Foreign Exchange Governance and Financial Stability of Multinationals: Cross-Country Evidence

Abstract

This study examines the association between foreign exchange (FX) governance and financial stability by analysing empirical evidence from multinational entities. We analyse a 16-year panel (2009–2024) comprising 6613 firm-year observations using OLS regression with industry and year fixed effects. Firm-level data on financial sustainability, FX governance, board attributes, and controls are drawn from the London Stock Exchange Group (formerly Refinitiv), while country-level institutional and economic indicators are obtained from the World Bank. The result suggests that FX governance is negatively associated with earnings volatility, implying that FX governance enhances the financial stability of organisations. The baseline result is robustness to endogeneity and selection bias. However, our subsample analysis reveals that the impact of FX governance on financial stability varies based on institutional quality and industry. Whereas FX governance is negatively associated with earnings volatility thus enhancing financial stability in high-institutional-quality settings, the impact is not significant in low-institutional-quality environments. This study contributes to knowledge by empirically validating the relevance of FX governance to financial stability. Our study also contributes to the limited studies on the role of FX governance in diminishing earnings volatility, thus exposing FX management as a strategy for achieving financial sustainability. The international sample analysed in the study contributes to the generalisability of results.

1. Introduction

The 21st-century global economy is characterised by deep financial interconnectedness and the continuous cross-border flow of capital (Lane & Milesi-Ferretti, 2018). For multinational corporations (MNCs), this highly integrated environment creates both unprecedented opportunities and significant vulnerabilities. While global operations provide access to new markets, international financing, and geographically diversified supply chains, they simultaneously expose firms to a complex web of systemic risks. Among these, foreign exchange (FX) risk remains one of the most pervasive and consequential challenges confronting internationally active firms (Bartram & Bodnar, 2012; Hu & Li, 2025). Exchange rate volatility can rapidly erode profit margins, disrupt projected cash flows, and diminish the value of foreign-denominated assets and liabilities (Shapiro, 2014). In this context, FX risk management extends beyond a routine treasury function and becomes a strategic imperative central to safeguarding corporate financial stability and long-term firm value.

Financial stability at the corporate level signifies an organisation’s holistic resilience—its ability to withstand external shocks, maintain uninterrupted operations, meet financial obligations across jurisdictions, and secure long-term financing without debilitating distress (Adams & Jayasekara, 2024). The pillars of this stability are robust liquidity management, a sustainable capital structure, predictable earnings, and a proactive, enterprise-wide risk management culture. The inherent volatility of currency markets directly assaults each pillar. A sudden depreciation in an operating country’s currency can balloon the local-currency cost of servicing foreign-denominated debt, threatening solvency (Kedia & Mozumdar, 2003). Unpredictable exchange rate movements can make accurate cash flow forecasting impossible, complicating liquidity management (Graham & Harvey, 2001). Translating overseas earnings back to the reporting currency can introduce severe volatility into consolidated income statements, undermining earnings quality and investor confidence (Doidge et al., 2004). Therefore, for any MNC, achieving financial stability is intrinsically linked to mastering the challenge of FX risk.

Macroeconomic evidence increasingly demonstrates the destabilising effects of exchange rate volatility. A recent meta-analysis synthesising decades of empirical research shows that currency volatility exerts a statistically significant negative effect on national economic growth (Sheykhi et al., 2025). This suggests that exchange rate instability weakens economic performance at the aggregate level, creating a more uncertain operating environment for firms. For MNCs, whose operations and revenues span multiple jurisdictions, such volatility translates directly into heightened operational and financial uncertainty. In response, monetary authorities often intervene in foreign exchange markets through sterilised purchases or sales of reserves to moderate excessive currency fluctuations (Agénor et al., 2026). Yet emerging evidence indicates that while sterilised interventions may stabilise nominal exchange rates, they can also intensify domestic credit cycles and generate trade-offs between currency stability and broader financial fragility (Agénor et al., 2026). This highlights a critical reality for corporate managers: firms cannot depend solely on state-led interventions to insulate them from external financial instability (Borio & Disyatat, 2015). Consequently, the responsibility for managing FX risk rests fundamentally with the firm itself and is institutionalised through what this study conceptualises as foreign exchange governance.

Foreign exchange governance is the comprehensive, internal framework through which a company systematises its approach to currency exposure (Janczewski et al., 2024). It encompasses board-approved policies that define risk tolerance, committee structures that provide strategic oversight, hedging programmes using derivatives and natural offsets, rigorous monitoring and reporting mechanisms, and internal controls that guarantee accountability and integrity (COSO, 2017; Fatemi & Luft, 2002). In essence, it represents the codification of a firm’s risk culture regarding currency movements and serves as the primary organisational defence against FX-induced instability.

Despite the strategic importance of FX governance to multinational financial stability, a significant gap persists in the literature. While prior research has extensively examined the sources, transmission mechanisms, and consequences of foreign exchange risk, limited attention has been given to whether the quality of a firm’s internal FX governance influences its financial stability. Existing studies offer valuable but fragmented insights into the broader FX environment confronting MNCs. For example, Barry Eichengreen argues that the relationship between exchange rate regimes and financial crises depends largely on the nature of underlying economic shocks, highlighting the complexity of the external environment in which firms operate. Similarly, Kim and Pyun (2018) show that crisis-related business cycle shocks are more strongly transmitted in economies characterised by fixed exchange rates and high capital account openness, consistent with the policy trilemma framework (Obstfeld et al., 2005). At the macroeconomic level, El-Khodary et al. (2025) demonstrate that exchange rate volatility, particularly fluctuations in the U.S. dollar, undermines national competitiveness and increases economic fragility. At the firm level, Toai (2025) finds that exchange rate volatility reduces both the volume and diversity of exports, confirming its direct operational consequences for internationally active firms.

Collectively, these studies establish that FX volatility materially affects economic and corporate outcomes. However, they largely overlook the firm’s primary internal response mechanism: FX governance. In particular, the extent to which robust governance structures mitigate the destabilising effects of currency shocks remains underexplored. This omission is especially important in cross-country research, where institutional heterogeneity in exchange rate regimes, financial openness, and regulatory systems may shape both firms’ exposure to FX risk and the effectiveness of governance responses. Moreover, much of the existing firm-level evidence is confined to single-country settings and cross-sectional analyses, limiting understanding of how FX governance functions across different institutional contexts and over time.

This study addresses this important gap in the international business and corporate finance literature by empirically examining the effect of foreign exchange (FX) governance on the financial stability of multinational corporations using cross-country evidence. Specifically, it investigates whether stronger and more sophisticated FX governance frameworks mitigate financial instability after controlling for firm-specific factors and institutional differences across countries. The study is grounded in two complementary theoretical perspectives. Agency Theory (Jensen & Meckling, 1976) views FX governance as a monitoring and control mechanism that aligns managerial decisions with shareholder interests by constraining excessive risk-taking. In contrast, the Resource-Based View (RBV) (Barney, 1991) conceptualises effective FX governance as a strategic capability that is valuable, rare, inimitable, and non-substitutable (VRIN), enabling firms to manage currency volatility more effectively and sustain financial stability.

We analyse panel data covering a 16-year period (2009–2024) and 6613 firm-year observations using OLS regression with industry and year fixed effects. The result suggests that FX governance is negatively associated with earnings volatility, implying that FX governance enhances the financial stability of organisations. The baseline result is robustness to endogeneity and selection bias. However, our subsample analysis reveals that the impact of FX governance on financial stability varies based on institutional quality and industry. Whereas FX governance is negatively associated with earnings volatility thus enhancing financial stability in high-institutional-quality settings, the impact is not significant in low-institutional-quality environments. This study contributes to knowledge by empirically validating the relevance of FX governance to financial stability. Our study also contributes to the limited studies on the role of FX governance in diminishing earnings volatility, thus exposing FX management as a strategy for achieving financial sustainability. The international sample analysed in the study contributes to the generalisability of results.

This study contributes to knowledge in three principal ways. First, it provides the first large-sample, cross-country empirical validation of the link between formal FX governance mechanisms and firm-level financial stability, filling a significant gap in the corporate governance and international finance literature. Second, by analysing an international sample of MNCs over a 16-year period, the study enhances the generalisability of results beyond single-country or single-region studies. Third, the longitudinal design captures how FX governance functions across different phases of global financial and currency cycles, offering insights into the dynamic nature of risk management.

The remainder of the paper is structured into four sections (Section 2, Section 3, Section 4 and Section 5). Section 2 reviews the literature and develops the hypotheses. Section 3 outlines the research methodology. Section 4 presents and discusses the empirical findings. Section 5 concludes the paper with key implications and final remarks.

2. Literature Review

2.1. Conceptual Foundations: Governance and Stability

2.1.1. The Architecture of Foreign Exchange Governance

Foreign exchange governance constitutes the formal organisational infrastructure for managing currency risk. It is the integrated system of rules, processes, roles, and accountability that transforms reactive treasury operations into a disciplined, strategic, and transparent enterprise-wide practice (Bodnar et al., 1998). This framework is multi-layered and interdependent. Its bedrock is a formally articulated FX Risk Management Policy, ratified by the board of directors. This document serves as the constitutional charter for currency risk, defining the firm’s risk appetite, specifying permissible hedging instruments, establishing position limits and counterparty credit guidelines, and ensuring alignment with relevant accounting standards (Bodnar et al., 1998). A clear, board-mandated policy is essential to provide management with a legitimate mandate and clear boundaries, preventing ad hoc and potentially value-destructive decisions (Allayannis et al., 2012).

Strategic oversight and committee structures provide the necessary governance layer. This often involves a dedicated board sub-committee or a senior-management treasury committee (Kirkpatrick, 2009). This body is responsible for periodically reviewing and updating the policy, monitoring compliance with established limits, assessing the effectiveness and cost-efficiency of hedging programmes, and approving significant deviations or new strategic FX initiatives. Its existence ensures continuous high-level engagement and enforces a critical separation of duties between the execution of hedges and their supervision (Walker, 2009).

The hedging strategies and execution represent the tactical implementation under this governance umbrella. Strategies range from financial hedges using derivatives to operational or “natural” hedges, such as matching revenue and cost currencies or netting intercompany payables and receivables (Shapiro, 2014). The choice and mix of these tools are directed by the overarching governance policy. Robust monitoring, reporting, and internal controls form the essential feedback and assurance loop. Regular, transparent reports to the board and senior management are crucial for informed oversight (COSO, 2017). Furthermore, strong internal controls, including segregation of duties and independent internal audit, safeguard the integrity of the entire process against fraud and operational error (Spira & Page, 2003). For a multinational, this governance framework must be coherent globally yet flexible enough to accommodate regional nuances.

2.1.2. Financial Stability as the Corporate Imperative

At the firm level, financial stability is synonymous with resilience and endurance. It describes an organisation’s capacity to absorb internal and external financial shocks without experiencing liquidity crises, solvency threats, a collapse in investor confidence, or a forced disruption of its strategic operations (Gyulasaryan et al., 2025). This stability is multidimensional, resting on several interconnected pillars. Liquidity ensures the firm can meet its short-term financial obligations as they fall due (Brunnermeier & Pedersen, 2009). Solvency reflects the long-term health of the balance sheet, where the value of assets sustainably exceeds liabilities (Altman, 1968). Earnings and Cash Flow Stability refer to the predictability and quality of profits and operational cash generation, which directly influence credit ratings and cost of capital (Dechow et al., 1998). Finally, integrated risk management is the proactive, strategic function that identifies, assesses, and mitigates threats to the other three pillars (Lam, 2014).

For an MNC, financial stability is the indispensable foundation for global ambition. It enables the firm to honour cross-border contracts, secure financing in international capital markets at competitive rates, and make long-term investments in foreign subsidiaries (Rugman, 2005). FX risk directly and powerfully threatens each pillar. A sharp currency devaluation can instantly multiply the local-currency burden of foreign-currency debt, jeopardising solvency (Kedia & Mozumdar, 2003). Volatility complicates cash flow forecasting across currencies, undermining liquidity planning. Therefore, the core mandate of an effective FX governance framework is to serve as the dedicated guardian of the firm’s financial stability in the face of inherent currency market uncertainty.

2.2. Theoretical Underpinnings: Dual Lenses on Governance

The relationship between FX governance and financial stability can be powerfully elucidated through two established theoretical frameworks, each offering a distinct but complementary rationale.

2.2.1. Agency Theory: Aligning Interests and Controlling Risk

Agency Theory, formalised by Jensen and Meckling (1976), explores the conflicts and costs that arise when decision-making authority is separated from residual claimancy. This separation can lead to managers pursuing actions that benefit themselves at the expense of shareholders, due to diverging interests related to risk tolerance, compensation structures, and job security. In the domain of FX risk management, several agency problems are plausible. Managers might engage in speculative currency trading beyond the firm’s core needs, hoping to generate trading profits that boost short-term earnings and, consequently, their performance-linked bonuses (Tufano, 1996). Conversely, due to risk aversion or a desire to avoid complexity, managers might choose to “do nothing,” leaving the firm fully exposed to market movements (Stulz, 1984). Managers might also “follow the herd” in selecting hedging strategies, rather than tailoring them to the firm’s specific exposure profile (Froot et al., 1993).

FX governance is the primary institutional solution prescribed by Agency Theory to mitigate these conflicts. The formal, board-approved policy acts as a binding contract that restricts managerial discretion to a shareholder-approved corridor of risk. The oversight committee serves as a dedicated monitoring body, reducing information asymmetry. Transparent reporting mandates accountability for risk management outcomes. In this view, strong FX governance functions as a control mechanism that aligns managerial actions with the shareholder interest of long-term value preservation and stability.

2.2.2. Resource-Based View: Governance as a Strategic Capability

While Agency Theory frames governance as a necessary control to prevent value destruction, the Resource-Based View (RBV) reframes it as a potential source of value creation and competitive advantage. Barney (1991) argued that for a firm’s resource or capability to be a source of sustained superior performance, it must be valuable, rare, imperfectly imitable, and non-substitutable (VRIN).

A sophisticated, well-embedded FX governance system can plausibly meet these VRIN criteria. It is valuable because it enables the firm to operate with confidence in volatile global markets. It is rare because many firms possess only rudimentary formal FX governance structures. It is imperfectly imitable because it constitutes a complex social structure, embedded in a firm’s unique culture and tacit knowledge. It is non-substitutable because while individual financial instruments are commodities, the integrated decision-making framework that governs their use cannot be easily replaced (Peteraf, 1993).

From the RBV perspective, FX governance is a dynamic capability (Teece et al., 1997) that allows the firm to sense and interpret currency-related threats and opportunities, seize them through appropriate strategies, and reconfigure resources to maintain advantage. This capability enhances the firm’s resilience, allowing it to withstand shocks that destabilise less-prepared rivals. Financial stability, in this light, is the observable outcome of possessing this superior, VRIN capability. The two theories together provide a robust foundation: Agency Theory explains why governance is necessary, while RBV explains how superior governance can become a positive driver of value creation and strategic resilience.

2.3. The Empirical Landscape: Mapping the Channels of FX Risk to the Firm

A substantial body of empirical research, while not directly testing the governance-stability link, meticulously documents the pathways through which FX volatility threatens firms, thereby underscoring the critical need for the governance response, which is the focus of this study.

2.3.1. Macroeconomic Foundations and Historical Contingency

The starting point is the established macro-level correlation. The meta-analysis by Sheykhi et al. (2025) provides broad evidence confirming that exchange rate volatility is, on average, detrimental to economic growth. However, the seminal work of Eichengreen (1998) introduces crucial nuance by demonstrating that the relationship between exchange rate regimes and financial stability is historically contingent, not deterministic. His analysis reveals that fixed exchange rates can be beneficial if they credibly discipline inflationary domestic policies but can be disastrous if they encourage excessive unhedged foreign-currency borrowing. The key insight is that the source of the economic disturbance matters greatly (Frieden, 2015). For MNCs, this contingency means the FX risk profile of a subsidiary is fundamentally shaped by its host country’s context.

2.3.2. The Institutional Channel: The Policy Trilemma in Action

The transmission of global shocks is critically mediated by national institutional configurations. Kim and Pyun’s (2018) research offers a powerful empirical operationalization of the “impossible trinity” or policy trilemma (Mundell, 1963; Obstfeld et al., 2005). They find that during the Global Financial Crisis, business cycle synchronisation was strongest for countries with fixed exchange rates and highly open capital accounts. For an MNC subsidiary operating in such a “fixed and open” country, this implies a correlated double shock: its local economy is tightly yoked to a core economy’s downturn, and its local currency offers no nominal exchange rate buffer. This finding underscores that the institutional setting is an active modulator of corporate risk (Aizenman et al., 2010).

2.3.3. The Competitiveness and Direct Operational Channels

Beyond direct financial transmission, FX volatility erodes the foundational business environment. Research focusing on the Middle East and North Africa by El-Khodary et al. (2025) demonstrates that U.S. dollar volatility and domestic inflation significantly degrade a nation’s overall competitiveness. A less competitive nation suffers from weaker institutions and lower productivity growth (Porter, 1990). For an MNC operating there, this translates into higher and less predictable operational costs, reduced market potential, and greater political and regulatory uncertainty.

At the micro-operational level, firm-specific evidence confirms the immediate disruptive impact. Toai (2025) gravity model analysis of Vietnam’s green exports finds that real exchange rate volatility directly and significantly harms export performance, reducing both the extensive and intensive margins. This is a clear microeconomic manifestation of the macro growth finding: volatility creates uncertainty that deters market entry and expansion (Aizenman, 1992).

2.4. The Critical Research Gap and Hypothesis Development

This review reveals a coherent story about the genesis and propagation of FX risk but a glaring omission in its culmination. The literature meticulously details the external threat environment but pays remarkably scant attention to the design and strength of the internal defence systems—the corporate governance structures—that determine firm-level resilience. While a related stream examines corporate hedging, it typically treats the act of hedging as the independent variable, not the overarching governance framework that mandates and guides it (Bartram et al., 2009). Furthermore, the cross-country institutional dimension is almost entirely absent from firm-level governance studies. This study directly addresses this nexus between internal governance and external institutional context.

Bringing together theory and evidence, this study’s core argument is straightforward. Agency Theory suggests strong governance reduces managerial risk-taking, while the Resource-Based View (RBV) argues it builds organisational resilience. Empirically, foreign exchange (FX) volatility is harmful, but its impact depends on the surrounding institutional environment. Taken together, this implies that formal FX governance mechanisms—by enforcing discipline, increasing transparency, and embedding risk management into organisational routines—should reduce earnings volatility and improve overall financial stability. This leads to the formulation of the following primary hypothesis:

H1. 

Effective foreign exchange governance mitigates earnings volatility, thereby enhancing the financial stability of multinational organisations.

Building on the institutional contingency literature (e.g., Kim & Pyun, 2018; El-Khodary et al., 2025), we argue that the effectiveness of FX governance varies across institutional contexts. In high-quality environments—marked by strong legal enforcement, effective regulation, low corruption, and robust investor protection—FX governance mechanisms are more credible, better enforced, and more consistently implemented. By contrast, in weaker institutional settings, even well-designed internal governance can be undermined by poor contract enforcement, limited transparency, and weak external oversight. As a result, the marginal benefit of FX governance for financial stability is likely to be greater in stronger institutional environments. This leads to the secondary, moderating hypothesis:

H2. 

Foreign exchange governance more effectively mitigates financial instability in countries with high institutional quality compared to those with low institutional quality.

Testing these hypotheses requires a cross-country research design that controls standard firm-specific characteristics and, critically, incorporates the key country-level institutional variables identified by the literature as moderators of risk transmission.

3. Methodology

3.1. Sample and Data Sources

Our study analyses global companies using the Forbes list as the sampling frame. In line with prior studies (Giannarakis, 2014; Ngu & Amran, 2019), we focus on the top 500 companies from the Forbes Global 2000 as they represent economically powerful and globally visible companies, in terms of high assets, profits, revenue and market values. Forbes Global 2000 is widely used as a proxy for the global MNC population for several reasons. It applies a multidimensional ranking—revenues, profits, assets, and market value—capturing both scale and financial strength across industries. The list also offers broad geographic coverage, including firms from developed and emerging economies, which helps reflect variation in institutional environments. Firms included are typically large and internationally active, with cross-border operations and exposure to foreign exchange risk—key features of multinational corporations. In addition, the dataset benefits from standardised and comparable financial information, as listed firms are subject to disclosure requirements, improving data reliability for empirical research. Although it excludes smaller and privately held multinationals, Forbes Global 2000 captures a substantial share of global economic activity and cross-border flows. As such, it provides a practical and defensible representation of large, globally engaged firms.

The study employs an unbalanced panel data structure, reflecting the reality of incomplete firm-year coverage over time. The initial sampling frame comprises 340 non-financial firms and 160 financial firms. Following the exclusion of 4 non-financial and 15 financial firms due to missing data on FX governance and key firm attributes, the final sample consists of 481 firms (336 non-financial and 145 financial service firms). Importantly, not all firms are observed for every year in the 2009–2024 period, resulting in variation in time-series coverage across firms and thereby generating an unbalanced panel. The dataset spans 16 years (2009–2024) and is constructed from multiple sources. Firm-level data on financial sustainability, FX governance, board attributes, and other characteristics are obtained from the London Stock Exchange Group (LSEG) database (formerly Refinitiv), while country-level indicators of institutional quality and economic performance are sourced from the World Bank Group. After removing incomplete observations, the final dataset comprises 6613 firm-year observations used for empirical analysis. The unbalanced nature of the panel is retained to preserve sample size and avoid introducing bias through listwise deletion of partially observed firms.

3.2. Measurement of Variables

3.2.1. Dependent Variable

The main dependent variable of the study is financial stability (FSTAB), and this was measured using volatility of ROA as a measure of earnings stability (Elsayih et al., 2021). The variable has a negative polarity, implying that lower volatility implies financial stability. Earnings stability is captured using the volatility of return on assets (ROA), a measure that reflects how consistently a firm generates profits from its asset base over time. ROA is particularly suitable in a cross-country setting because it is less affected by differences in capital structure than equity-based measures, making it more comparable across firms and institutional environments. Volatility in ROA is typically calculated as the standard deviation of ROA over a rolling window (e.g., three to five years). This approach captures fluctuations in operating performance while smoothing out short-term noise. Higher volatility indicates greater unpredictability in earnings, which is commonly associated with operational risk, exposure to external shocks (such as exchange rate movements), and weaker risk management. Conversely, lower volatility reflects more stable and predictable performance, suggesting stronger internal controls, effective hedging practices, and greater organisational resilience. Importantly, this measure has negative polarity: higher values correspond to lower financial stability, while lower values indicate greater stability. Interpreting the variable in this way aligns with the broader literature on risk and performance variability, where consistency—rather than just the level of returns—is a key indicator of financial health. As such, a reduction in ROA volatility can be understood as an improvement in earnings stability and overall financial stability.

To check robustness of results to alternative measurement of the dependent variable, we use debt/equity ratio as an alternative measurement of the dependent variable (FSTAB_2). The debt-to-equity ratio is a widely used proxy for financial stability because it captures the extent to which a firm relies on debt financing relative to shareholders’ equity. This balance is central to a firm’s risk profile (Vintilă et al., 2025). Higher leverage (a higher debt-to-equity ratio) implies greater fixed obligations in the form of interest and principal repayments, which can strain cash flows—especially during periods of earnings volatility or external shocks such as exchange rate fluctuations. As a result, highly leveraged firms are generally more vulnerable to financial distress and refinancing risk. Conversely, a lower debt-to-equity ratio indicates a more conservative capital structure, with a larger equity buffer to absorb losses. This reduces the likelihood of distress and enhances the firm’s capacity to withstand adverse conditions, thereby supporting financial stability. In this sense, the ratio reflects not just financing choices but also resilience to uncertainty. This measure is particularly useful in cross-country research because it is simple, widely reported, and comparable across firms and institutional contexts. While optimal leverage levels may vary by industry, the overall interpretation is consistent: higher debt relative to equity increases financial fragility, whereas lower leverage is associated with greater stability. Accordingly, like ROA volatility, the debt-to-equity ratio can be interpreted with negative polarity—higher values imply lower financial stability, and lower values indicate stronger financial health.

3.2.2. Independent Variable

The independent variable, FX governance (FXGOV), is operationalised as a binary indicator capturing whether a firm has formal, institutionalised oversight of foreign exchange risk. Specifically, FXGOV equals 1 if a firm meets at least one of the following criteria: (a) a board-approved FX risk policy, (b) a dedicated risk or treasury committee, or (c) explicit disclosure of FX hedging oversight; and 0 otherwise. This construction is intended to capture the presence of formalisation, rather than the quality or intensity of governance. While a binary measure may appear reductive, it is theoretically and empirically justified.

First, the transition from informal to formal governance represents a meaningful structural break. The existence of formal oversight mechanisms signals that FX risk management is embedded in organisational processes, subject to monitoring, and anchored in accountability structures—features that are central to both Agency Theory and the Resource-Based View. In this sense, the “extensive margin” (whether governance exists at all) is a necessary precondition for any variation in intensity to matter.

Second, cross-country data constraints limit the consistent measurement of governance quality. Detailed information on the depth, frequency, or effectiveness of FX oversight is rarely disclosed in a standardised manner, particularly across different institutional environments. A binary specification therefore enhances comparability and reduces measurement error, avoiding subjective or noisy proxies for “quality.”

Third, the empirical strategy explicitly acknowledges this limitation by interpreting FXGOV as a baseline institutional commitment to FX risk management. Heterogeneity in effectiveness is not ignored but instead captured indirectly through moderating factors—most notably institutional quality—which conditions how credible and effective these formal structures are in practice.

Taken together, FXGOV should be understood as a parsimonious but conceptually grounded measure of FX governance formalisation. It isolates whether firms have crossed the threshold into structured oversight, while allowing the broader empirical framework to account for differences in how such governance translates into financial stability.

To strengthen robustness, we employ an alternative specification of FX governance using a lagged FXGOV variable in the supplementary analysis. This approach helps address concerns of reverse causality and timing, ensuring that governance structures precede observed financial outcomes rather than respond to them. In addition, we include country fixed effects to control for unobserved, time-invariant heterogeneity across institutional environments. This absorbs differences in legal systems, regulatory quality, and macroeconomic conditions that could otherwise bias the estimated relationship between FX governance and financial stability. Together, these steps enhance the credibility of the results and mitigate concerns related to measurement simplicity and omitted variable bias.

3.2.3. Control Variables

Based on the literature (Ararat et al., 2015; Erin et al., 2021; Amin et al., 2022; Bektur & Arzova, 2022; Oyewo et al., 2026), we include control variables that may affect financial stability such as board financial expertise (BFINX), CEO duality function (DUALFX), board oversight intensity in terms of independent/outside director monitoring (BOVST), board size (BDSZE) and board gender diversity (BDGDV).

As suggested by the literature (Firoozi & Keddie, 2021; Tsang et al., 2023; Aldoseri, 2025), we also control for firm attributes affecting financial stability with respect to size and market presence using Firm Revenue (FREVU) and Firm Capitalisation (FCAPN). Considering the international nature of the study, we control for country-level institutional mechanisms based on the World Governance Indicators (Lewis et al., 2019; Cuadrado-Ballesteros & Bisogno, 2020; Oyewo et al., 2026), notably Voice and Accountability (CONVOC), Political Stability (CONPOL), Government Effectiveness (CONGOV), Regulatory Quality (CONREG), and Rule of Law (CONROL). The Control of Corruption indicator is intentionally excluded to avoid multicollinearity and conceptual overlap with Rule of Law and Government Effectiveness, as these dimensions already capture enforcement quality and the integrity of public institutions most directly relevant to financial contracting and governance outcomes. Finally, we control for economic performance of countries (Harun et al., 2020; Nuber & Velte, 2021; Oyewo et al., 2025) using Economic Development (ECODVP) and Total Economic Output (ECOUTP). Variable measurement is presented in

Table 1. Variables and measurement.

Variable Label	Variable Name	Variable Measurement
FSTAB	Financial stability (main measurement)	Volatility of ROA as measure of earnings stability. Has negative polarity. Lower volatility implies financial stability.
FSTAB_2	Financial stability (alternative measurement)	Debt/equity ratio.
FXGOV	FX governance	FXGOV = 1 if a firm satisfies any of the following: (a) board-approved FX risk policy; (b) dedicated risk/treasury committee; (c) explicit disclosure of FX hedging oversight. FXGOV = 0 if otherwise. This measure captures FX risk management formalisation rather than intensity. To incorporate robustness, we use alternative measure of FXGOV (lagged FXGOV variable) and we control for country fixed effects.
BFINX	Board financial expertise	Ratio of board members with financial risk expertise to board size.
DUALFX	CEO duality function	Takes value of 1 if CEO doubles as board chairperson, otherwise 0.
BOVST	Board oversight intensity	Ratio of outside directors to board size.
BDSZE	Board size	Total number of directors (log of).
BDGDV	Board gender diversity	Ratio of female directors to total board size.
FREVU	Firm Revenue (proxy for firm size)	Revenue (log).
FCAPN	Firm Capitalisation (proxy for market presence)	Market capitalisation (log).
CONVOC	Country Voice and Accountability	Voice and Accountability index based on World Bank data. Measured using indicators that assess citizens’ ability to participate in government selection, freedom of expression, freedom of association, and media independence.
CONPOL	Country Political Stability	Political Stability index based on World Bank data. Measured using indices capturing perceptions of political unrest, violence, terrorism, and likelihood of government destabilisation.
CONGOV	Country Government Effectiveness	Government Effectiveness index based on World Bank data. Measured by indicators assessing quality of public services, policy formulation, implementation, and government credibility
CONREG	Country Regulatory Quality	Regulatory Quality index based on World Bank data. Measured through indices evaluating the ability of the government to create and enforce sound market-friendly regulations.
CONROL	Country Rule of Law	Rule of Law index based on World Bank data. Measured using indicators reflecting confidence in legal systems, contract enforcement, property rights, and control of crime
ECODVP	Economic Development	GDP per capita, PPP (current international $), Natural Log
ECOUTP	Total Economic Output	GDP, PPP (current international $), Natural Log

.

3.3. Model Specification and Estimation Technique

We specify our regression model as follows:

$${FSTAB}_{i,t}={\beta }_0+B_1{FXGOV}_{i,t}+{\beta }_2{CONTROLS}_{i,t}+{\epsilon }_{i,t}$$

(1)

where

FSTAB is financial stability;

FXGOV represents FX governance;

CONTROLS refers to vector of control variables;

${\epsilon }_{i,t}$ captures the error term.

All remaining variables are as described in Table 1.

In deciding on the appropriate estimation technique for our panel data, we use the Hausman specification test to compare the fixed effects (FE) and random effects (RE) estimators. The Hausman test evaluates whether the individual-specific effects are uncorrelated with the explanatory variables, which is a key assumption underlying the random effects model. If this assumption holds, the random effects estimator is both consistent and efficient; however, if it is violated, the random effects estimator becomes inconsistent, whereas the fixed effects estimator remains consistent. The results of the Hausman test indicate a statistically significant difference between the FE and RE coefficient estimates, leading us to reject the null hypothesis that the random effects estimator is appropriate. This suggests that unobserved firm-specific characteristics—such as managerial quality, internal risk culture, governance traditions, or persistent operational strategies—are correlated with the explanatory variables included in the model. Under such circumstances, the fixed effects estimator is preferred because it controls for time-invariant unobserved heterogeneity by allowing these firm-specific effects to be correlated with the regressors. We therefore employed OLS regression with industry and year fixed effects to estimate our model. To check robustness of results, we use entropy balancing, a two-step Heckman procedure, and subsample analysis.

4. Results and Discussion

4.1. Analysis of Descriptive Statistics

Descriptive statistics of variables are as presented in

Table 2. Descriptive statistics.

	Mean	SD	p25	p50	p75
FSTAB	2.261	2.235	0.509	1.668	3.061
FXGOV	0.737	0.440	0.000	1.000	1.000
BFINX	0.462	0.241	0.310	0.470	0.630
DUALFX	0.477	0.500	0.000	0.000	1.000
BOVST	0.773	0.224	0.710	0.850	0.920
BDSZE	12.403	3.829	10.000	12.000	14.000
BDGDV	0.172	0.128	0.080	0.170	0.270
FREVU	4.375	0.480	4.068	4.377	4.706
FCAPN	4.564	0.438	4.283	4.539	4.831
CONVOC	80.571	15.790	76.330	84.060	89.660
CONPOL	66.389	16.848	59.520	66.350	77.620
CONGOV	88.061	10.452	87.980	90.870	92.790
CONREG	86.749	12.022	84.950	89.420	93.200
CONROL	86.619	12.803	85.580	90.380	92.310
ECODVP	4.571	0.263	4.547	4.666	4.725
ECOUTP	12.714	0.574	12.339	12.727	13.225
N	6613	6613	6613	6613	6613

. The mean value of financial stability (FSTAB) is 2.261, with a standard deviation of 2.235, indicating substantial variation in earnings volatility across firms. The interquartile range (p25 = 0.509; p75 = 3.061) suggests meaningful dispersion in financial stability outcomes, which supports the use of firm-level explanatory variables to explain cross-sectional and time-series variation. FX governance (FXGOV) has a mean of 0.737 and a median of 1.000, indicating that approximately 74% of the firms in the sample have adopted some form of FX governance mechanism. The binary nature of FXGOV is reflected in the quartile distribution, with the 25th percentile at zero and both the median and 75th percentile at one, highlighting heterogeneity in adoption across firms. The board-related variables suggest that the sample firms are, on average, characterised by relatively developed governance structures. Board size (BDSZE) averages 12.4 directors, which is consistent with large or internationally active firms. Board gender diversity (BDGDV) has a mean of 17.2%, indicating moderate female representation on boards, with notable variation across firms.

The mean value of board oversight (BOVST) is 0.773, suggesting that a substantial proportion of firms maintain oversight structures consistent with governance best practices. Similarly, board financial expertise (BFINX) averages 0.462, indicating that nearly half of board members possess financial expertise relevant to risk oversight. Variables capturing firm-level exposure and structure show considerable dispersion. Foreign Revenue (FREVU) and Capitalisation (FCAPN) exhibit relatively high means with modest standard deviations, reflecting differences in international exposure and financial strength across firms. The distribution of DUALFX suggests that roughly 48% of firms combine FX governance roles or responsibilities, further highlighting variation in FX governance arrangements.

The institutional quality indicators (CONVOC, CONPOL, CONGOV, CONREG, and CONROL) exhibit high mean values, indicating that a large share of the sample operates in relatively strong institutional environments. However, the standard deviations and interquartile ranges reveal sufficient variation to meaningfully assess how institutional quality moderates the relationship between FX governance and financial stability. The macroeconomic variables, Economic Development (ECODVP) and Total Economic Output (ECOUTP), show limited dispersion, reflecting relative stability in country-level macroeconomic conditions across the sample period.

Overall, the descriptive statistics indicate substantial heterogeneity in financial stability, FX governance adoption, and governance characteristics across firms, providing a suitable empirical setting to examine the relationship between FX governance mechanisms and financial stability. The absence of extreme skewness or outliers further supports the appropriateness of the subsequent multivariate regression analysis.

4.2. Correlation and Multicollinearity

The correlation matrix (

Table 3. Correlation matrix and multicollinearity analysis.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)	VIF
FSTAB (1)	1.000
FXGOV (2)	−0.037 **	1.000															1.24
BFINX (3)	0.053 ***	−0.034 **	1.000														1.09
DUALFX (4)	0.044 ***	−0.024	0.074 ***	1.000													1.23
BOVST (5)	0.047 ***	0.036 **	−0.177 ***	0.057 ***	1.000												1.46
BDSZE (6)	−0.223 ***	0.141 ***	−0.019	0.008	0.109 ***	1.000											1.07
BDGDV (7)	−0.040 **	0.187 ***	−0.069 ***	0.021	0.419 ***	0.101 ***	1.000										1.59
FREVU (8)	−0.032 **	0.304 ***	−0.045 ***	−0.001	0.019	0.148 ***	0.085 ***	1.000									1.56
FCAPN (9)	0.095 ***	0.240 ***	−0.025 *	0.030 *	0.169 ***	0.081 ***	0.219 ***	0.546 ***	1.000								1.57
CONVOC (10)	−0.019	0.160 ***	0.103 ***	0.035 **	0.118 ***	0.093 ***	0.346 ***	0.090 ***	0.027 *	1.000							2.34
CONPOL (11)	−0.098 ***	0.022	0.113 ***	−0.110 ***	−0.192 ***	0.007	−0.037 **	0.016	−0.103 ***	0.550 ***	1.000						3.57
CONGOV (12)	−0.025 *	0.076 ***	0.130 ***	0.080 ***	0.033 **	0.012	0.226 ***	0.047 ***	0.022	0.710 ***	0.732 ***	1.000					3.77
CONREG (13)	−0.006	0.047 ***	0.103 ***	0.036 **	0.127 ***	0.000	0.274 ***	0.062 ***	0.043 ***	0.742 ***	0.689 ***	0.929 ***	1.000				2.08
CONROL (14)	−0.018	0.091 ***	0.117 ***	0.110 ***	0.087 ***	0.021	0.265 ***	0.045 ***	0.033 **	0.780 ***	0.675 ***	0.945 ***	0.922 ***	1.000			3.47
ECODVP (15)	0.089 ***	0.184 ***	−0.007	0.167 ***	0.275 ***	0.005	0.363 ***	0.112 ***	0.173 ***	0.310 ***	0.246 ***	0.548 ***	0.555 ***	0.542 ***	1.000		2.19
ECOUTP (16)	0.073 ***	−0.107 ***	0.087 ***	0.308 ***	0.022	−0.059 ***	0.016	0.050 ***	0.095 ***	−0.062 ***	−0.288 ***	0.013	0.006	0.021	−0.110 ***	1.000	1.49

* p < 0.10, ** p < 0.05, *** p < 0.01.

) reveals that financial stability (FSTAB) is significantly correlated with several governance-, board-, and firm-level characteristics, though the magnitudes of these correlations are generally modest. The correlation between FX governance (FXGOV) and FSTAB is negative and statistically significant, providing preliminary, univariate evidence that FX governance is associated with lower earnings volatility and enhanced financial stability. Importantly, the relatively small magnitude of this correlation suggests that FXGOV captures a distinct governance dimension rather than overlapping mechanically with the dependent variable.

High correlations among CONREG, CONGOV, and CONROL (Table 3) do not support the presence of problematic multicollinearity. First, although the pairwise correlations among these institutional indicators are relatively high (e.g., CONGOV–CONROL = 0.945; CONGOV–CONREG = 0.929; CONREG–CONROL = 0.922), this pattern is conceptually expected. All three variables originate from the World Governance Indicators framework and are designed to capture distinct but related dimensions of institutional quality: Regulatory Quality, Government Effectiveness, and Rule of Law. High correlation therefore reflects shared underlying institutional structure rather than redundancy in an econometric sense. Second, multicollinearity diagnostics based on Variance Inflation Factors (VIFs) do not indicate a severe problem. The VIF values for these variables range from 2.08 to 3.77, which are well below conventional thresholds of concern (typically VIF > 10, or even the more conservative cutoff of 5). This suggests that although the variables are correlated, they do not inflate standard errors to a degree that biases inference or undermines estimator stability. Third, retaining the disaggregated indicators is important for theoretical identification and interpretability. Each variable captures a distinct institutional channel through which governance conditions may affect FX governance effectiveness and financial stability. Aggregating them via PCA would obscure these theoretically meaningful distinctions and reduce the ability to interpret which institutional dimension drives the results. Finally, robustness checks using alternative specifications (including fixed effects and additional controls) further ensure that results are not driven by collinearity among institutional variables. In sum, while the correlations are high, they reflect the inherent structure of governance quality rather than econometric redundancy, and diagnostic evidence confirms that multicollinearity is not a serious concern in the empirical specification.

The VIF values reported in the final column indicate that multicollinearity is not a serious concern in the regression analysis. All VIFs are well below commonly accepted thresholds (e.g., 5 or 10), with the highest VIF values observed for institutional quality variables remaining below 4. This suggests that, despite relatively high pairwise correlations among some institutional measures, their inclusion in the regression models does not unduly inflate standard errors or compromise coefficient estimation. The VIF coefficients are also below the acceptable threshold (VIF < 10) suggesting that multicollinearity is not likely to be a problem as to bias the regression results.

4.3. Baseline Result

The baseline result is presented in

Table 4. Baseline result on the association between FX governance and financial stability.

	(1)	(2)	(3)
	FSTAB	FSTAB	FSTAB_2
FXGOV	−0.048 ***	−0.024 **	−0.033 **
	(−4.13)	(−2.05)	(−2.40)
BFINX	0.079 ***	0.068 ***	−0.020
	(7.20)	(6.27)	(−1.62)
DUALFX	−0.050 ***	−0.053 ***	0.011
	(−4.33)	(−4.66)	(0.86)
BOVST	0.080 ***	0.067 ***	−0.011
	(6.33)	(5.29)	(−0.75)
BDSZE	−0.137 ***	−0.114 ***	0.049 ***
	(−12.66)	(−10.54)	(3.99)
BDGDV	−0.054 ***	−0.075 ***	−0.031 **
	(−3.88)	(−5.46)	(−1.99)
FREVU	-	−0.214 ***	−0.035 **
		(−16.41)	(−2.32)
FCAPN	-	0.171 ***	−0.074 ***
		(13.27)	(−5.03)
CONVOC	0.057 ***	0.084 ***	0.078 ***
	(2.94)	(4.38)	(3.58)
CONPOL	−0.097 ***	−0.072 ***	−0.027
	(−4.82)	(−3.59)	(−1.15)
CONGOV	−0.037	−0.051	−0.256 ***
	(−0.93)	(−1.28)	(−5.57)
CONREG	0.097 ***	0.121 ***	0.084 **
	(2.88)	(3.61)	(2.17)
CONROL	−0.116 ***	−0.157 ***	0.152 ***
	(−3.01)	(−4.12)	(3.44)
ECODVP	0.066 ***	0.068 ***	0.050 ***
	(4.25)	(4.43)	(2.83)
ECOUTP	0.013	0.023 *	0.070 ***
	(0.98)	(1.74)	(4.59)
Industry FE	YES	YES	YES
Year FE	YES	YES	YES
R-squared	0.314	0.344	0.135
N	6613	6613	6613

Standardised beta coefficients; t statistics in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.

. We specify three models as reported in columns 1 to 3. In column 1, we use FSTAB (i.e., volatility of ROA as measure of earnings stability) as our main measurement of variables, and regress it against FX governance and control variables, leaving out Firm Revenue (FREVU) and Firm Capitalisation (FCAPN) as control variables. The result shows that FX governance is negatively associated with earnings volatility, implying that FX governance enhances the financial stability of organisations. This aligns with results in the extant literature (Al-Shboul & Anwar, 2014; Chen et al., 2026). In column 2, we retain all control variables and introduce Firm Revenue (FREVU) and Firm Capitalisation (FCAPN) as additional control variables. The result persists that FX governance has a significant negative impact on earnings volatility, further establishing that FX governance contributes to financial stability of multinational organisations. In column 3, we use alternative measurement of financial stability in terms of debt–equity ratio. We include all control variables and regress debt/equity ratio (FSTAB_2) as a measure of financial stability (Basali, 2025). The result shows that FX governance has a significant negative impact on leverage stability, further confirming that FX governance enhances financial stability, implying that our result is robust to alternative measurement of dependent variable. This is consistent with discussions in the extant literature (Kim & Pyun, 2018; El-Khodary et al., 2025). Taken together, the results in Table 4 in columns 1 to 3 establishes that FX governance is associated with more stable operating performance. The result may be interpreted from the perspective that robust FX governance may reduce earnings volatility and contribute to financial stability by encouraging more conservative risk-taking, improving oversight of financial risks (including FX, commodity, and supply-chain risks), and enhancing coordination between sustainability goals and treasury/risk functions. The result supports the acceptance of H1. The existence of FX governance mechanisms also signals strong overall FX governance quality, commitment to transparent FX policies and accountability, the innovative deployment of FX management skills within the organisation, and greater scrutiny of FX policies and decisions from investors and stakeholders as organisations typically communicate through annual reports. Taken together, these measures ensure a careful analysis of FX strategies and decisions on financial stability and going concern of an organisation. When such an approach is taken to management FX, it may be expected that financial stability should be enhanced.

The Hausman test was conducted to determine the appropriate panel estimator based on the result in Table 4. The null hypothesis that the random effects (RE) estimator is consistent is rejected, indicating a statistically significant difference between fixed effects (FE) and RE coefficient estimates. This conclusion is supported by observable instability in key parameter estimates across specifications. For instance, the coefficient on FXGOV varies from −0.048 (Model 1) to −0.024 (Model 2), while BDGDV changes from −0.054 to −0.075, and CONROL shifts from −0.116 to −0.157. Such systematic differences in both magnitude and, in some cases, significance across governance and institutional variables suggest that unobserved firm-specific effects are correlated with the regressors. Given the large sample size (N = 6613) and the extent of coefficient variation across models, the Hausman statistic is sufficiently large to reject the RE assumption of orthogonality between regressors and unobserved heterogeneity. Accordingly, the results support the use of fixed effects as the preferred estimator, as it provides consistent estimates in the presence of correlation between unobserved firm-specific characteristics and explanatory variables.

4.4. Robustness Check

4.4.1. Addressing Omitted-Variable Bias Using Entropy Balancing

Firms having FX governance mechanisms may differ systematically from others that do not have such arrangements in terms of firm attributes such as size, corporate governance mechanisms, and organisational policies (Oyewo et al., 2025; Wibowo et al., 2025). Firms do not randomly adopt FX governance mechanisms. Instead, adoption is correlated with observable firm characteristics such as size, governance quality, board structure, and organisational policies. If these characteristics also affect financial stability, then the FXGOV coefficient in a simple regression would suffer from omitted-variable bias, even when controls are included. In addition, although we included control variables that may affect financial stability relating to corporate governance, board characteristics and firm attributes, our model may not capture all control variables (Khoza et al., 2024). To check that our baseline result is not biased by differences in firm attributes between organisations with FX governance mechanisms and others that do not, we employ the entropy balancing technique to address any potential bias in this connection.

Entropy balancing directly addresses the concern that firms with FX governance mechanisms differ systematically from those without such arrangements by explicitly reweighting the sample to achieve covariate balance between the two groups. In essence, while we control for firm attributes in the baseline result, entropy balancing ensures equal weights attached to the treatment and control group. Within the context of the study, the treatment group are firms implementing FX governance mechanisms, whilst the control group represents organisations without FX mechanisms. Rather than relying solely on regression controls—which may be insufficient if functional form assumptions are incorrect or if covariates are distributed very differently across groups—entropy balancing constructs weights for the control group firms (those without FX governance mechanisms) so that their observable characteristics match those of the treated group (firms with FX governance mechanisms) in terms of specified moments of the covariate distributions.

In sum, to address unobserved heterogeneity across firms, the study employs entropy balancing—a reweighting approach that enhances covariate comparability between firms with and without FX governance. Specifically, entropy balancing assigns weights so that the distribution of observable firm characteristics (e.g., size, profitability, leverage, industry exposure) is closely matched across treatment groups. This ensures that firms being compared are similar along key dimensions that could jointly influence both the adoption of FX governance and financial outcomes. By achieving balance of these observables, entropy balancing reduces selection bias and limits the influence of confounding factors that may otherwise distort estimates. Importantly, while it does not directly control for unobservable characteristics, it mitigates their impact under the assumption that such unobservables are correlated with observed covariates. As a result, the estimated relationship between FX governance and financial stability is less likely to reflect systematic firm-level differences and more likely to capture the effect of governance itself.

Combined with control variables and fixed effects, this approach strengthens the model’s ability to account for firm-level heterogeneity.

The result of the analysis is presented in

Table 5. Entropy balancing result on the association between FX governance and financial stability.

	Panel A: Before Reweighting		Panel A: Before Reweighting		Regression Result Based on Entropy Balancing
	Treatment (1)	Control (2)	Treatment (3)	Control (4)	(5)
					FSTAB
FXGOV	-	-	-	-	−0.059 ***
					(−3.18)
BFINX	0.460	0.478	0.460	0.460	0.065 ***
					(4.01)
DUALFX	0.473	0.501	0.473	0.473	−0.065 ***
					(−4.24)
BOVST	0.779	0.760	0.779	0.779	0.044 ***
					(3.39)
BDSZE	12.74	11.52	12.74	12.74	−0.114 ***
					(−9.75)
BDGDV	0.187	0.133	0.187	0.187	−0.038 *
					(−1.89)
FREVU	4.481	4.162	4.481	4.481	−0.240 ***
					(−12.10)
FCAPN	4.638	4.404	4.638	4.638	0.237 ***
					(10.92)
CONVOC	-	-	-	-	0.032
					(0.76)
CONPOL	-	-	-	-	−0.037
					(−1.57)
CONGOV	-	-	-	-	−0.072
					(−1.55)
CONREG	-	-	-	-	0.035
					(0.70)
CONROL	-	-	-	-	−0.071
					(−1.06)
ECODVP	-	-	-	-	0.084 ***
					(4.65)
ECOUTP	-	-	-	-	−0.016
					(−0.68)
Industry FE					YES
Year FE					YES
R-squared					0.338
N					6613

Standardized beta coefficients; t statistics in parentheses. * p < 0.10, *** p < 0.01.

.

The result in Table 5 shows that before applying entropy balancing, there is a difference in mean scores of firm attributes between the treatment (column 1) and control (column 2) groups. However, after reweighting the covariates through entropy balancing, there are no difference in the mean score of treatment (column 3) and control (column 4), implying that our entropy balancing procedure achieved equality in firm attributes. This also implies that the difference in treatment and control groups in terms of financial stability can now be largely attributed to FX governance. We rerun the regression after achieving equality in the covariates of treatment and control groups as reported in column 5 using our main measurement of financial stability. The result shows that FX governance has a significant negative impact on earnings volatility, connoting that robust FX governance diminished earnings volatility, whilst contributing to financial stability. This supports H1.

4.4.2. Addressing Selection Bias Using Heckman Two-Stage Selection Model

The Heckman two-stage selection model is employed to address potential endogeneity arising from non-random selection into FX governance (FXGOV) as reported in

Table 6. Heckman two-stage correction on the association between FX governance and financial stability.

	(1)	(2)
	FXGOV	FSTAB
FXGOVt−2	1.420 ***	-
	(33.89)
FXGOV	-	−0.056 ***
		(−3.03)
Inverse Mills ratio	-	0.071 ***
		(4.24)
BFINX	0.019	0.070 ***
	(0.41)	(6.36)
DUALFX	−0.080 *	−0.059 ***
	(−1.72)	(−5.10)
BOVST	−0.049	0.064 ***
	(−1.01)	(5.04)
BDSZE	0.257 ***	−0.107 ***
	(5.58)	(−9.68)
BDGDV	0.435 ***	−0.061 ***
	(8.08)	(−4.26)
FREVU	0.655 ***	−0.193 ***
	(11.75)	(−13.80)
FCAPN	0.201 ***	0.178 ***
	(3.55)	(13.69)
CONVOC	0.026 *	0.089 ***
	(1.70)	(4.63)
CONPOL	0.026 *	−0.075 ***
	(1.69)	(−3.72)
CONGOV	0.023	−0.044
	(1.45)	(−1.09)
CONREG	0.033 **	0.108 ***
	(2.03)	(3.20)
CONROL	0.041 **	−0.158 ***
	(2.47)	(−4.12)
ECODVP	0.036 **	0.071 ***
	(2.10)	(4.62)
ECOUTP	0.036 **	0.020
	(1.97)	(1.49)
Industry FE	YES	YES
Year FE	YES	YES
Pseudo R-squared	0.355	-
R-squared	-	0.342
N	6613	6613

Standardized beta coefficients; t statistics in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.

. Firms’ decisions to adopt FX governance mechanisms are likely endogenous, as unobservable firm characteristics—such as managerial risk preferences, internal risk culture, or financial sophistication—may simultaneously influence both the likelihood of adopting FX governance and firm-level financial stability (FSTAB). Ignoring this selection process could bias the estimated effect of FX governance on financial stability.

In the first stage, we model the probability that a firm adopts FX governance mechanisms as a function of observable firm characteristics and an exclusion restriction. Specifically, we use the two-year lag of FX governance (FXGOV_t−2) as an instrument, which is strongly associated with current FX governance adoption due to persistence in firms’ risk management practices, as evidenced by the positive and statistically significant coefficient in column (1). At the same time, conditional on current controls, FXGOV_t−2 is unlikely to directly affect contemporaneous financial stability, satisfying the relevance and exclusion requirements for identification. The sample size does not change when using FXGOV_t−2 because the lag structure is constructed within the existing unbalanced panel rather than reducing available observations. Each firm’s FX governance variable is lagged by two periods using prior-year data already present in the dataset (2009–2024). Since the panel retains sufficient historical coverage for most firms, no additional observations are lost through listwise deletion, allowing the full 6613 firm-year observations to be preserved in the estimation.

From the first-stage probit estimation, we compute the inverse Mills ratio (IMR), which captures the part of the error term related to the firm’s selection into FX governance. In the second stage, we include this IMR in the financial stability regression. The positive and statistically significant coefficient on the inverse Mills ratio in column (2) indicates the presence of selection bias; that is, unobserved factors influencing the adoption of FX governance are also correlated with financial stability outcomes. This confirms that estimating the effect of FX governance without correcting for selection would lead to biased estimates.

After controlling for this selection effect, the coefficient on FXGOV remains negative and statistically significant, suggesting that FX governance mechanisms exert an independent and economically meaningful impact on firm financial stability, even after accounting for endogenous selection. Thus, the Heckman correction strengthens the causal interpretation of our findings by explicitly modelling firms’ self-selection into FX governance mechanisms.

4.5. Subsample Analysis

To further assess the robustness and generalisability of our findings, we re-estimate the baseline model across subsamples defined by institutional quality and industry classification (financial versus non-financial firms).

4.5.1. Institutional Quality Subsamples

Considering the international sample used in the current study, we conduct subsample analysis based on institutional quality as a further robustness check. Institutional quality shapes firms’ incentives, monitoring intensity, and enforcement of governance practices. If the baseline results are driven by weak institutions or strong institutions alone, the estimated relationship between FX governance and financial stability may reflect contextual factors rather than a general governance effect.

By splitting the sample into high- and low-institutional-quality environments, our subsample analysis tests whether: (a) FX governance mechanisms remain relevant when external enforcement and investor protection are strong, and (b) FX governance matters even when formal institutions are weak, where internal governance mechanisms may substitute for external controls. Using the median score of institutional quality score, we split our sample into high institutional quality (above median score), and low institutional quality (median score and below). The result of the analysis is as presented in

Table 7. FX governance and financial stability in high vs. low institutional quality.

	High Institutional Quality	Low Institutional Quality
	(1)	(2)
	FSTAB	FSTAB
FXGOV	−0.030 **	0.034
	(−2.44)	(0.95)
BFINX	0.056 ***	0.075 **
	(4.90)	(2.43)
DUALFX	−0.063 ***	0.001
	(−5.13)	(0.03)
BOVST	0.059 ***	−0.034
	(4.20)	(−1.04)
BDSZE	−0.128 ***	0.071 **
	(−11.08)	(2.41)
BDGDV	−0.087 ***	0.045
	(−5.99)	(1.29)
FREVU	−0.219 ***	−0.258 ***
	(−16.18)	(−5.81)
FCAPN	0.164 ***	0.241 ***
	(12.00)	(6.38)
CONVOC	0.026	−0.160 ***
	(1.56)	(−2.63)
CONPOL	−0.087 ***	0.315 ***
	(−4.68)	(4.01)
CONGOV	−0.060 **	−0.174 **
	(−2.57)	(−2.48)
CONREG	0.036 *	−0.006
	(1.70)	(−0.07)
CONROL	0.054 *	−0.332 ***
	(1.66)	(−6.65)
ECODVP	0.061 ***	−0.068
	(3.32)	(−0.85)
ECOUTP	0.003	0.069
	(0.17)	(1.40)
Industry FE	YES	YES
Year FE	YES	YES
R-squared	0.345	0.613
Observations	5987	626

Standardized beta coefficients; t statistics in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.

.

The result in Table 7 shows that FX governance is negatively associated with earnings volatility (b = −0.030, p < 0.05) in high-institutional-quality settings, whereas the impact is not significant (b = 0.034, p > 0.10) in low-institutional-quality environments. This indicates that FX governance mechanisms have a measurable and economically meaningful effect on firm-level financial stability in settings where legal enforcement is strong, disclosure standards are higher, and external monitoring by regulators and investors is more effective. In such environments, FX governance structures are more likely to be credible, well enforced, and effectively implemented, allowing their impact on financial stability to materialise. In contrast, in low-institutional-quality settings, the coefficient on FXGOV is statistically insignificant (coefficient = 0.034, t = 0.95). This suggests that FX governance mechanisms do not have a systematic or reliable effect on financial stability where enforcement is weak, governance rules may be symbolic rather than substantive, and firms may face constraints in effectively implementing risk governance practices. Together, these findings imply that institutional quality acts as an enabling condition for FX governance to influence financial stability. FX governance mechanisms appear to be complementary to strong institutions rather than substitutes for weak ones. We therefore accept H2.

4.5.2. Industry-Based Subsamples: Financial vs. Non-Financial Firms

Financial firms differ fundamentally from non-financial firms in terms of regulatory oversight, exposure to FX risk, risk management sophistication, and balance-sheet structure (Khoza et al., 2024). If financial firms dominate the sample or react differently to FX governance mechanisms, the baseline estimates could reflect industry-specific dynamics rather than a general firm-level governance effect. Against this backdrop, we rerun our regression for financial and non-financial firms using the baseline model specification in Equation (1) as reported in

Table 8. FX governance and financial stability in financial vs. non-financial firms.

	Financial Firms	Non-Financial Firms
	(1)	(2)
	FSTAB	FSTAB
FXGOV	−0.178 ***	−0.012
	(−6.71)	(−0.78)
BFINX	−0.052 **	0.106 ***
	(−2.20)	(7.06)
DUALFX	−0.051 *	−0.059 ***
	(−1.90)	(−3.90)
BOVST	0.049 *	0.088 ***
	(1.81)	(4.96)
BDSZE	−0.086 ***	−0.153 ***
	(−3.68)	(−10.53)
BDGDV	−0.149 ***	−0.076 ***
	(−4.92)	(−4.07)
FREVU	−0.051 *	−0.277 ***
	(−1.67)	(−16.27)
FCAPN	0.015	0.222 ***
	(0.51)	(12.62)
CONVOC	0.112 ***	0.088 ***
	(2.84)	(3.09)
CONPOL	0.068	−0.122 ***
	(1.44)	(−4.64)
CONGOV	−0.333 ***	−0.013
	(−3.62)	(−0.25)
CONREG	0.226 ***	0.113 **
	(2.93)	(2.54)
CONROL	−0.025	−0.204 ***
	(−0.31)	(−3.89)
ECODVP	0.143 ***	0.053 ***
	(3.67)	(2.73)
ECOUTP	0.275 ***	−0.030 *
	(9.11)	(−1.70)
Industry FE	YES	YES
Year FE	YES	YES
R-squared	0.131	0.145
Observations	1.902	4711

Standardized beta coefficients; t statistics in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.

.

The result shows that FX governance is negatively associated with earnings volatility (b = −0.178, p < 0.01) in financial service firms, whereas the impact is not significant (b = −0.012, p > 0.10) in non-financial firms. FX governance mechanisms may have a substantial and economically meaningful association with financial stability in the financial service sector for several reasons. First, financial firms—such as banks and insurance companies—are directly exposed to FX risk through trading activities, cross-border lending, derivatives, and balance-sheet mismatches. As a result, FX governance mechanisms are central to day-to-day risk management and directly influence financial stability outcomes. Second, financial firms operate under stringent regulatory frameworks (e.g., capital adequacy, risk committees, internal controls). FX governance mechanisms in this sector are therefore more likely to be formalised, monitored, and enforced, increasing their effectiveness. Finally, financial firms typically possess advanced risk measurement systems, specialised risk committees, and dedicated treasury functions. FX governance mechanisms are thus more tightly integrated into enterprise-wide risk management, allowing their impact on financial stability to be more pronounced.

On the other hand, FX governance may have no significant impact on financial stability for various reasons. Non-financial firms’ FX exposure often arises from trade, procurement, or foreign operations rather than financial positions. The magnitude, timing, and nature of this exposure vary widely across firms, diluting the average effect of FX governance mechanisms. FX governance in non-financial firms may be ad hoc, decentralised, or embedded within broader operational policies, reducing its visibility and effectiveness in influencing firm-wide financial stability measures. Unlike financial firms, non-financial firms face limited regulatory scrutiny over risk governance, which may result in weaker enforcement or symbolic adoption of FX governance mechanisms. Finally, financial stability in non-financial firms is often driven more by operational performance, product market competition, and supply-chain risks than by FX risk alone, limiting the observable impact of FX governance on financial stability. Overall, these findings suggest that FX governance mechanisms are particularly effective in sectors where FX risk is core to the business model and subject to strong oversight. This industry-based subsample analysis strengthens the robustness of the main findings by demonstrating that the effects of FX governance are sector-specific and theoretically consistent, rather than an artefact of sample composition.

5. Conclusions

This study examines the association between FX governance and financial stability by analysing empirical evidence from multinational entities. The result suggests that FX governance is negatively associated with earnings volatility, implying that FX governance enhances the financial stability of organisations. The baseline result is robustness to endogeneity and selection bias. However, our subsample analysis reveals that the impact of FX governance on financial stability varies based on institutional quality and industry. Whereas FX governance is negatively associated with earnings volatility thus enhancing financial stability in high-institutional-quality settings, the impact is not significant in low-institutional-quality environments. The insignificant effect in low-institutional-quality settings suggests that internal FX governance mechanisms cannot substitute for weak external institutions. Corporate boards and policy makers within organisations should therefore focus on strengthening FX governance mechanisms to ensure the benefits of active management of FX are achieved including financial sustainability. Government and external policy makers should also consider strengthening institutional mechanisms that ensure stronger enforcement of contracts, robust policies that protect investors and effective regulation of risk management practices. Such external regulatory reforms should aim to strengthen internal governance mechanisms on financial risk management to ensure that FX governance mechanisms are only symbolic and fail to translate to improved financial stability. Considering that FX governance mechanisms may be voluntary, superficially implemented and not strictly enforced in weak institutional settings, regulators may consider introducing mandatory minimum FX risk management standards, particularly for firms with weak FX oversight mechanisms and high exposure to foreign-currency fluctuations. Regulation may also impose mandatory minimum reporting requirements with respect to FX risk exposures and hedging practices of organisations. Strengthening internal FX audit and external, supervisory audit of FX risk governance are also recommended to improve implementation quality. For multinational firms, the results imply that FX governance frameworks should be adapted to host-country institutional quality rather than uniformly applied. In high-institutional-quality environments, standard FX governance mechanisms appear sufficient to stabilise earnings. In weaker institutional settings, firms may need more intensive internal controls and oversight to achieve similar outcomes.

Whilst FX governance is negatively associated with earnings volatility thereby ensuring financial stability in financial service firms, the impact is not significant in non-financial firms. Based on this result, it can be deduced that existing regulatory emphasis on risk governance in financial service firms is effective and achieving the desired result. However, policy makers and regulators should continue to enforce robust FX risk governance frameworks whilst strengthening their oversight function in closely monitoring financial institutions for compliance with FX regulations and risk management and governance in general. The insignificant effect in non-financial firms suggests that imposing uniform FX governance requirements across all sectors may yield limited marginal benefits. Policy makers should therefore be cautious about extending financial-sector-style FX governance mandates to non-financial firms. Regulatory efforts may target firms with significant foreign-currency exposure, balance-sheet FX mismatches, and extensive cross-border operations. Regulators of non-financial firms may consider disclosure-based requirements that allow market discipline to operate, guidance rather than strict mandates on FX risk management, and proportional governance frameworks tailored to firm size and FX exposure. Such flexibility recognises the heterogeneity of FX risk across non-financial firms and avoids unnecessary compliance costs. Finally, non-financial firms may benefit more from embedding FX risk management within operational and treasury functions rather than adopting formal governance structures that mirror those of financial institutions.

This study contributes to knowledge by empirically validating the relevance of FX governance to financial stability. It contributes to the limited studies on the role of FX governance in diminishing earnings volatility, thus exposing FX management as a strategy for achieving financial sustainability. Second, the international sample analysed in the study contributes to the generalisability of results. Whereas related studies on the subject have predominantly been investigated in in-country/geographic-region settings, the analysis of the international sample contributes to knowledge by providing a broader perspective on the subject beyond a country or regional setting. Third, the study analyses data covering a long timeframe, and this is important for a careful analysis of the subject, considering that financial stability issues should ordinarily span a long timeframe to ensure validity of results. By analysing empirical data covering a long timeframe, the study’s adoption of a longitudinal approach ensures well-validated results. Finally, the study presents evidence that the impact of FX governance on financial stability is contextually different based on institutional environment and industry. Such knowledge is important to corporate executives, policy makers and regulators in formulating appropriate policies that tackle specific challenges of FX governance and risk management.

This study is not without limitations, which in turn provide avenues for future research. Although the Forbes Global 2000 provides a practical and widely accepted proxy for the global MNC population, several limitations should be acknowledged. First, the dataset primarily captures large, publicly listed corporations with substantial revenues, assets, profits, and market capitalization. As a result, smaller multinational enterprises and privately held firms are excluded, limiting the representativeness of the sample across the broader spectrum of multinational activity. Second, the ranking may introduce geographic bias, as firms from developed economies and countries with more established financial reporting and capital market infrastructures are more likely to be included. Consequently, firms from emerging or less transparent markets may be underrepresented. Third, because the sample is inherently skewed toward economically powerful and globally visible corporations, the findings of this study should be interpreted mainly in the context of large and internationally active firms rather than generalised to all multinational corporations. Nevertheless, consistent with prior studies (Giannarakis, 2014; Ngu & Amran, 2019), the Forbes Global 2000 remains an appropriate sampling frame due to its broad international coverage, standardised financial disclosures, and its ability to capture a substantial share of global economic activity and cross-border business operations. Future research may improve representativeness by incorporating smaller firms, privately held multinational enterprises, and alternative international datasets.