Fundamental Risk and Capital Structure Adjustment Speed: International Evidence

Abstract

This study investigates the impact of countries’ fundamental risk on the speed of adjustment (SOA) towards firms’ target capital structures. Using a dataset comprising 17,747 non-financial firms from 44 countries, this study finds that a reduction in country-specific fundamental risk significantly increases a firm’s rate of leverage adjustment. More specifically, we observe that a one standard deviation reduction in fundamental risk results in a substantial 12.79% increase in SOA for book leverage and a 4.81% increase for market leverage. The study also finds evidence of the influence of individual dimensions of fundamental risk on SOA. It implies that improved operational efficiency, high foreign accessibility, enhanced corporate transparency, and increased political stability expedite the pace of leverage adjustment within firms. Robustness checks using a machine learning random forest estimator predicted leverage targets to corroborate these findings. The results highlight the critical role of institutional quality in reducing financing frictions and promoting more efficient corporate capital adjustments. These insights have profound implications for policymakers, emphasising the need to strengthen institutional and regulatory frameworks to enhance capital market integrity and reduce friction, which could ultimately create value for the firm stakeholders.

1. Introduction

Even after decades of research on the implications of optimal capital structure on firm profitability and investments, the increasing level of leverage in the corporate capital structure has long been an intriguing question for practitioners and policymakers1. An extensive body of research suggests that firms seek a target leverage ratio and alter their leverage level once it deviates from the target (An et al., 2021; Cook & Tang, 2010; He et al., 2021; Huang & Ritter, 2009; Li et al., 2023; Öztekin & Flannery, 2012; Rawal et al., 2024a; Rawal et al., 2024b). Inconsistent with Modigliani and Miller’s (1958) irrelevant proposition, the partial-adjustment model of firm leverage advocates that firms do have target capital structures and firms would like to choose the speed of adjustment (SOA hereafter) towards target capital depending upon the trade-off between costs and the benefits of leverage adjustment. Such convergence is conditional on adjustment costs (e.g., transaction and debt agency costs) that limit and hinder complete adjustment toward target leverage (Fisher et al., 1989). As adjustment costs vary between firms, time, and countries, so does the convergence rate (Çam & Özer, 2021; He et al., 2021; Huang & Ritter, 2009; Öztekin & Flannery, 2012).

Initially, empirical findings focused on estimating SOA and shed light on various firm-specific factors influencing the speed of leverage adjustment. These factors encompass firm-level indicators such as investment opportunities, financial constraints, earnings volatility, corporate investment, crash risk exposure, and the sensitivity of equity costs to leverage deviations (Alter et al., 2015; Dang et al., 2012; Faulkender et al., 2012; Tan et al., 2021; Q. Zhou et al., 2016). In recent years, the focus has shifted from identifying firm-specific factors of adjustment cost to country-level determinants as country-specific factors impact the cost of capital and transaction costs (Belkhir et al., 2016; Çam & Özer, 2021). Prior research indicates that capital structure adjustment is influenced by a country’s legal tradition, financial system development, ease of capital access, shareholder rights enforcement, political and economic uncertainty, and macroeconomic condition (M. Baker, 2009; Çolak et al., 2018; Cook & Tang, 2010; Graham & Leary, 2011; Kang et al., 2018; Öztekin, 2015; Öztekin & Flannery, 2012; Rajan & Zingales, 1995). Appendix A comprehensively summarises the empirical research on the factors affecting capital structure adjustment speed in a tabular form.

No empirical evidence, however, has explored the impact of a country’s fundamental risk on the capital structure SOA. Nevertheless, economic reasoning suggests that a country’s risk could influence the firm’s adjustment costs when it seeks to modify its existing capital structure. This is due to several reasons. Firstly, a country’s fundamental risk signifies various vulnerabilities within its capital markets. These vulnerabilities arise from weaknesses in the institutions that are vital for the smooth operation of capital markets. A notable concern is the limited capacity of a country’s financial markets to finance its domestic growth opportunities. This constraint forces multinational corporations and foreign portfolio investors to actively participate in supporting domestic companies that encounter funding shortfalls; however, these markets face hindrances that impede the seamless flow of capital. These hindrances encompass operational inefficiencies, limitations on foreign involvement, issues pertaining to governance and transparency regulations, inadequate legal safeguards for resolving disputes, and political instability. These factors collectively discourage potential investors, both those seeking direct investments and those interested in portfolio investments, from engaging with these markets. Consequently, this situation results in a higher cost of capital for firms operating in countries with high fundamental risks. This, in turn, could translate to greater adjustment costs for firms seeking to modify their capital structure. Consequently, under the dynamic trade-off theory, it may lead to a slower SOA towards achieving the desired target capital structure. In other words, it is reasonable to argue that lower fundamental risk d a country would facilitate firms’ capital access by reducing the severity of market frictions such as agency conflicts, information asymmetry, and transaction costs, which may lead to a quicker convergence towards the target capital structure. Thus, we hypothesise that a company’s SOA to target capital structure will be more favourable in countries with lower fundamental risk.

To the best of our knowledge, no studies have investigated the influence of countries’ fundamental risk on the SOA; however, it has been argued that a country’s fundamental risk affects the cost of debt and equity (Çam & Özer, 2021). An advantage of examining a target adjustment model in the context of comprehensive fundamental risk indicators is that the costs and benefits of adjusting to targets will be a pure function of an inclusive measure of fundamental risk.

Employing a large sample of 17,747 non-financial firms across 44 economies spanning 15 years, our findings indicate that country-specific fundamental risk affects how corporations move towards their target capital structure. Using a partial adjustment model of firm leverage, results suggest that an economic environment supported by lower fundamental risk enhances capital market access and facilitates the rebalancing of capital structures. Consistent with related literature, lower fundamental risk in a country would facilitate firms’ access to appropriate long-term capital by reducing the severity of market frictions such as agency conflicts, information asymmetry, and transaction costs, which may lead to a quicker convergence towards the target capital structure (Belkhir et al., 2016; Çam & Özer, 2021; Öztekin & Flannery, 2012).

Our research adds to the existing body of knowledge in several ways. First, to our knowledge, no study examines the impact of countries’ fundamental risk in explaining disparities in adjustment speed amongst firms. By doing so, our study expands the growing corporate finance literature, which highlights the effects of country-specific factors on capital structure adjustment speed (Çolak et al., 2018; Kang et al., 2018; Öztekin, 2015; Öztekin & Flannery, 2012). Second, there is no empirical evidence on the implications of comprehensive fundamental risk indicators of (Karolyi, 2015) on SOA. Previous studies by (Goodell & Goyal, 2018) and (Çam & Özer, 2021), using the country risk dimensions of (Karolyi, 2015), showed that firms utilise less debt in countries that have lower fundamental risk. Our study further expands these findings in the context of target capital structure adjustment. Related strands of literature using machine learning techniques like random forest estimation (Amini et al., 2021; Smith, 2022) have been used to predict debt equity decisions with a reasonable level of accuracy (92% correct prediction). However, to the best of our knowledge, no one has used this technique in a dynamic capital structure framework to predict target capital structure before using these predicted targets to identify adjustment speed determinants. We fill this gap by doing the same as the robustness test.

The remainder of the paper is organised as follows: Section 2 presents the materials and methods, which include the data, variable description, and model specification; Section 3 presents the results, which also includes the robustness test; a discussion of the results is provided in Section 4; and Section 5 concludes the paper, as well as highlighting the limitations and scope of future research.

2. Materials and Methods

2.1. Materials (Data and Variables)

2.1.1. Data

The study sample represents 17,747 non-financial firms from 44 countries (Morgan Stanley Capital International developed and emerging market classification) from 2000 to 2015. We have excluded financial and utility firms as regulatory norms influence their capital structure decisions. The study period has been restricted due to the unavailability of countries’ fundamental risk (frisk) data. Sample selection criteria further require firms to have non-missing data and exclude any firm with fewer than 2 consecutive years of data. We collect firm-level financial information from the Thomson Reuters Eikon database. Data related to countries’ fundamental risk indicators have been collected from G. Andrew Karolyi’s “Cracking the emerging markets enigma” (Karolyi, 2015). In addition to this, data related to macroeconomic fundamentals are collected from the World Bank database. We winsorised all firm-level data at the 1% level to eliminate outliers. These procedures eventually result in 184,792 firm-year observations.

Table 1. Sample distribution, mean value of book and market leverage, and the mean value of the index of different dimensions of fundamental risk across countries.

Country Name	No of Firms	No. of Observation	Book Leverage	Market Leverage	Market Capacity	Operational Efficiency	Foreign Accessibility	Corporate Transparency	Political Stability	Fundamental Risk
Austria	43	573	0.33	0.30	−0.34	0.27	0.95	−0.05	1.11	0.74
Belgium	75	629	0.35	0.31	−0.21	0.52	1.13	−0.03	1.04	1.09
Brazil	113	1135	0.35	0.38	−0.40	−0.24	−0.96	0.37	−0.39	−0.71
Canada	488	4639	0.24	0.20	1.49	1.08	0.40	1.80	2.98	3.04
Chile	88	1087	0.28	0.34	−0.12	−1.96	0.15	0.38	0.39	−0.36
China	2666	22,165	0.29	0.17	−0.05	−0.45	−1.64	−0.85	−1.51	−2.22
Colombia	16	120	0.18	0.14	−0.79	−0.98	−0.98	−1.99	−2.70	−2.16
Denmark	61	762	0.29	0.25	1.05	0.97	0.63	0.18	1.46	1.95
Egypt	89	788	0.23	0.21	−0.73	−1.15	0.15	−0.27	−1.58	−2.01
Finland	87	1031	0.32	0.28	0.12	1.28	0.27	1.04	1.50	1.82
France	249	3039	0.33	0.27	0.12	1.08	0.98	1.42	0.43	1.64
Germany	350	4089	0.29	0.24	−0.13	0.94	0.45	0.70	1.04	1.26
Greece	101	1319	0.37	0.43	−0.12	0.60	0.86	−1.38	−0.21	−0.64
Hong Kong	709	8003	0.21	0.26	4.35	0.34	0.44	0.72	1.11	3.69
Hungary	11	121	0.27	0.26	−0.71	0.59	0.39	−0.09	0.09	−0.30
India	1590	13,282	0.34	0.34	−0.58	0.01	−1.86	−1.04	−0.86	−1.62
Indonesia	315	3248	0.32	0.30	−0.99	0.06	−0.79	−0.80	−0.92	1.63
Ireland	41	395	0.33	0.21	0.20	1.10	2.00	1.26	1.10	2.84
Israel	243	2302	0.40	0.37	0.40	−0.51	0.08	0.06	0.35	0.80
Italy	128	1389	0.39	0.34	−0.12	−0.50	0.93	0.26	−0.15	0.46
Japan	2640	33,284	0.29	0.30	1.60	−0.91	0.54	−0.87	0.55	0.64
S. Korea	1504	16,536	0.28	0.31	0.67	0.02	−0.16	−1.89	0.02	−0.43
Malaysia	595	7461	0.23	0.26	0.58	−0.48	−1.48	0.77	−0.29	0.35
Mexico	83	955	0.30	0.27	−0.98	−0.24	−0.27	0.75	−0.75	−0.99
Netherlands	60	749	0.37	0.29	0.50	−0.22	1.70	0.97	1.37	1.70
New Zealand	68	691	0.26	0.19	−0.11	0.29	0.12	1.09	1.23	1.97
Norway	76	793	0.37	0.34	−0.10	0.43	0.79	0.88	1.29	1.51
Peru	51	639	0.28	0.46	−1.00	−1.05	0.99	−0.21	−0.76	−1.05
Philippines	151	1182	0.22	0.21	−0.23	−0.42	−1.03	−0.85	−0.92	−2.09
Poland	286	1801	0.22	0.20	−0.64	0.23	−1.20	−0.12	0.19	−0.59
Portugal	19	255	0.52	0.45	−0.08	0.78	0.83	−0.53	0.47	0.52
Russia	76	444	0.38	0.35	−0.80	0.28	−0.57	−1.28	−1.14	−1.73
Saudi Arabia	107	968	0.23	0.15	−0.60	−0.46	−0.26	−0.06	−1.27	−1.22
Singapore	347	3778	0.25	0.27	1.22	0.77	0.16	0.84	0.72	2.36
South Africa	129	1535	0.27	0.21	0.61	−0.11	−0.40	1.36	−0.20	1.01
Spain	83	817	0.42	0.31	1.09	0.48	0.89	0.97	0.37	1.51
Sweden	291	2504	0.25	0.21	0.49	1.11	0.90	1.02	1.34	20.12
Switzerland	131	1712	0.27	0.22	1.20	0.81	0.79	0.67	1.45	1.81
Thailand	446	4758	0.27	0.25	−0.02	0.34	−1.35	−0.73	−0.98	−1.03
Turkey	238	2405	0.25	0.21	−0.73	−1.61	−0.44	−1.22	−0.66	−2.11
UAE	47	330	0.24	0.25	−0.61	−1.11	−0.50	0.36	−0.06	−1.07
UK	532	5729	0.24	0.19	0.97	1.80	0.95	1.58	0.75	3.22
USA	1582	18,127	0.24	0.17	1.42	2.14	0.11	1.21	0.49	2.87
Total	17,747	184,792

Note: Table 1 presents the distribution of firms across countries in our sample. It also presents the mean value of book and market leverage and the mean index value of different dimensions of fundamental risk in each country for the entire sample period.

represents the sample distribution.

2.1.2. Variables

Leverage: We measure leverage as the ratio of the total debt to total capital, where capital is defined as the sum of total debt and equity. This measure offers a comprehensive view of past financing decisions by accounting for the capital employed. This study used the book and market value leverage as a proxy for firms’ leverage. Book leverage (BLEV) is defined as total debt/(total debt + equity), and market leverage (MLEV) is defined as total debt/(total debt + market capitalisation).

Country fundamental risk: In order to measure country-level fundamental risk, we used the five dimensions of country risk: market capacity (MC), which measures the capital markets development; operational efficiency (OE), which measures the efficiencies of trading systems; foreign accessibility (FA), which measures the flow of capital; corporate transparency (CT), which measures the corporate reporting and governance; and political stability (PS), which captures the institutions and government efficiency of a country. MC measures the breadth, scope, and trading activity of financial markets and, therefore, minimises the costs of stock issuance and debt contracting (Belkhir et al., 2016; Çam & Özer, 2021; Demirgüç-Kunt & Maksimovic, 1999; Öztekin, 2015) and reduces transaction costs (Belkhir et al., 2016; Öztekin & Flannery, 2012). OE examines the technological efficiencies of trading systems and their contribution to reducing transaction costs. FA indicates the free movement of capital and access to external financing (An et al., 2021; Do et al., 2020). Efficient CT reduces information asymmetry and agency conflicts and lowers the cost of capital (Belkhir et al., 2016; North, 1990). PS represents the ability of institutions to safeguard investors, such as legal protection; thus, it reduces transaction costs (M. Baker, 2009; Goodell & Goyal, 2018). The country fundamental risk is measured as the first principal component (PC1) using the principal component analysis on all five risk dimensions mentioned above (MC, OE, FA, CT, PS). A higher value in the fundamental risk variable indicates lower risk.

Other variables: Consistent with related literature (Cook & Tang, 2010; Frank & Goyal, 2009; He et al., 2021; Kim & Xie, 2023; Öztekin & Flannery, 2012; Smith, 2022), this study has used several firm and industry-specific determinants in order to measure the target leverage of firms, which includes firm size (SIZE), tangibility (TANG), market-to-book ratio (MTB), profitability (PROF), depreciation (DEP), R&D expense (RDEXP), R&D dummy (RD_DUM), industry book leverage (BLEV_IND), and industry market leverage (MLEV_IND).

In addition, we have used financial constraints (FC), over-levered dummy (OL), GDP growth rate (GDPG), and inflation rate (INF) as control variables affecting a firm’s speed of adjustment (An et al., 2021).

Table 2. Variable definitions.

Variable	Variable Description	Variable Measurement
>BLEV	>Book leverage	>Total debt/(total debt + book value of equity)
>MLEV	>Market leverage	>Total debt/(total debt + market value of equity)
BDEV	>Book leverage deviation	>Deviation of book leverage from the target book leverage
MDEV	>Market leverage deviation	>Deviation of market leverage from target market leverage
PROF	>Profitability	>Ratio of operating income before depreciation to total assets
TANG	>Tangibility	>Ratio of net property, plant, and equipment to total assets
MTB	>Market-to-Book	>Ratio of total assets minus book equity plus market capitalisation to total assets
SIZE	>Size	>Natural logarithm of total assets
BLEV_IND	>Industry book leverage	>Median book leverage ratio by country, industry, and year
MLEV_IND	>Industry market leverage	>Median book leverage ratio by country, industry, and year
RDEXP	R&D expense	>Ratio of R&D expenses to total assets, where missing R&D values are equal to zero
RD_DUM	>R&D dummy	>Dummy variable equals one if R&D expenses are not reported; otherwise, it is zero
DEP	>Depreciation	>Captures non-debt tax shields measured by depreciation and amortisation divided by the book value of total assets
>OL	>Over leveraged	>Dummy variable equals one if the observed leverage ratio is higher than the target leverage; otherwise, it is zero
FC	Financial constraint	>Dummy variable equals one if the cash flow from the operation is negative; otherwise, it is zero
>GDPG	>GDP growth	>Rate of change in the gross domestic product (annual%)
>INF	>Inflation	>Rate of inflation
>MC	>High market capacity	>Dummy variable equals one if the observed score in the market capacity index is higher than the median score in that year for the sample; otherwise, it is zero
>OE	>High operational efficiency	>The dummy variable equals one if the observed score in the Operational Efficiency index is higher than the median score in that year for the sample; otherwise, it is zero.
>FA	>High foreign accessibility	>The dummy variable equals one if the observed score in the foreign accessibility index is higher than the median score in that year for the sample; otherwise, it is zero
>CT	>High corporate transparency	>The dummy variable equals one if the observed corporate transparency index score is higher than the sample median score in that year; otherwise, it is zero
>PS	>High political stability	>Dummy variable equals one if the observed score in the political stability index is higher than the median score in that year for the sample; otherwise, it is zero
>Frisk	>Low fundamental risk	>Dummy variable equals one if the observed score in the fundamental risk index is higher than the median score in that year for the sample; otherwise, it is zero (Higher value in the index means lower risk)

Note: Table 2 presents the description and measurement of the variable used in our study. Among the above variables, size (SIZE), tangibility (TANG), market-to-book ratio (MTB), profitability (PROF), depreciation (DEP), R&D expense (RDEXP), R&D dummy (RD_DUM), industry book leverage (BLEV_IND), and industry market leverage (MLEV_IND) are used for target leverage estimation. Overleveraged (OL), financial constraint (FC), GDP growth (GDPG), and inflation (INF) are used as control variables for speed of adjustment.

presents the variable definition.

Table 3. Descriptive Statistics.

Variables	Mean	Std Dev	Minimum	Maximum
BLEV	0.28	0.232	0	0.85
MLEV	0.26	0.249	0	0.90
BDEV	0.01	0.112	−0.847	0.733
MDEV	0.01	0.122	−0.896	0.814
PROF	0.08	0.112	−0.496	0.369
TANG	0.35	0.238	0.001	0.936
MTB	1.68	1.69	0.375	12.33
SIZE	19.23	1.95	14.78	24.41
BLEV_IND	0.26	0.15	0	1
MLEV_IND	0.23	0.18	0	1
RDEXP	0.01	0.074	−0.13	11.84
RD_DUM	0.27	0.45	0	1
DEP	0.03	0.03	0.0001	0.1458
OL	0.44	0.497	0	1
FC	0.20	0.402	0	1
GDPG	3.71	3.44	−10.14	25.17
INF	2.55	2.92	−4.47	44.96
MC	0.47	0.49	0	1
OE	0.48	0.50	0	1
FA	0.45	0.50	0	1
CT	0.47	0.49	0	1
PS	0.44	0.50	0	1
Frisk	0.40	0.49	0	1

Note: Table 3 reports descriptive statistics of the variables used in our study.

reports descriptive statistics of the variables used in our study. An average sample firm finances nearly 28% of its capital using debt (26% as per MLEV), earns around 8% return on assets, and has a tangibility ratio of 35%. The median BLEV and MLEV by country, industry, and year are 26% and 23%, respectively.

Table 4. Correlation Matrix.

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
(1) BLEV	1.000
(2) MLEV	0.800	1.000
(3) PROF	−0.005	−0.074	1.000
(4) DEP	0.068	0.016	0.246	1.000
(5) TANG	0.043	0.059	0.027	0.189	1.000
(6) MTB	−0.170	−0.379	0.044	0.015	−0.024	1.000
(7) SIZE	0.296	0.202	0.242	0.020	−0.050	−0.109	1.000
(8) BLEV_IND	0.513	0.453	0.100	0.017	−0.007	−0.167	0.247	1.000
(9) MLEV_IND	0.417	0.571	0.026	−0.027	0.012	−0.315	0.156	0.788	1.000
(10) RD	−0.107	−0.114	−0.291	0.018	0.036	0.130	−0.099	−0.162	−0.141	1.000
(11) RD_DUM	−0.109	−0.113	−0.113	0.027	−0.006	0.073	0.058	−0.192	−0.150	0.280	1.000

Note: Table 4 presents the variables’ correlation coefficients used for target leverage estimation.

presents the correlation coefficients of variables used for target leverage estimation.

2.2. Model Specification and Estimation Methods

Consistent with the existing literature (Çolak et al., 2018; Flannery & Rangan, 2006; Lemmon et al., 2008; Öztekin & Flannery, 2012) (see methodology column of Appendix A), we employ the following partial-adjustment model to estimate how fast a firm offsets the deviation from the target leverage:

$$LE{V}_{i,t+1,j} - LE{V}_{i,t,j}= \lambda \left(LE{V}_{i,t+1,j}^{\ast }- LE{V}_{i,t,j} \right)+ {\delta }_{i,t+1,j}\dots$$

(1)

where $LE{V}_{i,t,j}$ is measured by the book and market leverage (BLEV, MLEV) of firm i in country j at year t + 1, $LE{V}_{i,t+1,j}^{\ast }$ is the target leverage, and $\lambda$ is the speed of adjustment toward the target leverage. Since the target leverage $(LE{V}_{i,t+1,j}^{\ast }$) is unobservable, consistent with An et al. (2021), the fitted value of Equation (2) is used as a proxy for the target leverage. Equation (2) is estimated using fixed effects:

$$LE{V}_{i,t+1,j}= {\alpha }_i+ {\beta }_j{X}_{i,t,j}+f_{i,}+ t_{i, }+ {\upsilon }_{i,t+1,j}\dots$$

(2)

where, $X_{i,t,j}$ is a vector of firm- and industry-level variables, including firm SIZE, TANG, MTB, PROF, DEP, RDEXP, RD_DUM, and LEV_IND. To capture the unobserved heterogeneity across firms and years, we include firm and year fixed effects (i.e., f_i and t_i) in the model (Lemmon et al., 2008). Furthermore, Equation (1) can be rearranged as follows:

$$\Delta BLE{V}_{I,t+1,j} = \lambda \left( BDE{V}_{i,t,j}\right)+ {\delta }_{i,t+1,j}\dots$$

(3)

$$\Delta MLE{V}_{I,t+1,j} = \lambda \left( MDE{V}_{i,t,j}\right)+ {\delta }_{i,t+1,j} \dots$$

(4)

where $\Delta BLE{V}_{I,t+1,j}$ and $\Delta MLE{V}_{I,t+1,j}$ represent the change in book and market leverage of a firm i from year t to t + 1 in country j, $BDE{V}_{i,t,j}$ indicates the deviation of actual book leverage from target ($\mathrm{i}.\mathrm{e}., BLE{V}_{i,t+1,j}^{\ast }- BLE{V}_{i,t,j}$), and $MDE{V}_{i,t,j}$ represents the deviation of actual market leverage from target ($\mathrm{i}.\mathrm{e}., MLE{V}_{i,t+1,j}^{\ast }- MLE{V}_{i,t,j}$). The $\lambda$ coefficient estimates the adjustment speed towards the target leverage. A higher value of λ indicates a faster SOA. To examine the impact of the country’s fundamental risk dimensions on the SOA, we relax the assumption of a constant adjustment rate (Equations (3) and (4)) and allow the firm SOA to depend on the fundamental risk of the country (Öztekin & Flannery, 2012). Hence, we model $\lambda$ in Equations (3) and (4) as a function of the country’s fundamental risk dimension and other control variables:

$${\lambda }_{i,t,j}= {\beta }_0+ {\beta }_1{FRisk}_{i,t,j}+ {\beta }_2{{\rm Y}}_{i,t,j}\dots$$

(5)

where $FRisk$ is the country’s fundamental risk dimension and ${{\rm Y}}_{i,t,j}$ is the set of control variables. Substituting Equation (5) into Equations (3) and (4), and controlling for the industry (Ind), country (j), and year (t) fixed effects, results in the following specifications:

$$\Delta BLE{V}_{I,t+1,j}= {\beta }_{0}BDE{V}_{i,t,j}+ {\beta }_{1}FRis{k}_{i,t,j}\ast BDE{V}_{i,t,j}+ {\beta }_2 Z_{i,t,j}\ast BDE{V}_{i,t,j}+Ind.+j+t+{\delta }_{i,t+1,j}\dots$$

(6)

$$\Delta MLE{V}_{I,t+1,j} = {\beta }_{0}MDE{V}_{i,t,j}+ {\beta }_{1}FRis{k}_{i,t,j}\ast MDE{V}_{i,t,j}+ {\beta }_2 Z_{i,t,j}\ast MDE{V}_{i,t,j}+Ind.+j+t+{\delta }_{i,t+1,j}\dots$$

(7)

Equations (6) and (7) serve as our final empirical model for estimation, which is estimated with OLS with bootstrapped standard errors to account for generated regressor (Pagan, 1984).

A positive and significant ${\beta }_0$ in Equations (6) and (7) implies that the firm has a target leverage ratio and actively adjusts its capital structure to converge towards this target. When multiplied by 100, the coefficient’s magnitude denotes the proportional decrease in the deviation between the firm’s initial leverage and the target leverage over 1 year, known as SOA towards target capital structure.

The percentage change in the SOA of a firm due to a one standard deviation increase in $FRisk$ can be found by using the following equation:

$$SO{A}_{change}={\displaystyle \frac{\sigma Frisk\times {\beta }_1 }{{\beta }_0}}\dots$$

(8)

3. Results

3.1. Preliminary Evidence of SOA Heterogeneity Across Country Risk Profiles

Before presenting our main results, we provide preliminary evidence on heterogeneity in the speed of capital structure adjustments (SOA) across different country risk environments. Specifically, we partition the sample into high-risk and low-risk country groups based on several dimensions of national fundamental risk.

Countries are classified into high or low risk for each risk dimension using the sample median as the threshold. We then estimate the SOA separately for firms operating within each group using Equation (3).

Table 5. Average SOA of firms across different dimensions of country fundamental risk.

Risk Dimension	Group	Avg. SOA	Significant Difference (High vs. Low)
Fundamental risk	High	38%	−7% ***
	Low	45%
Market capacity	High	41%	1% *
	Low	40%
Operational efficiency	High	44%	7% ***
	Low	37%
Foreign accessibility	High	41%	0%
	Low	41%
Corporate transparency	High	44%	6% ***
	Low	38%
Political stability	High	43%	3% ***
	Low	40%

Note: Table 5 reports the average speed of capital structure adjustments (SOA) for firms operating in countries classified as high- or low-risk based on fivedimensions of national risk. SOA is estimated separately for each group using Equation (3). The last column reports the difference in SOA between high- and low-risk groups, with statistical significance denoted as ***, and * indicating statistical significance at the 1%, and 10% levels, respectively.

shows that firms in countries with low fundamental risk measured as the first principal component of five institutional quality dimensions (market capacity, operational efficiency, foreign accessibility, corporate transparency, and political stability) adjust significantly faster (45%) than those in high-risk countries (38%), with a 7% difference (p < 0.01). Corporate transparency, operational efficiency, and political stability are individually associated with faster SOA in low-risk settings. Foreign accessibility shows no difference, while market capacity shows a small but significant effect. These findings suggest that institutional quality and macro-level stability can facilitate firms’ more responsive capital structure decisions.

This heterogeneity underscores the importance of accounting for country-level risk characteristics when analysing capital structure dynamics. It strongly motivates the more detailed empirical analysis presented in the following sections.

3.2. Baseline Results

Table 6. Fundamental Risk and the Speed of Adjustment of Book Leverage.

	(1)	(2)	(3)	(4)	(5)	(6)
BDEV	0.2942 *** (35.51)	0.2904 *** (34.64)	0.3263 *** (40.20)	0.2923 *** (45.03)	0.3102 *** (48.63)	0.3422 *** (35.90)
Frisk * BDEV	0.0742 *** (10.49)
OE * BDEV		0.0734 *** (12.46)
FA * BDEV			0.0187 ** (2.41)
CT * BDEV				0.0727 *** (10.05)
PS * BDEV					0.0449 *** (6.85)
MC * BDEV						0.0034 (0.44)
FC * BDEV	0.0680 *** (8.27)	0.0655 *** (7.93)	0.0709 *** (8.92)	0.0682 *** (8.79)	0.0693 *** (8.74)	0.0714 *** (9.22)
OL * BDEV	0.0700 *** (6.88)	0.0715 *** (6.92)	0.0733 *** (9.18)	0.0703 *** (8.97)	0.0715 *** (8.12)	0.0740 *** (9.28)
GDPG * BDEV	0.0039 *** (4.01)	0.0034 *** (3.61)	0.0013 * (1.69)	0.0035 *** (3.43)	0.0026 *** (2.39)	0.0001 (0.17)
INF * BDEV	0.0050 *** (4.93)	0.004 *** (4.30)	0.0046 *** (3.85)	0.0045 *** (5.30)	0.0051 *** (4.78)	0.0036 *** (3.69)
Constant	0.0026 (0.84)	0.0029 (1.18)	0.0029 (0.99)	0.0026 (0.86)	0.0027 (1.02)	0.0029 (0.96)
Country FE	Yes	Yes	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
Adj.$R^2$	0.2247	0.225	0.224	0.225	0.224	0.22
N	184,792	184,792	184,792	184,792	184,792	184,792

Notes: Table 6 reports the regression results of Equation (6), which measures the impact of country fundamental risk (frisk) on SOA to target book leverage. It also reports the impact of individual dimensions of countries’ fundamental risk, such as the impact of OE, FA, CT, PS, and MC on SOA to target book leverage. The sample represents 17,747 companies from 44 countries (Table 1) from 2000 to 2015. Numbers in parentheses are t-statistics. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

and

Table 7. Fundamental Risk and the Speed of Adjustment of Market Leverage.

	(1)	(2)	(3)	(4)	(5)	(6)
MDEV	0.3396 *** (47.60)	0.3432 *** (50.98)	0.3474 *** (36.45)	0.3482 *** (51.74)	0.3500 *** (57.12)	0.3658 *** (35.77)
Frisk MDEV	0.0318 *** (5.41)
OE * MDEV		0.0221 *** (4.07)
FA * MDEV			0.0126 * (1.66)
CT * MDEV				0.0145 ** (2.26)
PS * MDEV					0.0114 * (1.63)
MC * MDEV						0.0112 (1.46)
FC * MDEV	0.0269 *** (3.45)	0.0267 *** (4.48)	0.0278 *** (3.91)	0.0275 *** (4.02)	0.0276 *** (3.94)	0.0284 *** (3.88)
OL * MDEV	0.1145 *** (14.20)	0.1151 *** (12.74)	0.1171 *** (16.26)	0.1152 *** (13.10)	0.1151 *** (15.40)	0.1156 *** (11.97)
GDPG * MDEV	0.0059 *** (5.93)	0.0054 *** (5.27)	0.0053 *** (4.45)	0.0052 *** (6.15)	0.0052 *** (4.52)	0.0042 *** (5.03)
INF * MDEV	0.0135 *** (14.56)	0.0132 *** (12.15)	0.0135 *** (11.08)	0.0131 *** (13.85)	0.0132 *** (10.08)	0.0122 *** (9.17)
Constant	0.0087 *** (3.21)	0.0088 *** (3.16)	0.0088 *** (3.44)	0.0088 *** (2.74)	0.0088 *** (3.81)	0.0088 *** (2.93)
Country FE	Yes	Yes	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
Adj.$R^2$	0.3314	0.3317	0.3312	0.3312	0.3312	0.3312
N	184,792	184,792	184,792	184,792	184,792	184,792

Notes: Table 7 reports the regression results of Equation (7), which measures the impact of country fundamental risk (frisk) on SOA to target market leverage. It also reports the impact of individual dimensions of countries’ fundamental risk, such as the impact of OE, FA, CT, PS, and MC on SOA to target market leverage. The sample represents 17,747 companies from 44 countries (Table 1) from 2000 to 2015. Numbers in parentheses are t-statistics. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

present our main partial-adjustment model estimation results for book leverage (∆BLEV) and market leverage (∆MLEV), respectively. Consistent with the literature, all independent variables are multiplied by the leverage deviation from the target (BDEV or MDEV).

Our results suggest that a lower country-level frisk influences SOA positively (Coef. = 0.071, t = 10.25). The result is not only statistically significant but also has economic significance. Regarding economic significance, a one standard deviation increase in frisk, which means a decrease in the fundamental risk as a higher value in the frisk index indicates lower risk, increases the SOA by 12.79% (=0.071 ×0.49/0.34) (calculated using Equation (8)).

In addition, Columns 2, 3, 4, and 5 show that the coefficient estimate for countries with high operating efficiency (Coef. = 0.073), high foreign accessibility (Coef. = 0.018), high corporate transparency (Coef. = 0.072), and high political stability (Coef. = 0.044) are positive and significant at a 1% level.

Using market leverage (as shown in Table 7), we observe that the Frisk * MDEV coefficient is also positive and significant (Coef. = 0.0318, t = 5.41), which means that a one standard deviation increase in frisk increases the SOA of market leverage by 4.81% (=0.0318 × 0.50/0.33) (using Equation (8)).

Except for MC, we find all other individual components of fundamental risk, i.e., OE, FA, CT, and PS, are positive and significant. This implies that a firm’s observed leverage can deviate from its target level in the presence of considerable adjustment costs due to higher frisk.

In Table 6 and Table 7, turning to the effect of control variables on SOA, we find that our result is consistent with the literature (An et al., 2021; Cook & Tang, 2010; Dang et al., 2012). In particular, levered firms (coefficients of OL * BDEV and OL * MDEV) and financially constrained firms (coefficients of FC * BDEV and FC * MDEV) show higher SOA (Drobetz & Wanzenried, 2006; Flannery & Rangan, 2006). We observe that firm SOA is higher when economic fundamentals are better (Cook & Tang, 2010).

3.3. Robustness Test

In our robustness tests, we use the random forest estimation (RFE) technique to estimate Equation (2) and to use the predicted target leverage for finding the impact of fundamental risk on SOA using Equations (6) and (7). (Graham & Leary, 2011) suggest that standard variables used to determine capital structure may have non-linear relations with dependent variables. Nevertheless, few empirical tests explicitly model these nonlinearities when predicting target leverage or quantifying a determinant’s importance. This is particularly appealing because accurately predicting unobservable leverage targets is essential for estimating leverage adjustment speed and accurately investigating factors affecting adjustment speed. Furthermore, leverage adjustments are discrete, infrequent, and large, posing significant challenges in assessing the targets. In recent years, machine learning techniques like RFE have been successfully used for accommodating non-linear functional relationships among estimators (Amini et al., 2021; Smith, 2022).

Table 8. Random Forest Estimation: Fundamental Risk and Speed of Adjustment.

$\mathbf{Panel} \mathbf{\left(}\mathbf{A}\mathbf{\right)}: \mathbf{Dependent} \mathbf{Variable}\mathbf{:} \mathbf{\Delta }\mathit{B}\mathit{L}\mathit{E}{\mathit{V}}_{\mathit{I}\mathbf{,}\mathit{t}\mathbf{+}\mathbf{1}\mathbf{,}\mathit{j}}$
	(1)	(2)	(3)	(4)	(5)	(6)
BDEV	0.517 4 *** (89.29)	0.5082 *** (75.87)	0.6153 *** (102.80)	0.5040 *** (71.72)	0.5665 *** (77.58)	0.6466 *** (72.30)
Frisk * BDEV	0.1636 *** (37.06)
OE * BDEV		0.1676 *** (29.25)
FA * BDEV			0.0088 * (1.66)
CT * BDEV				0.1776 *** (29.89)
PS * BDEV					0.0761 *** (11.63)
MC * BDEV						0.0113 (1.47)
FC * BDEV	0.1339 *** (28.80)	0.1267 *** (19.21)	0.1449 *** (22.33)	0.1337 *** (23.03)	0.1412 *** (25.62)	0.1463 *** (22.93)
OL * BDEV	0.1837 *** (22.99)	0.1821 *** (19.33)	0.1798 *** (20.24)	0.1832 *** (18.39)	0.1830 *** (23.68)	0.1775 *** (17.91)
GDPG * BDEV	0.0132 *** (15.73)	0.0122 *** (16.23)	0.0061 *** (8.12)	0.0129 *** (18.04)	0.0095 *** (10.72)	0.0044 *** (4.64)
Inf * BDEV	0.0070 *** (8.12)	0.0061 *** (6.69)	0.0048 *** (6.29)	0.0061 *** (8.31)	0.0069 *** (8.97)	0.0031 *** (3.34)
Constant	−0.0059 *** (−2.71)	−0.0055 *** (−2.82)	−0.0057 *** (−3.39)	−0.0058 *** (−3.60)	−0.005 *** (−3.26)	−0.005 *** (−3.08)
Country FE	Yes	Yes	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes	Yes	Yes
Adj.$R^2$	0.63	0.63	0.62	0.63	0.63	0.62
N	184,792	184,792	184,792	184,792	184,792	184,792
Panel (B): Dependent Variable: $\Delta \mathit{M}\mathit{L}\mathit{E}{\mathit{V}}_{\mathit{I}\mathbf{,}\mathit{t}\mathbf{+}\mathbf{1}\mathbf{,}\mathit{j}}$
MDEV	0.6284 *** (115.04)	0.6278 *** (123.04)	0.6968 *** (108.62)	0.6208 *** (131.78)	0.6733 *** (110.65)	0.7172 *** (104.23)
Frisk * MDEV	0.1093 *** (24.33)
OE * MDEV		0.0977 *** (23.41)
FA * MDEV			0.0281 * (1.61)
CT * MDEV				0.1141 *** (24.23)
PS * MDEV					0.0240 *** (5.20)
MC * MDEV						0.0351 *** (5.95)
FC * MDEV	0.0595 *** (11.72)	0.0567 *** (9.96)	0.0654 *** (14.23)	0.0591 *** (11.05)	0.0643 *** (12.50)	0.0665 *** (14.09)
OL * MDEV	0.1710 *** (22.57)	0.1714 *** (23.72)	0.1715 *** (23.56)	0.1697 *** (21.85)	0.1712 *** (23.74)	0.1701 *** (24.54)
GDPG * MDEV	0.0147 *** (29.83)	0.0136 *** (25.40)	0.0104 *** (14.36)	0.0143 *** (27.00)	0.0119 *** (19.43)	0.0093 *** (18.35)
Inf * MDEV	0.0150 *** (21.09)	0.0143 *** (21.07)	0.0117 *** (15.13)	0.0145 *** (20.11)	0.0133 *** (15.94)	0.0102 *** (11.42)
Constant	−0.0043 *** (−3.19)	−0.004 *** (−2.50)	−0.0042 ** (−2.41)	−0.0042 ** (−2.40)	−0.004 ** (−2.07)	−0.004 *** (−2.32)
Country FE	Yes	Yes	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes	Yes	Yes
Adj.$R^2$	0.70	0.70	0.70	0.70	0.70	0.70
N	184,792	184,792	184,792	184,792	184,792	184,792

Notes: Table 8 presents the robustness of our results. Here, the target leverage is estimated using the random forest technique by estimating Equation (2). Panel A of Table 8 reports the regression results of Equation (6), which measures the impact of country fundamental risk on SOA to target book leverage, and Panel B reports the results of Equation (7), which measures the impact of country fundamental risk on SOA to target market leverage. It also reports the impact of individual dimensions of countries’ fundamental risk, such as the impact of OE, FA, CT, PS, and MC on SOA to target book and market leverage. The sample represents 17,747 companies from 44 countries (Table 2) from 2000 to 2015. Numbers in parentheses are t-statistics. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

presents our estimation results.

Consistent with our main results, in Panels (A) and (B) of Table 8, we observe significant positive coefficients for Frisk * BDEV (Coef. = 0.1636, t = 37.06) and Frisk * MDEV (Coef. = 0.1093, t = 24.33) interaction terms. In terms of its economic significance, it means a one standard deviation increase in frisk increases the SOA of book leverage by 16.03% (=0.1636 ×.50/0.51) (using Equation (8)), and a one standard deviation increase in frisk increases the SOA of market leverage by 8.41% (=0.1093 * 0.50/0.62) (using Equation (8)). Furthermore, unlike our main results, in Panel (B), we observe a positive and significant effect on market capacity (MC). In line with the dynamic trade-off theory of capital structure, our results further support the view that a better country-specific institutional framework and lower fundamental risk can influence leverage adjustment speed favourably (Belkhir et al., 2016; Öztekin & Flannery, 2012). Furthermore, we conduct additional robustness tests by excluding countries with more significant firm-level observations (i.e., China, India, Japan, South Korea, USA) to confirm that our results are not biased due to a larger sample of selected countries. For brevity, we have not shown the results. Our results remain consistent with that sample, too.

3.4. Robustness to Unobserved Macroeconomic Forces

A central concern in our analysis is the potential for our findings to be confounded by unobserved macroeconomic forces. The period under study, which encompasses the global financial crisis (GFC) and the Eurozone sovereign-debt crisis, presents a significant identification challenge. These crises gave rise to several channels that could simultaneously affect both a firm’s capital structure adjustment speed and a country’s risk proxies, thus introducing endogeneity.

As highlighted in the literature, these macroeconomic shocks elicited significant fiscal and monetary policy responses, affecting bank credit and corporate deleveraging (Dantas et al., 2023; Silva & Volkova, 2018). The elevated policy uncertainty that followed the crises also influenced corporate real investment decisions and, by extension, the risk and composition of firms’ capital structures (Campello et al., 2021). Furthermore, the period saw a rise in geopolitical uncertainty and populism (Gyöngyösi & Verner, 2022), which has been shown to affect the real economy and ultimately spill over into firms’ financing decisions (Funke et al., 2023).

To address these concerns, we perform a robustness test using the bounding analysis developed by (Oster, 2019) and used by (Dantas et al., 2023). This approach allows us to formally assess the degree to which unobserved confounders would affect our results. We compare our baseline regression with an augmented model that explicitly controls for the macroeconomic channels discussed above. Specifically, the augmented model includes a dummy variable for the GFC (2008–2009) and another for the Eurozone sovereign-debt crisis (2010–2012). We also include proxies for broader uncertainty (Silva & Volkova, 2018), using the economic policy uncertainty (EPU) index (S. R. Baker et al., 2016; Davis, 2016) and the geopolitical risk (GPR) index (Caldara & Iacoviello, 2022).

The results in

Table 9. Bounding Analysis of the Fundamental Risk Coefficient.

	Base Model		Model with all Controlled		R2 Max		Bounded Value
Outcome	$\dot{\beta }$	$\dot{R}$2	$\tilde{\beta }$	$\tilde{R}$2	$\mathsf{\Pi }$= 1.3	$\mathsf{\Pi }$= 2	${\beta }_{\Pi =1.3}^{\ast }$	${\beta }_{\Pi =2}^{\ast }$
Frisk coefficient	0.0711	0.2235	0.0650	0.4200	0.546	0.8400	0.0611	0.0520

show that our coefficient of interest remains consistent across these specifications. The bounding analysis confirms that, for our key finding to be invalidated, the unobserved variables would need to have a selection effect significantly stronger than the comprehensive set of macroeconomic controls we have included. This provides strong evidence for our main conclusion that country fundamental risk is a significant determinant of a firm’s capital structure adjustment speed, not a spurious result driven by omitted macroeconomic factors.

Table 9 reports the bounding values for the fundamental risk coefficient estimate (coefficient of Frisk * BDEV) following the methodology proposed by Oster (2019). Following Oster (2019), we assume that selection on unobservables is proportional to selection on observables. The bounding value of the estimate (β*) is defined as ${\mathsf{\beta }}^{\ast }$= $\tilde{\mathsf{\beta }}$ − ${\displaystyle \frac{\left(\dot{\mathsf{\beta }}-\tilde{\mathsf{\beta }}\right)\left(\mathrm{R}2 \mathrm{Max} -\tilde{\mathrm{R}}2\right) }{(\tilde{\mathrm{R}}2-\dot{\mathrm{R}}2) }}$, where $\dot{\mathsf{\beta }}$ and $\dot{\mathrm{R}}$² is, respectively, the point estimate and R-squared for the base model (Table 3), and $\dot{\mathsf{\beta }}$ and $\tilde{\mathrm{R}}$² are the analogue values from the regression with all controls (GFC dummy, Eurozone crisis dummy, EPU index, and GPU index). The method assumes that the degree of proportionality between selection on unobservables and selection on observables is one (δ = 1), which entails assuming the maximum possible R² of the regression. We follow the calibration proposed by (Oster, 2019), which sets R² max = min (1, $\mathsf{\Pi }$ × R²) and $\mathsf{\Pi }$ = 1.3 as our benchmark. As in (Gyöngyösi & Verner, 2022), we also consider the conservative value of $\mathsf{\Pi }$ = 2.0 for robustness purposes.

4. Discussion

This study provides the first empirical evidence on how countries’ fundamental risk profiles influence the speed of adjustment (SOA) towards firms’ target capital structures. Consistent with dynamic trade-off theory, this study finds that firms operating in countries with lower fundamental risk adjust their leverage more rapidly. A one standard deviation improvement in the composite frisk index leads to a 12.8% faster SOA for book leverage and a 4.8% faster SOA for market leverage. Analyses of the individual risk dimensions reveal that enhanced operational efficiency, greater foreign accessibility, improved corporate transparency, and higher political stability contribute significantly to accelerating the pace of leverage rebalancing.

Our findings expand previous research on country-level determinants of capital structure dynamics by incorporating a comprehensive measure of fundamental risk. While (Öztekin & Flannery, 2012) and (Çam & Özer, 2021) demonstrate that legal systems, financial market development, macroeconomic uncertainty and institutional quality affect leverage adjustment and capital structure, respectively, our results underscore the integrative impact of multifaceted institutional attributes captured by the Frisk index. Specifically, the strong positive association between operational efficiency and SOA corroborates (Belkhir et al., 2016), who show that trading-system efficiencies lower transaction costs and facilitate financing. Likewise, our evidence on the roles of foreign accessibility and corporate transparency resonates with (An et al., 2021) and (Goodell & Goyal, 2018), highlighting the importance of cross-border capital flows and information disclosure in reducing agency conflicts.

The magnitudes of the coefficients suggest economically meaningful effects: firms in lower-risk countries can fine-tune their capital structures more nimbly, potentially lowering the costs of external financing and optimising debt–equity mixes in response to investment opportunities. For policymakers, our study emphasises the urgency of institutional reforms that enhance market capacity, streamline trading infrastructures, and bolster governance standards. Regulators can directly influence corporate financial behaviour and economic growth by mitigating operational friction and improving investor protection.

The robustness tests, incorporating random forest estimation (RFE) to predict unobservable leverage targets, confirm the baseline results, suggesting that non-linearities in the target estimation stage do not attenuate the fundamental risk effects. This methodological extension advances the literature by showing that machine learning-based target estimation can coexist with traditional partial-adjustment frameworks, yielding consistent policy-relevant insights.

Based on our empirical results, we find that, among the five dimensions of country fundamental risk, operational efficiency (OE), corporate transparency (CT), and political stability (PS) are the most influential in determining the speed of adjustment (SOA) towards target capital structure. These dimensions exhibit strong statistical and economic significance. They indicate that firms operating in environments with efficient trading systems, transparent governance, and stable political institutions face lower capital adjustment costs and can converge more rapidly on their desired capital structures. In contrast, foreign accessibility (FA) shows a notable economic impact, albeit lacking statistical significance, suggesting that capital openness still matters but may be mediated by other institutional factors. However, market capacity (MC) does not exhibit a statistically or economically significant effect on SOA in our context.

These findings have important policy implications. Policymakers aiming to support firms in achieving optimal capital structures should prioritise improvements in trading infrastructure, corporate governance practices, and institutional stability. These reforms can directly reduce friction and transaction costs in capital markets, thereby enhancing firms’ financial flexibility and speed of capital structure adjustment.

While extending the existing literature in the context of the impact of institutional environments on corporate finance decisions, our findings also support the trade-off theory of capital structure. The findings indicate that international differences in country-specific fundamental risk affect the cost of adjustment and, therefore, the speed with which corporations move towards their target capital structure in a partial-adjustment process.

5. Conclusions, Limitations, and Directions for Future Research

By demonstrating that lower fundamental risk environments expedite firms’ convergence on their optimal leverage ratios, our study contributes to a nuanced understanding of how institutional quality shapes corporate financing behaviour. Strengthening the building blocks of capital markets, i.e., transactional infrastructure, regulatory transparency, and political stability, can thus serve as levers for fostering more resilient and efficient corporate finance systems worldwide.

Although our analysis covers a large cross-country sample over 15 years, several limitations warrant caution. First, our reliance on the (Karolyi, 2015) fundamental risk data constrains the sample to 2000–2015 and may not capture more recent institutional shifts. Second, while the frisk index aggregates diverse dimensions, the potential for measurement errors in each component could bias estimates; future research could explore instrumental variables or high-frequency proxies. Third, we focus on non-financial firms; extending the framework to financial institutions or hybrid entities may reveal different dynamics. Fourth, we recognise that our study centers on the impact of fundamental risk on the speed of adjustment (SOA) towards target capital structure. It does not explicitly estimate how these risk dimensions shape the target capital structure. While prior research, such as (Çam & Özer, 2021), has already provided valuable insights into how institutional quality affects leverage and debt maturity, future studies could expand our findings by jointly examining how fundamental risk dimensions influence target levels and the dynamic path towards them. Finally, subsequent studies could investigate endogenous interactions between firm-level governance mechanisms and country risk to unpack micro–macro linkages more deeply.