers-use-it) - [The resilience process](#the-resilience-process) - [Worked example A (fully solved): Northfield Outdoor Co. network](#worked-example-a-fully-solved-northfield-outdoor-co-network) - [Worked example B (half solved): Tallow Creek Medical Center critical supplies](#worked-example-b-half-solved-tallow-creek-medical-center-critical-supplies) - [Exercise C (unsolved): Corvane Robotics controller chip](#exercise-c-unsolved-corvane-robotics-controller-chip) - [Connections to other cards](#connections-to-other-cards) - [Common mistakes and misconceptions](#common-mistakes-and-misconceptions) - [Self-test](#self-test) - [My notes and next steps](#my-notes-and-next-steps) - [Sources and further reading](#sources-and-further-reading) - [Answer key](#answer-key) - [Revision history](#revision-history) - [Standard and licence note](#standard-and-licence-note) ## When to use it **One-sentence rule:** use this card when you must decide **where** a supply chain is fragile and **how much** to spend on making it robust, for events that are rare, large and hard to put a probability on. | Question | Tool on this card | Key output | |---|---|---| | Which nodes would hurt most if they failed? | TTR, TTS and performance impact (Simchi-Levi et al. 2014, 2015) | Ranked exposure, risk exposure index | | What kinds of risk does the chain face? | Risk categories (Chopra and Sodhi 2004) | A complete risk list, not only the obvious ones | | Should we act before or after a disruption? | Mitigation versus contingency (Tomlin 2006) | Buffer, second source, flexible capacity or a response plan | | Is a mitigation worth its cost? | Expected loss plus worst-year view | Annual cost, break-even probability | | Do two suppliers really give two sources? | Common-cause probability | Probability both fail together | | Use when | Do not use alone when | |---|---| | A disruption would stop a flow for weeks | Variation is small and frequent: use safety stock | | TTR can be estimated with suppliers and engineers | The network is not mapped below tier 1: map it first | | Margin or service at risk per week is known | The main exposure is financial (currency, credit): use finance risk tools | | Management will choose between paid options | The decision is a regulatory obligation: follow the rule, then optimise | ## The idea in plain language **Risk matrices fail for rare events.** The usual way to rank risks is probability times impact. For a fire at a tier-2 plant or a port closure, nobody knows the probability to within a factor of five, so the ranking is driven by guesses. Simchi-Levi, Schmidt and Wei (2014) proposed starting from what can be estimated: **how long a node would take to recover** (time to recover, TTR) and **how long the chain can keep meeting demand without that node** (time to survive, TTS), using inventory and alternative sources. If TTS is longer than TTR, the disruption is absorbed. If it is shorter, sales are lost for the gap, and the **performance impact** is the margin lost in that gap. Ranking nodes by impact, before probability, reveals exposures that spend-based thinking hides: a cheap zipper from a single plant can threaten more profit than a large contract manufacturer with a backup. **Risk has many sources.** Chopra and Sodhi (2004) list disruptions, delays, systems failures, forecast errors, intellectual property, procurement, receivables, inventory and capacity risks, and warn that a remedy for one can worsen another: holding more inventory reduces disruption risk and raises inventory and obsolescence risk. **Act before or act after.** Tomlin (2006) separates **mitigation**, which is paid for in advance whether or not anything happens (holding a buffer, sourcing from two suppliers), from **contingency**, which is prepared in advance and used only when the disruption hits (rerouting, a backup supplier that ramps up, temporary capacity). The right mix depends on the cost of each, the length of disruptions and how fast the backup can ramp up. **Redundancy and flexibility.** Sheffi (2005) and Sheffi and Rice (2005) describe two ways to be resilient: redundancy (spare inventory, spare capacity, extra suppliers), which costs money all the time, and flexibility (standard parts, interchangeable plants, cross-trained people, postponement), which costs money to build but also helps in normal times. Culture matters too: firms that detect problems early and let people act without waiting recover faster. **Two suppliers are not always two sources.** If both buy from the same tier-2 mill, sit in the same flood plain or ship through the same port, a single event can stop both. Diversification only helps for independent failure causes. ## Formulas and notation **Time to survive** for node $i$, with inventory $I_i$ downstream of the node, demand rate $d_i$ served through it and alternative supply $a_i$ available while it is down: $$ TTS_i = \frac{I_i}{d_i - a_i} \qquad (a_i < d_i) $$ On this card TTS is given directly in weeks of cover. Simchi-Levi et al. (2015) compute it with an optimisation model that lets the whole network re-route; the single-path version here is the hand calculation. **Performance impact** (lost margin if node $i$ fails), **risk exposure index** and **expected annual loss**: $$ PI_i = m_i \cdot \max\big(0,\; TTR_i - TTS_i\big), \qquad REI_i = \frac{PI_i}{\max_j PI_j}, \qquad EL_i = \pi_i \cdot PI_i $$ **Choosing a mitigation** with annual cost $K_o$ that changes $TTS$ or $TTR$: $$ \text{Expected annual cost}_o = K_o + \pi_i \cdot PI_i^{(o)}, \qquad \text{Worst year}_o = K_o + PI_i^{(o)}, \qquad \pi^{*} = \frac{K_o}{PI_i^{(0)} - PI_i^{(o)}} $$ where $\pi^{*}$ is the break-even disruption probability above which option $o$ beats doing nothing. **Two suppliers with a common cause.** If each supplier fails independently with probability $q$ and a common cause (shared region, shared tier-2 source) stops both with probability $c$: $$ P(\text{both down}) = c + (1 - c)\,q^2 \qquad \text{against} \qquad q^2 \ \text{if independent} $$ | Symbol | Meaning | Unit | |---|---|---| | $TTR_i$ | Time for node $i$ to return to full function after a disruption | weeks | | $TTS_i$ | Time the chain can keep meeting demand without node $i$ | weeks | | $I_i$, $d_i$, $a_i$ | Inventory downstream of the node, demand through it, alternative supply | units, units per week | | $m_i$ | Margin at risk per week through node $i$ | USD thousand per week | | $PI_i$ | Performance impact: margin lost if node $i$ fails | USD thousand | | $REI_i$ | Risk exposure index, 1 for the worst node | none | | $\pi_i$ | Annual probability of a disruption at node $i$ | per year | | $EL_i$ | Expected annual loss | USD thousand per year | | $K_o$ | Annual cost of mitigation option $o$ | USD thousand per year | | $q$, $c$ | Independent failure probability, common-cause probability | per year | **Excel equivalents.** Performance impact: `=m*MAX(0,TTR-TTS)`. Risk exposure index: `=PI/MAX(PI_range)`. Expected loss: `=p*PI`. Break-even probability: `=K/(PI_0-PI_o)`. Both down: `=c+(1-c)*q^2`. Ranking: `=RANK.EQ(PI,PI_range)`. ## Assumptions and limits | Element | Assumption | What breaks it | What to do | |---|---|---|---| | TTR | Known, single value | Recovery depends on the scale of the event; suppliers are optimistic | Use a range (best, likely, worst) and ask for evidence from past incidents | | TTS | Inventory is where it is needed and usable | Stock in the wrong place, wrong sizes, expired | Count only usable stock; split by location and SKU | | Margin at risk | Sales lost during the gap are lost for good; no partial recovery | Some customers wait or substitute; ramp-up is gradual | Add a lost-sales share and a ramp curve | | Single path | One node fails, the rest of the network does not adapt | Re-routing and spare capacity elsewhere | Use a network optimisation (Simchi-Levi et al. 2015) | | Probability | A number exists and is stable | Rare events, climate trends, geopolitical shifts | Rank by PI first; use probability only to size spending | | Independence | Failures at different nodes are unrelated | Shared regions, ports, tier-2 sources, IT systems | Map shared dependencies and model common causes | | Expected value | The firm is risk-neutral | A large loss threatens covenants, cash or patients | Look at the worst year, not only the average | Business continuity management has its own management-system standard, ISO 22301 (Security and resilience, business continuity management systems). As of October 2026 the current edition is ISO 22301:2019 with Amendment 1:2024 (climate action changes, published February 2024). *Note on sources:* confirmed in the catalogues of national standards bodies (BSI, SIS, SCC), because iso.org could not be opened at checking time; check [iso.org](https://www.iso.org/) before you cite it. Due-diligence laws also touch supplier mapping: as of October 2026 the European Commission's page on the EU Corporate Sustainability Due Diligence Directive (Directive (EU) 2024/1760, amended by Omnibus I: Directive (EU) 2025/794 postponed the dates and Directive (EU) 2026/470, in force 18 March 2026, amended the substance) states that EU companies with at least 5,000 employees and EUR 1.5 billion net worldwide turnover are in scope, with transposition by 26 July 2028 and application from 26 July 2029; check the current text on [commission.europa.eu](https://commission.europa.eu/business-economy-euro/doing-business-eu/sustainability-due-diligence-responsible-business/corporate-sustainability-due-diligence_en) and take legal advice on what applies to you. ## How managers use it - **Finding hidden exposures.** A PI ranking shows the board where the money at risk sits. Low-spend, single-source parts and tier-2 sources usually move up; large suppliers with backups move down. - **Sizing buffers.** Strategic stock is set to close the gap between TTS and TTR for the nodes that matter, not as a blanket number of weeks for every item. - **Supplier contracts.** TTR becomes a contract term: suppliers commit to recovery plans, alternative sites and notification within hours. Sourcing teams add TTR to the scorecard (mba/mba-sourcing-procurement-card). - **Design for resilience.** Engineers standardise parts and qualify alternatives so that one board or one material is not tied to one plant. This lowers TTR permanently. - **Network design.** Where to put plants and DCs and how many lanes to use (mba/mba-network-design-card) trades cost against exposure; PI makes the exposure side measurable. - **Testing.** Continuity plans are tested in exercises and game days (compliance/business-continuity, ops/game-day-plan): a plan that has never been used has an unknown TTR. - **Reporting.** Risk committees track PI for the top nodes over time and record each mitigation in a risk register (compliance/risk-register). ## The resilience process The review moves from mapping to measuring to deciding, then back again as the network changes: map the network as far down the tiers as you can, collect TTR and TTS, compute PI, rank, choose mitigation or contingency, test and monitor. ```mermaid flowchart TD A["Map the network: tier 1, tier 2, lanes, DCs, systems"] --> B["Collect TTR from suppliers and engineers"] B --> C["Compute TTS from usable inventory and alternatives"] C --> D["Performance impact PI = margin x gap"] D --> E["Rank nodes by PI and risk exposure index"] E --> F{"Is PI above the risk appetite?"} F -->|"No"| G["Monitor and re-check each year"] F -->|"Yes"| H{"Act before or after a disruption?"} H -->|"Before: mitigation"| I["Buffer stock, second source, redesign"] H -->|"After: contingency"| J["Backup capacity, rerouting, response plan"] I --> K["Test with exercises and update TTR"] J --> K K --> G G --> A ``` ## Worked example A (fully solved): Northfield Outdoor Co. network **Situation (*Example*, fictional).** Northfield Outdoor Co. maps the nodes behind its outerwear and online sales. For each node, finance estimates the gross margin that flows through it per week, operations counts weeks of usable cover downstream (TTS), suppliers and engineers estimate TTR, and the risk team estimates an annual disruption probability from incident history. Both shell suppliers (Supplier F and Supplier N, as in mba/mba-sourcing-procurement-card) buy their membrane fabric from the same mill. The margins reflect today's split, about 70% of shell volume from Supplier F and 30% from Supplier N, before the 70/30 award to Supplier N recommended in that card's Worked example A3; re-run the table after the award. **A1. Performance impact ranking.** | Node *(Example)* | Margin at risk $m$ (USD k/week) | TTS (weeks) | TTR (weeks) | Gap | PI (USD k) | REI | $\pi$ | EL (USD k/yr) | |---|---|---|---|---|---|---|---|---| | Zipper supplier (single source) | 80 | 5 | 16 | 11 | 880 | 1.000 | 4% | 35.20 | | Membrane fabric mill (tier 2) | 60 | 9 | 20 | 11 | 660 | 0.750 | 3% | 19.80 | | Central distribution centre | 300 | 2 | 4 | 2 | 600 | 0.682 | 2% | 12.00 | | Supplier F assembly site | 42 | 8 | 12 | 4 | 168 | 0.191 | 5% | 8.40 | | Supplier N assembly site | 18 | 3 | 6 | 3 | 54 | 0.061 | 4% | 2.16 | | E-commerce platform | 110 | 0 | 0.3 | 0.3 | 33 | 0.038 | 15% | 4.95 | | Ocean lane and port of entry | 95 | 6 | 3 | 0 | 0 | 0.000 | 10% | 0.00 | | **Total** | | | | | | | | **82.51** | Excel for the zipper row: PI `=80*MAX(0,16-5)`, REI `=880/MAX(PI_range)`, EL `=4%*880`. ```vega-lite { "$schema": "https://vega.github.io/schema/vega-lite/v5.json", "description": "Northfield network: performance impact by node if it fails, USD thousand (fictional, computed)", "width": 360, "height": 220, "data": {"values": [ {"node": "Zipper supplier", "PI": 880, "absorbed": "No: TTS < TTR"}, {"node": "Fabric mill (tier 2)", "PI": 660, "absorbed": "No: TTS < TTR"}, {"node": "Central DC", "PI": 600, "absorbed": "No: TTS < TTR"}, {"node": "Supplier F site", "PI": 168, "absorbed": "No: TTS < TTR"}, {"node": "Supplier N site", "PI": 54, "absorbed": "No: TTS < TTR"}, {"node": "E-commerce platform", "PI": 33, "absorbed": "No: TTS < TTR"}, {"node": "Ocean lane and port", "PI": 0, "absorbed": "Yes: TTS >= TTR"} ]}, "mark": "bar", "encoding": { "y": {"field": "node", "type": "nominal", "sort": "-x", "title": "Node"}, "x": {"field": "PI", "type": "quantitative", "title": "Performance impact (USD thousand)"}, "color": {"field": "absorbed", "type": "nominal", "title": "Disruption absorbed?"} } } ``` **Reading A1.** Three findings change the conversation: 1. The **zipper supplier** tops the list although zippers are a small share of spend. It is single-source, used across many products and slow to replace (new zipper types must be tested and sewn into every pattern). Kraljic placed it in the bottleneck quadrant; PI puts a number on it. 2. The **fabric mill** is a tier-2 node that the tier-1 dual sourcing between Supplier F and Supplier N does not protect. It was invisible until the network was mapped. 3. The **ocean lane and port** has the highest probability but no impact, because six weeks of cover exceed a three-week recovery. A probability-times-impact matrix built on guesses might have put it near the top. The distribution centre has the largest weekly margin but a short TTR (a third-party warehouse can be activated in four weeks), so its PI is third. The e-commerce platform fails most often but recovers in about two days. **A2. Choosing a mitigation for the fabric mill.** Northfield spends USD 25k a week on membrane fabric (USD 1,300k a year) and holds inventory at 25% a year. Three options: | Option | Annual cost $K$ (USD k) | New TTS or TTR | PI (USD k) | Expected loss (USD k) | Expected annual cost | Worst year | Excel (expected) | |---|---|---|---|---|---|---|---| | Do nothing | 0.00 | TTS 9, TTR 20 | 660 | 19.80 | 19.80 | 660.00 | `=3%*60*MAX(0,20-9)` | | Strategic fabric buffer, +6 weeks (USD 150k of stock) | 37.50 | TTS 15 | 300 | 9.00 | 46.50 | 337.50 | `=6*25*25%+3%*60*MAX(0,20-15)` | | Qualify a second mill for 30% of volume, able to ramp to full volume within 6 weeks (USD 120k over 3 years plus a 2% premium) | 47.80 | TTR 6 | 0 | 0.00 | 47.80 | 47.80 | `=120/3+1300*30%*2%+3%*60*MAX(0,6-9)` | On expected cost alone, doing nothing wins: USD 19.80k against 46.50k and 47.80k. The break-even probability for the second mill is 7.24% a year (`=47.8/(660-0)`), and for the buffer 10.42% (`=37.5/(660-300)`). The estimate of 3% is a guess for a rare event, and the worst year tells a different story: without action a mill fire costs USD 660k of margin; with a second mill the worst year costs USD 47.80k. A risk-neutral analyst picks "do nothing"; a board that cannot accept a USD 660k hit in one season picks the second mill, which also gives Northfield bargaining power with the first mill and a source of innovation. The buffer is the weakest option here: it costs almost as much as the second mill and still leaves USD 300k exposed, because TTR is long. Buffers suit short TTR; second sources suit long TTR. **A3. How much does dual sourcing reduce the chance of being cut off?** Two suppliers, each down with probability 5% a year: | Case | Probability both down | Excel | |---|---|---| | Independent failures | 0.25% | `=0.05^2` | | Common cause 2% (same region or same tier-2 source) plus independent 3% each | 2.09% | `=0.02+(1-0.02)*0.03^2` | Each supplier alone is still down about 4.94% of years in the second case (`=0.02+(1-0.02)*0.03`), almost the same as before, but the chance of losing both rises 8.35 times. For Northfield's shells, the shared fabric mill is exactly such a common cause: dual sourcing at tier 1 gives much less protection than the 0.25% suggests. ## Worked example B (half solved): Tallow Creek Medical Center critical supplies **Situation (*Example*, fictional).** Tallow Creek Medical Center reviews three critical supplies. The cost of a shortage per day (USD thousand) covers substitutes bought at spot prices, cancelled or delayed procedures and staff time. TTS is days of usable stock on site and at the distributor; TTR is the supplier's recovery estimate. For saline, the 8% annual probability and 8-week (56-day) stoppage are the same estimates as in mba/mba-sourcing-procurement-card Exercise C. | Item *(Example)* | Shortage cost (USD k/day) | TTS (days) | TTR (days) | $\pi$ | Gap (days) | PI (USD k) | EL (USD k/yr) | |---|---|---|---|---|---|---|---| | IV saline bags | 9.0 | 12 | 56 | 8% | 44 | 396.0 | 31.68 | | Iodinated contrast media | 14.0 | 20 | 60 | 3% | [ ] | [ ] | [ ] | | Sterile procedure kits | 6.0 | 25 | 21 | 5% | [ ] | [ ] | [ ] | 1. Fill the blanks (Excel: `=cost*MAX(0,TTR-TTS)` and `=p*PI`). Which item has the largest PI, and which the largest expected loss? 2. Tallow Creek uses about 165 saline bags a day at USD 1.10 (60,000 a year, as in mba/mba-sourcing-procurement-card Exercise C). A buffer of 30 extra days is [ ] bags worth USD [ ], costing USD [ ] a year at a 25% holding rate (`=165*30*1.1*25%`). What is the new TTS, and the new PI? 3. Compare the buffer cost with the expected loss it removes. Is the buffer worth it on expected cost alone? 4. Saline has a shelf life. Which stock rule keeps a buffer usable, and who checks it? ## Exercise C (unsolved): Corvane Robotics controller chip **Situation (*Example*, fictional).** Every Corvane Robotics warehouse robot uses one motor-controller chip from a single fabrication plant. Margin at risk is USD 150k a week. Corvane holds 6 weeks of chips (TTS); if the plant fails, redesigning the board for another chip takes 26 weeks (TTR). The annual probability of a disruption is estimated at 5%. Chip purchases are USD 40k a week and the holding rate is 20% a year. a. Compute PI and expected annual loss with no action. b. Option 1: hold 10 more weeks of chips. Compute its annual cost, the new PI, the expected annual cost and the worst year. c. Option 2: redesign the board now so that it accepts a second chip, at USD 600k spread over 3 years; TTR falls to 8 weeks. Compute the same four numbers. d. Which option would you recommend to a venture-backed company with 18 months of cash, and would the answer change for a large, diversified firm? Use the break-even probabilities in your argument. e. Which contingency actions (not mitigation) could Corvane prepare at little cost today? ## Connections to other cards | Card or template | Connection | |---|---| | mba/mba-inventory-card | Safety stock for everyday variation; the holding cost of a strategic buffer | | mba/mba-sourcing-procurement-card | Kraljic supply risk, TCO of a second source, dual-sourcing contracts | | mba/mba-network-design-card | Number and location of plants, DCs and lanes; cost against exposure | | mba/mba-bullwhip-effect-card | Disruptions trigger order panics that amplify upstream | | mba/mba-decision-trees-card | Valuing a second source or a buffer as an option under uncertainty | | mba/mba-supply-chain-guide | Where this card sits in the elective | | mba/mba-stakeholder-esg-card | Supplier due diligence, climate and human-rights exposure in the network | | mba/mba-ai-risk-governance-card | Systems and AI vendors as nodes with their own TTR | | compliance/business-continuity, ops/game-day-plan | Continuity plans and exercises that test TTR | | compliance/risk-register | Recording the top nodes, owners and actions | ## Common mistakes and misconceptions 1. **Ranking by spend.** Spend measures purchasing effort, not exposure; the zipper in A1 shows the gap. 2. **Stopping at tier 1.** Shared tier-2 sources turn two suppliers into one point of failure. 3. **Trusting a single TTR.** Ask for a range and evidence; suppliers rarely volunteer bad news. 4. **Counting all stock as TTS.** Stock in the wrong place, size or state does not keep customers served. 5. **Using expected value only.** Rare, large losses can threaten a firm's survival; show the worst year. 6. **Treating buffers as the only answer.** When TTR is long, a buffer is expensive and still leaves a gap; second sources and redesign shorten TTR itself. 7. **Assuming independence.** Region, port, power grid, IT platform and tier-2 sources create common causes. 8. **Writing a plan and never testing it.** An untested plan has an unknown TTR. 9. **Confusing variability with disruption.** Safety stock covers normal fluctuations; resilience covers weeks of lost supply. One number cannot do both jobs. 10. **Forgetting the side effects of mitigation.** More stock raises obsolescence risk; more suppliers raise management cost and may dilute leverage (Chopra and Sodhi 2004). ## Self-test 1. Define TTR and TTS in one sentence each. 2. Write the formula for performance impact and explain why it uses $\max(0, \cdot)$. 3. Why can the node with the highest disruption probability have zero performance impact? Use A1. 4. What is the difference between mitigation and contingency in Tomlin (2006)? Give one example of each. 5. Name two kinds of redundancy and two kinds of flexibility. 6. In A2, why is the buffer a weak option for the fabric mill? 7. Compute the break-even probability for an option costing USD 30k a year that cuts PI from 500 to 100 (USD k). 8. Two suppliers each fail with probability 4% a year, independently. What is the probability both fail? What if a common cause stops both with probability 1% and each fails independently with 3%? 9. List five of the risk categories in Chopra and Sodhi (2004). 10. Why does Simchi-Levi et al.'s approach rank by impact before probability? 11. Give one reason a risk-neutral expected-cost analysis can be the wrong frame for a company. 12. How should TTR enter a supplier scorecard or contract? ## My notes and next steps | Item | My note | |---|---| | The node in my network I know least about | [ ] | | A TTR I would ask a supplier to confirm, and how | [ ] | | A mitigation and a contingency I would propose | [ ] | | Questions for my study group or tutor | [ ] | | Next card to read | [mba/mba-network-design-card / mba/mba-bullwhip-effect-card] | ## Sources and further reading - Simchi-Levi, D., Schmidt, W., and Wei, Y. (2014). From superstorms to factory fires: managing unpredictable supply-chain disruptions. *Harvard Business Review*, January-February 2014. - Simchi-Levi, D., Schmidt, W., Wei, Y., Zhang, P. Y., Combs, K., Ge, Y., Gusikhin, O., Sanders, M., and Zhang, D. (2015). Identifying risks and mitigating disruptions in the automotive supply chain. *Interfaces*, 45(5), 375–390. - Tomlin, B. (2006). On the value of mitigation and contingency strategies for managing supply chain disruption risks. *Management Science*, 52(5), 639–657. - Sheffi, Y. (2005). *The Resilient Enterprise: Overcoming Vulnerability for Competitive Advantage*. MIT Press. - Sheffi, Y., and Rice, J. B. (2005). A supply chain view of the resilient enterprise. *MIT Sloan Management Review*, Fall 2005. - Chopra, S., and Sodhi, M. S. (2004). Managing risk to avoid supply-chain breakdown. *MIT Sloan Management Review*, Fall 2004. - Kleindorfer, P. R., and Saad, G. H. (2005). Managing disruption risks in supply chains. *Production and Operations Management*, 14(1), 53–68. - Lee, H. L., Padmanabhan, V., and Whang, S. (1997). Information distortion in a supply chain: the bullwhip effect. *Management Science*, 43(4), 546–558. - European Commission, Corporate sustainability due diligence (page on Directive (EU) 2024/1760 and the Omnibus I amendments), checked 2026-10-28. - ISO 22301, Security and resilience: business continuity management systems, requirements (iso.org), 2019 edition with Amendment 1:2024. - Textbooks (further reading, by name only): Chopra and Meindl, *Supply Chain Management: Strategy, Planning, and Operation*; Simchi-Levi, Kaminsky and Simchi-Levi, *Designing and Managing the Supply Chain*; Sodhi and Tang, *Managing Supply Chain Risk*. - Open resource: OpenStax, *Principles of Management*, chapters on operations, openstax.org (check the licence shown on the book page). ## Answer key **Worked example B.** 1. Contrast media: gap 40 days, PI 560.0, EL 16.80. Sterile kits: gap 0 (TTS 25 exceeds TTR 21), PI 0, EL 0. Contrast media has the largest PI (560.0 against 396.0), but saline has the largest expected loss (31.68 against 16.80), because its disruption probability is higher. 2. 4,950 bags worth USD 5,445, costing USD 1,361.25 a year. New TTS 42 days, still 14 days short of the 56-day TTR, so PI falls to USD 126.0k (`=9*MAX(0,56-42)`) and expected loss to USD 10.08k. 3. The buffer removes an expected loss of USD 21,600 a year (`=31.68-10.08`, in USD k) for USD 1,361.25: worth it even on expected cost, and the worst year falls from USD 396k to about USD 127.36k (126 + 1.36). Closing the gap fully needs 44 extra days of cover, or a second source as in mba/mba-sourcing-procurement-card Exercise C. 4. First expiry, first out rotation with expiry dates tracked in the inventory system; pharmacy or materials management checks it monthly, and the buffer is used in normal operations and replenished, not left untouched. **Exercise C.** a. PI = $150 \times (26 - 6)$ = USD 3,000k; expected loss = $5\% \times 3{,}000$ = USD 150k a year. b. Annual cost $10 \times 40 \times 20\%$ = USD 80k; TTS 16, PI 1,500; expected annual cost 80 + 75 = USD 155k; worst year USD 1,580k. c. Annual cost USD 200k; gap 2 weeks, PI 300; expected annual cost 200 + 15 = USD 215k; worst year USD 500k. d. On expected cost, doing nothing (150) narrowly beats the buffer (155) and the redesign (215). Break-even probabilities: buffer 5.33% (`=80/(3000-1500)`), redesign 7.41% (`=200/(3000-300)`). A venture-backed company with 18 months of cash may not survive a USD 3,000k margin loss, so the worst year dominates: redesign (or both buffer and redesign) is defensible even at a higher expected cost. A large diversified firm can absorb the loss and might accept the risk or choose the buffer. e. Pre-agreed allocation with a distributor, a list of drop-in alternative chips already evaluated, a ready-to-run design change for the board, customer communication templates, and priority rules for which customers receive robots first. **Self-test.** 1. TTR: the time a node needs to return to full function after a disruption. TTS: the time the chain can keep meeting demand without that node, using inventory and alternatives. 2. $PI_i = m_i \max(0, TTR_i - TTS_i)$; if the chain survives longer than recovery takes, no sales are lost, so the impact cannot be negative. 3. The ocean lane and port: probability 10% but TTS 6 weeks exceeds TTR 3 weeks, so a disruption is absorbed. 4. Mitigation is paid in advance whatever happens (buffer stock, dual sourcing); contingency is prepared and used only when disruption hits (rerouting, a backup supplier ramping up). 5. Redundancy: spare inventory, spare capacity or extra suppliers. Flexibility: standard parts, interchangeable plants, cross-trained staff, postponement. 6. TTR is 20 weeks; six weeks of extra stock only moves TTS to 15, leaving USD 300k exposed for USD 37.5k a year, while the second mill removes the gap for USD 47.8k. 7. $30 / (500 - 100) = 7.5\%$. 8. Independent: $0.04^2 = 0.16\%$. With a common cause: $0.01 + 0.99 \times 0.03^2 = 1.0891\%$. 9. Any five of: disruptions, delays, systems, forecast, intellectual property, procurement, receivables, inventory, capacity. 10. Probabilities of rare events are unreliable, while TTR, TTS and margin can be estimated; impact reveals exposures that probability guesses hide. 11. A single large loss can breach covenants, exhaust cash or harm patients, costs that the average does not show. 12. As a scored criterion with evidence, and as a contract term: committed recovery plan, alternative site, notification time, and the right to verify. ## Revision history | Version | Date | Author | Change summary | |---|---|---|---| | 0.1 | YYYY-MM-DD | [Your name] | First version from the template | | 0.2 | YYYY-MM-DD | [Your name] | *Example:* Added tier-2 mapping, updated TTR from supplier survey | ## Standard and licence note Based on the supply chain risk and resilience literature (Chopra and Sodhi 2004, Kleindorfer and Saad 2005, Sheffi 2005, Sheffi and Rice 2005, Tomlin 2006, Simchi-Levi, Schmidt and Wei 2014, Simchi-Levi et al. 2015) and on the European Commission page on the Corporate Sustainability Due Diligence Directive, checked in October 2026; ISO 22301 is named only, nothing from it is reproduced. The text, examples and questions are original; no textbook or case text is copied or paraphrased, and Northfield Outdoor Co., Corvane Robotics and Tallow Creek Medical Center are fictional. The template is independent and not affiliated with or endorsed by any business school, authority, standards body or software vendor. Check the current official versions of every dated fact before you rely on it. Template licence: MIT.