Structural Measures of Resilience for Supply Chains

Arash Azadegan; M. Amin Rahimian; Marios Papachristou

arxiv: 2303.12660 · v4 · submitted 2023-03-22 · 💻 cs.SI · math.PR· math.ST· stat.TH

Structural Measures of Resilience for Supply Chains

Marios Papachristou , M. Amin Rahimian , Arash Azadegan This is my paper

Pith reviewed 2026-05-24 10:11 UTC · model grok-4.3

classification 💻 cs.SI math.PRmath.STstat.TH

keywords resiliencestructuralfailuresourcingsuppliersupplyarchitectureschain

0 comments

The pith

Percolation-based framework defines supply chain resilience metric and identifies four structural determinants plus two architectural regimes.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Supply chains form networks where one company's output feeds another's input. When suppliers fail, problems can spread. This work borrows math from physics on how networks break apart under random failures. It measures resilience by the highest failure rate the whole system can absorb before total output drops. The analysis points to four key features: how many raw materials enter, how many finished goods exit, how many sources each item needs, and how much influence each source has. Networks that are wide at the raw-material end but centralized turn out fragile. Networks with balanced widths and sparse links hold up better. The authors turn the math into a linear program that can run on large real networks and also handle cases where failures are correlated, such as from the same country or event.

Core claim

Using node percolation theory and branching processes, we identify four critical structural determinants of resilience: the number of raw materials, the number of finished goods, sourcing requirements, and sourcing influence. Our analysis reveals two distinct regimes: 'top hat' architectures... and 'rolling pin' structures.

Load-bearing premise

That supply-chain failure propagation can be faithfully captured by node percolation on a static directed graph whose parameters (sourcing requirements and influence) are known and fixed, allowing the linear program to approximate cascade sizes without needing dynamic re-routing or capacity constraints.

read the original abstract

Modern production systems are increasingly defined by dense networks of multi-tier sourcing dependencies, where localized upstream disruptions can cascade into system-wide collapses. While supply chain resilience has garnered significant managerial attention, we still lack theoretically-grounded, reliable, analytical metrics that can distinguish inherently resilient architectures from fragile ones. This paper addresses this gap by developing a structural resilience framework and a novel metric, defined as the maximum supplier failure rate that a network can sustain while maintaining an aggregate production level. Using node percolation theory and branching processes, we identify four critical structural determinants of resilience: the number of raw materials, the number of finished goods, sourcing requirements, and sourcing influence. Our analysis reveals two distinct regimes: "top hat" architectures, which are characterized by excessive raw materials and high centralization, making them inherently fragile; and "rolling pin" structures, which maintain controlled input/output widths and sparsity, allowing them to absorb non-trivial shocks. To operationalize these insights, we formulate resilience computation as a scalable linear program that approximates cascading failure sizes in large-scale networks with cycles, heterogeneous suppliers, and structural decoupling. Furthermore, we extend our framework to account for exogenous failure correlations, such as those arising from geographic or geopolitical factors that can undermine traditional supplier and input diversification strategies. We validate our theoretical results using multi-echelon supply chain data. These tools can inform network design, supplier diversification, and inventory planning to proactively reduce systemic risk.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper gives a percolation-derived metric for the max supplier failure rate a supply chain can take and splits architectures into top-hat (fragile) and rolling-pin (resilient) types, but the static-graph setup is the main limit.

read the letter

The core contribution is a resilience number defined as the highest upstream failure rate the network can sustain before aggregate output drops, plus the top-hat versus rolling-pin classification derived from percolation and branching processes. Four structural features drive the result: counts of raw materials and finished goods, plus sourcing requirements and influence. They turn cascade-size estimation into a linear program that handles cycles and heterogeneous nodes, and they add a term for exogenous correlations like geographic shocks. They report checking the framework on multi-echelon supply-chain data. That metric and the explicit regime split are not in the prior work the abstract cites, so the framing is new inside the subfield. The LP step is a reasonable move toward computation on realistic sizes. The main soft spot is the modeling premise itself. The analysis treats the directed graph and its sourcing parameters as fixed, so propagation follows node percolation without re-routing or binding capacities. If real firms adjust suppliers or if production limits kick in, the predicted regimes and the LP values could change rank order. The abstract does not model those adjustments, and the stress-test concern lands directly on that point. Validation is mentioned but the abstract gives no error bars, simulation comparisons, or sensitivity checks, so it is hard to judge how much the approximation distorts the metric. This is for readers who want analytical handles on supply-chain network design rather than purely empirical case studies. It is coherent on its own terms and engages the literature, so it deserves a serious referee even though the static assumption will need pushback in review.

Referee Report

2 major / 2 minor

Summary. The manuscript develops a structural resilience metric for supply chains, defined as the maximum supplier failure rate a network can sustain while maintaining aggregate production. Using node percolation theory and branching processes on directed graphs, it identifies four structural determinants (number of raw materials, number of finished goods, sourcing requirements, and sourcing influence) and distinguishes two regimes ('top hat' architectures that are fragile due to excessive raw materials and centralization, versus 'rolling pin' structures that are more resilient due to controlled widths and sparsity). Resilience computation is operationalized via a scalable linear program approximating cascade sizes in networks with cycles and heterogeneity; the framework is extended to exogenous failure correlations and validated on multi-echelon supply chain data.

Significance. If the static percolation model and LP approximation faithfully represent cascade dynamics, the work would supply analytically grounded metrics and architectural insights that could guide supplier diversification and network design in supply chain risk management, moving beyond purely simulation-based methods. The explicit identification of regimes and the handling of correlations represent potential advances, though their practical impact hinges on the modeling assumptions.

major comments (2)

[Abstract] Abstract: the central claim that node percolation on a static directed graph with fixed sourcing requirements and influence identifies robust 'top hat' vs. 'rolling pin' regimes is load-bearing, yet the model does not incorporate dynamic supplier re-routing or binding capacity constraints; if these features alter cascade sizes, the structural determinants and regime distinction may not generalize.
[Abstract] Abstract (LP formulation): the resilience metric is defined directly as the quantity solved by the percolation model, with the linear program presented only as an approximation to cascade sizes rather than an independent test or validation; this creates a circularity risk that weakens support for the four determinants and the claimed regimes.

minor comments (2)

[Abstract] Abstract: the terms 'sourcing requirements' and 'sourcing influence' are introduced without explicit definitions or formulas, which reduces clarity for readers unfamiliar with the branching-process setup.
[Abstract] Abstract: the validation statement ('we validate our theoretical results using multi-echelon supply chain data') lacks any indication of the specific metrics, baselines, or error measures used, making it difficult to assess how strongly the data support the regimes.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed and constructive report. We address the two major comments point by point below, clarifying the deliberate scope of the static percolation framework while remaining open to textual clarifications where they strengthen the presentation.

read point-by-point responses

Referee: [Abstract] Abstract: the central claim that node percolation on a static directed graph with fixed sourcing requirements and influence identifies robust 'top hat' vs. 'rolling pin' regimes is load-bearing, yet the model does not incorporate dynamic supplier re-routing or binding capacity constraints; if these features alter cascade sizes, the structural determinants and regime distinction may not generalize.

Authors: The model is intentionally restricted to static node percolation on directed graphs with fixed sourcing requirements and influence. This restriction is what permits the closed-form branching-process analysis that isolates the four structural determinants and distinguishes the two architectural regimes. The manuscript states these modeling choices explicitly in the introduction and methods; it does not assert that the regimes remain invariant once dynamic re-routing or capacity constraints are added. We can expand the limitations paragraph to reiterate that the reported regimes are properties of the static model and that extensions incorporating dynamics constitute separate research. revision: partial
Referee: [Abstract] Abstract (LP formulation): the resilience metric is defined directly as the quantity solved by the percolation model, with the linear program presented only as an approximation to cascade sizes rather than an independent test or validation; this creates a circularity risk that weakens support for the four determinants and the claimed regimes.

Authors: The resilience metric is defined as the largest failure rate for which the percolation process on the directed graph still preserves positive aggregate output; this definition precedes and is independent of the linear program. The LP is introduced solely as a scalable computational surrogate that approximates cascade sizes when cycles and heterogeneous degrees render exact branching-process recursion intractable. The four determinants themselves are obtained from the branching-process analysis on the theoretical model, not from the LP. We already compare the LP output to exact percolation results on acyclic subgraphs and to Monte-Carlo simulations on the empirical multi-echelon data; we can add a dedicated validation subsection if the referee considers the existing comparisons insufficiently prominent. revision: no

Circularity Check

0 steps flagged

No significant circularity; metric defined independently and analyzed via percolation model

full rationale

The paper explicitly defines its novel resilience metric as the maximum supplier failure rate sustainable while maintaining aggregate production level. It then applies node percolation theory and branching processes to identify four structural determinants of this metric and distinguishes two architectural regimes. The linear program is introduced strictly as a scalable approximation tool for cascade sizes in networks with cycles, not as a fitted prediction or self-referential quantity. No self-citation chains, uniqueness theorems imported from prior author work, ansatzes smuggled via citation, or self-definitional reductions appear in the abstract or described framework. The derivation remains self-contained, with external validation on multi-echelon supply chain data.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Framework rests on standard percolation assumptions applied to supply chains; no new free parameters or invented entities are introduced in the abstract.

axioms (1)

domain assumption Supply chain dependencies can be represented as a static directed graph on which node percolation and branching processes accurately describe failure cascades.
Invoked when the authors apply percolation theory to identify the four determinants and two regimes.

pith-pipeline@v0.9.0 · 5793 in / 1233 out tokens · 20225 ms · 2026-05-24T10:11:35.887129+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

We model failures using a node percolation process... four critical structural determinants... 'top hat' architectures... 'rolling pin' structures... linear program that approximates cascading failure sizes
IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

RG(ε) = sup {x : PG,x[S ≥ (1−ε)K] ≥ 1−1/K}

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

3 extracted references · 3 canonical work pages

[1]

The network origins of aggregate fluctuations

Acemoglu, Daron, Vasco M Carvalho, Asuman Ozdaglar, and Alireza Tahbaz-Salehi (2012). The network origins of aggregate fluctuations. Econometrica 80(5): 1977–2016. Acemoglu, Daron, Asuman Ozdaglar, and Alireza Tahbaz-Salehi (2015). Systemic risk and stability in finan- cial networks. American Economic Review 105(2): 564–608. Ahn, Dohyun and Kyoung-Kuk Kim...

work page arXiv 2012
[2]

IEEE: 10–pp

Proceedings of the. IEEE: 10–pp. Eisenberg, Larry and Thomas H Noe (2001). Systemic risk in financial systems. Management Science 47(2): 236–249. Elliott, Matthew, Benjamin Golub, and Matthew V Leduc (2022). Supply network formation and fragility. American Economic Review 112(8): 2701–47. Erol, Selman (2019). Network hazard and bailouts. Available at SSRN...

work page 2001
[3]

Incentive-Aware Models of Dynamic Financial Networks

Jalan, Akhil, Deepayan Chakrabarti, and Purnamrita Sarkar (2022). Incentive-Aware Models of Dynamic Financial Networks. preprint arXiv:2212.06808. Leskovec, Jure et al. (2007). Patterns of cascading behavior in large blog graphs. Proceedings of the 2007 SIAM international conference on data mining . SIAM: 551–556. Long Jr, John B and Charles I Plosser (19...

work page arXiv 2022

[1] [1]

The network origins of aggregate fluctuations

Acemoglu, Daron, Vasco M Carvalho, Asuman Ozdaglar, and Alireza Tahbaz-Salehi (2012). The network origins of aggregate fluctuations. Econometrica 80(5): 1977–2016. Acemoglu, Daron, Asuman Ozdaglar, and Alireza Tahbaz-Salehi (2015). Systemic risk and stability in finan- cial networks. American Economic Review 105(2): 564–608. Ahn, Dohyun and Kyoung-Kuk Kim...

work page arXiv 2012

[2] [2]

IEEE: 10–pp

Proceedings of the. IEEE: 10–pp. Eisenberg, Larry and Thomas H Noe (2001). Systemic risk in financial systems. Management Science 47(2): 236–249. Elliott, Matthew, Benjamin Golub, and Matthew V Leduc (2022). Supply network formation and fragility. American Economic Review 112(8): 2701–47. Erol, Selman (2019). Network hazard and bailouts. Available at SSRN...

work page 2001

[3] [3]

Incentive-Aware Models of Dynamic Financial Networks

Jalan, Akhil, Deepayan Chakrabarti, and Purnamrita Sarkar (2022). Incentive-Aware Models of Dynamic Financial Networks. preprint arXiv:2212.06808. Leskovec, Jure et al. (2007). Patterns of cascading behavior in large blog graphs. Proceedings of the 2007 SIAM international conference on data mining . SIAM: 551–556. Long Jr, John B and Charles I Plosser (19...

work page arXiv 2022