Figure 1: Share of debt maturing within the next year
This figure is a bar chart comparing the share of firm-quarters and share of maturing debt split by 10 percent. Roughly 0.4 firm-quarters have between 0-10 percent of the share in maturing debt. Between 10 to 90 percent, the share of firm-quarters steadily decreases from around 0.1 to 0.05. Share of firm-quarters spikes to 0.2 between 90-100 percent of the share in maturing debt.
Note: The figure shows the distribution of the share of debt which matures within the next twelve months across all firm-quarters of listed U.S. non-financial firms for 1995Q1–2017Q4 from Compustat.
Figure 2: Investment response to a contractionary monetary policy shock
This figure depicts two line graphs. The graph on the left depicts the average investment response, with percentage points on the y-axis from -5 to 1 and quarters since shock from 0 to 12 on the x-axis. The blue line steadily declines from 0 to -3 across the 12 quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, starting near 0 in quarter 0 and growing between -0.5 and -4.5 by quarter 12. The graph on the right shows the differential investment response, with percentage points on the y-axis from -0.5 to 0.2 and quarters since shock from 0 to 12 on the x-axis. The blue line depicts the estimated $$\beta_1^{h}$$ coefficients using the baseline specification in equation (2.3). This line starts at zero then slowly declines for the first 5 quarters to around -0.05, then declines more rapidly to -0.2 in quarter 8, finally increasing back to around -0.05 in quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 0.1 percentage points above and below the blue line. The thicker orange dashed line depicts the extended specification in equation (2.3). This line follows the lower end of the light blue band. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but grows to about 0.2 percentage points above and below the thick orange dashed line.
Note: Panel (a) shows the estimated $$\alpha_1^{h}$$ coefficients using the local projection in equation (2.2). The estimates are standardized to show the response of capital growth to a one standard deviation increase in $$\epsilon^{mp}_t$$. Panel (b) shows the estimated $$\beta_1^{h}$$ coefficients using the baseline specification in equation (2.3) and the extended specification in equation (2.4). The estimates are standardized to show the differential response of capital growth to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with a one standard deviation higher maturing bond share. Shaded areas (and outer dashed lines) indicate 95% confidence bands two-way clustered by firms and quarters.
Figure 3: Credit spread response to a contractionary monetary policy shock
This figure depicts two line graphs. The graph on the left depicts the average credit spread response, with basis points from -40 to 60 on the y-axis and quarters since shock from 0 to 12 on the x-axis. The blue line begins at zero then slightly decreases to around -5 in quarter 2, then steadily begins to increase up to 20 basis points by quarter 8, then remains around 20 basis points for the remaining quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but growing to about 20 basis points above and below the blue line. The graph on the right shows the differential credit spread response, with basis points from -4 to 8 on the y-axis and quarters since shock from 0 to 12 on the x-axis. The blue line depicts the estimated $$\beta_1^h$$ coefficients using the baseline specification in equation (2.3). This line starts at 1, increases to 3 by quarter 8 then fluctuates between 2 and 3 for the remaining quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 1 but grows to about 2 basis points above and below the blue line. The thicker orange dashed line depicts the extended specification in equation (2.3) and follows the same path as the blue line for the first 9 quarters, but steadies to 2 between quarters 9 and 12. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 1 but growing to about 4 basis points above and below the thick orange dashed line.
Note: Panels (a) and (b) are analogous to panels (a) and (b) of Figure 2 when the left-hand side of equations (2.2), (2.3), and (2.4) is replaced by changes in credit spreads, respectively.
Figure 4: Differential response associated with higher maturing bond share
This figure shows 4 line graphs. The graph on the top left shows debt, with percentage points from -1 to 0.5 on the y-axis and quarters since shock from 0 to 12 on the x-axis. The blue line begins at 0 then slowly declines to -0.5 in quarters 5 to 8, then grows to around -0.25 in quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 0.5 percentage points above and below the blue line. The graph on the top right shows sales, with percentage points from -0.6 to 0.2 on the y-axis and quarters since shock from 0 to 12 on the x-axis. The blue line stays at 0 until quarter 3, then steadily declines to -0.2 by quarter 6, where it remains through quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 0.2 percentage points above and below the blue line. The graph on the bottom left shows employment, with percentage points from -0.6 to 0.2 on the y-axis and quarters since shock from 0 to 12 on the x-axis. The blue line begins slightly below 0 and declines to -0.3 by quarter 7, then fluctuates between -0.2 and -0.4 for the remaining quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting around 0 but grows to about 0.3 percentage points above and below the blue line. The graph on the bottom right shows cost of goods sold, with percentage points from -0.6 to 0.2 on the y-axis and quarters since shock from 0 to 12 on the x-axis. The blue line begins slightly at 0 and steadily declines to around -0.3 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting around 0 but grows to about 0.3 percentage points above and below the blue line.
Note: The figure shows the estimated $$\beta_1^h$$ coefficients using the extended specification in equation (2.4), with the left-hand side being log changes in debt, sales, employment, and cost of goods sold, respectively. The $$\beta_1^h$$ estimates are standardized to show the differential response (approx. in p.p.) to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with a one standard deviation higher ($$\mathcal{M}_{it}$$ − $$\mathcal{M}_{i}$$). Shaded areas indicate 95% confidence bands two-way clustered by firms and quarters.
Figure 5: Timing
The flow chart demonstrates the timing of the firm problem. Starting from the left the figure starts at time t. An arrow that points right then begins, with the first point being “given $$z_{it}, k_{iy}, b^S_{it}, b_{it}$$.” The next point is “aggregate state $$S_t$$ is realized, choose $$l_{it}$$ and $$y_{it}$$.” The next point is “random draw of εit. The arrow then splits. The bottom arrow has a point for “default” and then ends at “endogenous exit”. The upper arrow has a point for “pay debt obligations” and then splits again. The bottom arrow ends at “exogenous exit.” The upper arrow has a point for “given $$z_{it+1}$$, choose $$\epsilon_{it}$$, $$k_{it+1}$$, $$b^S_{it+1}$$, $$b^L_{it+1}$$” and then points to $$t+1$$.
Figure 6: Firm variables conditional on size
The figure shows 4 bar charts. In each chart there are 4 sets of bar charts with 2 bars each. Each set represents the size quartiles of the firms, with the first quartile on the left. Each set has a blue bar that depicts the data and a red bar that depicts the model. The top left chart depicts firm leverage in percent, with the y-axis spanning from 0 to 100. The first 2 quartiles are similar with the blue bar being around 10 and the red bar being around 20. The third quartile has a slightly bigger blue bar at 20 and the red bar around 40. For the fourth quartile, both the blue and red bars are around 30. The top right chart depicts the credit spread on long-term debt in percent, with the y-axis spanning from 0 to 10. The blue bars for the first 3 quartiles are around 4, and the red bars are slightly lower, ranging between 2 and 3. For the fourth quartile, both the blue and red bars are near 2. The bottom left chart depicts share of debt due within a year in percent, with the y-axis spanning from 0 to 100. The blue bars decline about in size for each quartile, with the first quartile being around 50 and the third and fourth quartiles being around 10. The red bars follow the same pattern with the first two quartiles being around 40 and the last two quartiles being around 20. The bottom right chart depicts firm age in quarters since IPO, with the y-axis spanning from 0 to 100. The blue bars show the firm age for all firms is between 20 and 40. The red bars are close to each of the blue bars for the first 3 quartiles but is around 70 for the fourth quartile.
Note: For each variable, median values are shown by size quartile. The data sample is 1995–2017. Firm-level data on size (measured by capital), leverage, the share of debt due within a year, and age (quarters since IPO) is from Compustat. Firm-level credit spreads are computed using data from Compustat and FISD. Empirical median values are shown with 95% confidence intervals. Model moments are computed from the stationary distribution of the model using the 'post-IPO sample'. See Appendix D.1 and D.2 for details.
Figure 7: Aggregate response to a contractionary monetary policy shock
This figure shows 6 line charts. The top left chart shows the changes of GDP (blue line), consumption(dashed orange line), and investment (dashed green line), with the percent deviation on the y-axis ranging from -2 to 0 and the quarters since shock on the x-axis ranging from 0 to 12. All three lines follow the same pattern, starting negative, and quickly increasing to 0 where they remain after quarter 4. The top middle chart shows the nominal short-term rate $$i$$ (blue line) and the nominal long-term rate $$i^L$$ (dashed orange line), with the percentage point deviation on the y-axis ranging from 0 to 0.3 and the quarters since shock on the x-axis ranging from 0 to 12. Both lines begin positive and decline quickly to 0 where they remain after quarter 4, with the blue line starting at 0.3 and the dashed orange line starting near 0.05. The top right chart shows the real interest rate $$r$$ (blue line) and inflation $$\pi$$ (dashed orange line), with the percentage point deviation on the y-axis ranging from -2 to 1 and the quarters since shock on the x-axis ranging from 0 to 12. The blue line begins at 1 and quickly declines to zero by quarter 4, and the dashed orange line starts near -1.5 and quickly increases to zero by quarter 4. Both lines stay at zero after quarter 4. The bottom left chart shows the changes of $$p$$ (blue line), $$Q$$ (dashed orange line), and $$w $$(dashed green line), with the percent deviation on the y-axis ranging from -2 to 0 and the quarters since shock on the x-axis ranging from 0 to 12. All three lines follow the same pattern, starting negative, and quickly increasing to 0 where they remain after quarter 4. The bottom middle chart shows the Leverage (blue line) and the LTD-share (dashed orange line), with the percentage point deviation on the y-axis ranging from -0.05 to 0.2 and the quarters since shock on the x-axis ranging from 0 to 12. The blue line stays at zero for quarters 0 and 1, then increases to around 0.05 by quarter 4, then slowly begins to decline to slightly above 0 in quarter 12. The orange line jumps from 0 to 0.15 between quarters 0 and 1, then declines to 0.05 by quarter 4 and continues to decline to be slightly less than zero by quarter 12. The bottom right chart shows the Default rate (blue line), Average STD spread (dashed orange line), and Average LTD spread (dashed green line), with the percentage point deviation on the y-axis ranging from 0 to 0.06 and the quarters since shock on the x-axis ranging from 0 to 12. The blue and orange lines follow the same pattern, beginning between 0.04 and 0.06 and declining across the 12 quarter to near 0.01. The green line begins slightly above 0 and declines to 0 by quarter 12.
Note: The nominal short-term rate $$i$$, the nominal long-term rate $$i^L$$, the real interest rate $$r$$, and inflation $$\pi$$ are annualized. Leverage (debt over capital) and the long-term debt share (LTD-share) are cross-sectional averages. The default rate is annual. The short-term credit spread (STD spread) and the long-term credit spread (LTD spread) are cross-sectional averages. See Appendix D.1 for details.
Figure 8: Differential investment response associated with
$$\mathcal{M}_{it}$$
This figure is a line chart with two lines showing the differential investment response associated with $$\mathcal{M}_{it}$$ for the data and the model. On the y-axis is percentage points from -0.4 to 0.2, and on the x-axis is quarters since shock from 0 to 12. The blue line representing the data starts at zero then slowly declines for the first 5 quarters to around -0.05, then declines more rapidly to -0.2 in quarter 8, finally increasing back to around -0.05 in quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 0.1 percentage points above and below the blue line. The dotted red line representing the model starts at 0 and rapidly decline to -0.2 in quarter 1, then slowly increases to slightly below 0 in quarter 12.
Note: The red dotted line shows the estimated $$\beta_1^h$$ coefficients based on equation (4.2) using simulated model data. The $$\beta_1^h$$ estimates are standardized to capture the differential cumulative capital growth response (in p.p.) to a one standard deviation (30bp) increase in the nominal interest rate i associated with a one standard deviation higher $$\mathcal{M}_{it}$$. The blue solid line shows the empirical baseline estimates from Figure 2(b) together with 95% confidence bands.
Figure 9: Heterogeneous responses to a contractionary monetary policy shock
This figure shows 6 line charts. Each chart has a blue line that depicts Low $$\mathcal{M}_{it}$$ and a red dotted line that depicts High $$\mathcal{M}_{it}$$ over 12 quarters since a shock (x-axis).
The top left figure depicts capital, with the y-axis being the percentage point deviation from -0.4 to 0.1. The blue line jumps from 0 to 0.05 from quarter 0 to quarter 1, then declines to less than zero and remains there after quarter 4. The red line jumps down from 0 to -0.3 and slowly increases to -0.1 by quarter 12. The top middle figure depicts bond market revenue, with the y-axis being the percentage point deviation from -0.06 to 0. Both the blue line and the red dotted line begin negative and quickly increase, becoming close to zero in quarter 8 and remaining there. The red dotted line begins more negatively than the blue line.
The top right figure depicts debt issuance, with the y-axis being the percentage point deviation from -0.06 to 0. Both the blue line and the red dotted line begin negative and quickly increase, becoming close to zero in quarter 8 and remaining there. The red dotted line begins more negatively than the blue line.
The bottom left figure depicts average credit spread, with the y-axis being basis point deviation from 0.5 to 3.5. Both the blue line and the red line begin positive, with the red line starting at 3 and the blue line starting at 1.5. Both decline slowly throughout the 12 quarters with the red line ending slightly below 1 and the blue line ending near 0.5.
The bottom middle figure depicts outstanding long-term debt, with the y-axis being the percentage point deviation from 0 to 0.5. Both the blue line and the red dotted line begin at 0 and quickly increase to near 0.45. Both lines then decline slowly, with the red line declining to 0.3 and the blue line declining to 0.25. The bottom right figure depicts the default rate, with the y-axis being the basis point deviation from 0 to 2.5. Both the blue line and red line begin positive, with the blue line starting at 1 and the red line starting at 2.5, and slowly decline to near 0.5 by quarter 12.
Note: The panels show the effect of an unexpected one standard deviation (30bp) increase in the nominal interest rate i for firms below and above the median maturing bond share $$\mathcal{M}$$ at the time of the shock. The panels show average firm-level changes in (a) capital, (b) bond market revenue (relative to pre-shock firmlevel capital), (c) gross debt issuance, $$b^{S'} + (b^{L'}-(1-\gamma)b^{L}/\pi)$$ (relative to pre-shock firm-level capital), (d) the average of firms’ short-term and long-term credit spread (weighted by the firm-level share of short-term and long-term debt), (e) the stock of outstanding long-term debt b, and (f) the annualized default rate.
Figure 10: Decomposition of differential investment response associated with
$$\mathcal{M}$$
This figure contains two charts. The chart on the left is a line chart that shows the differential responses under experiments. The y-axis shows the percentage points between -0.2 and 0.05, and the x-axis shows the quarters since shock from 0 to 12. There are 4 lines: benchmark (dotted red), cash transfer (solid orange), real debt (dashed green), and cash transfer + real debt (dashed blue). Each line starts at zero. At quarter 1, each line jumps down to a negative value, with the red line being the most severe around -0.18 and blue line being the least severe around -0.25. Each line then increases and steadies around 0, with the blue and green lines becoming positive.
The chart on the right is a bar chart that shows the decomposition into channels, the y-axis ranges from 0% to 100%. The red bar is the benchmark and is 100%. The orange bar is the roll-over risk and is slightly below 40%. The green bar is the debt overhang and is slightly below 50%. The blue bar is both the roll-over risk and debt overhang and is 80%.
Note: In panel (a), the red dotted line shows the differential investment response to a contractionary monetary policy shock associated with a one standard deviation higher maturing bond share in the model (cf. Figure 8). The orange solid line shows the corresponding differential response in the cash transfer model experiment, which we use to offset roll-over risk. The green dashed line shows the corresponding differential response in the real debt model experiment, which we use to offset changes in debt overhang. The blue dash-dotted line shows the differential response in a model which combines both model experiments. In panel (b), we show the share of the peak impulse response function (at horizon h = 1) that can be attributed to roll-over risk, changes in debt overhang, or the combination of both. The shares are defined as relative reductions in the impulse responses under the model experiments compared to the benchmark.
Figure 11: Average investment response with exogenous variation in maturing bond share
This line chart contains a blue line representing the benchmark and a dotted red line representing the exogenously higher $$\mathcal{M}$$. The y-axis shows the percentage point deviation from -0.4 to 0, and the x-axis shows the quarters since shock from 0 to 12. The blue line declines from 0 to -0.15 percentage points in quarter 1, then slowly increases to slightly above -0.1 in quarter 12. The red dotted line declines from 0 to -0.35 percentage points in quarter 1, then quickly increases, becoming slightly above -0.1 in quarter 12.
Note: The Figure shows the average capital growth response to a contractionary monetary policy shock for the selected firm sample. The blue solid line shows the average response in the benchmark model. The red dashed line shows the average response given an exogenously higher level of $$\mathcal{M}$$ in the initial period.
Figure 12: Aggregate effects of unconventional and conventional monetary policy
This figure contains 6 line charts. Each chart shows conventional monetary policy in light blue and unconventional monetary policy in dark blue for 0 to 12 quarters since a shock.
The top left chart shows GDP, with the y-axis being percent deviation from 0 to 1. Both the light and dark blue lines start positive, with the light blue begging at 0.75 and the dark blue line beginning around 0.5. The lines decline to zero where they remain beginning near quarter 4.
The top middle chart shows investment, with the y-axis being percent deviation from 0 to 2. Both the light and dark blue lines start positive, with the light blue begging at 2 and the dark blue line beginning at 1. The lines decline to zero where they remain beginning near quarter 4.
The top right chart shows inflation, with the y-axis being percent point deviation from 0 to 1.5. Both the light and dark blue lines start positive, with the light blue beginning at 1.5 and the dark blue line beginning at 0.75. The lines decline to zero where they remain beginning near quarter 4.
The bottom left chart shows interest rates, with the y-axis being percentage point deviation from -0.3 to 0. The blue lines represent the nominal short-term rate i. The dark blue line stays at 0. The light blue line begins at -0.3 and quickly increases to around -0.05 in quarter 8 where it levels out until quarter 12. The graph additionally has dashed orange lines for the nominal long-term rate $$i^L$$, with the dark orange line being unconventional monetary policy and the light orange line being conventional monetary policy. Both orange lines follow the same pattern, starting negative at -0.05, then quickly returning to near 0 in quarter 3 and staying there. The light orange line is slightly lower than the dark orange line.
The bottom middle chart shows LTD share, with the y-axis being percentage point deviation from 0 to 6. The light blue line remains around 0. The dark blue line spike to slightly below 6 in quarter 1, then declines quickly to 2 around period 4, then slowly declines to around 1 by quarter 12.
The bottom right chart shows the default rate, with the y-axis being the basis point deviation from -20 to 5. The light blue line begins around -5 then slowly increases to slightly below 0 by quarter 12. The dark blue line starts at -5, then jumps to -18 in quarter 1, then rapidly increases to 5 by quarter 5. The line remains near 5 through quarter 12.
Note: This figure shows impulse responses to an expansionary unconventional monetary policy shock (dark lines) and compares it with impulse responses to an expansionary conventional monetary policy shock (light lines). The size of the unconventional monetary policy shock is chosen such that it generates the same fall in the long-term nominal interest rate $$i^L$$ as a conventional expansionary monetary policy shock in the benchmark model studied above. The nominal short-term rate $$i$$, the nominal long-term rate $$i^L$$, and inflation are annualized. LTD share is the cross-sectional average of the long-term debt share. The default rate is annual. See Appendix D.1 for details.
Figure 13: Comparison of unconventional and conventional monetary policy
The figure shows a bar graph that compares conventional monetary policy (lighter shade for each bar pairing) and unconventional monetary policy (darker shade for each bar pairing). The first 3 bar pairings show the percentage point from 0 to 2.5. The first pairing shows GDP (blue): the light bar is around 0.75 and the dark bar is around 0.5. The second pairing shows investment (orange): the light bar is slightly higher than 2 and the dark bar is slightly higher than 1. The third pairing shows inflation (green): the light bar is slightly lower than 1.5 and the dark bar is around 0.75. The last 2 bar pairings show the basis points from -40 to 0. The first pairing shows the short-term rate $$i_t$$ (green). The light bar is -30, and there is no dark bar. The final pairing shows the long-term rate $$i_t^L$$ (green). Both the light and dark bars are near -5.
Note: This figure shows peak impulse responses (at horizon h = 0) to an expansionary conventional monetary policy (MP) shock (light bars) and compares it with the impulse responses to an expansionary unconventional monetary policy shock (dark bars). See notes to Figure 12.
Figure A.1: Time series
This figure contains two line charts. The chart on the left depicts monetary policy shocks, with the y-axis being the basis points from -20 to 10 and the x-axis being the years 1995 to 2020. The line is very volatile, having many large negative changes, notably around 1996, 2002, 2007, 2009, and 2016. There are less positive changes with the largest growing to about 7 basis points in 2003. The chart on the right depicts credit spreads, with the y-axis being the percentage points from 0 to 10 and the x-axis being the years 1995 to 2020. The blue line represents the volume-weighted average spread and the orange dotted line represents the Gilchrist-Zakrajsek spread. Both lines are volatile, but mostly remain between 1 and 4. The exception is in 2008, where both lines reach above 8.
Note: Panel (a) shows the baseline monetary policy shock series at quarterly frequency as described in Section 2.1. In panel (b), the solid line shows the volume-weighted cross-sectional average credit spread based on Refinitiv data as described above. The dashed line shows the Gilchrist and Zakrajšek (2012) series of credit spreads.
Figure B.1: Differential response of other variables associated with higher
$$\mathcal{M}_{it}$$ using baseline local projection
This figure contains 4 line charts. Each chart shows the responses of a different variable between 0 and 12 quarters. The top left chart shows debt, with the y-axis being percentage points between -1 and 0.5. The blue line begins slightly below zero, lowers to around -0.25 by quarter 7, then increases to its quarter 0 value again by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, starting near 0 in quarter 0 and growing to be about 0.5 percentage points above and below the blue line by quarter 12.
The top right chart shows sales, with the y-axis being percentage points between -0.6 and 0.2. The blue line begins around zero for the first 2 quarters then lowers to -0.2 in quarter 6, then slightly increases to about -0.1 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, starting near 0 in quarter 0 and growing to be about 0.2 percentage points above and below the blue line by quarter 12.
The bottom left chart shows employment, with the y-axis being percentage points between -0.6 and 0.2. The blue line begins around zero for the first 2 quarters then lowers to below -0.2 in quarter 7, then fluctuates between -0.1 and -0.2 for the remaining quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, starting near 0 in quarter 0 and growing to be about 0.2 percentage points above and below the blue line by quarter 12.
The bottom right chart shows costs of goods sold, with the y-axis being percentage points between -0.6 and 0.2. The blue line begins around zero for the first 2 quarters then declines slowly to -0.2 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, starting near 0 in quarter 0 and growing to be about 0.2 percentage points above and below the blue line by quarter 12.
Note: The figure shows the estimated βh1 coefficients based on equation (2.3), but where the left-hand side is $$\Delta^{h+1} log (debt)_{it+h}$$ in panel (a), $$\Delta^{h+1} log (sales)_{it+h}$$ in panel (b), $$\Delta^{h+1} log (employment)_{it+h}$$ in panel (c), and $$\Delta^{h+1} log (cost of goods sold)_{it+h}$$ in panel (d). The $$\beta_1^h$$ estimates are standardized to capture the differential response to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with a one standard deviation higher maturing bond share. Shaded areas indicate 95% confidence bands two-way clustered by firms and quarters.
Figure B.2: Differential investment and credit spread responses associated with maturing bond share for alternative monetary policy shocks accounting for public and central bank information effects
This figure shows 4 line charts, each with two lines. The blue line represents the baseline specification, and the thick orange dashed line represents the extended specification. The top row uses the shocks of Miranda-Agrippino and Ricco (2021). The left chart shows the estimated coefficients of investment, with the y-axis being the percentage point from -0.6 to 0.2 and the x-axis being the quarters since shock from 0 to 12. The blue line begins at 0 and slowly declines to around -0.5 in quarter 9, then begins to increase, ending below 0 in quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 0.2 percentage points above and below the blue line. The thicker orange dashed line follows the lower end of the light blue band. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but grows to about 0.2 percentage points above and below the thick orange dashed line.
The right chart shows the estimated coefficients of the spread, with the y-axis being the basis points from -5 to 10 and the x-axis being the quarters since shock from 0 to 12. Both the blue and orange lines are slightly volatile but remain between 2 and 4 throughout the 12 quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 2 basis points above and below the blue line. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but grows to about 3 basis points above and below the thick orange dashed line.
The bottom row uses the shocks of Bauer and Swanson (2023). The left chart shows the estimated coefficients of investment, with the y-axis being the percentage point from -0.5 to 0.1 and the x-axis being the quarters since shock from 0 to 12. The blue line begins at 0 and slowly declines to around -0.2 in quarter 8, then begins to slightly increase, ending at -0.15 in quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 0.2 percentage points above and below the blue line. The thicker orange dashed line follows the same pattern as the blue line, but reaches -0.3 in quarter 8 and remains around this value until quarter 12. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but grows to about 0.2 percentage points above and below the thick orange dashed line.
The right chart shows the estimated coefficients of the spread, with the y-axis being the basis points from -4 to 6 and the x-axis being the quarters since shock from 0 to 12. Both the blue and orange lines are slightly volatile but decrease from around 2 to near 0 between quarters 0 and 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 2 basis points above and below the blue line. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but grows to about 3 basis points above and below the thick orange dashed line.
Note: The figure shows the estimated $$\beta_1^h$$ coefficients based on the baseline local projection (2.3) using various alternative monetary policy shocks $$\epsilon^{mp}_t$$ . In panels (a) and (b), we use the shocks of Miranda-Agrippino and Ricco (2021). In panels (c) and (d) we use the shocks of Bauer and Swanson (2023). The local projection for bond spreads interacts the regressors with a Great Recession (2008Q3-2009Q2) dummy and the figure plots the non-crisis coefficients. The $$\beta_1^h$$ estimates are standardized to capture the differential response to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with a one standard deviation higher $$\mathcal{M}_{it}$$. Shaded areas indicate 95% confidence bands two-way clustered by firms and quarters.
Figure B.3: Differential investment and credit spread response associated with maturing bond share based on alternative samples
This figure shows 4 line charts, each with two lines. The blue line represents the baseline specification, and the thick orange dashed line represents the extended specification. Each figure’s x-axis is the quarters since shock from 0 to 12.
The top row uses observations pre-Great Recession. The left chart shows the estimated coefficients of investment, with the y-axis being the percentage point from -0.6 to 0.2. The blue line begins at 0 for the first 5 quarters then declines to near -0.2 in quarter 9, then linearly increases back to 0 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 0.2 percentage points above and below the blue line. The thicker orange dashed line follows the lower end of the light blue band. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but grows to about 0.2 percentage points above and below the thick orange dashed line.
The right chart shows the estimated coefficients of the spread, with the y-axis being the percentage points from -5 to 10. Both the blue and orange lines have volatile increases from 0 to near 4 by quarter 8 then remains between 1 and 4 for the remaining quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 2 percentage points above and below the blue line. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but grows to about 3 percentage points above and below the thick orange dashed line.
The bottom left panel excludes observations from the great recession and shows the estimated coefficients of investment, with the y-axis being the percentage point from -0.6 to 0.2. The blue line begins at 0 and slowly declines to around -0.2 in quarter 8, then begins to slightly increase, ending at -0.15 in quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 0.2 percentage points above and below the blue line. The thicker orange dashed line follows the same pattern as the blue line but is about -0.05 below the blue line after quarter 0. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but grows to about 0.25 percentage points above and below the thick orange dashed line.
The bottom right chart has no great recession dummies and shows the estimated coefficients of the spread, with the y-axis being the percentage points from -5 to 15. Both the blue and orange lines begin around 2, increase to around 4 in quarter 1 then decrease to below 0 in quarters 4 through 9. They then increase to around 4 in quarter 10, then decrease to near 0 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line and the thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line. Both bands are around 4 percentage points below and above their respective lines.
Note: The figure shows the estimated $$\beta_1^h$$ coefficients based on the baseline local projection (2.3) and extended local projection (2.4), using alternative samples. Panels (a) and (c) use only observations until 2008Q2. Panel (b) excludes monetary policy shocks between 2008Q3 and 2009Q2. Panel (d) uses the full sample but does not include Great Recession dummies as in the specification in the main text. The $$\beta_1^h$$ estimates are standardized to capture the differential response to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with a one standard deviation higher maturing bond share. Shaded areas (and outer dashed lines) indicate 95% confidence bands two-way clustered by firms and quarters.
Figure B.4: Differential investment response associated of firms with and without maturing bonds compared to non-bond-issuing firms
This figure is a line chart with 3 lines, each with bands for confidence intervals. On the y-axis is percentage points from -4 to 2 and on the x-axis is quarters since shock from 0 to 12. The blue line represents $$\mathcal{M}_{it}$$ > 0. The blue line steadily declines from slightly below 0 in quarter 0 to -2 in quarter 9, where it levels out. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 1 percentage points above and below the blue line. The thick dashed orange line represents $$\mathcal{M}_{it}$$ = 0. The line begins slightly below 0, becomes slightly above 0 in quarter 4, then returns to below 0 in quarter 10. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but grows to about 1 percentage points above and below the thick orange dashed line. The thick dashed green line represents non-bond issuers. It declines steadily from below 0 to -1 by quarter 10, then has a steeper decline to near -2 in quarter 12. The thinner green dashed lines depicting the 95% confidence interval surround the thick green dashed line, following the same path starting at 0 but grows to about 2 percentage points above and below the thick green dashed line.
Note: The figure shows the estimated coefficients $$\beta_{·,1}^h$$ based on the local projection $$\Delta^{h+1} log k_{it+h} = \beta^h_{\mathcal{M}>0,0} 1 \{ \mathcal{M}_{it}>0 \} + \beta^h_{\mathcal{M}=0,0} 1 \{ \mathcal{M}_{it} = 0 \} + \beta^h_{\mathcal{M}>0,1} 1 \{ \mathcal{M}_{it} > 0 \}\epsilon^{mp}_t + \beta^h_{\mathcal{M}=0,1} 1 \{ \mathcal{M}_{it} = 0 \}\epsilon^{mp}_t + \beta^h_{non-issuer,1} 1 \{ Non-bond-issuer_i \}\epsilon^{mp}_t +\Gamma Z_{it} + \gamma_1^h\Delta$$gdp$$_{t-1} + \delta_i^h + \delta_{sq}^h + \upsilon_{it + h}^h. $$ Shaded areas (and outer dashed lines) indicate 95% confidence bands two-way clustered by firms and quarters.
Figure B.5: Differential investment response associated with maturing bond share including callable bonds or bonds with variable coupon
This figure shows 4 line charts, each with two lines. The blue line represents the baseline specification, and the thick orange dashed line represents the extended specification. Each figure’s x-axis is the quarters since shock from 0 to 12.
The top left chart shows $$\mathcal{M}_{it}$$ including callable and non-callable bonds, with the y-axis being the percentage point from -0.8 to 0.4. The blue line begins near 0 for the first 4 quarters then declines to near -0.1 in quarter 6 where it stays until quarter 9, then linearly increases back to 0 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 0.1 percentage points above and below the blue line. The thicker orange dashed line steadily lowers from 0 to -0.3 by quarter 6 where it stays around for the rest of the quarters. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but grows to about 0.2 percentage points above and below the thick orange dashed line.
The top right chart shows $$\mathcal{M}_{it}$$ including only callable bonds, with the y-axis being the percentage points from -0.8 to 0.4. The blue line remains near 0 then entire period. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 0.1 percentage points above and below the blue line. The orange line also remains near 0 the entire time period but is more volatile. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but quickly becomes about 0.4 percentage points above and below the thick orange dashed line in quarter 4.
The bottom left chart shows $$\mathcal{M}_{it}$$ including variable and fixed coupon bonds, with the y-axis being the percentage point from -0.6 to 0.4. The blue line stays near 0 for the first 5 quarters, then slowly declines to -0.1 in quarter 8 then recovers to 0 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 0.1 percentage points above and below the blue line. The thicker orange dashed line follows the same pattern as the blue line but is about -0.1 below the blue line after quarter 0. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but grows to about 0.2 percentage points above and below the thick orange dashed line.
The bottom right chart shows $$\mathcal{M}_{it}$$ including only variable coupon bonds, with the y-axis being the percentage points from -0.6 to 0.4. The blue line stays slightly above 0 for the first 8 quarters, then declines to slightly below 0. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 0.1 percentage points above and below the blue line. The thicker orange line starts around 0 for the first 3 periods then rises to around 0.1 where it remains until quarter 10, then decreases to near 0 by quarter 12. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but quickly becomes about 0.3 percentage points above and below the thick orange dashed line in quarter 4.
Note: The figure shows the estimated $$\beta_1^h$$ coefficients based on the baseline local projection (2.3) and extended local projection (2.4), for various alternative definitions of the maturing bond share $$\mathcal{M}_{it}$$ . In our main findings, $$\mathcal{M}_{it}$$ includes only non-callable fixed coupon bonds. In panel (a), we include both callable and non-callable (fixed coupon) bonds. In panel (b), we re-define $$\mathcal{M}_{it}$$ based only on callable (fixed coupon) bonds. In panel (c), we include both variable coupon and fixed coupon (non-callable) bonds. In panel (d), we re-define $$\mathcal{M}_{it}$$ based only on variable coupon (non-callable) bonds. The $$\beta_1^h$$ estimates are standardized to capture the differential response to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with a one standard deviation higher maturing bond share. Shaded areas (and outer dashed lines) indicate 95% confidence bands two-way clustered by firms and quarters.
Figure B.6: Differential credit spread response associated with maturing bond share including callable bonds or bonds with variable coupon
This figure shows 4 line charts, each with two lines. The blue line represents the baseline specification, and the thick orange dashed line represents the extended specification. Each figure’s x-axis is the quarters since shock from 0 to 12.
The top left chart shows $$\mathcal{M}_{it}$$ including callable and non-callable bonds, with the y-axis being the basis points from -10 to 15. The blue and orange lines are volatile but increase from above 0 in quarter 0 to 5 in quarter 10. They then decrease to above 0 in quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path starting at 0 but grows to about 5 basis points above and below the blue line. The thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, following the same path starting at 0 but grows to about 10 basis points above and below the thick orange dashed line.
The top right chart shows $$\mathcal{M}_{it}$$ including only callable bonds, with the y-axis being the basis points from -20 to 20. The blue and orange lines both decrease from 0 to around -8 in quarter 8, jump to below 0 in quarter 9 and 10, then return to around -5 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line and the thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, both following the same path starting at 0 but grows to about 12 basis points above and below their respective lines.
The bottom left chart shows $$\mathcal{M}_{it}$$ including variable and fixed coupon bonds, with the y-axis being the basis points from -5 to 5. The blue and orange lines fluctuate around 0 for the first 4 periods, then increase to around 2 in period 8, then fluctuate around 2 for the remaining periods. The light blue band depicting the 95% confidence interval surrounds the blue line and the thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, both following the same path starting at 0 but grows to about 2 basis points above and below their respective lines.
The bottom right chart shows $$\mathcal{M}_{it}$$ including only variable coupon bonds, with the y-axis being the basis points from -6 to 2. Both the orange and blue lines remain around -2 for the first 3 quarters, drop to -3 in quarter 4, then increase back to -2 in quarter 5. The lines then increase to below 0 and level out. The light blue band depicting the 95% confidence interval surrounds the blue line and the thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, both follow the same path about 2 basis points above and below their respective lines.
Note: The figure shows the estimated $$\beta_1^h$$ coefficients based on the baseline local projection (2.3) and extended local projection (2.4), for various alternative definitions of the maturing bond share $$\mathcal{M}_{it}$$. In our main findings, $$\mathcal{M}_{it}$$ includes only non-callable fixed coupon bonds. In panel (a), we include both callable and noncallable (fixed coupon) bonds. In panel (b), we re-define $$\mathcal{M}_{it}$$ based only on callable (fixed coupon) bonds. In panel (c), we include both variable coupon and fixed coupon (non-callable) bonds. In panel (d), we re-define $$\mathcal{M}_{it}$$ based only on variable coupon (non-callable) bonds. The local projections additionally control for a Great Recession dummy variable interacted with the regressors. The $$\beta_1^h$$ estimates are standardized to capture the differential response to a one standard deviation increase in εmpt associated with a one standard deviation higher maturing bond share. Shaded areas (and outer dashed lines) indicate 95% confidence bands two-way clustered by firms and quarters.
Figure B.7: Differential investment response associated with one-year maturing bond share
This figure contains two line charts, each with two lines. The blue line represents the baseline specification, and the thick orange dashed line represents the extended specification. Each figure’s x-axis is the quarters since shock from 0 to 12. The left figure shows the estimated coefficients for investment, with the y-axis being percentage points from -0.7 to 0.2. The blue and orange lines both decreases steadily from less than 0 to around -0.3 in quarter 9 and remain there through quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line and the thinner orange dashed lines depicting the 95% confidence interval surround the thick orange dashed line, both following the same path about 0.2 percentage points above and below their respective lines.
The right figure shows the estimated coefficients for credit spreads, with the y-axis being basis points between -10 and 6. The blue line is volatile around 0 for the first 9 quarters, then increases to between 1 and 2 for the final quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, follow the same path, growing to about 4 basis points above and below the blue line. The orange line decreases from 0 to around -3 in quarter 5, where it remains through quarter 12. The thinner orange dashed lines depicting the 95% confidence interval surrounds the thick orange dashed line, follow the same path, growing to about 3 basis points above and below the orange line.
Note: This figure shows the estimated $$\beta_1^h$$ coefficients based on the baseline local projection (2.3) and the estimated $$\beta_1^h$$ coefficients based on the extended local projection (2.4), using the maturing bond share $$\mathcal{M}_{it}^{1y}$$ defined over the next year (i.e., including maturing bonds in quarters t to including t + 3). The local projections with credit spreads as left-hand side in panel (b) additionally control for a Great Recession dummy variable interacted with the regressors. The $$\beta_1^h$$ estimates are standardized to capture the differential response to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with a one standard deviation higher maturing bond share. Shaded areas (and outer dashed lines) indicate 95% confidence bands two-way clustered by firms and quarters.
Figure B.8: Differential investment and credit spread responses associated with lagged maturing bond share
This figure contains two line charts, each with two lines. The blue line represents the baseline specification, and the thick orange dashed line represents the extended specification. Each figure’s x-axis is the quarters since shock from 0 to 12. The figure on the left shows the estimated coefficients for investment, with the y-axis being percentage points from -0.1 to 0.3. The blue line fluctuates above 0 for the first 4 quarters then steadily increases to near 0.07 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, follow the same path, growing to about 1.5 percentage points above and below the blue line. The orange line increases from 0 to just below 0.2 from quarter 0 to quarter 12. The thinner orange dashed lines depicting the 95% confidence interval surrounds the thick orange dashed line, follow the same path, growing to about 1.5 percentage points above and below the orange line.
The figure on the right shows the estimated coefficients for credit spreads, with the y-axis being basis points from -8 to 4. The blue line remains near -1 for the first 9 quarters, then decreases to -3 in quarter 10, then increases to -2 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, follow the same path, growing to about 4 basis points above and below the blue line. The orange line remains near -1 for the first 9 quarters, then decreases to -2 in quarter 10 and 11, then increases to -1 in quarter 12. The thinner orange dashed lines depicting the 95% confidence interval surrounds the thick orange dashed line, follow the same path, growing to about 4 basis points above and below the orange line.
Note: This figure shows the estimated $$\beta_1^h$$ coefficients based on the baseline local projection (2.3) and the extended local projection (2.4), using $$\mathcal{M}_{it-1}$$ instead of $$\mathcal{M}_{it}$$. The local projections with credit spreads as left-hand side in panel (b) additionally control for a Great Recession dummy variable interacted with the regressors. The $$\beta_1^h$$ estimates are standardized to capture the differential response to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with a one standard deviation higher maturing bond share. Shaded areas (and outer dashed lines) indicate 95% confidence bands two-way clustered by firms and quarters.
Figure B.9: Differential investment and credit spread response associated with Compustat maturing debt share
This figure contains two line charts, each with two lines. The blue line represents the baseline specification, and the thick orange dashed line represents the extended specification. Each figure’s x-axis is the quarters since shock from 0 to 12. The figure on the left shows the estimated coefficients for investment, with the y-axis being percentage points from -0.6 to 0.6. The blue and orange lines both fluctuate around 0 for all 12 quarters, though the orange line is mostly above 0 and the blue line is mostly below 0. The light blue band depicting the 95% confidence interval surrounds the blue line, follow the same path, growing to about 0.2 percentage points above and below the blue line, while the thinner orange dashed lines depicting the 95% confidence interval for the orange line follow the same path, but about 0.5 percentage points above and below the orange line.
The figure on the right shows the estimated coefficients for credit spreads, with the y-axis being basis points from -10 to 10. The blue and orange lines both increase from near 0 to near 2.5 between quarter 0 and 4. The lines stay near 2.5 until quarter 7, then decrease to below 0 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line and the thinner orange dashed lines follow the same path, but about 5 basis points above and below their respective line.
Note: This figure shows the estimated $$\beta_1^h$$ coefficients based on the local projections (2.3) and (2.4), using $$\overline{\mathcal{M}_{it}}$$ instead of $$\mathcal{M}_{it}$$. $$\overline{\mathcal{M}_{it}} = (\text{debt in current liabilities})_{it}/(\text{total debt})_{it−1} $$ measures maturing debt based on Compustat data only. The local projections with credit spreads as left-hand side in panel (b) additionally control for a Great Recession dummy variable interacted with the regressors. The $$\beta_1^h$$ estimates are standardized to capture the differential response to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with a one standard deviation higher maturing debt share. Shaded areas indicate 95% confidence bands two-way clustered by firms and quarters.
Figure B.10: Differential investment response associated with maturing bond share using alternative denominators
This figure contains four line charts. Each figure’s x-axis is the quarters since shock from 0 to 12. The first three figures each with two lines. The blue line represents the baseline specification, and the thick orange dashed line represents the extended specification. The top left figure shows the estimated coefficients of the capital denominator, with the y-axis being the percentage points between -0.8 to 0.2. The blue line decreases from 0 to near -0.2 between quarters 0 and 6, then remains near -0.2 for the remaining quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, follow the same path, growing to about 0.2 percentage points above and below the blue line. The orange dashed line follows the same path as the blue line but decreases to near -0.3 in quarter 6 where it stays for the remaining quarters. The thinner orange dashed lines depicting the 95% confidence interval surrounds the thick orange dashed line, follow the same path, growing to about 0.3 percentage points above and below the orange line.
The top right figure shows the estimated coefficients of the sales denominator, with the y-axis being the percentage points between -0.4 and 0.1. The blue and orange lines both start below 0 then decrease to around -0.2, then increase to between -0.1 and -0.15 by quarter 12. Each line has a 95% confidence interval band that surrounds it, following the same path as the line though roughly 0.1 percentage points above and below their respective line.
The bottom left figure shows the estimated coefficients of the assets denominator with the y-axis being the percentage points between -0.5 and 0.1. The blue line decreases from 0 to below -0.1 by quarter 8, then recovers to above -0.1 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, follow the same path, growing to about 0.1 percentage points above and below the blue line. The orange line decreases from below 0 to -0.25 in quarter 8 and increases to -0.2 by quarter 12. The thinner orange dashed lines depicting the 95% confidence interval surrounds the thick orange dashed line, follow the same path, growing to about 0.2 percentage points above and below the orange line.
The bottom right figure shows the estimated coefficients of the non-moving average denominator, with the y-axis being the percentage points between -0.5 and 0.1. This figure contains 4 lines. The blue line represents the maturing bond/debt. The orange dashed line represents the maturing bond/capital. The green dotted line represents the maturing bond/sales. The red dotted line represents the maturing bond/total assets. Each line starts near 0, then decreases to between -0.25 and -0.1 in quarter 8, then increases slightly by quarter 12. Each line has a 95% confidence interval band that surrounds it, following the same path as the line though roughly 0.1 percentage points above and below their respective line.
Note: In panels (a) to (c) the figure shows the estimated $$\beta_1^h$$ coefficients based on the baseline local projection (2.3) and extended local projection (2.4), for various alternative definitions of $$\mathcal{M}_{it}$$. In panel (a), we re-define $$\mathcal{M}_{it}$$ as the ratio of maturing bonds over the average capital stock in the preceding four quarters, in (b) the denominator is average sales, in (c) average assets. In panel (d) the figure shows the estimated $$\beta_1^h$$ coefficients based on the baseline local projection (2.3) using as denominator debt, capital, sales, or assets in the preceding quarter, instead of constructing a moving average. The $$\beta_1^h$$ estimates are standardized to capture the differential response to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with a one standard deviation higher maturing bond share. Shaded areas (and outer dashed lines) indicate 95% confidence bands two-way clustered by firms and quarters.
Figure B.11: Differential credit spread response associated with maturing bond share using alternative denominators
This figure contains four line charts. Each figure’s x-axis is the quarters since shock from 0 to 12. The first three figures each with two lines. The blue line represents the baseline specification, and the thick orange dashed line represents the extended specification. The top left figure shows the estimated coefficients of the capital denominator, with the y-axis being the basis points between -5 to 10. Both the orange and blue lines fluctuate around 0 until quarter 6, where they then increase to around 4 by quarter 12. Each line has a 95% confidence interval band that surrounds it, following the same path as the line though roughly 4 basis points above and below their respective line.
The top right figure shows the estimated coefficients of the sales denominator, with the y-axis being the basis points between -4 and 6. Both the blue and orange lines begin near 0, rise to around 2 by quarter 7, which they fluctuate around for the remaining quarters. Each line has a 95% confidence interval band that surrounds it, following the same path as the line though roughly 3 basis points above and below their respective line. The bottom left figure shows the estimated coefficients of the assets denominator with the y-axis being the basis points between -5 and 11. The blue and orange lines remain near 0 for the first 2 quarters, then increase to near 5 by quarter 8, which they fluctuate around for the remaining quarters. Each line has a 95% confidence interval band that surrounds it, following the same path as their respective line. The blue bands are roughly 5 basis points above and below the blue line and the small orange lines are roughly 7 basis points above and below the thick orange line.
The bottom right figure shows the estimated coefficients of the non-moving average denominator, with the y-axis being the basis points between -5 and 11. This figure contains 4 lines. The blue line represents the maturing bond/debt. The orange dashed line represents the maturing bond/capital. The green dotted line represents the maturing bond/sales. The red dotted line represents the maturing bond/total assets. Each line starts near 0 and fluctuates. The red, blue, and orange lines all increasing, being between 4 and 5 by quarter 12. The green line stays between 0 and 1 basis points. Each line has a 95% confidence interval band that surrounds it, following the same path as the line, all about 5 basis points above and below their respective line.
Note: In panels (a) to (c) the figure shows the estimated $$\beta_1^h$$ coefficients based on the baseline local projection (2.3) and extended local projection (2.4), for various alternative definitions of $$\mathcal{M}_{it}$$. In panel (a), we re-define $$\mathcal{M}_{it}$$ as the ratio of maturing bonds over the average capital stock in the preceding four quarters, in (b) the denominator is average sales, in (c) average assets. In panel (d) the figure shows the estimated $$\beta_1^h$$ coefficients based on the baseline local projection (2.3) using as denominator debt, capital, sales, or assets in the preceding quarter, instead of constructing a moving average. The local projections additionally control for a Great Recession dummy variable interacted with the regressors. The $$\beta_1^h$$ estimates are standardized to capture the differential response to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with a one standard deviation higher maturing bond share. Shaded areas (and outer dashed lines) indicate 95% confidence bands two-way clustered by firms and quarters.
Figure B.12: Differential investment and credit spread responses associated with maturing bond share using dummy specification of bond maturity
This figure contains 4 line charts, each with quarters since shock on the x-axis ranging from 0 to 12. The top charts show the differential effect of $$\mathcal{M}$$ >0. The left chart shows the estimated coefficients for investment, with the y-axis being the percentage point from -3 to 1. The blue line decreases from 0 in quarter 0 to above -2 in quarter 9, then begins to increase to around -1.5 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, follow the same path, growing to about 1 percentage point above and below the blue line. The right chart shows the estimated coefficients for the spread, with the y-axis being the basis points from -20 to 40. The blue line declines from 5 to just below 0 between quarters 0 and 1, then increases to around 15 in quarter 7 where it remains, though volatile for the remaining quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, follow the same path, growing to about 20 basis points above and below the blue line.
The bottom charts show the differential effect of $$\mathcal{M}$$ >15. The left chart shows the estimated coefficients for investment, with the y-axis being the percentage point from -7 to 2. The blue line stays near 0 for the first 4 quarters, then linearly declines to -4 in quarter 8, then linearly increases to above -2 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, follow the same path, growing to about 3 percentage point above and below the blue line. The right chart shows the estimated coefficients for the spread, with the y-axis being the basis points from -50 to 150. The blue line remains between 25 and 50 for the first 8 quarters, then decline to below 0 in quarter 8, then increases linearly to 50 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, follow the same path, growing to about 55 basis points above and below the blue line.
Note: The figure shows the estimated $$\beta_1^h$$ coefficients based on the baseline local projection (2.3), using $$1\{{\mathcal{M}_{it} > 0}\}$$ instead of $$\mathcal{M}_{it}$$ in panels (a) and (b), and $$1\{{\mathcal{M}_{it}> 15} \}$$instead of $$\mathcal{M}_{it}$$ in panels (c) and (d). The $$\beta_1^h$$ estimates are standardized to capture the differential response to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with, respectively, $$\mathcal{M}_{it}$$ > 0 and $$\mathcal{M}_{it}$$ > 15 (i.e., 15 % of debt). The local projections with credit spreads as left-hand side in panel (b) additionally control for a Great Recession dummy variable interacted with the regressors. Shaded areas indicate 95% confidence bands two-way clustered by firms and quarters.
Figure B.13: Grouping estimator for investment and credit spread responses
This figure contains 4 line charts, each with quarters since shock on the x-axis ranging from 0 to 12. There are dots at some quarters that denote statistical significance at the 5% level if filled at the 10% level if unfilled. The top charts have two lines, a blue line that represents $$\mathcal{M}_{it}$$ > 0 and a dashed orange line that represents $$\mathcal{M}_{it}$$ = 0.
The top left chart shows the estimated coefficients of investment, with the y-axis being the percentage points from -3 to 1. The blue line decrease to around -0.75 around quarter 3 then declines to around -1.5 in quarter 4, where it remains around for the remaining 12 quarters. The orange line starts negative around 0 then becomes positive near 0 after quarter 4 until quarter10, then returns to negative around 0. There are filled dots at every quarter starting at the 3rd quarter for both the blue and orange lines. The light blue band depicting the 95% confidence interval surrounds the blue line and the small orange lines depicting the 95% confidence interval surrounds the orange line, follow the same path as their respective line, growing to about 1 percentage point above and below their respective lines.
The top right chart shows the estimated coefficients of credit spreads, with the y-axis being the basis points from -40 to 60. Both the blue and orange lines follow the same path, starting at 0, decreasing through quarter 2 (blue to -20 and orange to -5), then increasing to near 30 by quarter 8, then decreasing slightly through quarter 12, with the blue line decreasing to near 10 and the orange line decreasing to near 20. There are no dots in this figure. The light blue band depicting the 95% confidence interval surrounds the blue line and the small orange lines depicting the 95% confidence interval surrounds the orange line, follow the same path as their respective line, growing to about 20 basis points above and below their respective lines.
The bottom charts have two lines, a blue line that represents $$\mathcal{M}_{it}$$ > 15 and a dashed orange line that represents $$\mathcal{M}_{it}$$ ≤15. The bottom left chart shows the estimated coefficients of investment, with the y-axis being the percentage points from -8 to 2. The blue line declines slightly from 0 in quarter 0 to around -1 for quarters 3 through 5, then declines quickly to -4 by quarter 8. The line then increases steadily to -2 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path as blue line, growing to about 3 basis points above and below the blue line. The orange line starts below 0 until quarter 4, then becomes slightly positive until quarter 10, then returns to just below 0 by quarter 12. The thinner orange dashed lines depicting the 95% confidence interval surrounds the thick orange dashed line, follow the same path, growing to about 1 basis point above and below the orange line. There are filled dots for both the orange and blue lines at quarters 6 through 9; there is an unfilled dot for both lines at quarter 10.
The top right chart shows the estimated coefficients of credit spreads, with the y-axis being the basis points from -200 to 200. The blue line increases from near 0 in quarter 0 to 100 in quarter 8, then declines to near 50 in quarter 9, which it fluctuates around for the remaining quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path as blue line, growing to about 150 basis points above and below the blue line. The orange line starts just below 0, until quarter3, then increases to near 25 by quarter 8 where it remains stable for the remaining quarters. The thinner orange dashed lines depicting the 95% confidence interval surrounds the thick orange dashed line, follow the same path, growing to about 25 basis point above and below the orange line. There are filled dots for both the orange and blue lines at quarter 3; there is an unfilled dot for both lines at quarter 6 and quarter 8.
Note: The figure shows the estimated $$\beta_{1,high}^h$$ and $$\beta_{1,low}^h$$,low coefficients based on the local projection $$\Delta^{h+1} y_{it+h} = \beta_{0,high}^h 1\{{\mathcal{M}_{it}}> \overline {\mathcal{M}} \} + \beta_{1,low}^h 1\{{\mathcal{M}_{it}}≤ \overline {\mathcal{M}} \} \times \epsilon^{mp}_t + \beta_{1,high}^h 1\{{\mathcal{M}_{it}}> \overline {\mathcal{M}} \} \times \epsilon^{mp}_t + \beta_{2,low}^h 1\{{\mathcal{M}_{it}}≤ \overline {\mathcal{M}} \} \times \Delta$$gdp$$_{t-1} + \beta_{2,high}^h 1\{{\mathcal{M}_{it}}> \overline {\mathcal{M}} \} \times \Delta$$gdp$$_{t-1} + \delta_i^h + \delta_{sq}^h + \upsilon_{it + h}^h, $$ where 1{·} denotes a dummy variable that equals one if the inequality is satisfied and zero else. In panels (a) and (c), we define $$\Delta^{h+1} y_{it+h}$$ as capital growth, in panels (b) and (d) as the change in credit spreads. In panels (a) and (b), the threshold maturing bond share is $$\overline {\mathcal{M}}$$ = 0, in panels (c) and (d) we set $$\overline {\mathcal{M}}$$ = 15. The local projection for bond spreads additionally interacts the regressors with a Great Recession (2008Q3-2009Q2) dummy and the figure plots the non-crisis coefficients. Shaded areas (and outer dashed lines) indicate 95% confidence bands based on standard errors clustered by firms and quarters. Filled (unfilled) dots indicate that the estimates of $$\beta_{1,high}^h$$ and $$\beta_{1,low}^h$$ are significantly different from each other at the 5% (10%) level.
Figure B.14: Differential investment responses due to maturing bond share conditional on firm age group
This figure contains 6 line charts, each with quarters since shock on the x-axis ranging from 0 to 12. The left charts show old firms and the right charts show young firms.
The top charts show these by level of $$\mathcal{M}_{it}$$, with the y-axis being percentage points between -3 and 2. These charts contain two lines, a blue line for $$\mathcal{M}_{it}$$ > 0 and an orange dashed line for $$\mathcal{M}_{it}$$ = 0. In the left figure, the blue line remains near 0 until quarter 2, then decreases to -1.5 by quarter 4, where it remains for the remaining quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path as blue line, growing to about 1.5 percentage points above and below the blue line. The orange line starts at 0, slowly increases to 0.5 in quarter 9, then declines to near 0 by quarter 12 The thinner orange dashed lines depicting the 95% confidence interval surrounds the thick orange dashed line, follow the same path, growing to about 1 percentage point above and below the orange line.
In the right figure, the blue line remains declines from -0.5 in quarter 0 to -1 in quarter 12, fluctuating to near 0 in quarter 4 and to just below -1 in quarter 9. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path as blue line, growing to about 1.5 percentage points above and below the blue line. The orange line starts at 0, slowly increases to 1 in quarter 10, then slowly declines to above 0.5 by quarter 12. The thinner orange dashed lines depicting the 95% confidence interval surrounds the thick orange dashed line, follow the same path, growing to about 1 percentage point above and below the orange line. The middle charts show the difference between $$\mathcal{M}_{it}$$ > 0 vs. $$\mathcal{M}_{it}$$ = 0 for each firm type. The left figure has a y-axis of percentage points from -4 to 1, where the blue line decreases from 0 to near -2 in quarter 4, remains there until quarter 9, then increases to near -1 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path and growing to about 1 percentage point above and below the blue line.
The right figure has a y-axis of percentage points from -5 to 1, where the blue line decreases from around -0.5 to -2 in quarter 9, then removers to near -1.5 in quarter 10 where it stays through quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path and growing to about 2 percentage points above and below the blue line. The bottom charts show the linear effect of $$\mathcal{M}_{it}$$ , with the y-axis being percentage points from -0.5 to 0.2. In the left figure, the blue line decreases from 0 to near -0.2 in quarter 4, remains there until quarter 9, then increases to near -0.1 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path and growing to about 0.1 percentage point above and below the blue line. In the right figure, the blue line starts near -0.1, increases to above 0 by quarter 4, declines to -0.2 in quarter 9, then increases to near 0 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path and growing to about 0.3 percentage point above and below the blue line.
Note: Panels (a) and (b) show in the solid lines the estimated coefficients $$\beta_{large,\mathcal{M}>0,1}^h$$ and $$\beta_{small,\mathcal{M}>0,1}^h$$, and in the dashed lines the analogous coefficients for $$\mathcal{M}$$ = 0, from the local projection $$\Delta^{h+1} log k_{it+h} = \beta_{young,\mathcal{M}>0,0}^hYoung_{it-1}1\{{\mathcal{M}_{it}}> 0 \} + \beta_{old,\mathcal{M}>0,0}^hOld_{it-1}1\{{\mathcal{M}_{it}}> 0 \} + \beta_{young,\mathcal{M}>0,1}^hYoung_{it-1}1\{{\mathcal{M}_{it}}> 0 \} \epsilon^{mp}_t + \beta_{old,\mathcal{M}>0,1}^hOld_{it-1}1\{{\mathcal{M}_{it}}> 0 \} \epsilon^{mp}_t +\Gamma Z_{it} + \gamma_1^h\Delta$$gdp$$_{t-1} + \delta_i^h + \delta_{sq}^h + \upsilon_{it + h}^h $$ where $$Young_{it}$$ and $$Old_{it}$$ are dummy variables capturing if the firm’s age is below or above the sector-quarter specific median age. The vector $$Z_{it}$$ here contains the interactions for $$1\{{\mathcal{M}_{it}} = 0 \}$$ as written out for $$1\{{\mathcal{M}_{it}} > 0 \}$$, and all interactions of age and maturity dummies with lagged GDP growth $$\Delta$$gdp$$_{t-1}$$. In panels (c) and (d) we plot the difference between the respective estimators. Panels (e) and (f) show the estimated coefficients $$\beta_{old,1}^h$$ and $$\beta_{young,1}^h$$ from the local projection $$\Delta^{h+1} log k_{it+h} = \beta_{young,0}^hYoung_{it-1} + \beta_{old,0}^hOld_{it-1} + \beta_{\mathcal{M},0}^h{\mathcal{M}_{it}} + \beta_{young,1}^hYoung_{it-1}{\mathcal{M}_{it}}\epsilon^{mp}_t + \beta_{old,1}^hOld_{it-1}{\mathcal{M}_{it}}\epsilon^{mp}_t +\Gamma Z_{it} + \gamma_1^h\Delta$$gdp$$_{t-1} + \delta_i^h + \delta_{sq}^h + \upsilon_{it + h}^h $$. The vector $$Z_{it}$$ contains all interactions of age dummies and maturity with lagged GDP growth $$\Delta$$gdp$$_{t-1}$$. 95% confidence bands based on standard errors clustered by firms and quarters.
Figure B.15: Differential investment responses due to maturing bond share conditional on firm size group
This figure contains 6 line charts, each with quarters since shock on the x-axis ranging from 0 to 12. The left charts show large firms and the right charts show small firms.
The top charts show these by level of $$\mathcal{M}_{it}$$. These charts contain two lines, a blue line for $$\mathcal{M}_{it}$$ > 0 and an orange dashed line for $$\mathcal{M}_{it}$$ = 0. For the left figure, the y-axis is percentage points between -3 and 2. The blue line remains near 0 until quarter 2, then decreases to -1.5 by quarter 4, where it remains for the remaining quarters. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path as blue line, growing to about 1 percentage point above and below the blue line. The orange line starts above 0, slowly increases to 1 in quarter 9, then declines to 0.5 by quarter 12 The thinner orange dashed lines depicting the 95% confidence interval surrounds the thick orange dashed line, follow the same path, growing to about 0.5 percentage points above and below the orange line.
For the right figure, the y-axis is percentage points between -3 and 3. The blue line increases from near 0 to below 1 by quarter 5, declines to around -0.5 in quarter 6, where it remains until quarter 8. The line then steadily increases to above 0 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path as blue line, growing to about 2 percentage points above and below the blue line. The orange line starts above 0, slowly increases to below 1 in quarter 10, then declines to 0.5 by quarter 12 The thinner orange dashed lines depicting the 95% confidence interval surrounds the thick orange dashed line, follow the same path, growing to about 1 percentage point above and below the orange line.
The middle charts show the difference between $$\mathcal{M}_{it}$$ > 0 vs. $$\mathcal{M}_{it}$$ = 0 for each firm type. The left figure has a y-axis of percentage points from -4 to 1, where the blue line declines from below 0 to -2 by quarter 4. The line stays near -2.5 for quarters 5 through 9, then returns to -2 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path and growing to about 1 percentage point above and below the blue line.
The right figure has a y-axis of percentage points from -4 to 2, where the blue line remains around 0 for the first 5 quarters, then declines to -1 in quarter 6, where it remains until quarter 10. It increases back to 0 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path and growing to about 2.5 percentage points above and below the blue line.
The bottom charts show the linear effect of $$\mathcal{M}_{it}$$ . The left figure has a y-axis of percentage points from – 0.9 to 0.2, where the blue line decreases from -0.1 to near -0.4 in quarter 5, remains there until quarter 9, then increases to near -0.3 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path and growing to about 0.1 percentage points above and below the blue line.
The right figure has a y-axis of percentage points from – 0.4 to 0.2. The blue line stays near 0 for the first 3 quarters, increases to near 0.05 in quarter 5, declines to -0.15 in quarter 8, then increases back to near 0 by quarter 12. The light blue band depicting the 95% confidence interval surrounds the blue line, following the same path and growing to about 0.1 percentage points above and below the blue line.
Note: Panels (a) and (b) show in the solid lines the estimated coefficients $$\beta_{large,\mathcal{M}>0,1}^h$$ and $$\beta_{small,\mathcal{M}>0,1}^h$$ , and in the dashed lines the analogous coefficients for $$\mathcal{M}$$ = 0, from the local projection $$\Delta^{h+1} log k_{it+h} = \beta_{small,\mathcal{M}>0,0}^hSmall_{it-1}1\{{\mathcal{M}_{it}}> 0 \} + \beta_{large,\mathcal{M}>0,0}^hLarge_{it-1}1\{{\mathcal{M}_{it}}> 0 \} + \beta_{small,\mathcal{M}>0,1}^hSmall_{it-1}1\{{\mathcal{M}_{it}}> 0 \} \epsilon^{mp}_t + \beta_{large,\mathcal{M}>0,1}^hLarge_{it-1}1\{{\mathcal{M}_{it}}> 0 \} \epsilon^{mp}_t +\Gamma Z_{it} + \gamma_1^h\Delta$$gdp$$_{t-1} + \delta_i^h + \delta_{sq}^h + \upsilon_{it + h}^h $$ where $$Small_{it}$$ and $$Large_{it}$$ are dummy variables capturing if the firm’s age is below or above the sector-quarter specific median age. The vector $$Z_{it}$$ here contains the interactions for $$1\{{\mathcal{M}_{it}} = 0 \}$$ as written out for $$1\{{\mathcal{M}_{it}} > 0 \}$$, and all interactions of age and maturity dummies with lagged GDP growth $$\Delta$$gdp$$_{t-1}$$. In panels (c) and (d) we plot the difference between the respective estimators. Panels (e) and (f) show the estimated coefficients $$\beta_{large,1}$$ and $$\beta_{small,1}$$ from the local projection $$\Delta^{h+1} log k_{it+h} = \beta_{small,0}^hSmall_{it-1} + \beta_{large,0}^hLarge_{it-1} + \beta_{\mathcal{M},0}^h{\mathcal{M}_{it}} + \beta_{small,1}^hSmall_{it-1}{\mathcal{M}_{it}}\epsilon^{mp}_t + \beta_{large,1}^hLarge_{it-1}{\mathcal{M}_{it}}\epsilon^{mp}_t +\Gamma Z_{it} + \gamma_1^h\Delta$$gdp$$_{t-1} + \delta_i^h + \delta_{sq}^h + \upsilon_{it + h}^h $$. The vector $$Z_{it}$$ contains all interactions of age dummies and maturity with lagged GDP growth $$\Delta$$gdp$$_{t-1}$$. 95% confidence bands based on standard errors clustered by firms and quarters.
Figure D.1: Steady state policy functions
This figure contains 9 three-dimensional color plots. For each firm, the x-axis is firm assets q, the y-axis is outstanding long-term debt b. Each column represents the level of z’: on the left is low z’, the middle is medium z’, and the right is high z’.
The first row shows the capital on the z-axis. The figure on the left has the x-axis ranging from 0 to 0.4, the y-axis ranging from 0 to 0.4, and the z-axis ranging from 0.25 to 0.3. As the x-axis increases, the z-axis increases, and the y-axis slightly increases. The middle figure has the x-axis ranging from -1 to 2, the y-axis ranging from 0 to 5, and the z-axis ranging from 0 to 2. The z-axis increases as the x-axis increases. The y-axis has the same value of the z-axis for each value of x, though the points between 4 and 5 are slightly larger. The right figure has the x-axis ranging from -2 to 6, the y-axis ranging from 0 to 10, and the z-axis ranging from 0 to 5. The figure looks similar to the middle figure.
The middle row shows the leverage on the z-axis. The figure on the left has the x-axis ranging from 0 to 0.4, the y-axis ranging from 0 to 0.4, and the z-axis ranging from 0 to 30. The z-axis increases as the y-axis increases. The x-axis has similar values of the z-axis and y-axis as the axes increase, with the lower values being slightly higher. The middle figure has the x-axis ranging from -1 to 2, the y-axis ranging from 0 to 5, and the z-axis ranging from 0 to 100. As the x-axis increases, the values of the y-axis decrease and the values of the z-axis decrease. They decrease quickly from the beginning on the x-axis until 0, where the values decrease more slowly. The right figure has the x-axis ranging from -2 to 6, the y-axis ranging from 0 to 10, and the z-axis ranging from 0 to 200. The figure looks similar to the middle figure.
The bottom row shows the long-term debt share on the z-axis. The figure on the left has the x-axis ranging from 0 to 0.4, the y-axis ranging from 0 to 0.4, and the z-axis ranging from 0 to 75. The z-axis increases, first rapidly then more slowly as the y-axis increases. The x-axis has similar values of the z-axis and y-axis as the axes increase. The middle figure has the x-axis ranging from -1 to 2, the y-axis ranging from 0 to 5, and the z-axis ranging from 0 to 100. For the y-axis near 0, the z-values spike from 0 to near 100, where they remain for all values of the y-axis. The values of the y-axis and z-axis are steady for values of the z-axis. The right figure has the x-axis ranging from -2 to 6, the y-axis ranging from 0 to 10, and the z-axis ranging from 0 to 100. The figure looks similar to the middle figure.
Note: On the x-axis are firm assets $$q = q(z, k, b^S, b^L, ε; S) $$normalized by average firm capital. On the y-axis is outstanding long-term debt $$b = (1 − γ)b^L $$ normalized by average firm debt. Policy functions for Capital (k′) are normalized by average firm capital. The remaining firm policies are in %. Leverage is total firm debt over assets $$((b^{S′} + b^{L′} )/k′)$$; the Long-term debt share is $$b^{L′}/(b^{S′} + b^{L′} )$$.
Figure D.2: Steady state policy functions (continued)
This figure contains 12 three-dimensional color plots. For each firm, the x-axis is firm assets q, the y-axis is outstanding long-term debt b. Each column represents the level of z’: on the left is low z’, the middle is medium z’, and the right is high z’.
The first row shows the equity issuance on the z-axis. The figure on the left has the x-axis ranging from 0 to 0.4, the y-axis ranging from 0 to 0.4, and the z-axis ranging from -25 to 100. As the x-axis increase, the y-axis and the z-axis decrease linearly. The middle figure has the x-axis ranging from -1 to 2, the y-axis ranging from 0 to 5, and the z-axis ranging from -50 to 200. As the x-axis increases, the z-axis decreases, quickly at first, then slowly. The x-axis and z-axis remain stable across the values of the y-axis. The right figure has the x-axis ranging from -2 to 6, the y-axis ranging from 0 to 10, and the z-axis ranging from -100 to 200. The figure looks similar to the middle figure.
The second row shows the default risk on the z-axis. The figure on the left has the x-axis ranging from 0 to 0.4, the y-axis ranging from 0 to 0.4, and the z-axis ranging from 0 to 10. As the y-axis increases, the values of x increase, slowly at first then more quickly. The values of the x-axis have a similar value of the z-axis for each value of the y-axis, though the values closer to -1 are slightly larger in the z-axis. The middle figure has the x-axis ranging from -1 to 2, the y-axis ranging from 0 to 5, and the z-axis ranging from 0 to 5. As both the x-axis and y-axis increase, the z-axis increases, first slowly, then more rapidly. The right figure has the x-axis ranging from -2 to 6, the y-axis ranging from 0 to 10, and the z-axis ranging from 0 to 8. The figure looks similar to the middle figure.
The third row shows the short-term credit spread on the z-axis. The figure on the left has the x-axis ranging from 0 to 0.4, the y-axis ranging from 0 to 0.4, and the z-axis ranging from 2 to 8. As the y-axis increases, the values of x increase, slowly at first then more quickly. The values of the x-axis have a similar value of the z-axis for each value of the y-axis, though the values closer to -1 are slightly larger in the z-axis. The middle figure has the x-axis ranging from -1 to 2, the y-axis ranging from 0 to 7, and the z-axis ranging from 0 to 5 As both the x-axis and y-axis increase, the z-axis increases, first slowly, then more rapidly. The right figure has the x-axis ranging from -2 to 6, the y-axis ranging from 0 to 10, and the z-axis ranging from 0 to 10. The figure looks similar to the middle figure. The bottom row shows the long-term credit spread on the z-axis. The figure on the left has the x-axis ranging from 0 to 0.4, the y-axis ranging from 0 to 0.4, and the z-axis ranging from 2 to 5. As the y-axis increases, the values of x increase, slowly at first then more quickly. The values of the x-axis have the same value of the z-axis for each value of the y-axis. The middle figure has the x-axis ranging from -1 to 2, the y-axis ranging from 0 to 7, and the z-axis ranging from 0 to 4. As the y-axis increases, the values of x increase quickly for the first 5th of the y-axis, then smooth for 2/5th and 3/5th through the y-axis, then increase quickly to near 3 for the final 5th. The values of the x-axis have a similar value of the z-axis for each value of the y-axis, though the values closer to -1 are slightly smaller in the z-axis. The right figure has the x-axis ranging from 0 to 5, the y-axis ranging from 0 to 10, and the z-axis ranging from 0 to 10. The figure looks similar to the middle figure.
Note: On the x-axis are firm assets $$q = q(z, k, b^S, b^L, ε; S) $$ normalized by average firm capital. On the y-axis is outstanding long-term debt $$ b = (1 − γ)b^L $$ normalized by average firm debt. Policy functions are in %. Equity issuance is relative to firm assets (e/k′). Default risk and credit spreads are annualized.
Figure D.3: Share of maturing debt at longer time horizons (in %)
This figure is a bar chart comparing the data (blue bar) to the model (red bar) for 1, 2, 3, 4, and 5 years on the x-axis. The y-axis is the share of maturing debt at longer time horizons in % and ranges from 0 to 80. For each year both the red and blue bars increase, going from 30% in 1 year to around 70% in 5 years. The red bars being slightly lower than the blue bars for 2 through 5 Years.
Note: The figure shows the cross-sectional average of firms’ share of debt maturing within one year, within two years, within three years, within four years, and within five years. The data sample is 1995–2017. Firmlevel data on maturing debt at time horizons of two to five years (dd2, dd3, dd4, dd5) is from Compustat.
Figure D.4: Price of long-term debt $$p^L$$
This figure is a line graph with three lines. On the y-axis is the price of long-term debt and on the x-axis is the choice of long-term debt. The green dashed line represents a high productivity firm and remains horizontal near the top of the chart. The dashed orange line represents a medium productivity firm and starts lower than the green dashed line then slowly declines. The blue line represents a low productivity firm and starts lower than the orange line then declines slowly though at a faster rate than the orange line.
Note: The price of long-term debt $$ p^L $$ in (3.12) is shown as a function of the firm’s choice of long-term debt $$ b^{L′} $$ for a given state of firm assets q and outstanding long-term debt b, and three different productivity levels z′. All firm-level choices besides $$b^{L′} $$ (i.e., capital k′ and short-term debt $$b^{S′} $$ ) are held at their steady-state values.
Figure D.5: Derivative of firm policies with respect to outstanding long-term debt
This figure contains 4 bar charts, each with two colors of bars. The blue bars represent Low $$\mathcal{M}_{it}$$ and the red bars represent High $$\mathcal{M}_{it}$$. The top left chart shows the long-term bond price, with the mass of firms from 0 to 0.35 on the y-axis and the $$∂p_t^L/∂b_t$$ from -0.2 to 0 on the x-axis. The red bars are primarily between -0.15 and -0.1 $$∂p_t^L/∂b_t$$, all with around 0.1 for the mass of firms. There are a few red bars between -0.05 and 0 $$∂p_t^L/∂b_t$$ with a mass of firms close to 0. There are a few blue between -0.2 and -0.15 $$∂p_t^L/∂b_t$$ with mass of firms near 0.05. Then near 0 $$∂p_t^L/∂b_t$$, there are taller blue bars, with the largest mass of firms being near 0.35.
The top right chart shows the capital, with the mass of firms from 0 to 0.2 on the y-axis and the $$∂k_(t+1)/∂b_t$$ from -0.5 to 1.5 on the x-axis. The red bars are concentrated between -0.5 and 0 $$∂k_(t+1)/∂b_t$$ with heights around 0.125 for mass of firms. The blue bars are concentrated throughout the y-axis with heights fluctuating between 0 and 0.1 for mass of firms. The bottom left chart shows the leverage, with the mass of firms from 0 to 0.4 on the y-axis and the $$∂LEV_(t+1)/∂b_t$$ from 0 to 0.5 on the x-axis. The red bars are concentrated between 0.4 and 0.5 $$∂LEV_(t+1)/∂b_t$$, with the largest height being 0.4 for mass of firms. The blue bars are concentrated between 0 and 0.2 and 0.4 and 0.5, with heights between 0 and 0.18 for mass of firms.
The bottom right chart shows the default risk, with the mass of firms from 0 to 0.6 on the y-axis and the $$∂\text{default risk}_(t+1)/∂b_t$$ from 0 to 0.04 on the x-axis. There is one red bar near 0 that has around 0.28 for mass of firms, then the rest of the red bars are between 0.015 and 0.035 $$∂\text{default risk}_(t+1)/∂b_t$$ with a mass of firms all less then 0.1. The large blue bars are concentrated near 0, with the largest having a mass of firms above 0.5. The rest of the blue bars are between 0.015 and 0.04 $$∂\text{default risk}_(t+1)/∂b_t$$ with a mass of firms below 0.05.
Note: The figure shows the stationary distribution in the calibrated model of the derivative of different firm policies with respect to outstanding long-term debt bt. Panel (a) shows the derivative of the long-term bond price $$p^L_t$$ with respect to bt, panel (b) the derivative of capital $$k_{t+1}$$ with respect to $$b_t$$, panel (c) the derivative of leverage $$LEV_{t+1}$$ with respect to bt, and panel (d) the derivative of default $$risk_{t+1}$$ with respect to $$b_t$$. Low $$\mathcal{M}_{it}$$ firms (blue) and High $$\mathcal{M}_{it}$$ firms (red) are firms with a maturing bond share below and above the median, respectively.
Figure D.6: Differential firm-level responses associated with
$$\mathcal{M}_{it}$$
This figure shows 4 line charts. The top left chart shows the differential response of the average credit spread, with the y-axis being the basis points from 0 to 0.4 and the x-axis being the quarters since shock from 0 to 12. The blue line starts near 0.35 and linearly declines to near 0.15 in quarter 5 where it remains for the remaining quarters.
The top right chart shows the differential response of debt, with the y-axis being the percentage point from -0.15 to 0 and the x-axis being the quarters since shock from 0 to 12. The blue line starts and 0 and jumps to -0.15 by the first quarter, then it recovers quickly, ending near -0.025 in quarter 12.
The bottom charts represent the differential response of sales (left chart) and employment (right chart). Both have the y-axis as percentage point from -0.1 to 0.02 and the x-axis as quarters since shock from 0 to 12. Both blue lines follow the same path, with the line starting at 0, jumping to around -0.09 in quarter 1, then quickly recovering, evening out near -0.02 in quarter 12.
Note: The lines show the differential response of the average credit spread, debt growth, sales growth, and employment growth associated with $$\mathcal{M}_{it}$$ in simulated model data. All values are standardized to capture the differential response to a one standard deviation (30bp) increase in the nominal interest rate it associated with a one standard deviation higher $$\mathcal{M}_{it}$$. The average credit spread in panel (a) is the average of a firm’s short-term and long-term credit spread weighted by firm-level shares of short-term and long-term debt. Debt in panel (b) is the sum of short-term and long-term debt. Sales in panel (c) is $$p_{it}y_{it}$$. Employment in panel (d) is lit.
Figure F.1: Differential investment and spread responses associated with higher maturing bond share using different sets of fixed effects
This figure shows 4 line charts, each with 5 lines. The blue line shows αi and αst FEs (the benchmark). The orange dashed line shows No FEs. The dotted green line shows only αi FEs. The dotted red line shows only αt FEs. The dashed green line shows only αst FEs. Each figure’s x-axis is quarters since shock from 0 to 12.
The top figures show the baseline specifications. The top left figure shows investment, with the y-axis in percentage points from -0.5 to 0.2. Each line decreases to near -0.2 by quarter 8, then increases to near -0.1 by the quarter 12. Each line has a 95% confidence interval band or smaller lines associated with it that follow the same path, but roughly 0.1 percentage points above or below the associated line.
The top right figure shows spreads, with the y-axis is basis points from -3 to 7. The blue and dotted green lines both increases from above 0 to around 4 by quarter 8, then decrease to near 2 by quarter 12. The dotted red line and dashed green line fluctuate, but overall decrease from just above 0 to just below 0 between quarter 0 and quarter 12. Each line has a 95% confidence interval band or smaller lines associated with it that follow the same path as the line, but roughly 2 basis points above or below the associated line.
The bottom figures show the extended specifications. The bottom left figure shows investment, with the y-axis in percentage points from -0.5 to 0.2. Each line decreases to -0.3 by quarter 9, then increases to -0.2 by quarter 12. Each line has a 95% confidence interval band or smaller lines associated with it that follow the same path, but roughly 0.2 percentage points above or below the associated line.
The bottom right figure shows spreads, with the y-axis is basis points from -3 to 7. The blue and dotted green lines both increases from above 0 to around 3 by quarter 8, then decrease to near 2 in quarter 9 where the remain for the remaining quarters. The dotted red line and dashed green line fluctuate, but overall decrease from just above 0 to -2 between quarter 0 and quarter 12. Each line has a 95% confidence interval band or smaller lines associated with it that follow the same path as the line, but roughly 4 basis points above or below the associated line.
Note: The figures show the estimated $$\beta_1^h$$ coefficients, when using different sets of fixed effects (FEs) as indicated in the legends, using the baseline specification in equation (2.3) in panel (a) and using the extended specification in equation (2.4) in panel (b). The local projections with the credit spread as left-hand side additionally control for a Great Recession dummy variable interacted with the regressors. The $$\beta_1^h$$ estimates are standardized to capture the differential response to a one standard deviation increase in $$\epsilon^{mp}_t$$ associated with a one standard deviation higher maturing bond share. Shaded areas (and outer dashed lines) indicate 95% confidence bands two-way clustered by firms and quarters.
Figure H.1: Firm dynamics
This figure shows 4 charts. The top left chart is a bar chart that shows the stationary distribution of firm size, with the mass of firms ranging from 0 to 0.7 on the y-axis and the capital relative to the average ranging from 0 to 6 on the x-axis. The first bar is the largest being around 0.65. The second bar is near 0.2. The next three bars are all near 0.05. There is no bar beyond 5.
The top right chart is a bar chart that shows the stationary distribution of firm age, with the mass of firms ranging from 0 to 0.7 on the y-axis and the age ranging from 0 to 500 on the y-axis. The bar with the age between 0 to 50 is the largest, being around 0.65 for the mass of firms, with each bar steadily quickly. There are no bars beyond 300 years.
The bottom left chart is a bar chart that shows the stationary distribution of capital growth rates, with the y-axis showing the mass of firms from 0 to 0.35 and the x-axis showing the capital growth rate from -1.5 to 1.5. The highest bars have a capital growth rate around 0, with the largest having a mass of firms near 0.35, with the bars quickly getting smaller on both sides of 0 and having no observations between -1.5 and -0.75 and 0.75 and 1.5.
The bottom right chart is a line chart that shows the average log of capital over the life cycle of firms, with the y-axis being the log of capital relative to the average and the x-axis being age from 0 to 200. The blue line begins at 0.4 at age 0, then linearly increases to 1.4 in year 150, then grows to 1.5 where it remains stable through year 200.
Figure H.2: Model experiment with compensating cash transfer
This figure shows 3 line charts, each showing the percentage point deviation on the y-axis ranging from 0 to 0.1 and showing the quarters since shock on the x-axis ranging from 0 to 12. Each chart two lines, with the blue line representing an increase in the nominal interest rate i for firms below the median maturing bond share $$\mathcal{M}$$ at the time of the shock and the red dashed line representing an increase in the nominal interest rate i for firms above the median maturing bond share $$\mathcal{M}$$ at the time of the shock. The left chart shows the total cash transfer, with both the red and blue lines beginning positive (red near 0.08 and blue near 0.03) and quickly returning to near zero by quarter 4 where they remain for the remaining quarters. The red line stays slightly above the blue line.
The middle chart shows the transfer bond market revenue, with both the red and blue lines beginning positive (red near 0.07 and blue near 0.02 and quickly returning to near zero by quarter 4 where they remain for the remaining quarters. The red line stays slightly above the blue line.
The right chart shows the transfer real value of nominal debt payments, with both the red and blue lines beginning positive (red near 0.02 and blue near 0.01 and quickly returning to near zero by quarter 3 where they remain for the remaining quarters. The red line stays slightly above the blue line.
Note: The panels show the effect of an unexpected one-standard deviation (30bp) increase in the nominal interest rate i for firms below (blue solid lines) and above (red dash-dotted lines) the median maturing bond share $$\mathcal{M}$$ at the time of the shock. The panels show (a) the average total cash transfer paid (relative to pre-shock firm-level capital), (b) the part of the transfer which compensates changes in bond market revenue (relative to pre-shock firm-level capital), and (c) the part of the transfer which compensates changes in the real value of nominal debt payments (relative to pre-shock firm-level capital).
Figure H.3: Model experiment with exogenously higher
$$\mathcal{M}$$ - Share of debt due in a year
This figure shows a line chart with the percentage from 20 to 100 on the y-axis and the quarters since shock from 0 to 12 on the x-axis. There are two lines, a benchmark line (blue) which remains around 30% for the entire time period, and a exogenously higher $$\mathcal{M}_{it}$$ line (dotted red) that begins at 100 and declines steadily to slightly below 40 by quarter 12.
Note: The lines show the average share of debt due in a year for the representative firm sample selected for the model experiment. The blue solid line shows the benchmark case without exogenous variation of debt maturity. The red dash-dotted line shows the case with an exogenously higher value of $$\mathcal{M}$$ in the initial period.