Figure 1: Financial Vulnerability Index and Financial Conditions
Indices. Indices are shown as the standard deviation of their own
historical distribution 1990-present. The Financial Conditions Indices
included in the range of FCIs are: Goldman Sachs U.S. Financial
Condition Index (GSUSFCI), Chicago National Financial Conditions Index
(NFCI), Financial Conditions Impulse on Growth (FCI-G), Kansas City
Financial Stress Index (KCFSI), and the St. Louis Financial Stress Index
(STLFSI). On top of this range is the CBOE Volatility Index (VIX) in red
and the Financial Vulnerability Index (FVI) in Black.
This is a line chart showing data from 1990 to present (approximately end of 2025). The chart displays multiple variables: a range of standardized Financial Conditions Indices (shown as a red shaded area), the CBOE Volatility Index or VIX (red line), and the Financial Vulnerability Index or FVI (black line). The range of FCIs includes the Goldman Sachs U.S. Financial Condition Index, Chicago National Financial Conditions Index, Financial Conditions Impulse on Growth, Kansas City Financial Stress Index, and St. Louis Financial Stress Index. The FVI ranges from 0 to 1, the standardized indices range from -1 to 1.8. The FVI displays a gradual build-up pattern before major crisis periods, peaking before 2000 and around 2007-2008, then declining as crises materialize. In contrast, the FCIs and VIX spike sharply during crisis periods (around 2000, 2008, and 2020), showing coincident behavior with financial stress. The FVI generally builds up gradually in pre-crisis periods and declines before the FCIs spike, while the FCIs remain relatively stable during pre-crisis periods and spike alongside the VIX during crises. By approximately 2025, the FVI shows moderate levels around 0-1 standard deviations.
Source: Goldman Sachs, Federal Reserve Bank of Chicago, Federal Reserve Bank of Kansas City, Federal Reserve Bank of St. Louis, Cboe Global Markets, Federal Reserve Board staff calculations.
Figure 2: Heatmaps of aggregate FVI and its subindices. Red: 0.81-1 percentiles
(Elevated); orange: 0.61-0.8 percentiles (Notable); yellow: 0.41-0.6
(Moderate); green: 0.21-0.4 (Low); blue: 0.01-0.2 (Extremely
Subdued).
(a) Aggregate FVI
(b) Valuation Pressures
(c) Funding Risk
(d) Financial Leverage
(e) Households and Business Borrowing
This figure contains five panels (a through e), each displaying a heatmap with quarterly data from 1990 to approximately 2025. Each panel uses a color scheme to represent different percentile ranges: red (0.81-1 percentiles, labeled "Elevated"), orange (0.61-0.8 percentiles, labeled "Notable"), yellow (0.41-0.6 percentiles, labeled "Moderate"), green (0.21-0.4 percentiles, labeled "Low"), and blue (0.01-0.2 percentiles, labeled "Extremely Subdued").
Panel (a) Aggregate FVI: Shows the overall FVI building up to elevated levels (red) before 2000 and around 2007-2008, then declining to moderate-to-low levels (yellow-green) through the 2010s. The index shows another increase to notable levels (orange) in 2025.
Panel (b) Valuation Pressures: Displays elevated pressures (red) around 2000 during the dot-com bubble period and between 2004 to 2007, subdued and low levels (blue-green) through much of the 2010s, then increasing to notable-elevated levels (orange-red) around 2017 and remained in this range until present.
Panel (c) Funding Risk: Shows elevated levels (red) around 2001-2002 and from 2006 to 2008, followed by a sharp decline to subdued-low levels (blue-green) through the 2010s and early 2020s, with a brief increase to moderate-notable around 2020, then returning to notable levels by 2025.
Panel (d) Financial Leverage: Exhibits elevated levels (red) around 2005-2007, followed by a sustained decline to subdued-low levels (blue-green) through most of the 2010s, an increase to moderate levels (yellow) around 2020 and to notable and elevated levels from 2023 to 2025. by 2025.
Panel (e) Households and Business Borrowing: Displays elevated levels (red) around 2001-2003 and 2007-2011, declining sharply afterward to low-subdued levels (green-blue) through the 2010s and early 2020s, with a temporary increase to moderate levels (yellow) around 2015-2017 and 2021-2022. Current readings are low (2025).
Source: Federal Reserve Board staff calculations.
Figure 3: Fitted skewed-t distributional forecasts of GDP growth conditional on
lagged GDP growth and FVI or NFCI. The vertical line represents the
realization of GDP growth for the targeted quarter. Legend values in
parenthesis report the estimated probabilities of GDP growth being
smaller or equal to the realized value under the corresponding fitted
distribution.
(a) 4-Quarter ahead forecast for 2008:Q1
(b) 4-Quarter ahead forecast for 2008:Q4
(c) 8-Quarter ahead forecast for 2008:Q1
(d) 8-Quarter ahead forecast for 2008:Q4
This figure contains four density plots (panels a through d) showing fitted skewed-t distributions of GDP growth forecasts. Each panel displays two overlapping probability density curves: one conditional on FVI (blue line) and one conditional on NFCI (orange line). A vertical line in each panel represents the realized GDP growth value.
Panel (a) 4-Quarter ahead forecast for 2008: Q1: Shows two density curves with peaks around 2-4 percent GDP growth. The FVI-conditional distribution (blue) assigns a probability of 0.31 to the realized outcome, while the NFCI-conditional distribution (orange) assigns a probability of 0.05. The realized value falls in the left tail of both distributions, around 1.8 percent.
Panel (b) 4-Quarter ahead forecast for 2008: Q4: Displays two density curves with the FVIconditional distribution (blue) assigning a probability of 0.06 and the NFCI-conditional distribution (orange) assigning 0.07 to the realized outcome. Distributions show peaks between 1 and 3percent, with the realized value falling in the extreme left tail around -2.5 percent.
Panel (c) 8-Quarter ahead forecast for 2008: Q1: Shows the FVI-conditional distribution (blue) assigning a probability of 0.33 and the NFCI-conditional distribution (orange) assigning 0.04 to the realized outcome. The distributions peak around 3 percent, with the realized value around 1.8 percent.
Panel (d) 8-Quarter ahead forecast for 2008: Q4: Displays the FVI-conditional distribution (blue) assigning a probability of 0.10 and the NFCI-conditional distribution (orange) assigning approximately 0.00 to the realized outcome. The distributions peak around 2-3 percent, with the realized value falling in the left tail around -0.5 percent.
Note: Realized value refers to actual GDP growth outcome. Legend values in parentheses report the estimated probabilities of GDP growth being smaller than or equal to the realized value under the corresponding fitted distribution.
Source: Bureau of Economic Analysis; Federal Reserve Board staff estimates based on quantile regression models.
Figure 4: Fitted skewed-t distributional forecasts of GDP growth conditional on
lagged GDP growth and FVI or NFCI. The fitted fitted skewed-t
distribution is estimated through quantile regressions.
(a) 4-Quarter ahead forecast: FVI
(b) 4-Quarter ahead forecast: NFCI
(c) 8-Quarter ahead forecast: FVI
(d) 8-Quarter ahead forecast: NFCI
This figure contains four heatmap-style contour plots (panels a through d) showing the evolution of fitted skewed-t distributional forecasts of GDP growth over time. The x-axis displays quarterly time periods, the y-axis shows GDP growth in percent, and the color intensity represents probability density. A vertical black line marks 2008:Q4 (the GFC trough).
Panel (a) 4-Quarter ahead forecast: FVI: Shows a heatmap from approximately 2000 to 2017, with the y-axis ranging from about -6 to 8 percent GDP growth. Higher probability densities (lighter colors) cluster around 1-3 percent growth through most periods. The distribution shifts dramatically downward around 2008-2009, with substantial probability mass appearing in negative growth territory (below -4 percent) during the crisis period. By 2010-2012, the distribution gradually shifts back toward positive growth rates.
Panel (b) 4-Quarter ahead forecast: NFCI: Displays a similar time range and scale as panel (a). The probability mass remains concentrated around 1-3 percent positive growth through most periods. The distribution shows less probability mass in the extreme negative growth region around the 2008:Q4 vertical line compared to panel (a), with lighter colors appearing less frequently in the -4 to -6 percent range during the crisis period.
Panel (c) 8-Quarter ahead forecast: FVI: Shows the same time period with y-axis ranging from -6 to 8 percent. The probability distributions display less dispersion compared to the 4-quarter forecasts. Substantial probability mass extends into negative growth territory during 2008-2010 and the vertical line lie in a region where substantial probability mass is assigned. The distributions gradually normalize toward positive growth by 2011-2012.
Panel (d) 8-Quarter ahead forecast: NFCI: Displays similar patterns to panel (b) but at the 8-quarter horizon. The distribution shows concentrated probability mass around 1-3 percent growth throughout most periods. Around the 2008:Q4 vertical line, the distribution shows no probability mass and much less probability mass around 2007-2008 compared to panel (c).
Note: The fitted skewed-t distribution is estimated through quantile regressions. The vertical line represents 2008:Q4, the trough of the Global Financial Crisis.
Source: Bureau of Economic Analysis; Federal Reserve Board staff estimates based on quantile regression models.
Figure 5: Responses of real consumption expenditure (first row) and fixed
long-term investment (second row) to a business cycle shock, %: a), d)
linear; b), e) high FVI; c), f) low FVI. Regimes are defined by FVI
being above (high) or below (low) 0.5. Shaded areas represent the 90%
(light) and 68% (dark) confidence bands.
(a) Consumption: Linear
(b) Consumption: High FVI
(c) Consumption:Low FVI
(d) LT fixed Inv.: Linear
(e) LT fixed Inv.: High FVI
(f) LT fixed Inv.:Low FVI
This figure contains six impulse response function plots (panels a through f) arranged in two rows, showing responses to a business cycle shock in percent. Each panel displays quarterly horizons on the x-axis (0 to 12 quarters) and percent deviation on the y-axis. Shaded areas represent 90 percent (light) and 68 percent (dark) confidence bands.
Panel (a) Consumption: Linear: Shows the linear response of consumption declining from 0 percent at impact to approximately -1 percent after 4 quarters, remaining around -0.5 to - 1 percent through quarter 8, then gradually returning toward 0 by quarter 12. The 68 percent confidence band (dark shading) indicates statistical significance through about quarter 8.
Panel (b) Consumption: High FVI: Displays a larger consumption response, declining from about -0.2 percent at impact to approximately -1.2 percent after 4 quarters, remaining significantly negative through quarter 8 at about -0.5 percent, then gradually recovering toward 0 by quarter 12. The response is statistically significant (dark shaded area excludes zero) through about quarter 10.
Panel (c) Consumption: Low FVI: Shows a smaller consumption response, declining to about -0.2 percent at quarter 2, then quickly returning toward 0 by quarter 2. The response becomes statistically insignificant (confidence bands include zero) after quarter 1, remaining near 0 through quarter 12.
Panel (d) LT fixed Inv.: Linear: Displays the linear response of long-term fixed investment declining sharply at impact to approximately -2 percent, reaching about -6.5 percent at quarter 5, then gradually recovering by quarter 12. The response is statistically significant throughout the horizon shown.
Panel (e) LT fixed Inv.: High FVI: Shows the largest investment response, declining from about -2 percent at impact to approximately -4 percent at quarter 2, remaining significantly negative around -2 to -4 percent through quarter 7, then gradually recovering toward 0. The response is highly significant throughout.
Panel (f) LT fixed Inv.: Low FVI: Displays a more muted investment response, declining to about -1.5 percent at quarter 1, recovering to around -0.5 percent by quarter 3. The response is not statistically significant (confidence bands include zero) throughout the entire horizon.
Source: Bureau of Economic Analysis; Federal Reserve Board staff estimates based on local projection methods.
Figure 6: Responses to a 100 bps contractionary monetary policy shock: a) FVI
(percentile points); b) NFCI (bps). Shaded areas represent the 90%
(light) and 68% (dark) confidence bands.
(a) Financial Vulnerability Index (FVI)
(b) National Financial Conditions Index (NFCI)
This figure contains two impulse response function plots (panels a and b) showing responses to a monetary policy shock in percentile points. The x-axis displays quarters from 0 to 16, and the y-axis ranges from approximately -60 to 20 percentile points. Shaded areas represent 90 percent (light) and 68 percent (dark) confidence bands.
Panel (a) Financial Vulnerability Index (FVI): Shows the response beginning near 0 at impact, declining gradually to about -10 percentile points by quarter 4, continuing to decline to approximately -25 percentile points by quarter 8, reaching a trough of about -35 percentile points around quarters 10-12. The 68 percent confidence band (dark shading) excludes zero from about quarter 7 onward, indicating statistical significance in the medium to long term.
Panel (b) National Financial Conditions Index (NFCI): Displays a different response pattern, with an immediate increase to about 80-100 basis points at impact (quarter 0), reaching a peak of approximately 120 bps at quarter 1, then declining sharply back toward 0 by quarter 4-5, and remaining near 0 through quarter 12. The response is statistically significant (dark shaded area excludes zero) only in the first 3-4 quarters, showing a shortlived effect.
Source: Bu, Rogers, and Wu (2021) monetary policy shock series; Federal Reserve Board staff estimates based on local projection methods.
Figure 7: Responses of PCE deflator to a 100 bps contractionary monetary policy
shock, log deviations: a) linear, b) high FVI, c) low FVI. Regimes are
defined by FVI being above (high) or below (low) 0.5. Shaded areas
represent the 90% (light) and 68% (dark) confidence bands.
(a) Linear
(b) High FVI
(c) Low FVI
This figure contains three impulse response function plots (panels a through c) showing log deviations of the PCE deflator. The x-axis displays quarters from 0 to 12, and the y-axis ranges from approximately -0.1 to 0.05 (log deviations). Shaded areas represent 90 percent (light) and 68 percent (dark) confidence bands.
Panel (a) Linear: Shows the PCE deflator response beginning -0.005 at impact, declining gradually to about -0.04 (approximately -4 percent) by quarter 4, continuing to decline to approximately -0.07 (about -7 percent) by quarter 8, then remaining at similar levels through quarter 12. The 68 percent confidence band (dark shading) excludes zero from about quarter 6 onward, indicating statistical significance.
Panel (b) High FVI: Displays a more muted response, with the PCE deflator fluctuating around 0 from horizon zero to 5 and from 9 to 12. The response is significant and reaches - 0.04 (approximately -4 percent) only in quarters 6-8. The confidence bands (both light and dark shading) include zero throughout most of the horizon, indicating the response is not statistically significant in the high vulnerability regime.
Panel (c) Low FVI: Shows a stronger response similar to the linear case, declining from - 0.005 at impact to about -0.05 (approximately -5 percent) by quarter 4, continuing to decline to approximately -0.07 (about -7 percent) by quarter 7. The 68 percent confidence band excludes zero from about quarter 5 onward, indicating strong statistical significance.
Source: Bureau of Economic Analysis; Bu, Rogers, and Wu (2021) monetary policy shock series; Federal Reserve Board staff estimates based on local projection methods.
Figure 8: Shapley value decomposition of the FVI. Stacked areas show
each subindex’s contribution (Shapley value \(\omega_{jt}\)) to the aggregate FVI over
time.
This is a stacked area chart showing quarterly data from 1990 to approximately 2025. The yaxis ranges from approximately -0.3 to 1, measuring the FVI level and contributions. The chart displays four stacked colored areas representing the Shapley value contributions of each subindex to the aggregate FVI: Valuation Pressures (red/pink area), Funding Risk (orange area), Financial Leverage (light blue/teal area), and Households and Business Borrowing (purple area). A black line traces the aggregate FVI level across the top of the stacked areas. The stacked areas show how each subindex contributes to the total FVI over time. Before 2000, House and Business Borrowing was the main contributor to FVI. Around 2007-2008, all four subindices show large positive contributions, with Financial Leverage, Funding Risk, and Households and Business Borrowing particularly elevated, driving the aggregate FVI to approximately 1 Following the crisis, House and Business Borrowing and Funding Risk contributions decline substantially, becoming slightly negative or near zero around 2010. Valuation Pressures show renewed increases around 2018. Starting in 2023, there is an increase in all contributions, particularly from Financial Leverage. The aggregate FVI stands at approximately 0.6 in 2025, with relatively balanced contributions across subindices and subdued contribution from Household and business borrowing.
Note: Stacked areas show each subindex's contribution (Shapley value) to the aggregate FVI over time. The black line represents the aggregate FVI level.
Source: Federal Reserve Board staff calculations based on Shapley value decomposition methods.
Figure 9: Shapley value decomposition of subindices - I. Stacked areas show
each variable’s contribution (Shapley value ϕitj)
to the aggregate FVI over time.
(a) Valuation Pressures
(b) Funding Risk
This figure contains two stacked area charts (panels a and b) showing quarterly data from 1990 to approximately 2025. Both panels display contributions of individual variables to their respective subindices using Shapley values.
Panel (a) Valuation Pressures: The y-axis ranges from approximately -0.6 to 1. Multiple colored stacked areas represent individual variable contributions to the Valuation Pressures subindex. A black line traces the subindex level. The chart shows substantial contributions from various valuation metrics building up to approximately 1 before 2000 during the dot-com period, declining afterward, then building again to about 1 around 2007, declining through the crisis. The subindex shows renewed increases around 2020- 2021, reaching approximately 0.8, with notable contributions from housing price ratios and commercial real estate prices and option implied volatility. Most recent readings, around 0.7 in 2025, are related with increased contributions of equity premia and cds spreads.
Panel (b) Funding Risk: The y-axis ranges from approximately -0.5 to 1. Stacked colored areas represent contributions from funding risk variables. The aggregate subindex (black line) shows low levels around 0-0.5 through most of the 1990s and early 2000s, then increases dramatically to approximately 1 around 2007-2008, with large contributions from short-term wholesale funding, runnable liabilities, and other funding measures. Following the crisis, contributions decline sharply, with the subindex remaining near or below 0 through most of the 2010s and early 2020s, indicating substantially reduced funding vulnerabilities. A persistent increase started 2023, associated with bank larger contributions coming from the market depth 2-year on the run Treasury securities. Readings of 0.6 in 2025 are associated with higher contributions from low-friction liquidity and gross short-term debt of financial sector over GDP.
Note: Stacked areas show each variable's contribution (Shapley value) to the respective subindex over time. The black line represents the aggregate subindex level.
Source: Federal Reserve Board staff calculations based on Shapley value decomposition methods.
Figure 10: Shapley value decomposition of subindices - II. Stacked areas show
each variable’s contribution (Shapley value ϕitj)
to the aggregate FVI over time.
(a) Financial Leverage
(b) Households and Business Borrowing
This figure contains two stacked area charts (panels a and b) showing quarterly data from 1990 to approximately 2025. Both panels display contributions of individual variables to their respective subindices using Shapley values.
Panel (a) Financial Leverage: The y-axis ranges from approximately -0.5 to 1. Multiple colored stacked areas represent contributions from financial leverage variables. The aggregate subindex (black line) remains between 0 and 0.5 through the 1990s and early 2000s, then increases sharply to approximately 0.8-1 around 2007-2008, with substantial contributions from measures such as tangible common equity ratios, fire-sale vulnerabilities, gross dealer borrowing. Following the crisis, contributions decline dramatically, with the subindex falling to around -0.5 to 0 through the 2010s, reflecting improved capital positions and deleveraging. Notable negative contributions appear from tangible common equity (indicating improved capital ratios), while other leverage measures moderate. A temporary increase to occurs around 2020 and a more persistent increase around 2023. The current largest contributions are primary dealers borrowing over financing, maturity gaps of banks and hedge funds leverage.
Panel (b) Households and Business Borrowing: The y-axis ranges from approximately -0.6 to 1.2. Stacked colored areas show contributions from household and business borrowing variables. The aggregate subindex (black line) builds gradually from about 0.6 in the 1990s to approximately 1 around 2007-2008, with large contributions from measures including mortgage debt to GDP, credit to GDP, household financial obligations, and lending standards. Following the crisis, contributions decline sharply, with the subindex falling to about 0-0.2 through the 2010s. Another decline occurs through 2020-2021, reaching approximately -0, reflecting increased household savings and improved business finances following pandemic-related deleveraging. By 2025, the subindex has risen back to approximately 0.2, with lending standards contributing positively while household debt measures remain relatively contained. Note: Stacked areas show each variable's contribution (Shapley value) to the respective subindex over time. The black line represents the aggregate subindex level.
Source: Federal Reserve Board staff calculations based on Shapley value decomposition methods.
Figure 11: IRFs of PCE deflator to a 100 bps contractionary monetary policy
shock, log deviations. Regimes defined on FVI subindices: a), b) Asset
Valuations; c), d) Funding Risk ; e), f) Financial Leverage; g), h) HH
and Business Leverage . Shaded areas are 90% (light) 68% (dark)
confidence bands.
(a) High Asset Valuations
(b) Low Asset Valuations
(c) High Funding Risk
(d) Low Funding Risk
(e) High Fin. Leverage
(f) Low Fin. Leverage
(g) High HH and Bus. Borrowing
(h) Low HH and Bus. Borrowing
This figure contains eight impulse response function plots (panels a through h) arranged in four rows, showing log deviations of the PCE deflator. Each panel displays quarters from 0 to 16 on the x-axis and log deviations ranging from approximately -0.05 to 0.03 on the yaxis. Shaded areas represent 90 percent (light) and 68 percent (dark) confidence bands. Each row corresponds to a different FVI subindex, with the left column showing high regime and the right column showing low regime.
Panel (a) High Asset Valuations: Shows a relatively flat response near 0, declining only slightly to about -0.05 to -0.005 by quarters 10-12. The confidence bands include zero throughout most of the horizon.
Panel (b) Low Asset Valuations: Displays a stronger decline from 0 to approximately -0.05 by quarter 8, continuing to about -0.01 by quarter 12, with the 68 percent band excluding zero throughout the entire horizon.
Panel (c) High Funding Risk: Shows negative response, declining to about -0.04 to 0.04 by quarters 8-12, with wide confidence bands that include zero through most periods.
Panel (d) Low Funding Risk: Exhibits a more pronounced decline to approximately -0.05 by quarter 7, reaching about -0.02 by quarter 12, with statistically significant responses throughout the entire horizon.
Panel (e) High Fin. Leverage: Displays a decline to about -0.05 in quarters 11, with confidence bands that marginally exclude zero in later quarters.
Panel (f) Low Fin. Leverage: Shows a stronger response, declining to approximately -0.05 by quarter 7 and reaching about -0.02 to by quarter 12, with clear statistical significance after quarter 2.
Panel (g) High HH and Bus. Borrowing: Exhibits a nearly flat response throughout, remaining very close to 0 with wide confidence bands that clearly include zero across all horizons, indicating no statistically significant effect.
Panel (h) Low HH and Bus. Borrowing: Displays the strongest response among all panels, declining from 0 to approximately -0.07 by quarter 7, reverting back to about -0.015 by quarter 12, with the 68 percent confidence band clearly excluding zero throughout the entire horizon, indicating highly significant and persistent disinflationary effects.
Source: Bureau of Economic Analysis; Bu, Rogers, and Wu (2021) monetary policy shock series; Federal Reserve Board staff estimates based on state-dependent local projection methods.
Figure 12: IRFs of fixed-investment to a negative business cycle shock, %.
Regimes defined on FVI subindices: a), b) Asset Valuations; c), d)
Funding Risk ; e), f) Financial Leverage; g), h) HH and Business
Leverage. Shaded areas are 90% (light) 68% (dark) confidence bands.
(a) High Asset Valuations
(b) Low Asset Valuations
(c) High Funding Risk
(d) Low Funding Risk
(e) High Fin. Leverage
(f) Low Fin. Leverage
(g) High HH and Bus. Borrowing
(h) Low HH and Bus. Borrowing
This figure contains eight impulse response function plots (panels a through h) arranged in four rows, showing percent deviations of fixed long-term investment. Each panel displays quarters from 0 to 12 on the x-axis and percent deviations ranging from approximately -4 to -6 to 2 on the y-axis. Shaded areas represent 90 percent (light) and 68 percent (dark) confidence bands. Each row corresponds to a different FVI subindex, with the left column showing high regime and the right column showing low regime.
Panel (a) High Asset Valuations: Shows investment declining from about -1.8 percent at impact to approximately -4 percent at quarter 2, reaching a trough of about -5 percent around quarter 4, then gradually recovering toward -2 to -3 percent by quarter 12. The response is statistically significant throughout.
Panel (b) Low Asset Valuations: Displays a decline to about -2 percent at quarters on impact, which reaches a though to -5 percent at horizon 5, after which the response reverts back to zero.
Panel (c) High Funding Risk: Exhibits a sharp decline from about -2 percent at impact to approximately -4 percent at quarter 2-4, then gradually recovering toward 0 by quarter 12. The 68 percent band excludes zero throughout most of the horizon.
Panel (d) Low Funding Risk: Shows a more modest decline to about -1.5 percent on impact and -2 percent at quarter 2, with confidence bands widening and approaching zero after quarter 6.
Panel (e) High Fin. Leverage: Displays investment declining to approximately -4 to -5 percent at quarters 2-5, then gradually recovering toward 0 percent by quarter 12, with persistent statistical significance.
Panel (f) Low Fin. Leverage: Shows a smaller initial decline to about -2 to -2.5 percent at quarters 1-2, reaching approximately -4 percent by quarter 6, with confidence bands including zero after quarter 2.
Panel (g) High HH and Bus. Borrowing: Exhibits a large response, declining from about -2 percent at impact, reaching a trough of about -4 percent at quarter 2, then gradually recovering toward 0 percent by quarter 9. The response is highly significant and persistent throughout.
Panel (h) Low HH and Bus. Borrowing: Displays a small response, declining to about -1.5 percent at quarter 2, quickly recovering to approximately -0.5 percent by quarter 4, with confidence bands clearly including zero after quarter 1, indicating the response becomes statistically insignificant in the low regime.
Source: Bureau of Economic Analysis; Federal Reserve Board staff estimates based on state-dependent local projection methods.