Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

Financial instability transition under heterogeneous investments and portfolio diversification

T0 review · 3 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read Uneven investment sizes push a rebalancing financial network across the stability boundary, even where equal-weight models predict safety.

desk verdict A well-executed extension of Corsi et al. that convincingly shows investment-size heterogeneity destabilizes the system, but the paper overstates the safety guarantee of its replica method and leans on an unvalidated matrix proxy. read the letter →

arxiv 2501.19260 v1 pith:NW636F5S submitted 2025-01-31 q-fin.RM cond-mat.dis-nnphysics.soc-ph

Preben Forer , Barak Budnick , Pierpaolo Vivo , Sabrina Aufiero , Silvia Bartolucci , Fabio Caccioli This is my paper · ORCID
classification q-fin.RMcond-mat.dis-nnphysics.soc-ph
keywords systemicriskfinancialcontagionportfoliorebalancingdiversificationrandommatrixtheorylargesteigenvaluereplicamethodheterogeneousinvestments
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

This paper extends an established random-matrix model of financial contagion through overlapping portfolios by letting each institution split its money unevenly between small and large positions, and asks whether this added realism changes the stability verdict. The answer it argues for is yes: the boundary between stable and unstable markets is set by the average largest eigenvalue of the matrix that drives the endogenous part of returns, and investment-size heterogeneity systematically raises that eigenvalue, so a market that looks calm under the usual equal-weight investment rule can be genuinely unstable. The paper also identifies two competing effects of diversification: widening the number of assets each bank holds deepens the market and stabilizes prices at low connectivity, but beyond a threshold it destabilizes the system by creating more overlapping portfolios. Finally, it shows that the standard shortcut of replacing the random matrix by its average badly underestimates the risk, whereas a replica-based calculation tracks the transition much more closely.

What carries the argument

The load-bearing object is the average largest eigenvalue $E[\lambda_{\max}]$ of the random matrix $\Phi = \frac{\eta-1}{\gamma \alpha^2} W W^T$, which encodes the feedback loop in which price changes force portfolio rebalancing, which in turn moves prices; the stability/instability transition is the crossing of the threshold $E[\lambda_{\max}] = 1$. Because the entries of the column-stochastic weight matrix $W$ are correlated through the column normalization, the paper computes this quantity two approximate ways: Method #2 takes the largest eigenvalue of the average matrix $E[\Phi]$, while Method #3 uses a replica representation of the top eigenvalue as the zero-temperature free energy of a partition function (the machinery of [17]) applied to a surrogate matrix in which $W$ is replaced by $c X$ with $c = (1-e^{-\alpha q})/(\alpha q)$, turning the problem into the top-eigenvalue statistics of a diluted Wishart matrix $X X^T$ with independent sparse entries. Population dynamics solves the resulting distributional equations, and direct numerical diagonalization of $\Phi$ serves as the reference standard.

What would settle it

Using the parameter values of Figures 2 and 3 ($q=8$, $\zeta=1.85$, $\sigma_s^2=0.009$, $\sigma_d^2=0.03$, $\gamma=50$), compute the replica method's predicted transition line and compare it cell-by-cell with $E[\lambda_{\max}]$ from numerical diagonalization of $10^4$ realizations of $\Phi$ over the $(\phi, p_B)$ grid; any cell in which the replica method predicts stability while diagonalization gives $E[\lambda_{\max}] > 1$ would falsify the claim that Method #3 never issues a false stable verdict.

Watch

Extended reading notes

Core claim

The central claim, stated on the paper's own terms, is that the typical stability of an investment network is governed by the average largest eigenvalue of $\Phi = \frac{\eta-1}{\gamma \alpha^2} W W^T$, the matrix appearing in the return dynamics $e_t = \Phi(e_{t-1}+\varepsilon_t)$; whenever $E[\lambda_{\max}] > 1$, at least one endogenous return process grows without bound and the market is unstable. With a binary Big/Small investment distribution constrained so that the mean investment per institution is unchanged, the paper finds that heterogeneity raises $E[\lambda_{\max}]$ and can convert a stable homogeneous market into an unstable one: small positions propagate shocks between institutions during rebalancing, while large positions magnify the resulting price moves. Diversification acts non-monotonically, deepening markets at low $q$ and increasing portfolio overlap at high $q$, and the paper shows that the largest-eigenvalue-of-the-average-matrix method (Method #2) underestimates this risk, whereas the replica approximation (Method #3), applied to a surrogate matrix with $W$ replaced by $cX$, $c = (1-e^{-\alpha q})/(\alpha q)$, tracks the numerical transition without ever issuing a false stable verdict.

Load-bearing premise

The replica method's phase boundary assumes that matching the average entry of the weight matrix $W$, whose columns each sum to one, with the scaled matrix $c X$ is enough to reproduce the largest eigenvalue of $W W^T$; the paper tests this only in aggregate, not at the $q = 8$ used for the phase diagrams.

Editorial extensions

If this is right

  • A market that the homogeneous 1/N rule classifies as stable can be unstable once banks are allowed to hold large and small positions at the same average total investment, so risk assessments built on the 1/N assumption are systematically over-optimistic.
  • Raising the diversification parameter $q$ first decreases $E[\lambda_{\max}]$, because more holders per asset means deeper markets and smaller price impact, then increases it through portfolio overlap, producing a U-shaped stability response with a minimum near $q \approx 10$ in the tested regimes.
  • Method #2 underestimates $E[\lambda_{\max}]$ substantially at low connectivity and can misclassify unstable markets as stable, while Method #3 may overestimate the risk but never labels an unstable market as stable, so a regulator using Method #3 would never receive a false all-clear.
  • In the asset-rich regime with $N > M$, representative of globally systemically important banks, each asset is held by very few institutions, so individual trades are a large fraction of the asset's volume and $E[\lambda_{\max}]$ reaches higher values than in the asset-poor regime.

Reading between the lines

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

  • An untested consequence of the model's mechanism is that the destabilizing effect should strengthen as the investment-size distribution becomes heavy-tailed: with the same mean constraint, rare very large positions would amplify the rebalancing sell-off effect, pushing the instability region toward lower heterogeneity values than the binary distribution does.
  • The paper verifies the Method #3 proxy only through an aggregate relative-gap curve; a sharper test is a pointwise comparison of the replica phase boundary and numerical diagonalization on the full $(\phi, p_B)$ grid at $q = 8$, since any cell where the proxy predicts stability while diagonalization gives $E[\lambda_{\max}] > 1$ would falsify the never-a-false-all-clear claim for that parameter se
  • Because the transition is defined by an average eigenvalue, a finite real market near the threshold could cross into instability through fluctuations of $\lambda_{\max}$ even when $E[\lambda_{\max}]$ is slightly below 1; the paper lists eigenvalue fluctuations as future work, and this observation suggests the safe regulatory region is inside, not merely on, the stable side of the boundary.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper studies a random-matrix model of portfolio rebalancing in which M financial institutions invest in N assets, with heterogeneous investment sizes drawn from a two-point distribution. The stability of the linear price dynamics is governed by the average largest eigenvalue E[lambda_max] of the matrix Phi = ((eta-1)/gamma) alpha^2 W W^T, where W is the column-stochastic weight matrix. The authors evaluate E[lambda_max] by direct numerical diagonalization, by the largest eigenvalue of E[Phi] (the "Corsi" method), and by a replica calculation applied to the approximate matrix c X X^T, with c fixed by matching first moments. They find that increasing investment-size heterogeneity (parameter phi) pushes the system toward instability, that increasing diversification q is stabilizing at low q and destabilizing at high q, and that the Corsi approximation underestimates systemic risk whereas the replica method is more accurate and, in the simulations, conservative in the sense of never misclassifying an unstable system as stable.

Significance. If the result holds, the paper makes a useful policy-relevant point: homogeneous-investment models can understate systemic risk, and the stability phase diagram depends non-trivially on the interaction between diversification and heterogeneity. The paper's main qualitative finding is supported by direct numerical diagonalization of the full random matrix, which does not rely on the two analytical approximations, and the comparison of three independent methods is a strength. The authors are also transparent about the failure of the Corsi approximation, which is itself a useful cautionary result. However, the quantitative reliability of the replica-based phase boundaries and the associated safety claim require additional validation, so the paper in its current form is not ready for acceptance.

major comments (3)
  1. [Section 5.2, Eq. (25) and Figure 7] The cX proxy used by method #3 is calibrated only by matching first moments, which does not control the top eigenvalue of W W^T. Figure 7 reports the relative gap in E[lambda_max] only for a single parameter set (B=3, s=0.3, pB=7/27) and does not state the gap at q=8, the connectivity used in the phase diagrams of Figures 2 and 3. Since B and s vary strongly across the (pB, phi) plane (for phi close to 1 and small pB, B becomes very large), the accuracy of the proxy at q=8 must be quantified across that plane before the method #3 transition lines can be used as a quantitative phase boundary.
  2. [Section 4.3] The statement "Using method #3 will never result in a false stable diagnosis" is a universal safety claim, but the evidence is empirical: Figure 3 shows no false-stable region for the finite set of parameters tested at q=8. Nothing in the first-moment matching of Eq. (25) guarantees that the proxy underestimates or overestimates lambda_max in a fixed direction. Please either prove a bound on the approximation error or replace the sentence with a weaker claim, e.g., "no false stable diagnoses were observed in our simulations."
  3. [Appendix C] The derivation of c in Eq. (C.9) uses only the unconditional mean E[W_ij]. However, the exact W is column-stochastic with column sums S_j = sum_i X_ij, and the top eigenvalue of W W^T is sensitive to the distribution of these column sums, because W_ij = X_ij / S_j takes large values when S_j is small. The paper does not check whether the distribution of S_j (or the second moment of W W^T) is reproduced by cX. A second-moment-matched proxy, or a direct report of the S_j statistics, would be needed to justify the replica-based phase boundary.
minor comments (5)
  1. [Table 1] The arrow meanings appear inconsistent: both rows for "system unstable" use a down arrow and both rows for "system stable" use an up arrow, regardless of whether the approximate method is correct. Clarify whether the arrow denotes the actual state or the approximate method's verdict, and adjust the entries so that the four cases are distinguishable.
  2. [Section 3.1 and Eq. (14)] The text says "each asset has E[sum_j X_ij] = q/alpha institutions investing in it on average," but sum_j X_ij is the total monetary investment in the asset, not a count of institutions. The count is sum_j delta_{c_ij,1}; the two quantities coincide only because the mean investment size is normalized to 1. Please rephrase.
  3. [Section 5.2, Eq. (26)] The same symbol Phi is used for the exact matrix in Eq. (7) and for the approximate matrix in Eq. (26); using a distinct notation (e.g., hat_Phi) would avoid confusion.
  4. [Figures 5 and 7] The number of Monte Carlo samples is not stated for Figure 5, although Figures 6 and 7 mention 10^4 realizations; please specify the sample size for all numerical averaging and indicate whether error bars are smaller than the marker size.
  5. [Equations (17)-(19)] The typeset formulas for the diagonal and off-diagonal entries of E[Phi] and for the largest eigenvalue are difficult to parse in the current version, with fractions and square roots not cleanly separated; please ensure the final version renders these formulas unambiguously.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central instability result is benchmarked against direct numerical diagonalization, and the method #3 proxy constant is derived from first moments rather than fitted to the predicted eigenvalue.

full rationale

The paper's claimed derivation chain is self-contained and does not reduce to its inputs by construction. The stability criterion is derived from the linear recursion Eq. (10)-(12): if any eigenvalue of Phi exceeds 1, return volatility diverges, so E[lambda_max] > 1 is a consequence of the stated dynamics, not an assumed conclusion. Heterogeneity is encoded in the parameters B, s, pB, ps under the mean-investment constraint pB B + ps s = 1 (Eq. 4), and the finding that heterogeneity destabilizes is obtained by direct numerical diagonalization (method #1) as well as by two analytical approximations; no parameter is fitted to force the instability boundary. Method #2 is an external Corsi et al. model [1]. Method #3 approximates the column-stochastic weight matrix W by cX, with c = E[W_ij]/E[X_ij] fixed by first-moment matching (Eq. 25 and Appendix C), and then applies the replica result of the companion paper [17]. This is a modeling approximation, not a fitted input called a prediction: c is not calibrated to E[lambda_max] or to the phase boundary. The load-bearing self-citation [17] is independently corroborated within the paper by numerical diagonalization in Figures 3 and 7, which compare the replica prediction against direct matrix diagonalization. The remaining concern that first-moment matching may not preserve top-eigenvalue statistics at q = 8 is a validity/accuracy question, not a circularity, because the central qualitative claim does not collapse when method #3 is set aside. Accordingly, no circular step can be exhibited from the paper's own equations, and the honest finding is no significant circularity.

Assumptions & free parameters 9 free parameters · 4 assumptions · 0 invented entities

The paper's central claim (heterogeneity increases instability) does not depend on fitting any parameter to data. All listed parameters are model inputs chosen for the numerical exploration, and the two analytical methods are checked against exact diagonalization. No new physical entities are introduced. The main relied-upon external input is the replica formalism of [17], a self-cited preprint.

free parameters (9)
  • q (connectivity/diversification parameter) = varies; q=8 in phase diagrams, 0-50 in Figure 5
    Sets the probability q/sqrt(NM) of a bank investing in an asset. Scanned by hand to explore the model; not fitted to any empirical target.
  • B (large investment size) = 3 in Figure 5; determined by phi and pB in phase diagrams
    One of the two investment sizes in the binary distribution p(K) = pB delta(K-B) + ps delta(K-s). Constitutive modeling choice, constrained by pB B + ps s = 1.
  • s (small investment size) = 0.3 in Figure 5; determined by phi and pB in phase diagrams
    The other investment size; constrained by pB B + ps s = 1.
  • pB (probability of a large investment) = 7/27 in Figure 5; scanned in phase diagrams
    Weight of the large investment in the binary mixture; ps = 1 - pB.
  • gamma (asset liquidity) = 50
    Liquidity constant in the price impact model of Eq. (A.1); set by hand following the numerical setting of Corsi et al.
  • zeta (risk appetite) = 1.85
    Constant in the Value-at-Risk constraint that pins down regulatory leverage eta in Eq. (8); chosen by hand.
  • sigma_s^2 (systematic volatility) = 0.009
    Systematic risk component in the portfolio volatility formula (A.26); chosen by hand.
  • sigma_d^2 (diversifiable volatility) = 0.03
    Idiosyncratic risk component in (A.26); chosen by hand.
  • N and M (numbers of assets and banks) = N=200 or 400, M=300
    Chosen to realize the two regimes alpha = sqrt(N/M) used in the study.
assumptions (4)
  • domain assumption The price impact model e_{i,t} = (1/gamma)(d_{i,t}/chi_{i,t}) and the Value-at-Risk based leverage rule (Eqs. A.1 and A.23-A.26) are valid descriptions of institutional trading.
    Inherited from Corsi et al. [1]; the paper extends the model without re-deriving or empirically validating these micro-foundations. Invoked in Appendix A.
  • domain assumption All institutions have comparable asset sizes, so the diagonal matrix Q_{t-1} can be replaced by the average market size, giving Phi = (eta-1)/gamma alpha^2 W W^T.
    Used in Appendix A (around Eq. A.18-A.22) to factor Q_{t-1}/\bar{A}* ~ 1. If institution sizes are heterogeneous, the evolution matrix is not simply W W^T.
  • standard math The replica result for the top eigenvalue of diluted Wishart matrices in [17] applies to the heterogeneous (binary K) XX^T ensemble treated here.
    Method #3 imports this result as a black box. [17] is a companion preprint by the same authors; the paper does not prove the equations for the binary case.
  • standard math The first-order Taylor approximations for the mean and variance of ratios (Appendix B) are accurate enough for Method #2.
    Standard approximation techniques; the paper tests their consequences against numerics and finds they underestimate risk.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Financial instability transition under heterogeneous investments and portfolio diversification." pith.science (2026). https://pith.science/paper/NW636F5S

@misc{pith2026250119260,
  author       = {Pith},
  title        = {Pith review of: Financial instability transition under heterogeneous investments and portfolio diversification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NW636F5S}},
  note         = {Machine review of arXiv:2501.19260}
}
read the original abstract

We analyze the stability of financial investment networks, where financial institutions hold overlapping portfolios of assets. We consider the effect of portfolio diversification and heterogeneous investments using a random matrix dynamical model driven by portfolio rebalancing. While heterogeneity generally correlates with heightened volatility, increasing diversification may have a stabilizing or destabilizing effect depending on the connectivity level of the network. The stability/instability transition is dictated by the largest eigenvalue of the random matrix governing the time evolution of the endogenous components of the returns, for which different approximation schemes are proposed and tested against numerical diagonalization.

Figures

Figures reproduced from arXiv: 2501.19260 by the authors.

Figure 1
Figure 1. Figure schematically illustrating the bipartite network setup of our model. [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Stability analysis of the financial model comparing the Corsi method (method #2) and direct diagonalization (method [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Stability analysis of the financial model comparing the replica method (method #3) and direct diagonalization [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Figure showing the evolution of the parameters [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Figures showing the evolution of the average largest eigenvalue of the matrix [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Graph showing the relative gap between λ˜max as predicted by Equation (19) and the numerical results from direct diagonalization, for different scales of matrices. The parameters are the same as [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]
Figure 7
Figure 7. Figure 7: Graphs showing the change in relative gap as a function of [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. The Extremity Premium: Sentiment Regimes and Adverse Selection in Cryptocurrency Markets

    q-fin.ST 2026-02 reject novelty 5.0 of 10

    Extreme sentiment regimes show higher estimated spreads and uncertainty than neutral ones in Bitcoin data, but the effect is sensitive to controls and overlaps mechanically with volatility.

Reference graph

Works this paper leans on

62 extracted references · 62 canonical work pages · cited by 1 Pith paper

  1. [17]

    Budnick B., Forer, P., Vivo P., Aufiero S., Bartolucci S., & Caccioli F. (2024). Top eigenvalue statistics of diluted Wishart matrices, Preprint

    work page 2024

  2. [1]

    Corsi, F., Marmi, S., & Lillo, F. (2013). When Micro Prudence Increases Macro Risk: The Destabilizing Effects of Financial Innovation, Leverage, and Diversification. Operations Research, Volume 64, Number 5, Pages 1073–88

    work page 2013

  3. [2]

    Gai, P., & Kapadia, S. (2010). Contagion in financial networks. Bank of England Working Papers , Number 383

    work page 2010

  4. [3]

    Caccioli, F., Catanach, T., & Farmer, J. (2011). Heterogeneity, correlations and financial contagion. Advances in Complex Systems. Volume 15, Issue Supp02, Pages 1250058–58 16

    work page 2011

  5. [4]

    A theory of systemic risk and design of prudential bank regulation

    Acharya, V., (2009). A theory of systemic risk and design of prudential bank regulation. Journal of Financial Stability , Volume 5, Issue 3, Pages 224–255

    work page 2009

  6. [5]

    Huang, X., Vodenska, I., Havlin, S., & Stanley, H. E. (2013). Cascading failures in bi-partite graphs: model for systemic risk propagation. Scientific reports, 3(1), 1219

    work page 2013

  7. [6]

    Caccioli, F., Shrestha, M., Moore, C., & Farmer, J. D. (2014). Stability analysis of financial contagion due to overlapping portfolios. Journal of Banking & Finance , Volume 46, Pages 233–245

    work page 2014

  8. [7]

    O., & Landerretche, O

    Gourinchas, P., Valdes, R. O., & Landerretche, O. M. (2001). Lending Booms: Latin America and the World. Econom ´ ıa, Volume 1, Number 2, Pages 47–99

    work page 2001

Show all 62 references
  1. [8]

    G., & Terrones, M

    Mendoza, E. G., & Terrones, M. E. (2012). An Anatomy of Credit Booms and Their Demise. National Bureau of Economic Research Working Paper Series , Number 18379

    2012

  2. [9]

    Borio, C., & Drehmann, M. (2009). Assessing the Risk of Banking Crises – Revisited. Bank for International Settlements Quarterly Review, March 2009

    2009

  3. [10]

    M., & Rogoff, K

    Reinhart, C. M., & Rogoff, K. S. (2009). The Aftermath of Financial Crises. American Economic Review , Volume 99, Number 2, Pages 466–72

    2009

  4. [11]

    Schularick, M., & Taylor A. (2012). Credit Booms Gone Bust: Monetary Policy, Leverage Cycles, and Financial Crises, 1870–2008. American Economic Review, Volume 102, Number 2, Pages 1029–61

    2012

  5. [12]

    Adrian, T., & Shin, H. S. (2009). Money, liquidity, and monetary policy. American Economic Review, Volume 99, Number 2, Pages 600–605

    2009

  6. [13]

    Greenwood, R., Landier, A., & Thesmar, D. (2015). Vulnerable banks. Journal of Financial Economics , Volume 115, Issue 3, Pages 471–485

    2015

  7. [14]

    J., & Gruber M

    Elton, E. J., & Gruber M. J. (1997). Modern portfolio theory, 1950 to date. Journal of Banking & Finance , Volume 21, Issue 11–12, Pages 1743-1759

    1997

  8. [15]

    S., & Zigrand, J.-P

    Danielsson, J., Shin, H. S., & Zigrand, J.-P. (2004). The impact of risk regulation on price dynamics. Journal of Banking & Finance, Volume 28, Issue 5, Pages 1069–1087

    2004

  9. [16]

    Wagner, W. (2011). Diversification at financial institutions and systemic crises. Journal of Financial Intermediation , Volume 20, Issue 3, Pages 163–182

    2011

  10. [18]

    S., & Zigrand, J.-P

    Danielsson, J., Shin, H. S., & Zigrand, J.-P. (2009). Risk appetite and endogenous risk. Financial Markets Group Discus- sion Paper, Number 647

    2009

  11. [19]

    Adrian, T., & Shin, H. S. (2010). Liquidity and leverage. Journal of Financial Intermediation , Volume 19, Issue 3, Pages 418–437

    2010

  12. [20]

    Adrian, T., & Shin, H. S. (2014). Procyclical leverage and value-at-risk. Review of Financial Studies , Volume 27, Issue 2, Pages 373–403

    2014

  13. [21]

    Adrian, T., Moench, E., & Shin, H. S. (2011). Financial intermediation, asset prices, and macroeconomic dynamics. Federal Reserve Bank of New York Staff Reports , Number 422

    2011

  14. [22]

    Adrian, T., & Boyarchenko, N. (2012). Intermediary leverage cycles and financial stability. Federal Reserve Bank of New York Staff Reports, Number 567

    2012

  15. [23]

    Shleifer, A., & Vishny, R. W. (1992). Liquidation values and debt capacity: A market equilibrium approach. Journal of Finance, Volume 47, Number 4, Pages 1343–1366

    1992

  16. [24]

    S., & Xiong, W

    Kyle, A. S., & Xiong, W. (2001). Contagion as a wealth effect. Journal of Finance , Volume 56, Issue 4, Pages 1401–1440

    2001

  17. [25]

    Cont, R., & Wagalath, L. (2011). Running for the exit: Distressed selling and endogenous correlation in financial markets. Mathematical Finance, Volume 23, Issue 4, Pages 718–741

    2011

  18. [26]

    Cont, R., & Wagalath, L. (2012). Fire sales forensics: Measuring endogenous risk. Mathematical Finance, Volume 23, Issue 4, Pages 773–792

    2012

  19. [27]

    D., & Geanakoplos, J

    Thurner, S., Farmer, J. D., & Geanakoplos, J. (2012). Leverage causes fat tails and clustered volatility. Quantitative Finance, Volume 12, Issue 5, Pages 695–707

    2012

  20. [28]

    D., Foti, N

    Caccioli, F., Farmer, J. D., Foti, N. J., & Rockmore, D. N. (2015). Overlapping portfolios, contagion, and financial stability. Journal of Economic Dynamics and Control , Volume 51, Pages 50–63

    2015

  21. [29]

    Duarte, F., & Eisenbach, T. M. (2013). Fire-sale spillovers and systemic risk. Federal Reserve Bank of New York Staff Reports, Number 645

    2013

  22. [30]

    & Battiston, S

    Tasca, P. & Battiston, S. (2011). Diversification and Financial Stability. CCSS Working Paper , Number 11–001

    2011

  23. [31]

    Lillo, F., & Pirino, D. (2015). The impact of systemic and illiquidity risk on financing with risky collateral. Journal of Economic Dynamics and Control , Volume 50, Pages 180-202

    2015

  24. [32]

    A., Hommes, C

    Brock, W. A., Hommes, C. H., & Wagener, F. O. (2009). More hedging instruments may destabilize markets. Journal of Economic Dynamics and Control , Volume 33, Issue 11, Pages 1912–1928

    2009

  25. [33]

    Caccioli, F., Marsili, M., & Vivo, P. (2009). Eroding market stability by proliferation of financial instruments. European Physical Journal B , Volume 71, Issue 4, 467–479

    2009

  26. [34]

    G., & May, R

    Haldane, A. G., & May, R. M. (2011). Systemic risk in banking ecosystems. Nature, Volume 469, Pages 351–355

    2011

  27. [35]

    Stein, J. C. (1998). An adverse-selection model of bank asset and liability management with implications for the trans- mission of monetary policy. The RAND Journal of Economics , Volume 29, Number 3, Pages 466–486

    1998

  28. [36]

    S., & Gertler, M

    Bernanke, B. S., & Gertler, M. (1989). Agency costs, net worth, and business fluctuations. American Economic Review, Volume 79, Issue 1, Pages 14–31

    1989

  29. [37]

    S., Gertler, M., & Gilchrist, S

    Bernanke, B. S., Gertler, M., & Gilchrist, S. (1996). The financial accelerator and the flight to quality.Review of Economics and Statistics , Volume 78, Number 1, Pages 1–15. 17

    1996

  30. [38]

    S., Gertler, M., & Gilchrist, S

    Bernanke, B. S., Gertler, M., & Gilchrist, S. (1999). The financial accelerator in a quantitative business cycle framework. Handbook of Macroeconomics, Volume 1, Part C, Pages 1341–1393

    1999

  31. [39]

    Kiyotaki, N., & Moore, J. (1997). Credit cycles. Journal of Political Economy , Volume 105, Number 2, Pages 211–248

    1997

  32. [40]

    C., Pichler, A., & Wozabal, D

    Pflug, G. C., Pichler, A., & Wozabal, D. (2012). The 1 /N investment strategy is optimal under high model ambiguity. Journal of Banking & Finance , Volume 36, Issue 2, Pages 410–417

    2012

  33. [41]

    DeMiguel, V., Garlappi, L., & Uppal, R. (2009). Optimal Versus Naive Diversification: How Inefficient is the 1/N Portfolio Strategy? Review of Financial Studies , Volume 22, Issue 5, Pages 1915–1953

    2009

  34. [42]

    Duarte, F., & Eisenbach, T. M. (2015). Fire-sale Spillovers and Systemic Risk. The Journal of Finance , Volume 76, Issue 3, Pages 1251–1294

    2015

  35. [43]

    Shin, C., & White, M. (2020). Fire-Sale Vulnerabilities of Banks: Bank-Specific Risks under Stress and Credit Drawdowns. FEDS Notes , 2020-10-08

    2020

  36. [44]

    Feinstein, Z., & Halaj, G. (2023). Interbank Asset-Liability Networks with Fire Sale Management. European Central Bank Working Paper Series , Number 2806

    2023

  37. [45]

    Coen, J., Lepore, C., & Schaanning, E. (2019). Taking Regulation Seriously: Fire Sales Under Solvency and Liquidity Constraints. Bank of England Research Paper Series , Number 793

    2019

  38. [46]

    Cont, R., & Schaanning, E. (2017). Fire Sales, Indirect Contagion and Systemic Stress Testing. Norges Bank Working Papers, Number 2

    2017

  39. [47]

    Mazzarisi, P., Lillo, F., & Marmi, S. (2018). When Panic Makes You Blind: A Chaotic Route to Systemic Risk. Journal of Economic Dynamics and Control , Volume 100, Pages 176–199

    2018

  40. [48]

    Globally Systemically Important Banks: updated assessment methodology and the higher loss absorbency requirement, Documents, Number 59

    Bank for International Settlements (2013). Globally Systemically Important Banks: updated assessment methodology and the higher loss absorbency requirement, Documents, Number 59

    2013

  41. [49]

    Nagao, T., & Tanaka, T. (2006). Spectral density of sparse sample covariance matrices.Journal of Physics A: Mathematical and Theoretical, Volume 40, Number 19, Pages 4973–4987

    2006

  42. [50]

    K¨ uhn, R. (2008). Spectra of sparse random matrices. Journal of Physics A: Mathematical and Theoretical , Volume 41, Number 29, Page 295002

    2008

  43. [51]

    A., Vivo, P., & K¨ uhn, R

    Susca, V. A., Vivo, P., & K¨ uhn, R. (2021). Cavity and replica methods for the spectral density of sparse symmetric random matrices. SciPost Physics , Lecture Notes 33

    2021

  44. [52]

    Bouchaud, J., & Cont, R. (1998). A Langevin approach to stock market fluctuations and crashes. The European Physical Journal B - Condensed Matter and Complex Systems , Volume 6, Pages 543–550

    1998

  45. [53]

    Delpini, D., Battiston, S., Caldarelli, G., & Riccaboni, M. (2019). Systemic risk from investment similarities. PLoS ONE, 14(5), Pages 1–15

    2019

  46. [54]

    Bardoscia, M., Battiston, S., Caccioli, F., & Caldarelli, G. (2015). DebtRank: A Microscopic Foundation for Shock Propagation. PLoS ONE, 10(7), Pages 1–13

    2015

  47. [55]

    Aymanns, C., & Farmer, J. D. (2014). The Dynamics of the Leverage Cycle. Journal of Economic Dynamics and Control , Volume 50, Pages 155–179

    2014

  48. [56]

    Basel III Leverage Ratio Framework and Disclosure Requirements, Documents, Number 63

    Bank for International Settlements (2014). Basel III Leverage Ratio Framework and Disclosure Requirements, Documents, Number 63

    2014

  49. [57]

    Aramonte, S., Schrimpf A., & Shin H. S. (2023). Margins, debt capacity, and systemic risk. BIS Working Papers, Number 1120

    2023

  50. [58]

    Aramonte, S., Schrimpf, A., & Shin, H. S. (2021). Non-bank financial intermediaries and financial stability. BIS Working Papers, Number 972

    2021

  51. [59]

    Mokkelbost, P. B. (1971). Unsystematic Risk Over Time. Journal of Financial and Quantitative Analysis , Volume 6, Number 2, Pages 785–796

    1971

  52. [60]

    Ben-Horim, M., & Levy, H. (1980). Total Risk, Diversifiable Risk and Nondiversifiable Risk: A Pedagogic Note. Journal of Financial and Quantitative Analysis , Volume 15, Number 2, Pages 289–297

    1980

  53. [61]

    Markowitz, H. M. (1959). Portfolio Selection: Efficient Diversification of Investments. Yale University Press

    1959

  54. [62]

    backstop

    Harville, D. A. (1998). Matrix Algebra From a Statistician’s Perspective. Springer. 18 Appendix A. Derivation of Eqs. (6) and (7) In order to study how trading affects asset prices, one models the endogenous component of price move- ment ei,t through a simple linear price impa...

    1998

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.