Eduardo Maqui, Nicholas Vause and Márcia Silva-Pereira
In recent decades, the corporate bond market has grown from a relatively niche source of finance for UK corporations to a central pillar alongside bank loans. This transition raises an important question: as with bank credit conditions, have supply conditions in the corporate bond market come to significantly affect UK economic activity? Our recent research suggests the answer is a resounding yes. We show that a measure of corporate bond financing conditions − the Excess Bond Premium (EBP) − not only anticipates macroeconomic outturns in the UK, but also influences investment by UK firms, especially those that are highly leveraged and more reliant on bond finance.
The rise of bond financing
To motivate our analysis, Chart 1 shows how the composition of UK corporate debt has changed over the past 35 years. A key feature is the rising share of debt securities (mainly corporate bonds), which increased from just 15% in the early 1990s to over 40% by the mid-2020s. Indeed, UK corporations now raise as much finance from bonds as bank loans.
Chart 1: Composition of UK corporate debt
Notes: Non-bank loans includes finance leasing and peer-to-peer lending as well as direct and syndicated loans from non-bank financial institutions. Debt securities is mainly (>90%) corporate bonds but also includes commercial paper.
Source: Bank of England calculations.
Measuring bond financing conditions: the excess bond premium
To study how financing conditions in the corporate bond market affect economic activity, we first require a summary measure of those conditions. Thus, we follow Gilchrist and Zakrajšek (2012) by decomposing corporate bond spreads − the additional compensation required by investors to buy corporate bonds rather than government bonds − into two components. One component reflects ‘fundamentals’ relating to the riskiness of the borrowers or the specific bonds. The residual component is known as the EBP and reflects risk appetite of investors. Specifically, when the EBP rises, it signals that investors require more compensation to hold corporate bonds, over and above what is justified by borrower default risk or bond-specific risks such as illiquidity.
We compute the EBP for the UK by identifying the bond obligations of individual UK firms over time, taking into account mergers and acquisitions. We then combine various sources of data on these matched firms and bonds in order to regress corporate bond spreads on measures of obligor-specific default risk (in particular the distance to default) and bond-specific market and liquidity risks (such as modified duration and size of issue). We retain the residuals from this regression and aggregate them across firms to form the EBP.
Chart 2 shows our results. Investor willingness to invest in UK corporate bonds at lower rates of compensation generated a negative EBP for much of the decade preceding the 2007−08 global financial crisis (GFC) − a period of low macroeconomic uncertainty (The Great Moderation). The EBP then swung sharply positive during the GFC, when investors required substantially more compensation to invest in bonds than suggested by fundamentals. The EBP was also distinctly positive in other periods of financial stress or economic uncertainty, in particular following the dot-com crash (2000−01), during the euro-area sovereign debt crisis (2010−12), ahead of the Brexit referendum (2016), at the outbreak of the Covid-19 pandemic (2020), and following the Russian invasion of Ukraine (2022).
Chart 2: Decomposition of UK corporate bond spreads
Notes: The chart shows an index of corporate bond spreads constructed from 1,680 bonds issued by 149 UK private non-financial corporations (black line) and how it decomposes into a component explained by borrower and bond-specific fundamentals (dark blue) and the excess bond premium (light blue).
Source: Authors calculations. Based on Gilchrist and Zakrajšek (2012).
What happens when bond financing conditions tighten?
Equipped with our measure of bond financing conditions, we first study the consequences of changes in conditions for macroeconomic indicators, including GDP, investment and the unemployment rate. We take two approaches. First, we employ local projections, regressing changes in the macroeconomic indicators from 1 to 16 quarters ahead on contemporary changes in the EBP. In these regressions, we include the policy interest rate and the term spread, as well as several other control variables, to isolate changes in the macroeconomic indicators already anticipated by these other predictors. As shown in Chart 3, a one standard deviation increase in the EBP (of 53 basis points) is associated with a decline in GDP of as much as 2 percentage points, a reduction in investment of as much as 4 percentage points, and an increase in the unemployment rate of as much as 0.5 percentage points. These peak effects all occur about 1.5 years after the shock.
Chart 3: Impulse response of macroeconomic outcomes to an EBP shock
Notes: The panels show estimates of the effects of a one standard deviation EBP shock on macroeconomic outcomes up to four years after the shock. The solid lines show the expected effects, while the darker and lighter shaded areas respectively show ranges in which we are 90% and 95% confident that the effects lie. Refer to the staff working paper for details of the methodology. Investment is gross fixed capital formation.
Source: Authors calculations. Based on Gilchrist and Zakrajšek (2012).
While these effects are sizeable, note that the estimates come with a significant range of uncertainty (blue-shaded areas of the chart). They also depend on our assumption of being able to infer shocks to bond financing conditions from changes in the EBP, which may to an extent be confounded by other macroeconomic drivers. As a sensitivity check, we compute impulse response functions based on a vector auto-regression model and find weaker responses, of around half the magnitudes reported above (refer to Appendix D in the paper), although these effects remain economically significant. Potential limitations to our identification should bite less at the firm level, which we explore below, since firm-level outcomes are less likely to be correlated with confounding aggregate dynamics.
Digging into these aggregate economic responses with similar analyses at sector level, we find that the impact of changes in the EBP is not uniform across different parts of the economy. Notably, investment in capital-intensive assets − like machinery, equipment, and buildings − declines much more than investment in intellectual property. Similarly, investment in manufacturing and production industries is hit harder than investment in services. Interestingly, public-sector investment tends to move countercyclically, increasing when private investment falls, which helps to stabilise capital formation in aggregate.
Firm-level effects: who gets hit hardest?
Finally, we study the effects of shocks to bond financing conditions, as captured by changes in the EBP, on individual firms. Here, we allow for different responses depending on both the level and composition of firms’ debt. Specifically, we allow for different responses for firms in each of the four groups shown in Table A. We estimate these responses through separate local projections for each group, where we regress firm-level outturns − such as growth in investment, assets, sales and profits − over various future horizons on contemporaneous changes in the EBP.
Table A: Firm groups by leverage and share of bond financing
| Group | Leverage (long-term debt/total assets) |
Bond share (bond debt/long-term debt) |
| Low leverage and low bond share (LL) | Below median | Below median |
| Low leverage and high bond share (LH) | Below median | Above median |
| High leverage and low bond share (HL) | Above median | Below median |
| High leverage and high bond share (HH) | Above median | Above median |
Chart 4 shows the results for investment, which is one of our key findings. While the first three panels show no statistically significant response of investment − as measured by capital expenditure − by LL, LH or HL firms to changes in the EBP; the final panel shows that HH firms cut investment aggressively, with a peak decline in investment of almost 10 percentage points around 1.5 years after a one standard deviation shock. Hence, it appears to be the behaviour of these firms − which are not only highly leveraged but have a high share of bonds in their debt − that drives the response of aggregate investment (shown in Chart 3).
Chart 4: Impulse response of firm-level investment to an EBP shock
Notes: The panels show estimates of the effects of a one standard deviation EBP shock on the capital expenditure of UK firms up to four years after the shock for firms with low leverage and low bond share (LL), high leverage and low bond share (HL), low leverage and high bond share (LH) and high leverage and high bond share (HH). The solid lines show the expected effects, while the darker and lighter shaded areas respectively show ranges in which we are 90% and 95% confident that the effects lie. Refer to the staff working paper for details of the methodology.
Source: Authors calculations.
This evidence is consistent with a financial accelerator mechanism in which highly leveraged firms cut investment especially sharply when the cost of finance increases, thereby amplifying the sensitivity of aggregate investment to changes in EBP compared to an economy with a more-even distribution of debt. Our results add a new dimension to this mechanism, as we show the amplification of the investment response to changes in the EBP depends not only on a firm’s leverage but also on the share of bonds in its debt. The results therefore characterise a specifically market‑based finance propagation channel, in which the structure of corporate debt shapes the transmission of financing shocks to real economic activity.
Why does this matter for policy?
Our findings have several important implications. First, the EBP provides a timely signal of changes in bond financing conditions that can foreshadow changes in economic activity. Hence, it may serve policymakers as a useful complement to other business-cycle indicators. Second, the amplified response to changes in the EBP for highly leveraged, bond-reliant firms highlights the importance of diversified funding sources for economic resilience. Third, having shown how changes in bond financing conditions ripple through investment, employment and growth, future research on what in turn determines these conditions seems particularly valuable.
Eduardo Maqui works in the Bank’s RegTech, Data and Innovation Division, Nicholas Vause works in the Bank’s Market-Based Finance Division and Márcia Silva-Pereira is an Economist at Banco de Portugal.
If you want to get in touch, please email us at bankunderground@bankofengland.co.uk or leave a comment below.
Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.
Share the post “Bond financing conditions and economic activity in the UK: aggregate and firm-level evidence”
