DebtStreamness: An Ecological Approach to Credit Flows in Inter-Firm Networks

Anah\'i Rodr\'iguez-Mart\'inez; Fabio Caccioli; Francesco Caravelli; Pierpaolo Vivo; Silvia Bartolucci; Victoria Landaberry

REVIEW 4 major objections 7 minor 41 references

DebtStreamness: An Ecological Approach to Credit Flows in Inter-Firm Networks

T0 review · 4 major / 7 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read DebtStreamness ranks firms by credit distance from banks; in Uruguay the average is 1.67 steps, but small loops push some firms past 20.

desk verdict DebtStreamness is downstreamness with a relabeled boundary; the empirical short-chain result may be an artifact of incomplete rows for non-surveyed firms. read the letter →

arxiv 2505.01326 v1 pith:PRBQHUF7 submitted 2025-05-02 econ.GN physics.soc-phq-fin.EC

Anahí Rodríguez-Martínez , Silvia Bartolucci , Francesco Caravelli , Victoria Landaberry , Pierpaolo Vivo , Fabio Caccioli This is my paper · ORCID

classification econ.GNphysics.soc-phq-fin.EC

keywords DebtStreamnesstrophiclevelsinter-firmcreditnetworkschainssystemicrisknetworkreconstructioninput-outputlinkagesUruguay

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

DebtStreamness is a new network statistic that tells how many layers of credit lie between a firm and the banking sector: a firm that borrows only from banks has value 1, and each extra intermediary adds the debt-weighted average of its creditors' values. The paper introduces it as the economic analogue of a trophic level, treats bank credit as the energy entering the system, and shows the statistic solves the linear system $DS_i = 1 + \sum_j A_{ij}DS_j$. Applied to an inter-firm credit network of over a thousand firms built from a 2018 central-bank survey in Uruguay, the metric puts the average firm 1.67 credit steps from banks, with clear tiers near 1, 2, and 3. The paper argues that this position-based measure captures a financial hierarchy that production-linkage classifications miss, and that a firm's score can be inflated dramatically by a small reciprocal loop even when its direct bank borrowing does not change. If the claims hold, the metric offers a way to spot hidden financial intermediaries from partial data, which matters for monitoring systemic risk.

What carries the argument

The load-bearing object is the recursive identity $DS_i = 1 + \sum_j A_{ij}DS_j$ (equivalently $\vec{DS}=(I-A)^{-1}\vec{1}$), in which $A_{ij}$ is the fraction of firm $i$'s debt owed to firm $j$. This is the credit-network analogue of a trophic level, with banks as the primary producers: every firm starts at distance 1 and adds the debt-weighted average distance of the firms it borrows from. The identity turns an infinite sum over credit paths into a solvable linear system, with $(I-A)^{-1}$ playing the role of the Leontief inverse, and it is the mechanism that lets a sparse, partial survey produce a complete ordering of firms.

What would settle it

Obtain a complete bilateral inter-firm credit registry and compute DebtStreamness twice, once from the full exposure matrix and once from the truncated matrix containing only each firm's top three creditors; if the two rankings correlate much worse than the reported Spearman values near 0.99, or if the registry reveals longer hidden chains that produce new high-DebtStreamness firms, the paper's robustness-based short-chain conclusion is falsified.

Watch

Extended reading notes

Core claim

The central discovery is that a firm's position in an inter-firm credit network can be compressed into one recursive number, DebtStreamness, defined by $DS_i = 1 + \sum_j A_{ij}DS_j$, where $A_{ij}=L_{ij}/D_i$ is the share of firm $i$'s borrowing that comes from firm $j$; collecting all firms gives $\vec{DS}=(I-A)^{-1}\vec{1}$. The metric counts every path by which bank-originated credit can reach a firm, weighted by the credit share carried along the path, so it is the expected number of credit links separating the firm from the banking sector. In the partial network reconstructed from a 2018 survey of large Uruguayan firms, the average is 1.67, most firms occupy tiers near 1, 2, and 3, and three firms exceed 20. The paper traces the three outliers to a component with a two-firm borrowing loop: deleting that one link drops the component's average from 9.28 to 1.52. It then shows that sector-level aggregation hides the tiers, that production-based upstream/downstream labels do not reproduce the credit hierarchy, and that two maximum-entropy allocations of the unreported half of credit leave DebtStreamness rankings almost unchanged.

Load-bearing premise

The empirical conclusions rest on the premise that the roughly half of each firm's inter-firm borrowing that is not reported because it lies beyond the top three creditors would not, if fully known, rearrange the DebtStreamness rankings.

Editorial extensions

If this is right

Under the paper's robustness result, a regulator with only a firm's top-three creditors and its total inter-firm borrowing, roughly half the credit known, can still order firms by distance from the banking sector with near-perfect rank stability.
A one-link deletion experiment shows that feedback loops can carry most of a firm's measured distance: removing a single two-firm cycle cut the outlier component's average DebtStreamness from 9.28 to 1.52, so stress tests should count reciprocal links as amplifiers.
Because DebtStreamness compounds through creditors, a small, peripheral firm can inherit a large score from a loop upstream, so systemic screening by this metric will flag entities that size-based screens would miss.
Aggregation hides the effect: sector-level DebtStreamness collapses toward 1 for every sector, meaning the tiered credit chains visible at firm level are lost when data are grouped into productive sectors.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The linear definition is not tied to banking: the same $\vec{DS}=(I-A)^{-1}\vec{1}$ could rank any node's distance from any anchor source in any directed weighted network, so the metric can be carried over to supply-chain layers, payment systems, or interbank networks.
A sharper test of the robustness claim would apply the top-three truncation to a complete bilateral credit registry and check whether Spearman correlations stay near 0.99; the current paper only varies how the missing half of credit is allocated, not whether the missing half follows longer chains.
One could decompose each firm's DebtStreamness into the contribution from paths that visit no repeated firm and the excess created by cycles; firms whose score is mostly excess would be flagged as loop-driven, a distinction the outlier analysis suggests matters.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

You should know two things up front. First, the math is not new: DebtStreamness as defined in Eq. (3) is exactly downstreamness from the banking sector, and the paper says so in Appendix A. Second, the central empirical claim—that credit chains are short (mean DS 1.67) with a few loop-driven outliers—is built on a network where only the 240 surveyed firms have full borrowing rows. The other ~800 counterparties have rows that are mostly empty, so their DS is pinned to 1 (if they have bank borrowing) or they are excluded entirely. That truncates chains that pass through non-surveyed intermediaries and likely biases the mean downward. The robustness checks in Sec. 3.2 only redistribute residual credit among surveyed rows; they leave the non-surveyed rows untouched. So the high correlations in Fig. 11 do not address the gap that matters most.

What the paper does well: the derivation is clean, the writing is transparent about data limitations, and the application to a real inter-firm credit network from the Central Bank of Uruguay is a useful contribution. The sector-level comparison with production-based classifications is a nice touch, and the visualizations help. The authors also deserve credit for acknowledging the equivalence to downstreamness in the appendix, even if the abstract still calls it a "novel metric."

The soft spots are proportionate. The loop-amplification result and the negative correlation between bank-borrowing share and DS are basically consequences of the defining equations, not independent discoveries—the paper presents them as findings, which overstates their significance. The novelty overclaim is real but mild, since the application domain is new. The bigger issue is the incomplete-row bias, which is not cured by the reconstruction exercises. The data are also one country and one year, so external validity is limited.

Who is this for? Researchers working on financial networks or firm-level credit data who want a concrete example of how a known positional metric transfers to a new domain. It is a reasonable starting point, not a finished regulatory tool. I would send it to peer review because the question is worth asking and the data are rare, but the authors should be pushed to address the non-surveyed rows, either by reconstructing them or by clearly bounding how much the short-chain result could change under plausible missing structures.

Referee Report

4 major / 7 minor

Summary. The paper introduces DebtStreamness (DS), a network metric defined by DSi = 1 + ∑_j A_ij DS_j (Eq. 3), which measures a firm's average distance from the banking sector in a directed inter-firm credit network, adapting the input-output downstreamness concept to credit flows with banks as the 'primary energy source'. The authors apply DS to a 2018 Central Bank of Uruguay survey of the top three creditors/debtors of 240 large firms, yielding an analyzed network of 843 nodes. They report that credit chains are short (mean DS ≈ 1.67), that a tiered structure exists with peaks near 1, 2, and 3, that three firms are outliers with DS > 20 driven by a 2-cycle loop, that bank-borrowing share correlates negatively with DS, and that sector-level aggregation obscures these patterns. Robustness is claimed via two network reconstruction methods that redistribute residual credit for surveyed firms, giving high Spearman/Kendall correlations with the original DS values.

Significance. If the empirical findings hold, DS provides a simple, interpretable index of a firm's position in credit chains that could complement production-based measures for monitoring systemic risk, and the paper demonstrates an interesting application of ecological trophic-level thinking to granular credit data. The mathematical derivation in Appendix A is correct and the metric is well defined. The main value of the paper is the application to a unique firm-level dataset, not the novelty of the metric itself, which closely follows existing downstreamness/upstreamness definitions. However, the empirical support is weakened by the incomplete nature of the survey and by the fact that several 'findings' (loop amplification, negative correlation with bank share) are direct consequences of the defining equations rather than data-driven discoveries.

major comments (4)

[Sec. 2.3, Eq. (6)-(7); Fig. 11] The robustness analysis reconstructs only the borrowing rows of the 240 surveyed firms; the rows of the approximately 600 non-surveyed firms are left with only the edges reported by surveyed counterparties, and their total debt D_i is not survey-based. The central empirical claims—mean DS of 1.67 and the three outliers with DS>20—could change materially if the unreported credit of non-surveyed firms flows through hidden intermediaries or forms longer chains. The high Spearman/Kendall correlations in Fig. 11 do not address this, because they keep the same node skeleton and only redistribute residual credit in surveyed rows. Please either extend the reconstruction to non-surveyed rows under explicit assumptions or restrict the central claims to the surveyed subset of firms.
[Sec. 3 and Fig. 8] The loop-induced increase in DS and the near -0.99 correlation between bank-borrowing share and DS follow directly from the defining equations (2)-(3): DS is a path sum over the same matrix A, so a 2-cycle adds infinitely many path terms, and a lower bank-borrowing share mechanically raises the row sum of A and hence DS. The paper presents these as empirical findings (e.g., 'we find that local network motifs such as loops can substantially increase a firm's DebtStreamness'). They should be labeled as analytic properties of the metric; the genuinely empirical content is the estimated distribution of DS values and the firm rankings, not these qualitative implications of the definition.
[Sec. 2.2 and Sec. 2.3] The manuscript does not state how D_i and A_ij are computed for non-surveyed nodes, despite reporting DS for 843 nodes. If D_i is approximated as the sum of observed incoming edges for those nodes, then the row sums of A equal 1 by construction, which changes the interpretation of DS and biases the chain-length statistics. Please provide the exact network construction procedure, including how 'incoming paths from the banking sector' is determined and how the 843-node network is derived from the 240 surveyed firms and their counterparties. This is necessary for reproducibility and for assessing the truncation bias identified above.
[Abstract; Sec. 2.3] The reconstruction methods are described in the abstract as 'two maximum-entropy network reconstruction methods', but the fully connected and sparse reconstructions are simple heuristics (uniform allocation and sparse random allocation) with no maximum-entropy objective or constraint optimization. Please either replace the term 'maximum-entropy' with an accurate description or provide the maximum-entropy derivation, since the abstract's validation claim depends on this characterization.

minor comments (7)

[Fig. 7 caption] There is a typo 'newtork' for 'network', and the caption does not clarify that the yellow triangle is the banking-sector node distinct from the DS color coding of firms.
[Eq. (5)] The identity matrix and the vector of ones are both denoted by '1', which is confusing; please use, for example, I and 1 (bold) or e.
[Sec. 3, second paragraph] 'Debstreamness' should be 'DebtStreamness' for consistency with the rest of the paper.
[Fig. 8, top right panel caption] 'T op' should be 'Top'.
[References] The reference 'Bacilieri, A., & Austudillo-Estevez, P.' misspells 'Astudillo' and differs from the in-text citation 'Bacilieri & Austudillo-Estevez, 2023'; please harmonize the spelling and citation style.
[Sec. 3.1, Fig. 10] The claim that DebtStreamness 'captures distinct financial structures not visible through production data' is based on a single figure with roughly similar proportions across DS ranges; a statistical test or a more quantitative comparison would strengthen this conclusion.
[General] The paper does not include a data or code availability statement; given that the empirical claims depend on a specific partial survey and the reconstruction methods, making the code available would improve reproducibility.

Circularity Check

2 steps flagged · score 5.0 of 10

Two of the paper's headline findings—that loops inflate DebtStreamness and that bank-borrowing share predicts DebtStreamness—are direct consequences of the defining equation, while the empirical DS distribution and sector comparisons are not circular.

self definitional [Section 3, 'Component with outliers' (Eqs. 2–3), and Abstract]
"Two main factors contribute to their increased DebtStreamness: i) paths from the banking sector to these three firms are longer due to intermediaries, and ii) the sum in Equation 3 will contain infinite terms, corresponding to paths that repeatedly go through the loop."

Equation (3) defines DSi as 1 plus the weighted sum of creditors' DS values, and Eq. (2) defines it as a sum over all paths weighted by path length. A 2-cycle between two firms generates infinitely many repeated paths, so the presence of a loop is an input to the metric, not an independent empirical discovery. Removing the loop is therefore guaranteed to reduce the computed DS; the paper reports this counterfactual as a finding ('local network motifs such as loops can substantially increase...') when it follows from the definition alone.
self definitional [Section 3, 'Component with outliers' (Eq. 2, Fig. 8, top right)]
"We observe a strong negative correlation between the two quantities, which is around −0.99. This seems to suggest that, for the system at hand, the network structure is not that important; rather, in agreement with the results of Bartolucci et al. (2023, 2025b, 2021) obtained for related centrality measures, knowing the fraction of money that a firm directly borrows from the banking sector is enough to estimate, on average, its DebtStreamness with good approximation."

The first term of Eq. (2) is exactly B_i/D_i, the direct bank-borrowing share, and the remaining terms are nonnegative contributions of longer inter-firm paths; equivalently, summing Eq. (3) over creditors gives DS_i = 1 + (1 − B_i/D_i) times the average DS of creditors. Thus a large bank share mechanically pins DS_i down near 1, so the negative correlation and the statement that the bank share 'is enough to estimate' DS are properties of the defining equation, not an independent empirical regularity. The supporting self-citation to Bartolucci et al. reinforces a conclusion already contained in the metric's definition.

full rationale

Most of the paper is not circular: the path-sum definition (Eq. 2), the equivalent matrix form (Eq. 3), and the derivation in Appendix A are internally consistent mathematical manipulations, and the computed DS values for the Uruguayan network (mean 1.67, peaks near 1, 2, 3, outliers above 20) are empirical quantities obtained from the data. The comparison with sector classifications and the sector-level aggregation are independent of the definition. However, two of the three headline 'findings' are direct consequences of the definition: loops increase DS because Eq. 3 sums repeated loop paths, and bank-borrowing share is the first term of the path sum that mechanically anchors DS. Presenting these as empirical discoveries is self-definitional. The reconstruction robustness checks are not circular, but they only perturb surveyed rows, so the high correlations do not address missing rows for non-surveyed counterparties; this is a data limitation, not an additional circularity. Self-citations to Bartolucci et al. are used for context and are not load-bearing for the main derivation. Overall score reflects partial circularity in the framing of definitional consequences, while the empirical distribution and sector analysis retain independent content.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The central claim relies on the accounting identity that banks and inter-firm lending are the only credit sources, and on the assumption that the partial survey data are representative. No new entities are introduced; DebtStreamness is a derived quantity, not a postulated thing.

free parameters (1)

Log-normal distribution parameters for inter-firm credit amounts = mu approx 11, sigma approx 2
Fitted to the empirical distribution in Fig. 3; used only for descriptive comparison, not in the computation of DebtStreamness.

assumptions (4)

domain assumption All non-bank credit flows are captured by the inter-firm lending matrix L; bonds, equity, and government loans are omitted.
Introduced in Eq. (1) and Sec. 2.1; this accounting identity defines the network and the boundary conditions for DebtStreamness.
domain assumption Banks are the sole primary originator of credit; all credit ultimately flows from the banking sector.
Stated in Sec. 1 as the 'primary energy source' analogy; DebtStreamness measures distance from banks, so any non-bank origin would change the interpretation.
domain assumption The matrix (1-A) is invertible on the analyzed network; firms without any path from the banking sector are excluded.
Sec. 2.2 excludes such firms because their DebtStreamness is undefined; this selection may bias the empirical distribution.
ad hoc to paper The top-three creditors and debtors reported in the survey provide sufficient information to rank firms by DebtStreamness.
This is tested via reconstruction in Sec. 3.2, but the reconstruction methods themselves assume specific distributions for missing credit, which are not guaranteed.

how reviews work

0 comments

Cite this review

Pith. "Pith review of DebtStreamness: An Ecological Approach to Credit Flows in Inter-Firm Networks." pith.science (2026). https://pith.science/paper/PRBQHUF7

@misc{pith2026250501326,
  author       = {Pith},
  title        = {Pith review of: DebtStreamness: An Ecological Approach to Credit Flows in Inter-Firm Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PRBQHUF7}},
  note         = {Machine review of arXiv:2505.01326}
}

read the original abstract

Understanding how credit flows through inter-firm networks is critical for assessing financial stability and systemic risk. In this study, we introduce DebtStreamness, a novel metric inspired by trophic levels in ecological food webs, to quantify the position of firms within credit chains. By viewing credit as the ``primary energy source'' of the economy, we measure how far credit travels through inter-firm relationships before reaching its final borrowers. Applying this framework to Uruguay's inter-firm credit network, using survey data from the Central Bank, we find that credit chains are generally short, with a tiered structure in which some firms act as intermediaries, lending to others further along the chain. We also find that local network motifs such as loops can substantially increase a firm's DebtStreamness, even when its direct borrowing from banks remains the same. Comparing our results with standard economic classifications based on input-output linkages, we find that DebtStreamness captures distinct financial structures not visible through production data. We further validate our approach using two maximum-entropy network reconstruction methods, demonstrating the robustness of DebtStreamness in capturing systemic credit structures. These results suggest that DebtStreamness offers a complementary ecological perspective on systemic credit risk and highlights the role of hidden financial intermediation in firm networks.

Figures

Figures reproduced from arXiv: 2505.01326 by the authors.

**Figure 1.** DebtStreamness (DS) of firm i. This diagram shows the gross debt value of firm i, positioned downstream in the final stage of the debt chain within the production process, relative to banks, which are upstream and act as the credit originators. We show three different toy scenarios from the firm i’s perspective. Scenario I refers to the situation when firm i borrows from the bank (initial node). Scenario II reflects… view at source ↗

**Figure 2.** Main connected components of the inter-firm credit network. The distribution of inter-firm credits is well-fitted by a log-normal distributions of parameters µ ≈ 11 and σ ≈ 2 as shown in [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

**Figure 3.** Distribution of inter-firm credits. The data are well-fitted by a log-normal distributions of parameters µ ≈ 11 and σ ≈ 2. In [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

Figures from the paper (9 more)

**Figure 4.** Figure 4: Fraction of inter-firm credit associated with the top three creditors of surveyed firms. The distribution is quite heterogeneous, but on average the top three creditors of each firm account for about 50% of their inter-firm credit. 7 [PITH_FULL_IMAGE:figures/full_fig_…

**Figure 5.** Figure 5: DebtStreamness of firm network. DebtStreamness (DS) of firms in the Uruguayan interfirm credit network. The average value of DebtStreamness is 1.67, with three outliers with DS > 20. As mentioned earlier (see [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]

**Figure 6.** Figure 6: Left panel: Histogram of the share of inter-firm credit for firms in the largest component. Approximately 55% of firms primarily rely either on inter-firm credit or direct borrowing from banks, while the rest use a mix of both. Right panel: Histogram of DebtStreamness …

**Figure 7.** Figure 7: Visualisation of the inter-firm credit network. [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]

**Figure 8.** Figure 8: Top left panel: Network representation of the component with the three outliers. A node’s size represents its total borrowing, and links indicate lender-borrower relationships. Two of the outliers do not borrow directly from the banking sector and form a loop in the ne…

**Figure 9.** Figure 9: In the left panel of [PITH_FULL_IMAGE:figures/full_fig_p013_9.png]

**Figure 9.** Figure 9: Sector aggregation. Left Panel: DS and bank-debt ratio of sectors. Right Panel Network component visualisation aggregated by sector with the bank node as originator [PITH_FULL_IMAGE:figures/full_fig_p015_9.png]

**Figure 10.** Figure 10: Sector classification by DS. This plot represents the sector classification of firms and their DebtStreamness (DS), categorized into UpStream, Key sector, DownStream, and Others with no classification available across three DS ranges: DS < 1.5, 1.5 ≤ DS ≤ 2.5, and DS …

**Figure 11.** Figure 11: Comparison of DebtStreamness computed from the network with only top three creditors and the partially reconstructed ones. Left panel: comparison with partially reconstructed fully connected network. Right panel: comparison with partially reconstructed sparse network.…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

41 extracted references · 36 canonical work pages

[1]

\ Ulanowicz, R E

abarca2002effects APACrefauthors Abarca-Arenas, L G. \ Ulanowicz, R E. APACrefauthors \ 2002 . The effects of taxonomic aggregation on network analysis The effects of taxonomic aggregation on network analysis . Ecological Modelling 149 3 285--296

work page 2002
[2]

, Carvalho, V M

acemoglu2012network APACrefauthors Acemoglu, D. , Carvalho, V M. , Ozdaglar, A. \ Tahbaz-Salehi, A. APACrefauthors \ 2012 . The network origins of aggregate fluctuations The network origins of aggregate fluctuations . Econometrica 80 5 1977--2016

work page 2012
[3]

\ Pascual, M

allesina2008network APACrefauthors Allesina, S. \ Pascual, M. APACrefauthors \ 2008 . Network structure, predator--prey modules, and stability in large food webs Network structure, predator--prey modules, and stability in large food webs . Theoretical Ecology 1 55--64

work page 2008
[4]

\ Pascual, M

allesina2009googling APACrefauthors Allesina, S. \ Pascual, M. APACrefauthors \ 2009 . Googling food webs: can an eigenvector measure species' importance for coextinctions? Googling food webs: can an eigenvector measure species' importance for coextinctions? PLoS computational biology 5 9 e1000494

work page 2009
[5]

, Chor, D

Antras2012 APACrefauthors Antras, P. , Chor, D. , Fally, T. \ Hillberry, R. APACrefauthors \ 2012 02 . Measuring the Upstreamness of Production and Trade Flows Measuring the upstreamness of production and trade flows . American Economic Review 102 . APACrefDOI doi:10.1257/aer.102.3.412 APACrefDOI

doi:10.1257/aer.102.3.412 2012
[6]

\ Austudillo-Estevez, P

Bacilieri2023 APACrefauthors Bacilieri, A. \ Austudillo-Estevez, P. APACrefauthors \ 2023 03 . Reconstructing firm-level input-output networks from partial information Reconstructing firm-level input-output networks from partial information \ Papers\ \ 2304.00081 . arXiv.org . APACrefURL https://ideas.repec.org/p/arx/papers/2304.00081.html APACrefURL

arXiv 2023
[7]

, Borsos, A

bacilieri2023firm APACrefauthors Bacilieri, A. , Borsos, A. , Astudillo-Estevez, P. \ Lafond, F. APACrefauthors \ 2023 . Firm-level production networks: what do we (really) know Firm-level production networks: what do we (really) know . INET Oxford Working Paper 2023

work page 2023
[8]

Spectrally gapped

Bartolucci_2021 APACrefauthors Bartolucci, S. , Caccioli, F. , Caravelli, F. \ Vivo, P. APACrefauthors \ 2021 11 . “Spectrally gapped” random walks on networks: a Mean First Passage Time formula “spectrally gapped” random walks on networks: a mean first passage time formula . SciPost Physics 11 5 . APACrefURL http://dx.doi.org/10.21468/SciPostPhys.11.5.08...

work page doi:10.21468/scipostphys.11.5.088 2021

Show all 41 references

[9]

, Caccioli, F

Bartolucci2020 APACrefauthors Bartolucci, S. , Caccioli, F. , Caravelli, F. \ Vivo, P. APACrefauthors \ 2023 . Ranking influential nodes in networks from aggregate local information Ranking influential nodes in networks from aggregate local information . Physical Review Resear...

2023
[10]

, Caccioli, F

Silvia2023 APACrefauthors Bartolucci, S. , Caccioli, F. , Caravelli, F. \ Vivo, P. APACrefauthors \ 2025 1 . Correlation between upstreamness and downstreamness in random global value chains Correlation between upstreamness and downstreamness in random global value chains . Jo...

2025
[11]

, Caccioli, F

bartolucci2025upstreamness APACrefauthors Bartolucci, S. , Caccioli, F. , Caravelli, F. \ Vivo, P. APACrefauthors \ 2025 2 . Upstreamness and downstreamness in input--output analysis from local and aggregate information Upstreamness and downstreamness in input--output analysis...

2025
[12]

\ Meli \'a n, C J

bascompte2005simple APACrefauthors Bascompte, J. \ Meli \'a n, C J. APACrefauthors \ 2005 . Simple trophic modules for complex food webs Simple trophic modules for complex food webs . Ecology 86 11 2868--2873

2005
[13]

, Puliga, M

Battiston2012 APACrefauthors Battiston, S. , Puliga, M. , Kaushik, R. , Tasca, P. \ Caldarelli, G. APACrefauthors \ 2012 08 . DebtRank: Too Central to Fail? Financial Networks, the FED and Systemic Risk Debtrank: Too central to fail? financial networks, the fed and systemic ri...

2012 doi
[14]

APACrefauthors \ 2023

BCU2023 APACrefauthors BCU, S. APACrefauthors \ 2023 . Informe sobre el colch\'on de capital contrac\'iclico en Uruguay Informe sobre el colch\'on de capital contrac\'iclico en Uruguay . Banco Central del Uruguay 21 . Analisis para la evaluacion del nivel de capital contracicl...

2023
[15]

APACrefauthors \ 1952 09

Carter1952 APACrefauthors Carter, C F. APACrefauthors \ 1952 09 . Activity Analysis of Production and Allocation Activity Analysis of Production and Allocation . The Economic Journal 62 247 625-628 . APACrefURL https://doi.org/10.2307/2226909 APACrefURL APACrefDOI doi:10.2307/...

1952 doi
[16]

, Mastrandrea, R

cimini2021reconstructing APACrefauthors Cimini, G. , Mastrandrea, R. \ Squartini, T. APACrefauthors \ 2021 . Reconstructing networks Reconstructing networks . Cambridge University Press

2021
[17]

APACrefauthors \ 2005 Spring

Coleman2005 APACrefauthors Coleman, S. APACrefauthors \ 2005 Spring . Free and Costly Trade Credit: A Comparison of Small Firms Free and Costly Trade Credit: A Comparison of Small Firms . Journal of Entrepreneurial Finance 10 1 75-101 . APACrefURL https://ideas.repec.org/a/pep...

2005
[18]

APACrefauthors \ 2022

demirel2022network APACrefauthors Demirel, G. APACrefauthors \ 2022 . Network science for the supply chain: theory, methods, and empirical results Network science for the supply chain: theory, methods, and empirical results . The Digital Supply Chain The digital supply chain \...

2022
[19]

, Williams, R J

dunne2002network APACrefauthors Dunne, J A. , Williams, R J. \ Martinez, N D. APACrefauthors \ 2002 . Network structure and biodiversity loss in food webs: robustness increases with connectance Network structure and biodiversity loss in food webs: robustness increases with con...

2002
[20]

, Golub, B

Elliott2022 APACrefauthors Elliott, M. , Golub, B. \ Leduc, M V. APACrefauthors \ 2022 August . Supply Network Formation and Fragility Supply network formation and fragility . American Economic Review 112 8 2701-47 . APACrefURL https://www.aeaweb.org/articles?id=10.1257/aer.20...

2022 doi
[21]

\ Reddy, K

Gopalan2022 APACrefauthors Gopalan, S. \ Reddy, K. APACrefauthors \ 2022 . What Determines Interfirm Trade Credit? Empirical Evidence from the ASEAN What determines interfirm trade credit? empirical evidence from the asean . ERIA Discussion Paper Series No. 432

2022
[22]

\ Sasidharan, S

Gopalan2020 APACrefauthors Gopalan, S. \ Sasidharan, S. APACrefauthors \ 2020 . Financial liberalization and access to credit in emerging and developing economies: A firm-level empirical investigation Financial liberalization and access to credit in emerging and developing eco...

2020 doi
[23]

\ May, R M

haldane2011systemic APACrefauthors Haldane, A G. \ May, R M. APACrefauthors \ 2011 . Systemic risk in banking ecosystems Systemic risk in banking ecosystems . Nature 469 7330 351--355

2011
[24]

APACrefauthors \ 1973

hannon1973structure APACrefauthors Hannon, B. APACrefauthors \ 1973 . The structure of ecosystems The structure of ecosystems . Journal of theoretical biology 41 3 535--546

1973
[25]

\ Soltas, E

Helper2021 APACrefauthors Helper, S. \ Soltas, E. APACrefauthors \ 2021 . Why the Pandemic Has Disrupted Supply Chains Why the Pandemic Has Disrupted Supply Chains \ Written Materials . The White House . APACrefURL https://www.whitehouse.gov/cea/written-materials/2021/06/17/wh...

2021
[26]

, de Valk, C

ialongo2022reconstructing APACrefauthors Ialongo, L N. , de Valk, C. , Marchese, E. , Jansen, F. , Zmarrou, H. , Squartini, T. \ Garlaschelli, D. APACrefauthors \ 2022 . Reconstructing firm-level interactions in the Dutch input--output network from production constraints Recon...

2022
[27]

, Caccioli, F

LANDABERRY2021100032 APACrefauthors Landaberry, V. , Caccioli, F. , Rodriguez-Martinez, A. , Baron, A. , Martinez-Jaramillo, S. \ Lluberas, R. APACrefauthors \ 2021 . The contribution of the intra-firm exposures network to systemic risk The contribution of the intra-firm expos...

2021
[28]

APACrefauthors \ 1951

Leontief1951 APACrefauthors Leontief, W W. APACrefauthors \ 1951 . Input-Output Economics Input-output economics . Scientific American 185 4 15-21

1951
[29]

APACrefauthors \ 1980

levine1980several APACrefauthors Levine, S. APACrefauthors \ 1980 . Several measures of trophic structure applicable to complex food webs Several measures of trophic structure applicable to complex food webs . Journal of theoretical Biology 83 2 195--207

1980
[30]

, Rockmore, D

LUMSDAINE2021100804 APACrefauthors Lumsdaine, R. , Rockmore, D. , Foti, N. , Leibon, G. \ Farmer, J. APACrefauthors \ 2021 . The intrafirm complexity of systemically important financial institutions The intrafirm complexity of systemically important financial institutions . Jo...

2021
[31]

APACrefauthors \ 1972

may1972will APACrefauthors May, R M. APACrefauthors \ 1972 . Will a large complex system be stable? Will a large complex system be stable? Nature 238 5364 413--414

1972
[32]

, Savoie, C

mcnerney2022production APACrefauthors McNerney, J. , Savoie, C. , Caravelli, F. , Carvalho, V M. \ Farmer, J D. APACrefauthors \ 2022 . How production networks amplify economic growth How production networks amplify economic growth . Proceedings of the National Academy of Scie...

2022
[33]

\ Temurshoev, U

Miller2015 APACrefauthors Miller, R. \ Temurshoev, U. APACrefauthors \ 2015 11 . Output Upstreamness and Input Downstreamness of Industries/Countries in World Production Output upstreamness and input downstreamness of industries/countries in world production . International Re...

2015 doi
[34]

, Pimm, S L

montoya2006ecological APACrefauthors Montoya, J M. , Pimm, S L. \ Sol \'e , R V. APACrefauthors \ 2006 . Ecological networks and their fragility Ecological networks and their fragility . Nature 442 7100 259--264

2006
[35]

, Heesterbeek, J A

neutel2002stability APACrefauthors Neutel, A M. , Heesterbeek, J A. \ De Ruiter, P C. APACrefauthors \ 2002 . Stability in real food webs: weak links in long loops Stability in real food webs: weak links in long loops . Science 296 5570 1120--1123

2002
[36]

APACrefauthors \ 2021

orenstein2021changing APACrefauthors Orenstein, P. APACrefauthors \ 2021 . The changing landscape of supply chain networks: an empirical analysis of topological structure The changing landscape of supply chain networks: an empirical analysis of topological structure . INFOR: I...

2021
[37]

\ Rajan, R G

Petersen2015 APACrefauthors Petersen, M A. \ Rajan, R G. APACrefauthors \ 2015 04 . Trade Credit: Theories and Evidence Trade credit: Theories and evidence . The Review of Financial Studies 10 3 661-691 . APACrefURL https://doi.org/10.1093/rfs/10.3.661 APACrefURL APACrefDOI do...

2015 doi
[38]

APACrefauthors \ 2021

rasmussen1956studies APACrefauthors Rasmussen, P. APACrefauthors \ 2021 . Studies in inter-sectoral relations Studies in inter-sectoral relations . North-Holland; København, Amsterdam, Einar Harcks

2021
[39]

, Johnson, S

sansom2021trophic APACrefauthors Sansom, B. , Johnson, S. \ MacKay, R S. APACrefauthors \ 2021 . Trophic incoherence drives systemic risk in financial exposure networks Trophic incoherence drives systemic risk in financial exposure networks \ . Working Paper

2021
[40]

, Bondavalli, C

scotti2009using APACrefauthors Scotti, M. , Bondavalli, C. , Bodini, A. \ Allesina, S. APACrefauthors \ 2009 . Using trophic hierarchy to understand food web structure Using trophic hierarchy to understand food web structure . Oikos 118 11 1695--1702

2009
[41]

APACrefauthors \ 2012

ulanowicz2012growth APACrefauthors Ulanowicz, R E. APACrefauthors \ 2012 . Growth and development: ecosystems phenomenology Growth and development: ecosystems phenomenology . Springer Science & Business Media

2012

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [1]

\ Ulanowicz, R E

abarca2002effects APACrefauthors Abarca-Arenas, L G. \ Ulanowicz, R E. APACrefauthors \ 2002 . The effects of taxonomic aggregation on network analysis The effects of taxonomic aggregation on network analysis . Ecological Modelling 149 3 285--296

work page 2002

[2] [2]

, Carvalho, V M

acemoglu2012network APACrefauthors Acemoglu, D. , Carvalho, V M. , Ozdaglar, A. \ Tahbaz-Salehi, A. APACrefauthors \ 2012 . The network origins of aggregate fluctuations The network origins of aggregate fluctuations . Econometrica 80 5 1977--2016

work page 2012

[3] [3]

\ Pascual, M

allesina2008network APACrefauthors Allesina, S. \ Pascual, M. APACrefauthors \ 2008 . Network structure, predator--prey modules, and stability in large food webs Network structure, predator--prey modules, and stability in large food webs . Theoretical Ecology 1 55--64

work page 2008

[4] [4]

\ Pascual, M

allesina2009googling APACrefauthors Allesina, S. \ Pascual, M. APACrefauthors \ 2009 . Googling food webs: can an eigenvector measure species' importance for coextinctions? Googling food webs: can an eigenvector measure species' importance for coextinctions? PLoS computational biology 5 9 e1000494

work page 2009

[5] [5]

, Chor, D

Antras2012 APACrefauthors Antras, P. , Chor, D. , Fally, T. \ Hillberry, R. APACrefauthors \ 2012 02 . Measuring the Upstreamness of Production and Trade Flows Measuring the upstreamness of production and trade flows . American Economic Review 102 . APACrefDOI doi:10.1257/aer.102.3.412 APACrefDOI

doi:10.1257/aer.102.3.412 2012

[6] [6]

\ Austudillo-Estevez, P

Bacilieri2023 APACrefauthors Bacilieri, A. \ Austudillo-Estevez, P. APACrefauthors \ 2023 03 . Reconstructing firm-level input-output networks from partial information Reconstructing firm-level input-output networks from partial information \ Papers\ \ 2304.00081 . arXiv.org . APACrefURL https://ideas.repec.org/p/arx/papers/2304.00081.html APACrefURL

arXiv 2023

[7] [7]

, Borsos, A

bacilieri2023firm APACrefauthors Bacilieri, A. , Borsos, A. , Astudillo-Estevez, P. \ Lafond, F. APACrefauthors \ 2023 . Firm-level production networks: what do we (really) know Firm-level production networks: what do we (really) know . INET Oxford Working Paper 2023

work page 2023

[8] [8]

Spectrally gapped

Bartolucci_2021 APACrefauthors Bartolucci, S. , Caccioli, F. , Caravelli, F. \ Vivo, P. APACrefauthors \ 2021 11 . “Spectrally gapped” random walks on networks: a Mean First Passage Time formula “spectrally gapped” random walks on networks: a mean first passage time formula . SciPost Physics 11 5 . APACrefURL http://dx.doi.org/10.21468/SciPostPhys.11.5.08...

work page doi:10.21468/scipostphys.11.5.088 2021