Skip to content
LB Research
Black-and-white view looking up between Wall Street buildings, with the New York Stock Exchange facade and American flags.
Back to insights

Research insight · Issue 009

Too Big to Fail Became Too Complex to Fail

Systemic importance is not one ranking.

Published
Data through
16 September 2026
Download canonical PDF
ContentsExecutive summary

Executive summary

Systemic importance is not one ranking.

  1. 01HY OAS ranks 9th in structural strength but 3rd in robust bridge importance.
  2. 02HYG is the strongest shock transmitter.
  3. 03VIX ranks far higher on downside consequence than on transmission.
Structural strength rank
HY OAS: 9
Robust bridge importance rank
HY OAS: 3
HYG transmission rank
1
VIX downside consequence rank
2

Why it matters

A node can be important because it connects markets, bridges clusters, transmits shocks or responds most strongly to stress. Those roles do not always belong to the same asset.

Analysis 01

Systemic importance has more than one dimension

The same assets do not dominate every dimension of systemic importance. HYG ranks highly on structure and transmission, while SPY combines strong bridge importance with high downside consequence. Percentiles are calculated cross-sectionally across the 11-instrument universe for each dimension. HY OAS shows the clearest divergence. Its structural importance is relatively low, yet its bridge importance is high. The result suggests that systemic importance is better understood as a set of distinct roles than as a single ranking.

Figure 1. Systemic importance across structure, bridge position, transmission and downside consequence.

HYG has structure and transmission percentiles of 100, bridge importance of 91 and consequence of 82. SPY has bridge importance and consequence percentiles of 100. HY OAS has structure 27, bridge 82, transmission 27 and consequence 27. VIX has transmission 64 and consequence 91.

Figure 1

Systemic importance has more than one dimension

Caption & downloads

Figure 1. Systemic importance across structure, bridge position, transmission and downside consequence.

HYG has structure and transmission percentiles of 100, bridge importance of 91 and consequence of 82. SPY has bridge importance and consequence percentiles of 100. HY OAS has structure 27, bridge 82, transmission 27 and consequence 27. VIX has transmission 64 and consequence 91.

Download PNG

Analysis 02

Being connected and being a bridge are different roles

Most highly connected assets also rank relatively high in bridge importance. HY OAS is the exception. It sits much lower on structural connectedness but remains one of the most important bridges in the system. A node can matter because of where it sits, not just how many strong links it has.

Figure 2. Structural connectedness versus bridge importance across the cross-market network.

SPY has structural connectedness percentile 82 and bridge importance 100; HYG has 100 and 91. HY OAS has structural connectedness 27 but bridge importance 82, above many more strongly connected instruments.

Figure 2

Being connected and being a bridge are different roles

Caption & downloads

Figure 2. Structural connectedness versus bridge importance across the cross-market network.

SPY has structural connectedness percentile 82 and bridge importance 100; HYG has 100 and 91. HY OAS has structural connectedness 27 but bridge importance 82, above many more strongly connected instruments.

Download PNG

Analysis 03

Some bridge positions survive the network definition

SPY and HYG remain bridges across every threshold tested. HY OAS retains a bridge role across three of four network definitions, despite ranking much lower on ordinary structural strength. A bridge is more convincing when its importance survives changes in network specification.

Figure 3. Persistence of bridge importance across alternative correlation thresholds.

SPY and HYG have non-zero betweenness at 100% of tested thresholds, HY OAS at 75%, LQD at 50% and the 2-year yield at 25%. All other instruments are at 0%.

Figure 3

Some bridge positions survive the network definition

Caption & downloads

Figure 3. Persistence of bridge importance across alternative correlation thresholds.

SPY and HYG have non-zero betweenness at 100% of tested thresholds, HY OAS at 75%, LQD at 50% and the 2-year yield at 25%. All other instruments are at 0%.

Download PNG

Analysis 04

Bridge position does not imply shock transmission

HYG is the strongest current net transmitter, followed by SPY and LQD. HY OAS is different. Despite its importance as a structural bridge, it currently receives more transmission than it sends. Structural position and directional transmission capture different forms of systemic importance.

Figure 4. Current net shock transmission across the cross-market system.

HYG is the largest current net transmitter at +35.1, followed by SPY at +22.0, LQD at +21.2 and the 10-year yield at +11.6. HY OAS is a net receiver at −13.6, while IG OAS is the largest net receiver at −24.2.

Figure 4

Bridge position does not imply shock transmission

Caption & downloads

Figure 4. Current net shock transmission across the cross-market system.

HYG is the largest current net transmitter at +35.1, followed by SPY at +22.0, LQD at +21.2 and the 10-year yield at +11.6. HY OAS is a net receiver at −13.6, while IG OAS is the largest net receiver at −24.2.

Download PNG

Analysis 05

The transmission hierarchy survives the forecast horizon

The transmission ordering is unchanged across all three forecast horizons. HYG remains first, followed by LQD, SPY and the 10-year yield. The directional ranking is not sensitive to the forecast horizon used.

Figure 5. Shock-transmission ranks across 5-, 10- and 20-step GFEVD horizons.

The rank order is identical at H=5, H=10 and H=20: HYG, LQD, SPY, 10-year yield, VIX, 2-year yield, TLT, DXY, HY OAS, GLD and IG OAS.

Figure 5

The transmission hierarchy survives the forecast horizon

Caption & downloads

Figure 5. Shock-transmission ranks across 5-, 10- and 20-step GFEVD horizons.

The rank order is identical at H=5, H=10 and H=20: HYG, LQD, SPY, 10-year yield, VIX, 2-year yield, TLT, DXY, HY OAS, GLD and IG OAS.

Download PNG

Analysis 06

Transmission and downside consequence identify different nodes

High transmission does not automatically imply high downside consequence. Percentiles are calculated cross-sectionally across the 11-instrument universe for each dimension. VIX ranks much higher on downside consequence than on transmission, while HYG remains prominent across both dimensions. How a node spreads stress and how it responds to stress are different systemic roles.

Figure 6. Shock transmission versus downside consequence across the cross-market system.

SPY has transmission percentile 82 and downside consequence 100; VIX has 64 and 91, while HYG has 100 and 82. DXY has transmission 36 and consequence 9, and HY OAS has 27 on both dimensions.

Figure 6

Transmission and downside consequence identify different nodes

Caption & downloads

Figure 6. Shock transmission versus downside consequence across the cross-market system.

SPY has transmission percentile 82 and downside consequence 100; VIX has 64 and 91, while HYG has 100 and 82. DXY has transmission 36 and consequence 9, and HY OAS has 27 on both dimensions.

Download PNG

Methodology

How the analysis is constructed

Sample period
19 September 2023–16 September 2026
Coverage
11 cross-market instruments spanning equities, credit, rates, volatility, gold and the U.S. dollar.
Data source
Yahoo Finance, accessed through yfinance. Federal Reserve Economic Data (FRED).
Data through
16 September 2026
Cross-market universe
  • DXY, GLD, HYG, LQD, SPY, TLT and VIX — Yahoo Finance.
  • HY OAS, IG OAS, the 2-year Treasury yield and the 10-year Treasury yield — FRED.

Transformations

Market prices and indices are converted to daily log returns. Credit spreads and Treasury yields are converted to daily changes.

Network structure

Rolling 120-observation correlation networks are estimated using absolute correlations. Structural importance is measured using connection strength and betweenness centrality. Bridge robustness is tested at correlation thresholds of 0.35, 0.40, 0.45 and 0.50. A persistent bridge is defined as having non-zero betweenness at three or more of the four tested thresholds.

Shock transmission

Directional connectedness is estimated using a VAR(1) and row-normalised generalised FEVD. The baseline forecast horizon is 10 steps, with robustness checks at 5 and 20 steps.

Downside consequence

Stress days are defined using the lower tail of SPY returns. Asset responses are standardised so downside consequence can be compared across different market variables. Robustness is tested at the 10th, 15th and 20th percentiles of SPY returns.

Research workflow

Data
Daily market data for 11 cross-market instruments
Transformations
Log returns for market prices and indices; daily changes for spreads and yields
Rolling window
120 observations
Structural network
Absolute-correlation network
Network threshold
Baseline |ρ| ≥ 0.40
Structural importance
Connection strength and betweenness centrality
Threshold robustness
Repeated at 0.35, 0.40, 0.45 and 0.50
Transmission model
VAR(1), no deterministic trend
Shock decomposition
Row-normalised generalised FEVD
Forecast horizon
Baseline H = 10
Horizon robustness
H = 5, 10 and 20
Downside state
Lower tail of SPY returns
Consequence measure
Standardised asset response on downside days
Stress robustness
10th, 15th and 20th percentiles
Synthesis
Structure, bridge position, transmission and downside consequence treated as separate systemic roles

Structural connectedness, bridge position, shock transmission and downside consequence are treated as distinct dimensions of systemic importance. For cross-dimensional comparison, each measure is converted to a cross-sectional percentile rank across the 11-instrument universe.

Research implications

What the evidence shows

The evidence in this report shows that systemic importance cannot be reduced to one ranking.

  • 01Structural connectedness, bridge position, shock transmission and downside consequence identify different roles.
  • 02HY OAS ranks relatively low on structural strength but remains an important bridge.
  • 03HYG is the strongest current shock transmitter.
  • 04VIX ranks far higher on downside consequence than on transmission.
  • 05Bridge importance persists for some assets across alternative network thresholds.
  • 06Transmission rankings remain stable across 5-, 10- and 20-step GFEVD horizons.

Together, these results suggest that systemic importance is role-dependent. A node may matter because it connects many parts of the system, because it links otherwise distinct clusters, because it transmits shocks, or because it responds most strongly when stress arrives. Those roles do not always belong to the same asset. Monitoring several dimensions together therefore gives a more complete view of systemic structure than relying on a single centrality or transmission ranking.

Publication record

Sources and disclosures

Data

  • Yahoo Finance, accessed through yfinance.
  • Federal Reserve Economic Data (FRED), Federal Reserve Bank of St. Louis.
  • Cross-market universe comprising DXY, GLD, HYG, LQD, SPY, TLT, VIX, HY OAS, IG OAS, the 2-year Treasury yield and the 10-year Treasury yield.
  • Data through 16 September 2026.

Methodology reference

  • Diebold, F. X. and Yılmaz, K. (2014). “On the network topology of variance decompositions: Measuring the connectedness of financial firms.” Journal of Econometrics, 182(1), 119–134.
  • Pesaran, H. H. and Shin, Y. (1998). “Generalized impulse response analysis in linear multivariate models.” Economics Letters, 58(1), 17–29.
  • Freeman, L. C. (1977). “A set of measures of centrality based on betweenness.” Sociometry, 40(1), 35–41.
  • Kou, G., Chao, X., Peng, Y., Alsaadi, F. E. and Herrera-Viedma, E. (2019). “Machine learning methods for systemic risk analysis in financial sectors.” Technological and Economic Development of Economy, 25(5), 716–742.

Software

Python, pandas, NumPy, statsmodels, NetworkX, matplotlib, fredapi and yfinance.

This publication has been prepared by LB Research for informational and educational purposes only. It does not constitute investment advice, investment research as defined under applicable regulation, a recommendation, or an offer to buy or sell any financial instrument.

The analysis is based on publicly available market data and the methodologies described within this publication. While reasonable care has been taken in preparing this report, no representation or warranty is made regarding the accuracy, completeness or timeliness of the information presented.

Any opinions expressed reflect the author's judgement at the publication date and may change without notice. Past performance is not indicative of future results.

Readers remain solely responsible for their own investment decisions.

© 2026 LB Research. All rights reserved. Questions or feedback: Contact LB Research

Download canonical PDF
Explore LB Analyst

In development: extending this research into cross-market structure and risk intelligence.