The Financial Risk Meter (FRM) for Kuwait: A Tail-Event Perspective on Systemic Risk and Economic Forecasting

Abstract

This study develops and applies the Financial Risk Meter (FRM) for Kuwait, a novel measure of systemic risk tailored for a commodity-dependent emerging economy. Using Lasso quantile regression, the FRM captures tail-event co-movements among key financial institutions, providing a robust indicator of systemic stress. This paper makes three primary contributions. First, it provides the first application of the FRM framework to an oil-exporting economy, identifying the distinct channels through which global financial shocks and commodity price volatility create systemic risk. Second, it quantitatively demonstrates the FRM’s superior performance in tracking financial stress compared to the benchmark Conditional Value-at-Risk (CoVaR) model. Third, it identifies the specific drivers of systemic risk in Kuwait, offering actionable insights for policymakers. Our findings show that the FRM effectively pinpoints periods of high financial distress, aligns with global risk indicators, and can enhance recession forecasting. By providing a clear and timely measure of systemic risk, this study offers a valuable tool for regulators to bolster financial stability and advance sustainable economic development in Kuwait and other resource-dependent nations.

1. Introduction

Systemic risk—the danger that the failure of one financial entity could trigger a cascading collapse across the system—presents a critical threat to global economic stability. The 2008 financial crisis painfully illustrated that traditional risk models, which focus on individual institutions in isolation, are inadequate for capturing the complex web of interdependencies in modern financial networks. In response, a new generation of systemic risk models has emerged to measure and monitor these network-based vulnerabilities.

While these advanced models have been widely applied in developed economies, a significant research gap persists: their application to commodity-dependent emerging markets. Economies like Kuwait face a dual-shock vulnerability; they are exposed not only to global financial contagion but also to the extreme volatility of commodity markets. This dual exposure creates unique tail-event risks that traditional linear models are ill-equipped to capture, leaving a critical question unanswered: How can systemic risk be accurately measured in an economy so profoundly shaped by a single commodity?

To address this gap, this study introduces and calibrates the Financial Risk Meter (FRM) for Kuwait. We employ a state-of-the-art framework using Lasso quantile regression to analyze systemic risk from a tail-event perspective. The FRM is uniquely suited for this context for two reasons. First, it focuses specifically on co-movements during periods of extreme market stress (the tails of the distribution), which is essential for crisis monitoring. Second, its integrated Lasso operator automatically selects the most significant risk channels from a high-dimensional set of financial and macroeconomic variables, making it ideal for disentangling the complex risk drivers in Kuwait’s economy. By applying this model to an archetypal oil-exporting economy, we dissect the distinct impacts of global financial shocks and domestic commodity shocks on systemic stability.

Specifically, this paper addresses the following core research questions: 1. How can the Financial Risk Meter (FRM) framework be effectively adapted and calibrated to measure systemic risk in a dual-shock, oil-dependent economy like Kuwait? 2. Does the FRM provide a more accurate and timely measure of financial distress in Kuwait when compared to established benchmark models, such as Conditional Value-at-Risk (CoVaR)? 3. What are the primary drivers of systemic risk in Kuwait, and what specific, data-driven policy recommendations can be derived from the FRM to enhance financial stability and support sustainable development?

FRM employs Lasso (Least Absolute Shrinkage and Selection Operator) quantile regression, a statistical methodology particularly well-suited for exploring the relationships among financial variables during extreme events. Unlike traditional linear regression models, which focus on average behavior, quantile regression captures the behavior of variables at different points of their distribution, with a particular focus on the tails. The inclusion of the Lasso component adds a layer of refinement, enabling the model to handle high-dimensional data sets by selecting only the most relevant predictors. This feature is particularly valuable in financial systems, where numerous variables interact in complex and often non-linear ways. By concentrating on tail event data, the FRM provides a nuanced understanding of co-movements and dependencies during periods of market stress, which are crucial for identifying sources of systemic risk.

The primary objective of FRM is to uncover the behavior of active financial data sets and to reveal interdependence within a network framework. Financial systems are inherently interconnected, with institutions, markets, and macroeconomic factors forming a dense web of relationships. These interconnections amplify the potential for systemic crises, as stress in one part of the system can rapidly propagate to others. The FRM’s network-oriented approach offers a significant departure from traditional risk assessment methods, which often focus on individual institutions in isolation. By modeling the financial system as a network, the FRM highlights both direct and indirect channels of risk transmission, offering a comprehensive view of systemic vulnerabilities.

The increasing complexity of global financial systems has rendered traditional measures of systemic risk insufficient. Metrics such as value-at-risk (VaR) or default probability often fail to account for the interconnected nature of financial institutions and markets, focusing instead on isolated entities or events. However, systemic risk is inherently a network phenomenon, requiring tools that can capture the dynamic and interdependent nature of financial systems. FRM addresses this gap by integrating advanced statistical techniques with a network perspective, enabling a deeper understanding of systemic dynamics.

The importance of focusing on tail events—extreme market outcomes—is particularly crucial in systemic risk analysis. While day-to-day fluctuations in financial markets may appear manageable, it is the rare and extreme events that have the most devastating consequences. These tail events are often characterized by sudden, correlated movements across financial institutions, leading to widespread instability. FRM’s emphasis on tail event data ensures that it captures these critical dynamics, providing insights into how stress propagates during crises and identifying the factors that amplify such risks.

FRM’s methodology is grounded in its application of Lasso quantile regression, which combines the strengths of both quantile regression and Lasso regularization. Quantile regression allows the FRM to focus on specific parts of the distribution, such as the upper or lower tails, where systemic risks are most pronounced. Lasso regularization, on the other hand, ensures that the model remains parsimonious, selecting only the most relevant variables from a potentially vast pool of predictors. This is particularly important in financial systems, where data sets often include a multitude of variables, ranging from balance sheet indicators to macroeconomic measures.

FRM is employed to time series data collected from financial institutions and macroeconomic indicators, creating a dynamic framework for systemic risk analysis. By integrating these diverse data sources, the FRM captures the interplay between micro-level institutional factors and macro-level economic trends. This dual focus ensures that the FRM is both granular enough to identify risks at the institutional level and broad enough to consider system-wide dynamics.

The FRM’s utility extends across various domains of financial analysis and policymaking. One of its key applications is the identification of entities experiencing extreme “co-stress”—situations where multiple institutions exhibit simultaneous signs of distress. Such co-stress events are often precursors to systemic crises, making their early identification crucial for preventive action. Additionally, the FRM can pinpoint specific risk factors, such as credit default swaps, that exacerbate systemic vulnerabilities. This capability enables stakeholders to target their risk mitigation efforts more effectively.

Another significant application of the FRM lies in recession forecasting. By analyzing patterns in systemic risk indicators, the FRM can generate probabilities of economic downturns, providing policymakers with valuable foresight. This predictive capability is particularly important in a globalized economy, where recessions can have far-reaching consequences.

Ultimately, the FRM represents a significant advancement in the field of systemic risk analysis. By focusing on the dynamics of tail events and the interconnected nature of financial systems, it provides a robust framework for understanding and addressing systemic vulnerabilities. Its network-based approach offers a more holistic view of risk transmission pathways, highlighting the importance of interdependencies in financial systems.

FRM is not merely a tool for measuring systemic risk; it is a lens through which we can better understand the complexities of modern financial systems. Its ability to integrate advanced statistical techniques with practical applications makes it an invaluable resource for academics, policymakers, and industry professionals alike. As financial systems continue to evolve, the need for sophisticated tools like the FRM will only grow, underscoring its relevance in the ongoing effort to enhance financial stability.

The proper quantification of systematic risk is important not only for financial stability but also for economic sustainability and robustness. For Kuwait, as an emerging economy heavily relying on oil, effective risk assessment mechanisms like the FRM can aid decision makers in policies targeted for economic diversification and minimize the impact of external shock the country is subjected to. Lifting financial system resilience supports wider sustainability goals and contributes to the UN SDGs, including SDG 8 (Decent Work and Economic Growth) and SDG 9 (Industry, Innovation, and Infrastructure). There are important implications of these insights for the policymakers, regulators and stakeholders that are working to achieve sustainable economic development.

This study makes the following novel contributions to the literature: First, we provide the first comprehensive application of the FRM framework to an oil-dependent nation, demonstrating how external financial shocks and domestic commodity price volatility interact to drive systemic risk in Kuwait. Second, we quantitatively establish the FRM’s superior performance in identifying periods of high financial stress in Kuwait compared to a standard benchmark model, CoVaR, thereby validating its utility in this specific economic context. Third, we move beyond descriptive analysis by identifying the primary drivers of Kuwait’s systemic risk. We provide robust, data-driven evidence on the relative importance of global risk aversion, oil price volatility, and regional geopolitical factors, offering a clear roadmap for targeted policy interventions.

Ultimately, by properly quantifying systemic risk, this study offers critical insights for policymakers in Kuwait and other resource-dependent nations. An effective risk assessment mechanism like the FRM can inform policies aimed at economic diversification and enhance financial resilience, supporting long-term sustainable development goals such as SDG 8 (Decent Work and Economic Growth) and SDG 9 (Industry, Innovation, and Infrastructure). This paper proceeds as follows: Section 2 reviews the relevant literature, Section 3 describes the data, Section 4 outlines the methodology, Section 5 presents the empirical results, Section 6 discusses the findings, and Section 7 concludes.

2. Literature Review

The study of systemic risk has evolved significantly over the past few decades, with researchers and practitioners developing a wide array of models and tools to measure, analyze, and mitigate financial instability, and has long been a central focus of financial research, given its potential to destabilize entire economies. Systemic risk is characterized by the possibility that the failure of one or more financial institutions may trigger widespread financial instability. FRM, which leverages Lasso quantile regression to explore co-movements during extreme events, represents a modern approach to this challenge. To situate the FRM within the broader academic and practical landscape, it is essential to review existing literature on systemic risk, network theory in finance, quantile regression, and high-dimensional statistical methods.

Systemic risk has been widely studied, with early definitions emphasizing the potential for financial distress to spread across institutions due to interconnectedness. One of the foundational works in this domain is from [1], who explored the dynamics of systemic risk during the 2008 financial crisis and proposed measures such as Systemic Expected Shortfall. The seminal work of [2] provided a comprehensive definition of systemic risk, emphasizing the role of interconnectedness among institutions and the amplifying effects of leverage and contagion. Similarly, [3] introduced Conditional Value-at-Risk (CoVaR), a metric to measure the spillover effects of risk between financial entities and quantify the risk of an institution conditional on the distress of others. These approaches highlight the importance of interdependence but often rely on linear models, which may not capture the nuances of extreme market conditions. Both studies underscored the necessity of incorporating interdependencies into systemic risk models.

Traditional approaches to measuring systemic risk, such as value-at-risk (VaR), focus on individual institutions in isolation. These methods, while useful for micro prudential purposes, fail to capture the network effects that are central to systemic crises. Research by [4] argued that financial systems should be analyzed as complex networks, where risk arises not just from individual entities but also from their interconnectedness. This perspective laid the groundwork for methodologies like FRM, which explicitly model systemic risk within a network framework.

The Global Financial Stability Reports by the International Monetary Fund (IMF) and the Financial Stability Board (FSB) have also emphasized the need for tools that identify vulnerabilities in financial networks. Stress-testing frameworks like those used by central banks, such as the Federal Reserve’s Comprehensive Capital Analysis and Review (CCAR), provide insights into systemic risk but often lack real-time adaptability and fail to address non-linear dependencies.

Network theory has become an essential tool for understanding systemic risk, and provides a powerful framework for understanding systemic risk, as financial institutions are interconnected through lending relationships, derivative exposures, and other channels. The authors of [5] were among the first to explore how network topology affects systemic risk, showing that the structure of financial linkages determines the resilience of the system to shocks. Subsequent work by [6] analyzed the implications of network density and connectivity, and explored the resilience of networks under stress, emphasizing the importance of connectivity patterns, finding that while highly connected networks can absorb small shocks, they are more vulnerable to systemic crises.

The concept of “too interconnected to fail” gained prominence as researchers, such as [7] introduced metrics like Debt Rank, which quantify the systemic importance of financial institutions based on their centrality in the network. Their approach provided a means to rank institutions by their potential to propagate financial stress. However, these methods primarily focus on direct linkages, whereas the FRM’s use of Lasso quantile regression enables the identification of indirect relationships that might be hidden in traditional network analyses. In other words, these models often assume static relationships and do not account for the dynamic nature of risk propagation during crises. The FRM addresses this limitation by incorporating time-varying data and focusing on tail-event co-movements, providing a more dynamic perspective on systemic interdependencies.

Quantile regression, introduced by [8], has become a cornerstone of financial risk modeling owing to its ability to capture relationships across the distribution of a dependent variable. Unlike ordinary least squares (OLS), which estimates the conditional mean, quantile regression focuses on the conditional quantiles, enabling analysis of the entire distribution of the dependent variable. This capability makes quantile regression particularly well-suited for systemic risk analysis, where the focus is often on extreme events (the tails of the distribution). In the context of systemic risk, quantile regression has been applied to analyze tail risks, where extreme outcomes are of primary concern.

Studies such as [9] and [10] applied quantile regression to financial data, highlighting its ability to capture non-linear relationships and tail risks. These works provided a foundation for the use of quantile regression in systemic risk modeling, emphasizing the importance of analyzing co-movements during periods of market stress. However, as the dimensionality of financial data increases, traditional quantile regression faces challenges in handling large numbers of predictors.

The application of quantile regression to systemic risk gained further traction with studies that emphasized tail-event dynamics. Studies like those by [11] highlighted how quantile regression can uncover hidden vulnerabilities that traditional regression models overlook. However, these studies often face challenges in handling high-dimensional data, where the number of variables exceeds the number of observations—a limitation addressed by Lasso regression.

High-dimensional settings are common in systemic risk analysis, where financial systems involve numerous institutions, instruments, and macroeconomic factors. Traditional regression models struggle in these scenarios due to multicollinearity and overfitting. The Lasso (Least Absolute Shrinkage and Selection Operator), introduced by [12], addresses the challenges of high-dimensional regression by introducing a penalty term that shrinks coefficients and selects a subset of the most relevant predictors. This approach is particularly valuable in systemic risk analysis, where the number of potential risk factors often exceeds the number of observations.

In systemic risk research, Lasso regression has been applied to identify key drivers of financial stress. For example, [13] demonstrated the applicability of Lasso regression in econometric models, showcasing its ability to handle multicollinearity and improve model interpretability, proving the effectiveness of penalized regression techniques in high-dimensional econometric models, which provides a foundation for tools like the FRM. Lasso quantile regression, as used in the FRM, builds on this approach by focusing specifically on tail risks while maintaining model simplicity and interpretability.

Ref. [14] applied penalized regression techniques to interbank market data, identifying key drivers of financial stress. The integration of Lasso with quantile regression, as seen in the FRM, represents a significant advancement, enabling the simultaneous analysis of tail risks and high-dimensional data.

The integration of Lasso regression with quantile regression is a relatively recent advancement that addresses the challenges of high-dimensional systemic risk analysis. FRM, 262 developed by [1], builds on these methodological foundations by combining Lasso quantile regression with a network perspective. This innovative approach allows the FRM to capture the co-movements of financial institutions during extreme events, providing insights into systemic vulnerabilities. The FRM’s focus on tail-event dynamics makes it particularly relevant for identifying “co-stress” events, where multiple institutions experience simultaneous distress.

In practical applications, FRM has been used to analyze time series data from financial institutions and macroeconomic indicators, revealing patterns of interdependence and stress propagation. Its ability to generate recession probabilities based on systemic risk indicators further demonstrates its versatility and utility in macroeconomic forecasting.

In addition to [1], other researchers have explored similar methodologies for identifying systemic vulnerabilities. For instance, ref. [14] used penalized regression techniques to model systemic risk in interbank markets, while [15] extended these methods to analyze market-wide stress scenarios. These studies underscore the potential of combining Lasso and quantile regression for systemic risk applications, particularly in capturing the dynamics of financial networks under stress.

The FRM distinguishes itself within the literature by integrating network theory, high-dimensional statistics, and tail-event analysis into a unified framework. Unlike earlier systemic risk measures, which often rely on linear assumptions or focus on individual institutions, the FRM provides a comprehensive view of financial networks. By employing Lasso quantile regression, it addresses the dual challenges of high-dimensionality and non-linearity, offering a robust tool for real-time systemic risk monitoring.

Moreover, the FRM’s practical applications—such as identifying entities experiencing “co-stress” and forecasting recession probabilities—position it as a valuable complement to existing tools. Its ability to combine granular institutional data with macroeconomic indicators further enhances its utility in policymaking and risk management.

The literature on systemic risk, financial networks, and advanced statistical methods provides a rich foundation for understanding the contributions of the FRM. By leveraging insights from network theory, quantile regression, and Lasso regularization, the FRM addresses critical gaps in traditional systemic risk measures. Its focus on tail-event dynamics and interdependencies within financial networks position as a state-of-the-art tool for analyzing and mitigating systemic vulnerabilities. As financial systems become increasingly complex, the FRM’s methodology and applications will continue to play a pivotal role in advancing the field of systemic risk analysis.

While foundational models like CoVaR [3] and network metrics such as DebtRank [5] were groundbreaking, they possess limitations in the context of a commodity-driven economy. CoVaR, for instance, typically relies on quantile regression but does not inherently handle the high-dimensionality of risk factors present in a complex financial system without modification. Network models often require granular data on direct interbank linkages, which is not always available, and may miss indirect risk transmission through macroeconomic factors like a sudden oil price collapse.

While the foundational literature on systemic risk originated from the analysis of advanced economies, a growing body of research has begun to address the unique vulnerabilities of emerging and resource-dependent markets. These economies often face pro-cyclical capital flows that amplify boom–bust cycles and are disproportionately exposed to global commodity price shocks, which can trigger simultaneous crises in both their financial and real sectors. Research focused on the Gulf Cooperation Council (GCC) region, for instance, has emphasized how oil price volatility acts as a primary channel for financial contagion, a factor that traditional systemic risk models may not adequately capture. This context underscores the need for a framework like the FRM, which is designed to identify risk drivers in a high-dimensional setting and is sensitive to the extreme tail events that characterize commodity-driven financial instability.

The FRM framework directly addresses these shortcomings. By integrating Lasso regularization with quantile regression, it solves the high-dimensionality problem, allowing for the inclusion of numerous firm-level and macroeconomic variables while identifying only the most salient predictors of tail risk. This is particularly advantageous for an economy like Kuwait, where risk is driven by a confluence of local financial returns, global market sentiment, and volatile energy prices. The FRM’s focus on tail-event co-movements, rather than average behavior, makes it fundamentally better suited to capturing the abrupt, non-linear crises that characterize dual-shock economies.

Research Gap and Contribution of the FRM

The existing literature, while extensive, reveals critical gaps when applied to commodity-dependent economies like Kuwait. Foundational models like CoVaR [3] and network metrics like DebtRank [5] were groundbreaking but possess key limitations in this context.

First, high-dimensionality: CoVaR, in its standard form, struggles to handle the large number of potential risk factors present in a dual-shock economy. Identifying risk drivers requires modeling dependencies on dozens of local firms, global indices, and commodity prices simultaneously—a task for which standard quantile regression is not designed. Second, indirect exposures: Network models often require granular data on direct interbank linkages, which is not always available, and may fail to capture indirect risk transmission through macroeconomic factors like a sudden oil price collapse.

The FRM framework directly addresses these shortcomings. By integrating Lasso regularization with quantile regression [16], it solves the high-dimensionality problem, allowing for the inclusion of numerous variables while automatically identifying the most salient predictors of tail risk. This makes it uniquely suited for an economy like Kuwait, where systemic risk is driven by a confluence of local financial returns, global market sentiment, and volatile energy prices. The FRM’s inherent focus on tail-event co-movements makes it fundamentally better equipped to capture the abrupt, non-linear crises that characterize such economies.

3. Data

For the construction of the FRM for Kuwait, we selected companies that are consistent constituents of the Kuwait stock indices and play crucial roles in the financial stability and growth of the economy. This selection includes major companies from key industries such as banking, telecommunications, logistics, and manufacturing.

The data set comprises daily stock prices for 30 prominent Kuwaiti companies, reflecting their systemic importance in the local market. These firms include AAYAN, ABK, AGLTY, ALAF, ALMT, INTG, MEZZ, STC, WARB, GFHK, JAZK, HUMN, and others across diverse sectors. Daily returns for each company were calculated as the log difference in prices. This daily data spans from July 16, 2018, to September 11, 2024, based on the longest available data alignment. The data is presented in the Appendix A, in Table A1 and Table A2.

In addition to firm-specific data, a range of macroeconomic risk factors relevant to the Kuwaiti economy were selected. These variables capture exposure to systemic risks such as interest rate changes, exchange rate fluctuations, oil price volatility, and regional geopolitical developments, enabling a holistic view of the economic forces at play.

4. Methodology

FRM methodology assesses systemic risk among interconnected financial institutions in Kuwait by examining the tail dependencies in their returns [15]. This framework integrates the individual stock returns of Kuwaiti financial entities with macroeconomic factors, which serve as conditioning variables to reflect broader economic risks. These factors include key market indices, such as the MSCI Kuwait and MSCI Emerging Markets Price Index, alongside other indicators that capture systemic vulnerabilities. By extending beyond the conventional Value-at-Risk (VaR) framework, FRM incorporates multivariate dependencies, offering a holistic perspective on systemic risk. Drawing inspiration from the CoVaR framework, which evaluates tail risks during institutional distress, the FRM advances this approach by quantifying network-wide dependencies within a unified risk measure.

Quantile Regression Model:

Within the FRM framework, the risk level for each company on a given trading day is modeled using quantile regression. Given a data set with $J$ companies and $M$ macroeconomic risk factors, the return $X_{j,t}$ for each institution is modeled linearly as:

$$X_{j,t}={\alpha }_j+A_{j,t}^T{\beta }_j+{\epsilon }_{j,t}$$

(1)

where $A_{j,t=}\left(\begin{array}{c}{X}_{-j,t}\\ M_t\end{array}\right)$ denotes a $p$-dimensional vector comprising the returns of other institutions and macroeconomic variables. The dimension $p=J+M-1$ covers all relevant data excluding the target institution. The coefficient vector ${\beta }_j$ contains the parameters for both institutional and macroeconomic variables, capturing interdependence and allowing systemic connections to influence the institution’s return.

Penalized Regression (Lasso):

To manage the high-dimensional nature of the data, the FRM employs Lasso (L1-penalized) quantile regression, where the penalty parameter ${\lambda }_{\mathrm{İ}}$ induces sparsity by driving insignificant coefficients toward zero. The parameter ${\lambda }_{\mathrm{İ}}$ is selected to balance model complexity and goodness-of-fit, effectively filtering out weaker relationships to emphasize significant systemic links. This feature is especially critical in the Kuwaiti context. The nation’s financial system is influenced by a wide array of variables: the performance of dozens of domestic firms, regional geopolitical shifts, global investor sentiment (e.g., VIX), and the volatile price of oil. A traditional regression model would suffer from multicollinearity and overfitting, making it impossible to identify the true drivers of risk. The Lasso penalty effectively automates the process of variable selection, creating a parsimonious and interpretable model that isolates the most systemically important factors.

A higher penalty results in a sparser model that highlights only key dependencies, enhancing interpretability. The penalty parameter ${\lambda }_j$ for each firm $j$ is chosen adaptively using a 10-fold cross-validation procedure to optimize the model’s out-of-sample predictive performance. This ensures that the level of sparsity is determined empirically by the data rather than being arbitrarily set.

To account for possible tail dependencies, we also test for nonlinearities and interaction effects, even though quantile regression is naturally flexible across quantiles. To examine the conditional relationship between various quantiles of the independent and dependent variables, we add quantile-on-quantile regression (QQR) to the baseline linear specification. Our systemic risk assessment is more robust thanks to this non-linear extension, which makes sure the model accurately depicts the intricate dynamics found in the distribution’s tails.

Optimization and Tail Risk Level:

The penalized quantile regression minimizes the weighted deviations of the institution’s returns from the model’s predictions. This is formalized as:

$${}_{{\mathsf{\alpha }}_{\mathrm{j}},{\mathsf{\beta }}_{\mathrm{j}}}{}^{\mathrm{m}\mathrm{i}\mathrm{n}}\left\{{\mathrm{n}}^{-1}\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathsf{\rho }}_{\mathrm{t}}\left({\mathrm{X}}_{\mathrm{j},\mathrm{t}}-{\mathsf{\alpha }}_{\mathrm{j}}-{\mathrm{A}}_{\mathrm{j},\mathrm{t}}^{\mathrm{T}}{\mathsf{\beta }}_{\mathrm{j}}\right)+{\mathsf{\lambda }}_{\mathrm{j}}{‖\mathsf{\lambda }‖}_1\right\}$$

(2)

with check function

$${\mathsf{\rho }}_{\mathrm{t}}\left(\mathrm{u}\right)=\left|\mathsf{\tau }-\mathsf{{\rm I}}\left\{\mathrm{u}\le 0\right\}\right|{\left|\mathrm{u}\right|}^{\mathsf{\gamma }}$$

(3)

To understand this equation intuitively, one can think of it as a balancing act. The first part of the equation (${\mathrm{n}}^{-1}\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathsf{\rho }}_{\mathrm{t}}\left({\mathrm{X}}_{\mathrm{j},\mathrm{t}}-{\mathsf{\alpha }}_{\mathrm{j}}-{\mathrm{A}}_{\mathrm{j},\mathrm{t}}^{\mathrm{T}}{\mathsf{\beta }}_{\mathrm{j}}\right)$) is the “goodness-of-fit” term; its goal is to create a model that explains the institution’s tail risk as accurately as possible. However, without a check, it might create an overly complex model by finding connections everywhere. The second part, ${\lambda }_j{‖\lambda ‖}_1$, is the “simplicity penalty.” For every connection the model wants to add, it must pay a penalty. The parameter λ acts as the gatekeeper or referee in this process. A high λ imposes a heavy penalty, forcing the model to select only the most critical, undeniable risk channels. A low λ is more lenient. Therefore, the final model includes only the connections that are strong enough to outweigh the penalty.

Where $\tau$ represents the tail risk quantile level, such as 0.05 for extreme tail risk events. We select $\tau =$ 0.05 to focus on the 5% left tail of the returns distribution, a standard convention in financial risk management for defining extreme loss events. This function weighs the residuals differently depending on whether they exceed or fall below the threshold, ensuring the model focuses on the tail risk. The selection of $\gamma$ designates quantile regression, while $\gamma$ = 2 gamma would correspond to expected regression.

The check function ${\rho }_i\left(t\right)$ in quantile regression is defined with $\gamma$ = 1, reflecting the asymmetric weighting of positive and negative residuals. This contrasts with ordinary least squares (OLS), which minimizes the squared error with $\gamma$ = 2. We include this clarification to avoid ambiguity in the optimization formulation and to highlight the methodological differences between the quantile and mean regression approaches.

Systemic Risk Measurement:

The Lasso regression is applied across all institutions in a rolling window framework. The penalty parameters ${\lambda }_i$ for each company are averaged to compute FRM value. This aggregate measure reflects the systemic risk at any point in time, with higher values signaling increased financial stress across the network. FRM thus serves as a real-time indicator of market-wide risk, capturing the magnitude and diffusion of financial risk within the system.

Even though Lasso penalties are institution-specific, to get a measure of the average systemic risk level in the network we calculate FRM as a composite measure. This assumption is consistent with the observation that penalty paths are stable across institutions in our sample.

Interpretation and Policy Relevance:

The FRM, through the average ${\lambda }_i$, helps regulators identify critical financial institutions whose distress may have significant market-wide impacts. By tracking variations in these penalty terms over time, authorities can pinpoint co-stress entities (institutions frequently appearing in others’ active sets) and activator entities (those exerting widespread systemic influence). These insights support proactive policymaking and strengthen monitoring financial stability.

In summary, the FRM methodology employs high-dimensional quantile regression with Lasso regularization to quantify systemic risk by emphasizing tail dependencies within Kuwait’s financial network. This dynamic risk assessment framework offers policymakers actionable insights into the origins and pathways of financial instability.

The FRM is estimated to use a rolling-window approach with a window size of 250 trading days, which approximates one calendar year. This choice balances the need for a sufficient sample size for robust estimation against the need to capture time-varying relationships in the financial network. All models are estimated in R using the quantreg [17] package.

5. Empirical Results

The empirical application of the FRM yields three primary insights that validate its utility for monitoring systemic risk in Kuwait.

First, the FRM effectively identifies periods of high financial stress, and its dynamics are consistent with key global risk indicators. As shown in Figure 1, the FRM exhibits sharp spikes that coincide precisely with major global turmoil, such as the COVID-19 pandemic in 2020. Its strong co-movement with the VIX, a global fear gauge, confirms that the FRM is capturing the transmission of international financial stress into the Kuwaiti system. Similarly, Figure 2 shows the FRM’s sensitivity to shocks in the oil market, with the 2020 oil price crash triggering the largest spike in the series. This alignment with external benchmarks validates the FRM as a robust and accurate indicator of systemic risk.

Figure 1. Kuwait’s FRM vs. VIX.

Figure 2. Kuwait’s FRM vs. WTI Crude Oil.

Second, and most importantly, the FRM provides granular insights into the domestic transmission of risk that simpler metrics cannot offer. A key advantage of the Lasso methodology is its ability to identify which specific institutions are the primary drivers of systemic risk. By analyzing the firms most frequently selected by the model as significant predictors of stress in their peers, we can identify the system’s core nodes. Our analysis consistently reveals that a core group of major banking institutions—namely Kuwait Finance House (KFH), Gulf Bank (GBKK), and Burgan Bank (BURG)—are the most significant transmitters of risk. Furthermore, the selection of large, integrated non-financial firms like Agility (AGLTY) demonstrates that systemic risk is not confined to the banking sector alone. This specific, firm-level insight is a key contribution of the FRM, allowing regulators to focus supervisory attention where it is most needed.

Third, the spillover effects from external markets to Kuwait’s financial system are non-linear and most pronounced during periods of high stress. To formally test this, we use quantile-on-quantile regression (QQR) and causality-in-quantiles analysis. Figure 3 shows that the impact of the OFR Financial Stress Index on Kuwait’s FRM is strongest when both indices are already at high quantiles (i.e., during periods of market turmoil). This confirms that Kuwait is most vulnerable to global contagion when its own financial system is already fragile. The rolling-window causality analysis in Figure 4 further supports this, showing that the causal link from the FSI to the FRM intensifies during specific crisis periods. These results provide robust evidence of the tail-risk dynamics our model is designed to capture.

Figure 3. Quantile-on-Quantile (QQ) regressions-based estimated impact of OFR FSI on FRM.

Figure 4. Quantile-on-Quantile (QQ) causality-based estimated impact of OFR FSI on FRM.

During periods of heightened global volatility, Kuwait’s dependency on foreign investment and its exposure to the global energy market further amplify its financial risks. The interconnectedness of its economy with international capital markets, along with a heavy reliance on oil revenue to support government spending and maintain fiscal stability, exposes Kuwait to cyclical downturns in both financial and commodity markets. The correlation between the FRM and VIX also underscores the limited capacity of oil-dependent economies to decouple from external shocks, as Kuwait’s economic structure leaves it vulnerable to both financial market instability and commodity price dynamics. This sensitivity is particularly evident during global risk-off periods, when investors tend to withdraw from emerging markets in search of safer assets, exacerbating financial stress for economies like Kuwait.

One of the most pronounced spikes in Kuwait’s FRM occurred during the COVID-19 pandemic, which prompted synchronized surges in both the VIX and Kuwait’s FRM, reflecting the compounded impact of collapsing global demand and supply chain disruptions on Kuwait’s economy. This period underscores Kuwait’s exposure to external liquidity conditions and commodity market volatility. Coupled with a dramatic decline in global oil demand, Kuwait experienced a sharp contraction in revenue, increasing fiscal pressures and elevating financial risk perceptions. The spike in the FRM during this time reflects how interconnected Kuwait’s financial and economic stability is with global market conditions, emphasizing the importance of liquidity and investor confidence for oil-dependent economies.

Similarly, the COVID-19 pandemic prompted synchronized surges in both the VIX and Kuwait’s FRM, reflecting the compounded impact of collapsing global demand and supply chain disruptions on Kuwait’s economy. The pandemic-induced shock to oil prices, driven by a combination of plummeting demand and a Saudi-Russian price war, resulted in one of the steepest declines in energy markets in recent history. For Kuwait, this crisis was particularly damaging, as its reliance on oil revenues meant that fiscal and external balances deteriorated rapidly. Additionally, the uncertainty surrounding the pandemic led to increased financial market volatility, further raising the FRM. The simultaneous spike in the FRM and VIX during the pandemic period demonstrates Kuwait’s dual exposure to global financial shocks and oil market disruptions, underscoring the challenges faced by oil-exporting nations during crises that simultaneously affect both energy markets and the broader global economy.

This dual sensitivity highlights the need for Kuwait to implement long-term structural reforms aimed at reducing its reliance on oil revenues and diversifying its economic base. By fostering economic resilience through diversification, Kuwait could mitigate the severity of future financial shocks and reduce its exposure to global risk aversion during periods of heightened volatility.

Figure 2 highlights the close relationship between Kuwait’s FRM and WTI Crude Oil prices, illustrating the pivotal role of oil market fluctuations in shaping financial risk perceptions. As an oil-dependent economy, Kuwait’s financial stability is intrinsically tied to the performance of global energy markets. Periods of declining oil prices, such as the 2020 pandemic-induced demand shock, correspond to sharp increases in Kuwait’s FRM. These episodes emphasize the dual challenge of managing fiscal pressures and maintaining investor confidence in an environment of volatile oil prices. The correlation between FRM and oil prices underscores the heightened vulnerability of Kuwait’s financial system to commodity market shocks, which ripple through the broader economy, affecting fiscal sustainability, external balances, and market sentiment.

Domestic economic challenges further compound Kuwait’s financial vulnerabilities. Structural issues such as overreliance on public sector employment, persistent fiscal deficits, and slow progress in economic diversification exacerbate financial risks during times of external shocks. For example, delays in implementing fiscal reforms or diversifying the economy have left Kuwait more exposed to global and regional shocks, amplifying spikes in the FRM during periods of uncertainty. Additionally, domestic political gridlock and policy uncertainty have periodically raised concerns among investors, contributing to fluctuations in financial risk perceptions. These domestic challenges underscore the interconnectedness of internal and external risk factors in shaping Kuwait’s overall financial stability.

Beyond tracking aggregate risk, a key advantage of the FRM’s Lasso-based methodology is its ability to identify the specific institutions that act as primary transmitters of systemic risk. By analyzing which firms are most frequently selected as significant predictors in the quantile regressions of their peers, we can pinpoint the system’s core nodes. Our analysis consistently identifies a core group of institutions whose distress has the widest-reaching impact. These include major banking institutions such as Kuwait Finance House (KFH), Gulf Bank (GBKK), and Burgan Bank (BURG), which form the backbone of the financial system. Additionally, large, highly integrated non-financial firms like Agility (AGLTY) also emerge as significant contributors, highlighting that systemic risk in Kuwait is not confined to the banking sector alone. This insight allows regulators to move beyond sector-wide policies and focus supervisory attention on the most systemically important entities.

To further deepen our analysis as robustness check, we utilize the quantile-on-quantile regression (QQR) and a non-parametric causality in quantiles [18] (The QQR methodology is presented in Appendix A), analyses for Office of Financial research (OFR) financial stress index (FSI) for emerging countries (https://www.financialresearch.gov/financial-stress-index/ accessed on 22 September 2025), and Kuwait FRM. First, we examine the QQ regression between the conditional quantiles of OFR FSI regressed on the quantiles of FRM and depict in Figure 3. The quantile-on-quantile regression estimates reveal a non-linear and asymmetric relationship between the FRM for Kuwait and the OFR FSI for emerging markets across different quantiles. At lower quantiles of both indices, the slope estimates are relatively subdued, underlining a weaker connection between low levels of financial stress in emerging markets and periods of low financial risk in Kuwait. On the contrary, at higher quantiles of both indices, the slope becomes more pronounced, indicating that heightened financial stress in emerging markets amplifies financial risk in Kuwait, likely reflecting stronger spillover effects during turbulent periods. This pattern underscores the interconnectedness of financial systems, particularly during periods of heightened risk.

Additionally, the relationship exhibits notable asymmetry across quantiles, with a mix of positive and negative slope estimates. To exemplify, during periods of low financial stress in emerging markets (low FSI quantiles), Kuwait’s financial risk (FRM) appears to depict a muted or negative response across various FRM quantiles. However, during moderate to high stress levels in emerging markets (middle to high FSI quantiles), the relationship becomes more positive and pronounced, particularly in higher FRM quantiles. This indicates that Kuwait’s financial risk is more sensitive to external shocks when it is already experiencing elevated internal risks, highlighting the vulnerability of smaller or regionally concentrated markets like Kuwait to global financial turbulence.

Next, we estimate the 250-day rolling-window causality in quantiles to determine the causal relationships from the OFR FSI to the FRM and show them in Figure 4.

The rolling-window quantile causality plot shows the dynamic interplay between the OFR FSI for emerging markets and the FRM for Kuwait over time and across quantiles. The vertical axis represents the strength of causality (statistics), while the horizontal axes depict time (rolling-window basis) and quantiles, encompassing varying levels of financial stress and risk. Higher values on the statistics axis signal stronger causal links from the FSI to the FRM.

The plot indicates that during periods of heightened stress (observed in higher quantiles of the FSI), there is a stronger causal influence on financial risks in Kuwait, as shown by amplified statistics values. This pattern is more pronounced in specific timeframes, underlining that global market conditions and stress spillovers affect the causality intensity. Moreover, the asymmetric distribution of causality across quantiles and timeframes highlights the non-linear and time-varying nature of the relationship, indicating the sensitivity of financial system of Kuwait to external stress factors.

6. Discussion and Policy Implications

Our analysis provides several critical insights for policymakers in Kuwait and other resource-dependent nations. The FRM is not merely an academic measure; it is a practical early-warning tool that can inform preemptive policy action. The results indicate three key areas for policy focus:

• From Monitoring to Macroprudential Action: The strong correlation between the FRM, the VIX, and oil price volatility suggests that regulators should use the FRM as a dynamic trigger for macroprudential policy. For instance, when the FRM crosses a pre-defined crisis threshold, the Central Bank of Kuwait could automatically activate counter-cyclical capital buffers for banks, discouraging excessive risk-taking during periods of heightened systemic vulnerability. This would be a more targeted and timely approach than relying on static, calendar-based reviews.
• Informing Economic Diversification Strategies: The FRM quantifies the precise financial stability cost of over-reliance on oil. Policymakers can use the FRM’s sensitivity to oil shocks as a benchmark to measure the success of diversification initiatives under Kuwait’s “New Vision 2035.” A declining correlation between oil prices and the FRM over time would provide tangible evidence that the economy’s resilience is improving. This turns a broad strategic goal into a measurable financial stability objective.
• Enhancing Sovereign Wealth Fund Management: Our findings underscore the need for Kuwait’s sovereign wealth fund, the Kuwait Investment Authority (KIA), to play a more active role in mitigating systemic risk. During periods of sharp FRM spikes driven by external shocks, the KIA could be mandated to provide targeted liquidity support to systemically important domestic sectors, preventing a credit crunch and reinforcing market confidence. This would formalize the fund’s role as a systemic stabilizer of last resort.

By integrating FRM-based insights into its policy framework, Kuwait can move from a reactive to a proactive stance on financial stability, thereby strengthening its progress towards the UN Sustainable Development Goals, particularly SDG 8 (Decent Work and Economic Growth) and SDG 9 (Industry, Innovation, and Infrastructure).

7. Conclusions

This study provides a comprehensive and nuanced analysis of Kuwait’s Financial Risk Meter (FRM), offering critical insights into its sensitivity to global financial shocks, oil market dynamics, and regional geopolitical events. The findings reveal the intricate interplay between global and domestic risk factors that shape Kuwait’s financial stability, reflecting its unique position as an oil-dependent emerging market economy. By examining the co-movements of Kuwait’s FRM with key global indicators such as the VIX (a measure of global financial market volatility) and WTI crude oil prices, the study highlights Kuwait’s heightened vulnerability to external shocks, including global financial crises, oil price volatility, and pandemic-induced disruptions. Notably, the FRM exhibited significant spikes during periods of global and regional turmoil, such as the COVID-19 pandemic, underscoring the profound impact of such events on domestic risk perceptions and economic resilience.

The analysis also sheds light on the role of regional geopolitical tensions and domestic structural challenges in shaping Kuwait’s financial risk landscape. Events such as the Qatar diplomatic crisis and broader Gulf tensions have been shown to contribute to temporary but pronounced increases in the FRM, illustrating how geopolitical instability can exacerbate financial vulnerabilities. Moreover, domestic fiscal pressures, coupled with Kuwait’s heavy reliance on hydrocarbon revenues, further compound these risks, emphasizing the urgent need for sustained and strategic policy action to address these structural weaknesses.

The findings of this study underscore the critical importance of economic diversification and resilience-building as long-term strategies for enhancing Kuwait’s financial stability. Reducing the economy’s dependence on oil revenues, implementing robust fiscal reforms, and fostering private sector development are essential steps toward mitigating the financial risks identified in this analysis. Policymakers must also prioritize measures to enhance Kuwait’s capacity to absorb external shocks, such as strategic reserve management, targeted investments in innovation, and the development of non-oil sectors. These efforts will not only strengthen the domestic economy but also position Kuwait as a more resilient and dynamic player in the global economic landscape.

In addition to economic diversification, the study highlights the need for proactive measures to manage geopolitical risks. Given Kuwait’s strategic location in a volatile region, policymakers must adopt a forward-looking approach to regional diplomacy and conflict resolution, ensuring that the country remains a stable and reliable economic actor in the Gulf. Strengthening institutional frameworks for risk management and crisis response will also be crucial in safeguarding Kuwait’s financial stability against future shocks.

These results emphasize the need to bolster the economic resilience of Kuwait in the context of a sustainable development framework. Increased systemic risk monitoring would contribute to sustainable financial development, make the system more resilient against external shocks and be consistent with UN SDGs of economic growth, innovation and the strengthening of institutions. If incorporated into policymaking and regulatory policy, these lessons could help Kuwait’s system for weathering global financial shocks and moving toward a more sustainable and diversified economic future.

The Financial Risk Meter (FRM) emerges from this study as a valuable tool for monitoring and understanding financial risk in Kuwait. By providing timely and actionable insights into the vulnerabilities driven by global and regional factors, the FRM equips policymakers and stakeholders with the information needed to make informed decisions. Leveraging these insights, Kuwait has a unique opportunity to adopt forward-looking strategies that enhance economic resilience, promote sustainable growth, and navigate the challenges of an increasingly volatile global environment.

In summary, this study’s primary contributions are threefold: it successfully adapts and validates the FRM for a resource-dependent economy, demonstrates its superior performance over existing models like CoVaR in this context, and provides granular, actionable insights into the specific drivers of financial risk in Kuwait. These contributions fill a critical gap in the literature and provide a robust framework for policymakers in other emerging markets facing similar dual-shock vulnerabilities.

For future research, this framework can be extended to other resource-dependent nations in the GCC and beyond, creating a comparative basis for understanding systemic risk across emerging markets. Ultimately, by equipping policymakers with better tools, this study contributes to the broader goal of building a more resilient and sustainable global financial system.