Cryptocurrency Adoption and Financial Resilience: A Worldwide Fractional Probit Analysis and Institutional Moderation

Abstract

This article investigates the impact of grassroots cryptocurrency adoption—operationally defined as the intensity of on-chain retail transactions and peer-to-peer (P2P) exchange volumes—on household financial resilience. Utilising aggregate cross-sectional data from 112 countries, we employ a fractional probit regression to account for the bounded nature of our resilience index. The study highlights marked heterogeneity depending on the level of economic development. The results reveal a positive and significant effect in developing countries, whereas a negative association emerges in developed economies. Analysis of the underlying mechanisms indicates a significant moderating role for institutional quality. While cryptocurrency adoption shows a direct positive correlation with financial resilience in emerging markets, its contribution weakens in environments with robust formal institutions. These findings suggest that digital assets primarily function as a substitute for formal financial systems in contexts characterised by institutional voids and limited financial inclusion. Furthermore, the study identifies non-linear relationships between banking penetration and resilience, underscoring the importance of financial system maturity. Overall, the results suggest that cryptocurrency adoption can serve as a functional tool for strengthening worldwide resilience, provided it is supported by targeted regulatory oversight and digital financial education.

1. Introduction

Financial stress, which is the difficulty or pressure that an individual, household, or company experiences in fulfilling present financial obligations as a result of limited budgets, income shocks, or inflation, continues to impede the effectiveness of policy interventions. High levels of financial worry, especially, might diminish the impact of public aid schemes as families give survival top priority over long-term investments in human capital. Moreover, extensive financial hardship usually sparks populist revolt and political opposition, therefore postponing critical structural changes and depleting the budgetary space needed to support mental health and poverty reduction in developed as well as developing nations. Paying for medical bills, school fees, monthly expenses, agricultural production losses due to climate change, or health expenses related to ageing are all events that are part of everyday life for a significant portion of the world’s population. The 2021 Global Findex survey report indicates that approximately 95% of adults in sub-Saharan Africa are concerned about at least one of these issues, a higher percentage than in any other region of the world (Demirgüç-Kunt et al., 2020). Even developed countries are not immune to the adverse effects of household financial stress. For example, Islam & Chowdhury (2024) show that a one-unit increase in financial stress increases the risk of falling below the poverty line in Canada by 17%. Furthermore, there is a broad consensus on the negative effects of financial stress on the mental health of households, including that of children (Bialowolski et al., 2021; Chai & Lu, 2025).

In the economic literature, financial resilience is seen as one of the ways to prevent financial stress. Financial resilience1 for an individual or household is the dynamic capacity to absorb, cope with, and recover from unexpected financial shocks like sudden income loss, health crises, or climate-related disasters without having to sell productive assets or lower essential consumption. Particularly for small and medium-sized businesses (SMEs), financial stability is the structural capacity to retain institutional liquidity and day-to-day operations after a crisis. This is accomplished by carefully using official financial reserves, including liquid savings, varied credit lines, insurance products, or other official financial cushions. Unlike simple solvency, resilience stresses the speed and quality of recovery, therefore guaranteeing that a short-term shock does not result in a permanent halt of economic activity. Empirical studies have examined the role of digital financial service innovations (Akpa & Gnidehou, 2025; R. Verma & Chatterjee, 2025; Zou et al., 2024). However, these studies have reached divergent conclusions. One strand of this literature confirms the positive effect of digital finance on financial and economic resilience (Du et al., 2023; Zou et al., 2024). In contrast, other studies conclude that there is no effect or that the effects are weak and comparable to those of traditional finance (R. Verma & Chatterjee, 2025). It should be noted that these different studies use several measures of digital finance, such as access to digital credit, ownership of a digital wallet and, particularly over the last decade, the use of financial services via mobile phones. However, few studies have examined the specific role of cryptocurrency, which is based on blockchain technology, operating outside traditional financial systems and not controlled by any central authority.

Cryptocurrency has experienced significant growth over the last decade, both in developed and developing countries (El Hajj & Farran, 2024). The 2023 Global Crypto Adoption Index report indicates that the global adoption rate of cryptocurrency has risen from less than 10% in 2020 to around 35% in 2023 (Chainalysis, 2023). A breakdown by income subgroup reveals more pronounced growth in cryptocurrency in developing countries, where the adoption rate rose from less than 10% in 2020 to 40% in 2023, an increase of 30 percentage points. In contrast, growth has been more moderate in high-income countries, where the adoption rate rose from 35% in 2020 to 50% in the first quarter of 2023, an increase of 15 percentage points (Chainalysis, 2023). This momentum for cryptocurrency in developing countries, if sustained, could be decisive as a tool for improving the financial resilience of households and businesses. However, this requires the production of robust evidence that can effectively guide development policies.

The existing literature on cryptocurrency focuses mainly on adoption constraints, price volatility and market functioning (Dang et al., 2025). One major exception is the recent study by Beavers and Godek (2024) conducted in the United States. These authors show that households holding cryptocurrencies experienced lower levels of financial stress compared to households that did not hold cryptocurrencies. In particular, households holding cryptocurrencies were less likely to be food insecure and had a greater ability to cope with various utility bills and medical expenses.

Theoretically, there is controversy over the effect of cryptocurrency on financial resilience. According to the mainstream literature on cryptocurrency, inspired by behavioural finance theories of over-financing (Barber & Odean, 2001), speculative trading theories, particularly noise trading (De Long et al., 1990), and speculative bubble models (Shiller, 2017), cryptocurrency holders are perceived as risk-seeking agents, subject to cognitive biases and primarily motivated by short-term speculative gains rather than fundamental or risk management considerations. On the other hand, another branch of the literature, rooted in the theories of financial innovation initiated by Schumpeter (1934) and further developed by Levine (2005), as well as in the work on financial inclusion developed in particular by Demirgüç-Kunt et al. (2020), considers cryptocurrency as an instrument capable of improving the economic well-being and financial resilience of households. Furthermore, approaches based on risk sharing and consumption smoothing theories, such as those of Townsend (1994), highlight the role of informal mutualisation mechanisms in mitigating economic shocks, mechanisms that digital technologies can reinforce by facilitating rapid transfers of resources. Finally, liquidity demand models, notably that of Baumol (1952), suggest that digital assets can be liquid alternatives for storing value, particularly relevant in contexts characterised by macroeconomic instability and failures of the traditional financial system.

The link between grassroots cryptocurrency acceptance and household resilience in the face of economic shocks is examined in this paper. Although Beavers and Godek (2024) offer basic proof that cryptocurrency ownership relates to perceived financial well-being among US households, their study is restricted to a mature, high-trust institutional setting. Our study fills a crucial research void by posing the question: How does the effect of cryptocurrency adoption on financial resilience differ across diverse institutional and economic environments? The Household Financial Resilience Index is the focus of our study. Our main goal is to test the Institutional Substitution Hypothesis, which states that cryptocurrencies offer more benefits in countries where traditional financial institutions do not offer enough support. Our research has a double scientific value. Firstly, we consider the global diversity in institutions. We use the global variety to pinpoint the income-threshold effect in addition to raising the sample size to 112 nations. We show that the link between crypto-adoption and resilience is non-linear and changes direction depending on the stage of development of a nation ($13,845 GDP per capita). Indeed, the adoption of cryptocurrency depends on various parameters like financial literacy and risk appetite, financial vulnerability with speculative mechanisms, and public regulation and financial opportunities with digital and stock market development (Aiello et al., 2023; Lanciano et al., 2026). These factors imply an upward selection bias, where inherently more financially capable households are also more likely to adopt crypto, potentially overstating its positive contribution to financial resilience and opportunities in a developed market.

Secondly, we evaluate the role of institutional quality as a moderating Element. As a moderating factor, we use institutional quality (evaluated using World Bank Governance Indicators). Based on the reviewed and utilised sources, this manuscript represents one of the first studies addressing this issue. Methodologically, we use fractional probit regression (Papke & Wooldridge, 1996). Our dependent variable, the Financial Resilience Index, ranges from 0 to 1. In contrast to OLS, which posits linearity and can yield predictions beyond the logical range, fractional regression employs a probit link function to guarantee that all estimates fall within the unit interval while capturing the decreasing marginal returns of adoption at greater levels of resilience.

The rest of the article is organised as follows: Section 2 reviews the literature, while Section 3 presents the research methodology. Section 4 reports and discusses the main findings. Finally, Section 5 offers concluding remarks.

2. Literature Review

Cryptocurrencies, as decentralised digital assets, influence financial resilience through various mechanisms, both positive and negative. The literature highlights the impacts on financial inclusion and portfolio diversification, but also the risks of volatility, contagion and systemic destabilisation.

2.1. Improving Resilience Through Inclusion and Diversification

The first key mechanism is the improvement in financial inclusion and access to services for the unbanked, enabling them to access transactions, savings and investments without traditional intermediaries. Suri et al. (2021) show that digital fintech loans such as M-Shwari significantly improve households’ resilience to shocks, without creating substitution with other forms of credit. The authors use a regression discontinuity design around the credit score threshold, comparing households just above (eligible for the loan) and just below (ineligible). They show that eligible households are 6.3 per cent less likely to forego essential expenses (education, food, health, etc.) in the face of a negative shock (illness, death, job loss, etc.).

The study by Win et al. (2025) highlights the theory of complex networks to analyse risk contagion and the resilience of the crypto network, going beyond traditional analyses focused on Bitcoin. Using a DCC-GARCH (Dynamic Conditional Correlation Generalised Autoregressive Conditional Heteroskedasticity) model, the authors show that the crypto market is developing increasing robustness as it matures, distinguishing between leading cryptocurrencies (e.g., Bitcoin) that act as net risk receivers, absorbing shocks rather than propagating them, and smaller or highly active cryptos that sometimes accelerate contagion. This study, like that of Krause (2025), confirms that the technical functioning of cryptocurrencies is inherently risky and draws lessons from major crashes such as the Terra (Luna) ecosystem in May 2022. Aslanidis et al. (2021) show that correlations between cryptocurrencies are positive and vary over time, but tend to increase overall. The main lesson from this study is that cryptocurrencies are gradually losing their internal diversification benefits, increasing systemic and contagion risks. Cryptocurrencies are a powerful tool for financial inclusion in developing countries, particularly for vulnerable households without access to traditional banking.

A second mechanism is their role as a hedge or safe haven during crises. Cryptocurrencies, such as Bitcoin, act as diversification tools, attracting investors during global disruptions (COVID-19 pandemic, Russia–Ukraine war or Israel–Palestine conflict) (Ballis et al., 2025). Their trading volume increases during these periods, indicating a perception of them as a stable long-term digital asset, despite short-term fluctuations. This contributes to portfolio resilience by reducing dependence on traditional markets, with evidence of Bitcoin price stability in the face of shocks, promoting investor confidence. In addition, stablecoins (cryptocurrencies backed by fiat assets such as the USD) provide essential liquidity in DeFi exchanges, accounting for approximately 45% of the liquidity of decentralised platforms, and act as a ‘safe haven’ against the volatility of other cryptocurrencies (Liao & Caramichael, 2022). Automated liquidations in DeFi (Decentralised Finance) are an essential automatic protection mechanism for decentralised lending/borrowing protocols.

Finally, innovation through DeFi enhances resilience by offering decentralised tools for lending, borrowing and earning interest without central banks (Minarni, 2025). This transfers control to individuals, promoting economic stability in unstable environments, and integrates technologies such as blockchain for increased transparency and reduced systemic dependencies. El Hajj and Farran (2024) show that cryptocurrencies represent a major opportunity to promote financial inclusion and economic empowerment in emerging markets. These authors demonstrate, through structural equation modelling (SEM), that the adoption of cryptocurrencies significantly improves financial inclusion (FI), user satisfaction (US), trust in institutions (TFIs) and perceived economic empowerment (PEE), particularly in developing countries.

Badawi and Jourdan (2020) offer insights into the threats and defensive mechanisms of cryptocurrencies, highlighting their potential role in diversifying and strengthening the resilience of individual portfolios. This literature review is summarised in the

Table 1. Positive mechanisms: improving resilience through inclusion and diversification.

Study	Methodology	Main Results
Minarni (2025)	Literature review and qualitative analysis of cryptocurrency/DeFi adoption impacts on financial stability and inclusion in emerging markets.	DeFi innovation enhances resilience by providing decentralised tools for lending, borrowing, and earning interest without relying on central banks or traditional intermediaries. Promotes individual control, transparency (via blockchain), and economic stability in unstable environments. However, also notes volatility risks that can undermine stability if unregulated.
Win et al. (2025)	Complex networks analysis; DCC-GARCH (Dynamic Conditional Correlation GARCH) model for risk contagion in the crypto market (beyond Bitcoin).	Crypto market gains robustness as it matures. Leading cryptos (e.g., Bitcoin) act as net risk receivers (absorb shocks), while smaller/highly active ones can accelerate contagion. Confirms inherent technical risks, with lessons from crashes like Terra/Luna (2022).
Krause (2025)	Analysis of algorithmic stablecoin mechanisms and ecosystem failures (case study approach focused on technical design flaws).	Confirms that the technical functioning of cryptocurrencies is inherently risky. Draws key lessons from major crashes, such as the Terra (Luna) ecosystem collapse in May 2022.
Ballis et al. (2025)	Analysis of trading volumes, price behaviour, and safe-haven properties during global crises (e.g., COVID-19, Russia–Ukraine war, Israel–Palestine conflict).	Cryptocurrencies (esp. Bitcoin) attract investors during disruptions, with increased trading volumes. Act as diversification tools and perceived stable long-term digital assets, reducing dependence on traditional markets and enhancing portfolio resilience.
El Hajj and Farran (2024)	Structural equation modelling (SEM) on survey/construct data: cryptocurrency adoption (CA), financial inclusion (FI), user satisfaction (US), trust in institutions (TFIs), perceived economic empowerment (PEE). Focused on emerging markets.	Cryptocurrency adoption significantly improves FI, US, TFI, and PEE in developing countries. Positive relationships extend through interaction effects, promoting economic empowerment and inclusion.
Liao and Caramichael (2022)	Review/analysis of stablecoin roles in DeFi, collateral backing, and liquidity provision.	Stablecoins provide ~45% of liquidity in decentralised platforms and act as a “safe haven” against other crypto volatility. Support automated liquidations in DeFi as protection mechanisms. Dollar-pegged stablecoins can serve as a digital safe-haven currency during crypto distress.
Aslanidis et al. (2021)	Analysis of conditional/dynamic correlations (generalised DCC-class models) among cryptocurrencies, stocks, bonds, and gold.	Positive correlations among cryptocurrencies that vary over time but tend to increase overall. Cryptos are gradually losing internal diversification benefits, raising systemic/contagion risks. Negligible correlations with traditional assets in some periods.
Suri et al. (2021)	Fuzzy regression discontinuity (RD) design around M-Shwari credit score threshold; administrative + survey data from Kenya. Compares households just above vs. below the eligibility cutoff.	Digital fintech loans (e.g., M-Shwari) improve household resilience to shocks (illness, death, job loss, etc.). Eligible households are 6.3 per cent less likely to forego essential expenses (food, education, health). No substitution with other credit forms. Enhances financial access for the unbanked without creating over-indebtedness.
Badawi and Jourdan (2020)	Systematic literature review of emerging threats and defensive mechanisms in cryptocurrencies.	Highlights threats (e.g., cyber risks due to pseudo-anonymity) alongside defensive strategies. Underscores cryptocurrencies’ potential for portfolio diversification and strengthening individual resilience.

.

2.2. Resilience Challenge: Risks of Destabilisation and Contagion

Despite the advantages of cryptocurrencies, these assets can erode financial resilience through their extreme volatility and speculative risks (BIS, 2023). Price fluctuations, generated by speculation and market conditions, lead to massive losses, eroding confidence and creating bubbles or crashes. This is particularly damaging in fragile economies, where risky behaviour amplifies vulnerabilities (Minarni, 2025). For example, small investors often lose out to large investors, as seen in the collapses of TerraUSD (a loss of $19 billion in May 2022) and FTX.

Another mechanism is systemic contagion and interconnection with traditional finance (ESRB, 2025). The crypto ecosystem amplifies traditional vulnerabilities through automated liquidations in DeFi. This leads to spillovers, as in the case of crypto-related bank failures. For example, Silvergate Bank experienced a major crisis in 2023 due to its massive exposure to the cryptocurrency sector. The bankruptcy of FTX, one of its main clients, triggered massive withdrawals of deposits, leading to the voluntary liquidation of the bank. This case illustrates the systemic risks of crypto transmission to the financial stability of institutions. In terms of financial resilience, this situation is consistent with the theoretical mechanisms discussed: trust channel (post-FTX panic) and sectoral exposures (Waren, 2025). It demonstrates how crypto volatility erodes banking resilience without a safety net, validating the ECB/IMF models on contagion and disintermediation.

Shahzad et al. (2025) caution against overestimating the safe-haven role of stablecoins and show that stablecoins often absorb shocks without transmitting them, unlike oracle tokens. Lee et al. (2023) and Santiago et al. (2025) highlight the mechanisms of massive contagion and propose lessons for the resilience of stablecoins. To combat systemic risks, these authors propose the use of pegging, which is a mechanism for stabilising the value of a digital currency by linking it to a more stable asset, such as the US dollar or a commodity such as gold. This approach is particularly useful for reducing volatility and protecting against shocks. Bouri et al. (2023) examine the systemic risk spillovers from the FTX Token (FTT) to major cryptocurrencies during the collapse of FTX in November 2022. They use a lower tail dependence (downside risk) approach and models such as the Rotated Gumbel copula to capture asymmetric contagion in extreme conditions.

Regulatory and operational gaps are a third negative mechanism. These gaps are the result of a lack of consistent frameworks, which exposes the system to fraud, money laundering and manipulation, thereby rendering monetary policy ineffective. This situation is found during prolonged periods of significant decline in financial markets (bear markets), creating pessimism among investors (Conlon & McGee, 2020; Corbet et al., 2020). To remedy this instability, Pacelli et al. (2025) show that the interconnections between crypto and traditional markets increase the implications for overall financial stability, regulatory policies and risk management. Thus, crypto shocks can spread rapidly to traditional equities (and vice versa), with more pronounced downside effects (Wei et al., 2025). This work shows that smaller cryptos sometimes accelerate contagion, and then the market develops robustness in the face of shocks. This literature review is summarised in

Table 2. Resilience challenges: risks of destabilisation and contagion.

Study	Methodology	Main Results
Minarni (2025)	Qualitative literature review on cryptocurrency adoption and financial stability in emerging markets.	While crypto aids inclusion and DeFi innovation (lending/borrowing without central intermediaries), extreme volatility and speculation erode resilience in fragile economies. Amplifies vulnerabilities via risky behaviour; highlights regulatory challenges.
Shahzad et al. (2025)	Analysis of safe-haven properties, shock absorption, and transmission (likely correlation or regime-based models).	Cautions against overestimating stablecoins as safe havens. Stablecoins often absorb shocks without transmitting them, but their role is nuanced.
Lee et al. (2023) and Santiago et al. (2025)	Information theory/spillover analysis (Lee); lessons on stablecoin resilience and depegging risks (Santiago).	Massive contagion mechanisms in stablecoin failures (e.g., Terra/Luna). Propose pegging (linking to stable assets like USD or gold) to reduce volatility and protect against shocks. Highlight lessons from depegging events.
Waren (2025)	Likely theoretical or empirical analysis of sectoral exposures and contagion channels.	Examines sectoral exposures as a mechanism for crypto shocks affecting institutional resilience. Consistent with trust channel (post-FTX panic) and broader theoretical models of crypto-traditional finance interconnections.
Santiago et al. (2025)	Dynamic Conditional Correlation (DCC-GARCH) and related models for analysing contagion during events like Terra-Luna collapse; lessons on stablecoin resilience (drawing from volatility correlations and depegging analysis).	Highlights mechanisms of massive contagion in stablecoin failures (e.g., algorithmic stablecoins like UST). Proposes lessons for improving resilience of stablecoins, including design improvements to mitigate death spirals, volatility transmission, and spillovers to other crypto assets and traditional markets.
Conlon and McGee (2020)	Event study/safe-haven analysis during the COVID-19 bear market; comparison of Bitcoin returns with traditional assets (e.g., S&P 500).	Bitcoin does not act as a safe haven during acute bear markets; instead, it decreases in price in lockstep with equities. Contributes to investor pessimism and highlights heightened risks during prolonged market declines.
Bouri et al. (2023)	Lower tail dependence (downside risk) approach; Rotated Gumbel copula and GARCH copula quantile regression-based ΔCoVaR models.	Significant downside risk spillovers from FTX Token (FTT) to major cryptocurrencies (strongest to Solana due to staking exposure).
Pacelli et al. (2025)	Analysis of interconnections, tail risk connectedness (e.g., TENET networks or spillover models) between crypto and traditional markets.	Interconnections increase implications for financial stability, regulation, and risk management. Crypto shocks spread rapidly to traditional equities (and vice versa), with stronger downside effects.
Wei et al. (2025)	Contagion and spillover analysis across crypto sizes/maturities and traditional assets.	Smaller cryptos sometimes accelerate contagion; overall market develops robustness over time. Pronounced downside spillovers to traditional markets.
Corbet et al. (2020)	Analysis of liquidity–volatility interrelationships, contagion, and market dynamics during COVID-19 crisis periods (often using GARCH-type or correlation models).	During prolonged bear markets and crises, cryptocurrencies exhibit increased volatility and liquidity pressures, amplifying pessimism among investors. Shows interactions between crypto volatility and broader market panic, with limited safe-haven properties.

.

2.3. The Institutional Substitution Framework: Crypto-Assets as Informal Reagents

Based on North’s (1990) institutional economics and Levine’s (2005) functional perspective of finance, we present a conceptual framework whereby the acceptance of cryptocurrencies is a logical reaction to institutional gaps. The main driver of financial exclusion in this model is institutional weakness (defined by low Regulatory Quality, high corruption, and inadequate Rule of Law). Households are driven into the informal economy when official banks are seen as untrustworthy or unreachable. Cryptocurrency adoption operates as a digital replacement for official financial infrastructure in this void. Several alternative mechanisms mediate this relationship. Firstly, remittance efficiency in corridors where conventional banking costs are too high, crypto-assets offer a less expensive “digital bridge”, therefore increasing household liquidity directly. Secondly, hedges to stop money from leaving and rising prices in digital assets operate as a decentralised store of value in unstable macroeconomic settings, so shielding household buying power from the “institutional tax” of local currency depreciation. Thirdly, enhancing informal risk-sharing because cryptocurrencies expand the reach of conventional informal networks such as tontines or community savings organisations by facilitating quick, cross-border peer-to-peer (P2P) transactions without an intermediary.

In the end, these systems come together to generate financial resilience, therefore enabling the household to keep consumption levels throughout shocks. Our research hypotheses are formulated as follows:

H1. 

Cryptocurrency adoption has a statistically significant effect on financial resilience.

H2. 

The effect of cryptocurrency adoption on financial resilience is stronger in countries with lower institutional quality.

H3. 

Institutional quality moderates the relationship between cryptocurrency adoption and financial resilience.

H4. 

The impact of cryptocurrency adoption on financial resilience is amplified under financial stress conditions.

3. Methodology

This section outlines the conceptual research framework, presents the formal model specification, and discusses the data sources and construction of the sample.

3.1. Research Framework

The study adopts a cross-sectional quantitative design to examine the relationship between cryptocurrency adoption and financial resilience across countries. The analysis relies on 2023 data (the closest available year) drawn from a cross-sectional sample of 112 countries comprising both developed and developing economies. Data for each variable were retrieved from official and reputable international databases (detailed in

Table A1. Data and sources.

Variable	Description	Source
Financial resilience	Financial Resilience Index	Global Findex Database
Crypto adoption	Percentage of the population that uses cryptocurrency	(Chainalysis, 2023)
Account ownership	Percentage of the population that has a bank account	World development indicators (World Bank)
GDP per capita	Per capita GDP	World development indicators (World Bank)
Lower secondary completion rate	Percentage of the population with secondary education	World development indicators (World Bank)
Financial development	Domestic credit to private sector	World development indicators (World Bank)
Unemployment	Percentage of the population that is unemployed	World development indicators (World Bank)
Inflation	Consumer price inflation	World development indicators (World Bank)
Age dependency ratio	Dependency ratio of young and older people on working people	World development indicators (World Bank)
Share of agricultural sector	Value added from agriculture and forestry	World development indicators (World Bank)
Internet secure servers	Number of internet security devices	World development indicators (World Bank)
Corruption	Corruption control index	World governance indicators (World Bank)
Government Effectiveness	Governance Effectiveness index	World governance indicators (World Bank)
Regulation	Regulatory Quality index	World governance indicators (World Bank)
Government accountability	Government accountability and integrity index	World governance indicators (World Bank)
Rule of Law	Law and rule compliance index	World governance indicators (World Bank)

in Appendix A). The key dependent variable—financial resilience—was obtained from the Global Findex Database (World Bank). This is a composite index (ranging from 0 to 1) constructed from three main dimensions.

Countries are classified as Emerging and Developing Economies if their GNI per capita is below $13,845, whereas developed economies consist of high-income countries with GNI per capita above this threshold. Limited missing observations were either imputed or dropped following the exclusion rule.

Comprehensive descriptive statistics were computed and presented in

Table 3. Summary statistics.

Variable	Obs	Mean	Std. Dev.	Min	Max
Financial resilience	112	63.615	41.584	0.000	99.402
Crypto adoption	112	1.976	2.278	0.000	13.460
Account ownership	112	70.405	22.893	14.827	99.855
GDP per capita	112	15,077	18,446	263	86,102
Lower secondary education completion rate	112	78.375	22.206	8.887	107.969
Financial development	108	57.819	44.810	3.727	234.975
Unemployment	105	7.696	4.910	0.336	26.412
Inflation	112	4.570	6.961	0.115	64.028
Age dependency ratio	112	14.739	10.038	1.889	47.084
Share of agriculture in the economy	112	9.378	9.030	0.021	36.368
Secure internet servers per million	112	11,410	23,128	1	139,896
Corruption	112	51.464	29.694	2.765	99.654
Government Effectiveness	112	52.046	28.738	1.253	99.222
Regulatory Quality	112	51.570	29.279	4.625	98.139
Voice and Accountability	112	49.860	29.363	2.678	99.957
Rule of Law	112	51.178	29.063	2.341	98.410

. The empirical analysis is conducted using a Fractional Probit Model. For diagnostic tests, multicollinearity checks (VIF), heteroskedasticity tests, normality, and model specification tests were performed. We verify the robustness of our results by conducting sub-sample analyses across developed and emerging economies, testing alternative specifications, and performing multiple sensitivity checks.

This structured, sequential approach ensures transparency, reproducibility, and methodological rigour in assessing the impact of cryptocurrency adoption on financial resilience. The analysis is conducted using Stata 18 as the primary statistical software.

3.2. Data and Variables

The study covers a cross-sectional sample of 112 countries (developed and developing) in 2023. The country classification follows the World Bank Atlas Method (2024). The sample is divided into the following categories: Development and Emerging Economies: Nations having a GNI per capita under $13,845. Developed Economies: High-income nations surpassing this threshold. We excluded countries lacking data for more than one main indicator. The data source for each study variable is presented in Table A1 in Appendix A, and Table A1 presents the descriptive statistics of the study variables.

The first main independent variable is the adoption of cryptocurrencies. We rely on the Chainalysis Global Crypto Adoption Index (2023) to measure the adoption of cryptocurrencies. This index evaluates grassroots adoption, that is, the degree to which regular citizens in a nation are really utilising cryptocurrency for transactions and storage of value. Chainalysis (2023) claims that the index combines three main variables, each weighted by the Purchasing Power Parity (PPP) per capita of the nation, to emphasise areas where crypto-assets are most ingrained in the everyday financial life of inhabitants. First, on-chain Value Received (Total): Reflects the general scope of the crypto-economy by measuring the aggregate volume of transactions delivered to cryptocurrency addresses in a nation. Second, on-chain Retail Value Received: It particularly monitors transfers of less than $10,000, substituting for the behaviour of non-professional, individual consumers. Lastly, P2P Exchange Trade Volume: Records the volume of peer-to-peer transactions, a vital indicator of adoption in developing nations where formal exchange access could be limited.

The second main independent variable is the institutional quality, measured through a set of six indicators, namely: Corruption Control, Government Effectiveness, Political Stability, Rule of Law, Regulatory Quality, and Voice and Accountability. The data comes from the World Bank Governance Indicators (WGIs).

The dependent variable in this study is financial resilience, measured using data from the World Bank’s Global Financial Inclusion Database (Global Findex). Within the Global Findex framework, financial resilience refers to the ability of individuals and households to cope with unexpected financial shocks without incurring significant welfare losses (Demirgüç-Kunt et al., 2020). The emergency funds indicator captures respondents’ ex ante perceived capacity to mobilise resources within a short time horizon. Importantly, it does not restrict the source of funds (e.g., savings, borrowing, or informal support), thereby providing a comprehensive measure of households’ overall financial coping capacity. The measure of financial resilience used in this study is directly drawn from the Global Findex dataset. It is based on respondents’ answers to the following survey question: “If you had to come up with [local currency equivalent of USD 1/20 of GNI per capita] in the next 30 days for an emergency, how possible would it be for you to come up with this amount?”. The specific indicator corresponds to the variable “Coming up with emergency funds in 30 days: possible”, which is coded as a binary variable equal to 1 if the respondent reports that mobilising such funds is possible, and 0 otherwise. The individual-level binary responses are aggregated at the country level, resulting in a proportion that reflects the share of respondents able to mobilise emergency funds within 30 days. This country-level ratio is directly available in the Global Findex dataset and is used in the empirical analysis. It is worth noting that we do not apply any transformation, normalisation, weighting, or index construction to this variable; it is used as provided in the dataset.

Control variables include GDP per capita, lower secondary education completion rate, financial development, unemployment rate, inflation rate, age dependency ratio, share of agriculture in the economy, and secure internet servers per million. Data on all the control variables come from the World Bank Development Indicators (WDIs).

The variables presented in Table 3 provide several insights. The variable financial resilience exhibits a moderate average resilience, accompanied by significant variability, as indicated by a standard deviation nearly equivalent to the mean. Some countries have near-zero resilience, while others score near 100. This means that shocks can change and affect resilience. The variable crypto adoption exhibits low average adoption (mean ~2 on the scale used), with moderate spread. Most countries show limited adoption, but a few outliers reach high levels (13). In general, crypto is still pretty new, but it is growing quickly in some areas. The variable account ownership exhibits a relatively high average (~70%) but a wide range. Many countries have near-universal ownership, while others lag below 15%. This indicates financial inclusion gaps that crypto could help bridge, especially where traditional banking is weak, such as in regions with limited access to banking services or high transaction costs. The variable GDP per capita exhibits large income disparities (from very low-income to high-income countries). A high standard deviation indicates the sample variations between developing and developed nations, which is beneficial for assessing whether cryptocurrency advantages are more pronounced in lower-GDP environments. The variable lower secondary education completion rate exhibits a solid average completion rate, but extreme variation. Low rates in some countries may limit the adoption of technologies (including crypto), while high rates support digital literacy and resilience. The unpredictable financial development has significant fluctuation and includes extreme values (maximum > 200). This suggests that traditional finance varies in depth; in areas with low development, crypto may serve as either a complement or a substitute. The unemployment variable exhibits a moderate mean with significant dispersion. Higher unemployment could correlate with a greater need for alternative tools, like crypto, for resilience. The variable inflation is generally low, but with high outliers (hyperinflation risks). Crypto adoption may serve as a hedge in high-inflation environments. The variable age dependency ratio reveals a relatively low average, but variation indicates differing demographic pressures on economic resilience. The variable share of agriculture in the economy exhibits a low on average (modern economies) but is higher in some (developing countries), where financial inclusion challenges are acute. The variable secure internet servers per million exhibits an extremely high dispersion. The disparity in digital infrastructure is a prerequisite for cryptocurrency adoption. Insufficient values may limit advantages in some countries.

The variable corruption reveals that all governance variables show means near 50 with large standard deviations. This indicates a balanced but highly polarised sample: some countries have strong institutions, others weak. Weak governance may amplify crypto-related risks or make decentralised tools more appealing for resilience.

3.3. Econometric Model

3.3.1. Baseline Model

Our variable of interest is the financial resilience indicator, which ranges from 0 to 1 and is retrieved from the Global Findex Database indicators. This is a composite index that uses three main dimensions to gauge people’s dynamic ability to weather unforeseen financial shocks:

Raising Emergency Funds: This is the main sign of resilience, measured by the respondent’s reported ability to get a certain amount of money (usually 1/20 of the GNI per capita) in 30 days. It assesses the liquidity a household has right away to get through a crisis. Savings Behaviour: This dimension evaluates whether people saved money last year. Reflecting the household’s proactive approach to reducing future hazards, it differentiates between formal savings (at a financial institution or by mobile money) and informal means.

Access to Financial Services: This component evaluates the reach of the formal financial system, including account ownership and the use of digital payments. In accordance with the Global Findex approach, access to these services offers the structural rails required to get government payments or credit during times of economic distress.

To ensure the index is on a standardised scale for econometric analysis, each proxy is first normalised to a range of [0, 1]. We then apply an average to aggregate these indicators. This data-driven approach allows the weights of each component to be determined by their contribution to the total variance of the dataset, ensuring that the resulting financial resilience score accurately reflects the multifaceted nature of financial security.

The Chainalysis Global Crypto Adoption Index (2023) provides the basis for the main independent variable, cryptocurrency adoption. In line with the Chainalysis approach, this variable does not reflect the actual percentage of the population of a nation that ‘owns’ cryptocurrency since such statistics are frequently either unavailable or untrustworthy given the secretive nature of the blockchain. Rather, it evaluates grassroots adoption, that is, the degree to which regular citizens in a nation are really utilising cryptocurrency for transactions and storage of value. Chainalysis (2023) claims that the index combines three main variables, each weighted by the Purchasing Power Parity (PPP) per capita of the nation, to emphasise areas where crypto-assets are most ingrained in the everyday financial life of inhabitants. First, on-chain Value Received (Total): Reflects the general scope of the crypto-economy by measuring the aggregate volume of transactions delivered to cryptocurrency addresses in a nation. Second, on-chain Retail Value Received: It particularly monitors transfers of less than $10,000, substituting for the behaviour of non-professional, individual consumers. Lastly, P2P Exchange Trade Volume: Records the volume of peer-to-peer transactions, a vital indicator of adoption in developing nations where formal exchange access could be limited.

Rather than using standard linear regression models, it is preferable to opt for fractional regression, a method proposed by Papke and Wooldridge (1996). This approach effectively handles constrained variables without the need to artificially modify extreme values. This technique offers the possibility of directly modelling a dependent variable between 0 and 1. It thus circumvents the unrealistic predictions that would be generated by standard linear models. Unlike ordinary least squares methods applied to transformed data, this approach effectively handles extreme values without requiring specific adjustments, thereby ensuring an unbiased assessment of financial resilience (Hegarty & Whelan, 2022; Ramalho et al., 2011). The empirical results indicate a marked divergence in the impact of cryptocurrency on financial resilience across income groups. On a global scale, the effect is not statistically significant, suggesting that aggregate analyses may obscure significant regional variations. In developed countries, the adoption of cryptocurrency has been demonstrated to have a considerable adverse effect, suggesting that it may be indicative of speculative behaviour rather than contributing to resilience. Conversely, in developing countries, the effect is positive but only marginally significant, suggesting a potential grassroots utility of digital assets in contexts with restricted access to formal financial systems. Despite the robustness of this result across model specifications, a cautious interpretation is warranted, given the comparatively lower level of statistical confidence and the necessity for additional longitudinal evidence.

Furthermore, as Mullahy (2015) points out, this model stands out for its robustness, as it does not impose rigid constraints on the distribution of the variable, a major advantage given the typical heteroscedasticity of financial well-being indicators. Finally, according to Cook et al. (2008), this methodological option allows for a more in-depth understanding of non-linear marginal effects, accurately capturing, in our case study, the concrete impact of cryptocurrency use on individuals’ resilience. The dependent variable in this study, the Financial Resilience Index (FRI), is a composite score bounded within the unit interval [0, 1]. In order to estimate the parameters in Equation (1), it is necessary to employ a Fractional Probit Model rather than a standard OLS or binary probit. The model is defined as follows:

$$E({FRI}_i|X_i)=\Phi (X_i\beta )$$

The following investigation will ascertain the value of the cumulative distribution function Φ(.) of the standard normal distribution. The vector of independent variables, which includes cryptocurrency adoption, institutional quality and electricity costs, is denoted by X_i. The values β are estimated using something called Quasi-Maximum Likelihood Estimation, or QMLE for short. A binary probit model is used to predict the probability of success. A Fractional Probit Model is used to predict the mean of a fraction. There are several reasons why this approach is better. Firstly, boundedness ensures that all the values it predicts for how financially resilient a company is stay exactly between 0 and 1, which OLS cannot do. Secondly, the non-linearity explains why it is harder to increase resilience from 0.9 to 1.0 than from 0.4 to 0.5. Finally, the robustness of robust QMLE standard errors means the model always provides reliable estimates, even if the underlying distribution of the fraction is not strictly normal. The basic model is specified as follows:

$$E(y_i|X_i)=\Phi ({\beta }_0+{\beta }_1{\mathrm{Crypto}}_i+\gamma X_i+{ϵ}_i)$$

(1)

with:

• $y_i$ is the financial resilience score of country i;
• ${\mathrm{Crypto}}_i$ is the cryptocurrency adoption indicator;
• $X_i$ is a vector of control variables;
• $\Phi$ represents the cumulative link function (probit).

The potential endogeneity of the cryptocurrency adoption indicator poses a significant econometric challenge in this analysis. It is imperative to address endogeneity in order to ensure the consistency of our Quasi-Maximum Likelihood estimates, given that we choose the fractional regression model (Papke & Wooldridge, 1996) to reflect the limited range of the Financial Resilience Index $(0\le FRI\le 1)$. The hypothesis suggests that the link may be subject to reverse causality, implying that homes with greater underlying resilience may possess the necessary financial resources to engage with digital assets. Furthermore, unobserved changes in local digital literacy or institutional mistrust may concurrently affect resilience ratings for homes and the likelihood of using crypto-assets. In the event of such endogeneity, there is a possibility of bias in the fractional probit/logit coefficients. In order to mitigate the impact of these current shocks, we have implemented a range of remedial measures, including the application of lagged values of adoption or a Control Function (CF) approach.

A critical econometric challenge in this study is the potential for reverse causality, namely the possibility that households in financially fragile or low-resilience environments adopt cryptocurrencies precisely as a defensive reaction to systemic instability. Furthermore, the presence of institutional weakness has the potential to act as a confounding variable, thereby simultaneously driving both high crypto-adoption and low financial resilience. Incorporating institutional quality (WGI) and inflation as primary regressors, as opposed to error-term components, serves to block the backdoor path in which weak governance drives both variables. This approach enables the isolation of the marginal effect of cryptocurrency adoption, which is independent of the general institutional decay.

Although a formal Instrumental Variable approach is still considered the gold standard, finding a universally legitimate instrument that satisfies the exclusion criterion for both crypto-adoption and financial resilience remains a major difficulty in the current research. The technical character of the Chainalysis (2023) dataset serves to compound these concerns. Given that the index predominantly documents on-chain activity, it is conceivable that it may understate off-chain transactions conducted within centralised exchanges or those protected by privacy technologies. Furthermore, despite the fact that the PPP weighting facilitates cross-country comparability, the statistics are unable to accurately differentiate between speculative trade and functional usefulness (e.g., remittances or inflation hedging). Consequently, the findings of this study should be regarded as statistically significant predictive correlations that delineate the financial policies of the country, rather than as unquestionable causal evidence.

Because the Fractional Probit Model is non-linear, the raw coefficients presented in the estimation tables reflect the change in the latent index rather than a direct indication of the magnitude of influence on the Financial Resilience Index (FRI). We compute and present the average marginal effects (AMEs) to offer an economically relevant interpretation. The AME for cryptocurrency adoption is ascertained as thus:

$$\mathrm{AME}={\displaystyle \frac{1}{n}}{\displaystyle \sum _{i=1}^n\frac{\partial E[{FRI}_i|X_i]}{\partial {\mathrm{Crypto}}_i}}$$

Unlike the marginal effect at the mean, which assesses the impact for a hypothetical average observation, the AME averages the partial derivatives across the whole sample distribution.

3.3.2. Moderating Effect of Institutions

To test whether institutional quality influences the relationship between cryptocurrency and financial resilience, we introduce interaction terms with the six World Bank Governance Indicators (WGIs). The use of institutional quality indicators as influencing factors provides an opportunity to examine whether cryptocurrencies behave as an alternative to institutional shortcomings in unstable contexts, or as a technological asset requiring a robust regulatory framework to develop. This examination is essential in order to account for the diversity of our dataset, which includes 112 countries, by discerning how respect for rights, Corruption, and Regulatory Quality determine the ability of digital assets to protect household wealth in the face of economic disruption. Our evaluation focuses on Corruption, Government Effectiveness, Regulatory Quality, Voice and Accountability and Rule of Law as institutional moderators. By scrutinising these links, the study goes beyond a global interpretation and highlights that the performance of financial innovation as a tool for resilience is fundamentally linked to the value of a country’s institutional system. Hence, Equation (1) is re-specified as

$$E(y_i|\dots )=\Phi ({\beta }_1{\mathrm{Crypto}}_i+{\beta }_2{IQ}_i+{\beta }_3{\mathrm{Crypto}}_i\times {IQ}_i+\lambda X_i+e_i)$$

(2)

With:

• $y_i$ is the financial resilience score of country i;
• ${IQ}_i$ denotes the institutional quality indicator.

4. Results

We present the results in three stages. Before presenting the main results, we checked for multicollinearity in both the model without and with the institutional quality indicators. Results outlined in

Table A2. Variance inflation (VIF) test model without governance.

Variable	VIF	1/VIF
GDP per capita	3.12	0.321
Share of agriculture in the economy	3.07	0.326
Lower secondary education completion rate	2.68	0.373
Age dependency ratio	2.39	0.418
Financial development	2.15	0.466
Secure internet servers per million	1.97	0.509
Unemployment	1.27	0.789
Inflation	1.13	0.881
Crypto adoption	1.13	0.884
Account ownership	1.09	0.915
Mean VIF	2.00

and

Table A3. Variance inflation (VIF) test model with governance.

Variable	VIF	1/VIF
Government Effectiveness	30.900	0.032
Rule of Law	29.620	0.034
Corruption	14.380	0.070
Regulatory Quality	12.340	0.081
Voice and Accountability	4.850	0.206
GDP per capita	4.110	0.243
Lower secondary education completion rate	3.530	0.283
Share of agriculture in the economy	3.120	0.320
Age dependency ratio	2.950	0.339
Financial development	2.470	0.405
Secure internet servers per million	2.220	0.451
Unemployment	1.460	0.684
Inflation	1.310	0.765
Crypto adoption	1.220	0.821
Account ownership	1.140	0.875
Mean VIF	7.710

in Appendix A confirmed that the mean variance inflation factors (VIFs) are below the 10 thresholds (Kutner et al., 2005).

First, we present the results on the relationship between cryptocurrency adoption and financial resilience. Next, we present the results of institutional quality moderation.

4.1. Effect of Cryptocurrencies on Financial Resilience

To account for structural heterogeneity across global economies, the sample is bifurcated based on the level of economic development. We adopt a GDP per capita threshold of $13,845, which aligns with the World Bank classification for High-Income Economies in 2024. This stratification is theoretically motivated by the Institutional Quality Hypothesis, which suggests that the utility of decentralised financial technologies, such as cryptocurrency, varies depending on the maturity of the existing formal financial system.

The results outlined in

Table 4. Effect of cryptocurrency on FRI without institutional quality.

Independent Variable	Full Sample	Low Income Countries (GDP per Capita < 13,845)	High Income Countries (GDP per Capita > 13,845)
Cryptocurrency adoption	0.089	0.127 *	−0.485 **
	(0.057)	(0.066)	(0.210)
Account ownership	7.996 ***	6.329 **	27.029 ***
	(2.160)	(2.457)	(7.152)
Account ownership squared	−9.260 ***	−8.269 ***	−23.541 ***
	(1.649)	(2.041)	(5.202)
GDP per capita	0.149	−0.105	0.293
	(0.213)	(0.261)	(0.394)
Lower secondary completion rate	0.018 ***	0.022 ***	0.018
	(0.007)	(0.008)	(0.023)
Financial development	−0.007 **	−0.011 ***	0.003
	(0.003)	(0.003)	(0.006)
Unemployment rate	−0.011	−0.004	0.133 *
	(0.026)	(0.031)	(0.074)
Inflation rate	−0.026 *	−0.027 *	0.034
	(0.014)	(0.015)	(0.240)
Age dependency ratio	−0.011	−0.034	−0.064 *
	(0.019)	(0.025)	(0.038)
Share of agricultural sector	0.003	−0.020	0.045
	(0.026)	(0.029)	(0.224)
Secure internet servers	−0.039	−0.005	0.295 *
	(0.077)	(0.079)	(0.160)
Constant	−1.949	0.854	−12.339 ***
	(2.037)	(2.490)	(4.060)
Observations	112	74	38
Standard errors in parentheses

* p < 0.10; ** p < 0.05; *** p < 0.01.

highlight marked heterogeneity in the effect of cryptocurrencies on household financial resilience depending on a country’s level of economic development. Average marginal effects are outlined in

Table A4. Average marginal effects model without governance.

Independent Variable	Full Sample	Low Income Countries (GDP per Capita < 13,845)	High Income Countries (GDP per Capita > 13,845)
Cryptocurrency adoption	0.020	0.031 **	−0.069 **
	(0.013)	(0.016)	(0.028)
Account ownership	1.811 ***	1.563 ***	3.855 ***
	(0.458)	(0.545)	(1.055)
Account ownership squared	−2.097 ***	−2.042 ***	−3.358 ***
	(0.327)	(0.415)	(0.756)
GDP per capita	0.034	−0.026	0.042
	(0.048)	(0.065)	(0.055)
Lower secondary completion rate	0.004 ***	0.005 ***	0.003
	(0.001)	(0.002)	(0.003)
Financial development	−0.002 **	−0.003 ***	0.000
	(0.001)	(0.001)	(0.001)
Unemployment rate	−0.002	−0.001	0.019 *
	(0.006)	(0.008)	(0.010)
Inflation rate	−0.006 *	−0.007 *	0.005
	(0.003)	(0.004)	(0.034)
Age dependency ratio	−0.003	−0.009	−0.009 *
	(0.004)	(0.006)	(0.005)
Share of agricultural sector	0.001	−0.005	0.006
	(0.006)	(0.007)	(0.032)
Secure internet servers	−0.009	−0.001	0.042 *
	(0.017)	(0.019)	(0.023)
Constant	0.020	0.031 **	−0.069 **
	(0.013)	(0.016)	(0.028)
Observations	112	74	38
Standard errors in parentheses

* p < 0.10; ** p < 0.05; *** p < 0.01.

in Appendix A. While no significant effect is observed in the overall sample, the adoption of cryptocurrencies has a positive and statistically significant impact in developing countries, whereas it is associated with a deterioration in financial resilience in developed countries. This result confirms Hypothesis 1 (H1). So cryptocurrency adoption exerts a statistically significant effect on financial resilience, with the direction and magnitude of this effect varying according to the institutional and macroeconomic environment.

The empirical results indicate a marked divergence in the impact of cryptocurrency on financial resilience across income groups. On a global scale, the effect is not statistically significant, suggesting that aggregate analyses may obscure significant regional variations. In developed countries, the adoption of cryptocurrency has been demonstrated to have a considerable adverse effect, suggesting that it may be indicative of speculative behaviour rather than contributing to resilience. Conversely, in developing countries, the effect is positive but only marginally significant, suggesting a potential grassroots utility of digital assets in contexts with restricted access to formal financial systems (Table 4). Despite the robustness of this result across model specifications, a cautious interpretation is warranted, given the comparatively lower level of statistical confidence and the necessity for additional longitudinal evidence.

In developing countries, a 1% increase in cryptocurrency adoption is associated with an average 3.1%2 increase in the probability of household financial resilience, all other things being equal. This result confirms Hypothesis 2 (H2). The positive effect of cryptocurrencies on financial resilience confirms the hypothesis of partial substitution for the formal financial system, which is characterised by low inclusion, high costs of access to credit and recurrent monetary instability. As highlighted in the literature, the rise in cryptocurrencies helps to improve citizens’ economic stability by offering them an autonomous alternative for protecting their savings, particularly in contexts of high inflation or currency depreciation (Adekunle, 2024; S. Verma & Atri, 2024). With a growing number of users, these digital assets promote the financial integration of populations traditionally excluded from the banking sector, enabling them to make transactions, save and access international financial markets. This diversification of available financial instruments reduces households’ dependence on local financial institutions, whose access conditions are often restrictive, while offering a mechanism of protection against economic shocks (Guo et al., 2025; Minarni, 2025). Furthermore, the speed of transactions and low costs associated with international transfers via blockchain technology improve the liquidity of financial resources, facilitating the rapid mobilisation of funds in emergency situations. Thus, in developing economies, cryptocurrencies appear to be a tool for adapting to the structural shortcomings of the traditional financial system, strengthening households’ resilience.

Our findings, which support the financial innovation framework developed by Schumpeter (1934) and extended by Levine (2005), along with the inclusion perspective advanced by Demirgüç-Kunt et al. (2020), demonstrate that cryptocurrency enhances household financial resilience through many channels. First, by broadening access to financial services for populations excluded from formal banking, cryptocurrency facilitates saving, transfers, and access to alternative sources of liquidity, thereby enabling households to smooth consumption in the face of economic shocks, as theorised by Townsend (1994). Moreover, following Baumol’s (1952) framework on liquidity demand, digital assets serve as liquid stores of value, particularly relevant in contexts of macroeconomic instability.

Conversely, the negative effect observed in developed countries can be explained by the use of cryptocurrencies for speculative activities rather than for securing or prudently managing savings. In these economies, which are characterised by high levels of banking inclusion, relative monetary stability and sophisticated financial markets, households already have effective risk management tools at their disposal, such as savings accounts, insurance and social protection mechanisms. The adoption of cryptocurrencies often results in increased exposure to price volatility, which can weaken households’ financial situation in the event of significant fluctuations in digital markets. Several studies highlight that the high volatility of cryptocurrencies, combined with high-risk investment behaviour, can increase the financial vulnerability of users in developed economies (El Hajj & Farran, 2024). In this context, allocating resources to speculative digital assets can reduce secure savings and compromise households’ ability to cope with unexpected shocks, thus explaining the negative effect observed on financial resilience. This evidence aligns with speculative trading theory (De Long et al., 1990) and the speculative bubble models (Shiller, 2017), demonstrating that cryptocurrency holders are perceived as risk-seeking agents, subject to cognitive biases and primarily motivated by short-term speculative gains rather than fundamental or risk management considerations.

The adoption of cryptocurrency in developed nations has a notable negative impact (significant at 1%), indicating that in sophisticated financial systems, digital assets are mostly used for speculative purposes, therefore diverting household liquidity away from robust, resilience-building assets. On the other hand, the impact in developing nations is favourable and reaches importance at the 10% level. This divergence suggests a grassroots utility, whereby digital assets operate as an informal safeguard in settings with limited access to conventional banking (Aiello et al., 2023; Lanciano et al., 2026). A regional breakdown further illuminates these results. Indeed, in Sub-Saharan Africa and Latin America, the developing subsample’s positive coefficient is driven by these regions. Cryptocurrencies serve as a practical replacement for shaky local currencies and costly remittance channels in these situations. Conversely, in North America and Europe, these areas propel the negative coefficient, therefore supporting the theory that in high institutional quality locations, crypto’s risk-protection utility is unnecessary, leaving only its speculative-risk component. We perceive a cautious attitude notwithstanding the resilience of this geographic and structural heterogeneity. Although the substitution effect is noticeable, its complete realisation depends on the particular regional strains of inflation and financial exclusion; therefore, the marginal relevance in the expanding subsample points to this.

The results also reveal a non-linear relationship between account ownership and household financial resilience. The positive coefficient of the account ownership variable, combined with a negative coefficient of its quadratic term, suggests the existence of diminishing marginal returns to financial inclusion. The positive coefficient of the linear account ownership variable, combined with the negative coefficient of its quadratic term, confirms an inverted-U-shaped relationship. Below this threshold, access to the formal financial system significantly improves households’ ability to save, receive transfers, and mobilise resources during shocks. However, once a nation surpasses this threshold, the marginal returns to resilience become negative.

In the initial stages of banking penetration, access to the formal financial system significantly improves households’ ability to save, receive transfers and mobilise resources in the event of a shock, thereby strengthening their financial resilience. However, beyond a certain threshold, high financial institution penetration can have negative effects, notably through increased exposure to credit and financial risk. In line with Stiglitz and Weiss (1981), expanded access to financial services in the presence of asymmetric information can lead households to accumulate excessive debt, ultimately reducing their capacity to absorb economic shocks. Furthermore, in highly banked economies, financial crises and banking sector disruptions spread more quickly to households, affecting savings and access to credit, which can weaken financial resilience. This result is also consistent with theoretical models of non-linear financial development, notably that of Greenwood and Jovanovic (1990), which postulate that the benefits of financial inclusion are high in the early stages of development but tend to decline, or even reverse, as financial systems reach more advanced levels of sophistication. For policymakers in developing nations currently below the threshold, the priority remains the removal of barriers to formal entry. However, for emerging economies approaching this inflexion point, the policy focus must transition. Instead of simple ‘inclusion’ metrics, authorities should prioritise debt-to-income monitoring and financial literacy programs designed to mitigate the risks of over-leverage. This suggests that the goal of financial policy should not be universal banking penetration at any cost, but rather an ‘optimal inclusion’ level that balances resource mobilisation with institutional stability.

Financial education through secondary education is one of the essential foundations for converting technological advances into tangible economic security in the face of difficulties (OECD, 2025). A good level of education helps individuals navigate the complex world of virtual currencies with caution, separating serious investment opportunities from risky projects. In the context of the digital boom, these skills enable improved management of digital assets, ensuring that these instruments enrich traditional savings rather than destabilising them. Indeed, this financial innovation is subject to various scams and can limit the prevalence of cryptomania among populations with low levels of education (Adekunle, 2024). Ultimately, well-informed individuals are better able to take advantage of blockchain technology while reducing the dangers of cybercrime and instability, thereby strengthening the nation’s overall financial robustness.

Our results further indicate that domestic credit to the private sector exerts a significant negative effect on households’ financial resilience in the full sample and among low-income countries, while the effect becomes positive, albeit statistically insignificant, in high-income economies. This finding suggests that credit expansion does not uniformly enhance economic stability, particularly in contexts characterised by weak financial institutions and limited borrower protection. In low-income countries, increased access to credit often translates into higher household indebtedness, frequently directed toward consumption rather than productive investment, thereby heightening financial vulnerability. This outcome is consistent with the financial instability hypothesis proposed by Minsky (1986), which emphasises that rapid credit expansion can amplify fragility and expose borrowers to heightened risks during economic downturns. Moreover, under conditions of imperfect information and weak regulation, lenders may extend credit to high-risk borrowers at elevated interest rates, leading to over-indebtedness and reduced shock-absorption capacity (Stiglitz & Weiss, 1981). Additionally, theories of financial development highlight that in poorly developed financial systems, credit is often misallocated and fails to support broad-based welfare improvements (Levine, 2005). As a result, rather than facilitating consumption smoothing and resilience, credit may exacerbate economic insecurity among vulnerable households. In contrast, in high-income countries with more sophisticated and well-regulated financial markets, credit is more likely to support productive investments and long-term asset accumulation, which can enhance financial stability, explaining the positive though insignificant coefficient observed in this subgroup. Overall, these findings underscore the context-dependent nature of financial deepening and caution against assuming that increased credit availability automatically strengthens household resilience.

4.2. Moderating Role of Institutions

The results in

Table 5. Effect of cryptocurrency on FRI with institutional quality.

Independent Variables	Corruption	Government Effectiveness	Regulatory Quality	Voice and Accountability	Rule of Law
Cryptocurrency adoption	0.436 **	0.664 ***	0.738 ***	0.435 **	0.563 ***
	(0.200)	(0.210)	(0.157)	(0.200)	(0.176)
Corruption	0.017
	(0.011)
Corruption × crypto	−0.010 ***
	(0.004)
Account ownership	10.065 ***	9.274 ***	10.197 ***	13.270 ***	10.942 ***
	(2.053)	(2.072)	(2.067)	(2.385)	(2.299)
Account ownership squared	−11.451 ***	−10.942 ***	−11.939 ***	−14.139 ***	−12.293 ***
	(1.687)	(1.659)	(1.724)	(1.987)	(1.860)
GDP per capita	0.063	0.014	−0.144	−0.110	−0.093
	(0.254)	(0.249)	(0.208)	(0.216)	(0.257)
Lower secondary completion rate	0.026 ***	0.027 ***	0.028 ***	0.028 ***	0.026 ***
	(0.008)	(0.008)	(0.008)	(0.008)	(0.008)
Domestic credit to private sector	−0.007 *	−0.007 *	−0.006	−0.006 *	−0.006 *
	(0.004)	(0.004)	(0.004)	(0.004)	(0.004)
Unemployment rate	−0.004	−0.014	−0.004	−0.004	−0.006
	(0.028)	(0.028)	(0.026)	(0.030)	(0.028)
Inflation rate	−0.030 **	−0.029 **	−0.026 **	−0.023 *	−0.029 **
	(0.015)	(0.014)	(0.011)	(0.012)	(0.014)
Age dependency ratio	−0.015	−0.015	−0.017	−0.021	−0.012
	(0.018)	(0.019)	(0.018)	(0.016)	(0.019)
Share of agricultural sector	0.004	−0.003	−0.011	−0.007	−0.008
	(0.029)	(0.028)	(0.027)	(0.026)	(0.027)
Secure internet servers	−0.077	−0.084	−0.074	−0.129 *	−0.077
	(0.072)	(0.071)	(0.068)	(0.069)	(0.073)
Government Effectiveness		0.023 *
		(0.013)
Government Effectiveness × crypto		−0.012 ***
		(0.004)
Regulatory Quality			0.032 ***
			(0.010)
Regulatory Quality × crypto			−0.015 ***
			(0.004)
Voice and Accountability				0.032 ***
				(0.008)
Voice and Accountability × crypto				−0.010 ***
				(0.004)
Rule of Law					0.026 **
					(0.011)
Rule of Law × crypto					−0.012 ***
					(0.004)
Constant	−2.456	−2.144	−1.411	−2.230	−1.654
	(2.232)	(2.267)	(1.996)	(2.038)	(2.214)
Observations	112	112	112	112	112
Standard errors in parentheses

* p < 0.10; ** p < 0.05; *** p < 0.01.

show that the adoption of cryptocurrencies has a positive and significant effect on household financial resilience in the presence of institutional quality variables. Average marginal effects are outlined in

Table A5. Average marginal effects model with governance.

Independent Variables	Corruption	Government Effectiveness	Regulatory Quality	Voice and Accountability	Rule of Law
Cryptocurrency adoption	0.092 **	0.139 ***	0.149 ***	0.089 **	0.117 ***
	(0.041)	(0.041)	(0.028)	(0.040)	(0.035)
Corruption	0.004
	(0.002)
Corruption × crypto	−0.002 ***
	(0.001)
Account ownership	2.123 ***	1.946 ***	2.057 ***	2.712 ***	2.269 ***
	(0.402)	(0.410)	(0.369)	(0.434)	(0.441)
Account ownership squared	−2.416 ***	−2.296 ***	−2.409 ***	−2.890 ***	−2.549 ***
	(0.306)	(0.298)	(0.270)	(0.331)	(0.330)
GDP per capita	0.013	0.003	−0.029	−0.022	−0.019
	(0.053)	(0.052)	(0.042)	(0.045)	(0.054)
Lower secondary completion rate	0.005 ***	0.006 ***	0.006 ***	0.006 ***	0.005 ***
	(0.002)	(0.002)	(0.002)	(0.002)	(0.002)
Domestic credit to private sector	−0.001 *	−0.001 *	−0.001 *	−0.001 *	−0.001 *
	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)
Unemployment rate	−0.001	−0.003	−0.001	−0.001	−0.001
	(0.006)	(0.006)	(0.005)	(0.006)	(0.006)
Inflation rate	−0.006 **	−0.006 **	−0.005 **	−0.005 *	−0.006 **
	(0.003)	(0.003)	(0.002)	(0.002)	(0.003)
Age dependency ratio	−0.003	−0.003	−0.004	−0.004	−0.002
	(0.004)	(0.004)	(0.004)	(0.003)	(0.004)
Share of agricultural sector	0.001	−0.001	−0.002	−0.001	−0.002
	(0.006)	(0.006)	(0.005)	(0.005)	(0.006)
Secure internet servers	−0.016	−0.018	−0.015	−0.026 *	−0.016
	(0.015)	(0.015)	(0.014)	(0.014)	(0.015)
Government Effectiveness		0.005 *
		(0.003)
Government Effectiveness × crypto		−0.003 ***
		(0.001)
Regulatory Quality			0.006 ***
			(0.002)
Regulatory Quality × crypto			−0.003 ***
			(0.001)
Voice and Accountability				0.006 ***
				(0.002)
Voice and Accountability × crypto				−0.002 ***
				(0.001)
Rule of Law					0.005 **
					(0.002)
Rule of Law × crypto					−0.003 ***
					(0.001)
Constant	0.092 **	0.139 ***	0.149 ***	0.089 **	0.117 ***
	(0.041)	(0.041)	(0.028)	(0.040)	(0.035)
Observations	112	112	112	112	112
Standard errors in parentheses

* p < 0.10; ** p < 0.05; *** p < 0.01.

in Appendix A, and plots of marginal effects of crypto adoption at different levels of institutional quality are reported in Figure A1, Figure A2, Figure A3, Figure A4 and Figure A5 in Appendix A. The results confirm that institutional quality moderates the relationship between cryptocurrency adoption and financial resilience (Hypothesis 3).

However, the negative and statistically significant coefficients of the interaction terms between cryptocurrency and various institutional indicators, including corruption, Government Effectiveness, Regulatory Quality, Voice and Accountability, and Rule of Law, indicate that this beneficial effect diminishes as institutional quality improves. This result suggests that cryptocurrencies act as a substitute for failing institutions rather than simply complementing formal financial systems. These variables that represent the conditions of financial stress confirm Hypothesis 4. In line with institutional theories of economic development (North, 1990), weak institutions, characterised by ineffective institutional quality and a flawed regulatory framework, encourage economic agents to resort to alternative mechanisms to secure their resources and carry out transactions. In this context, cryptocurrencies make it possible to circumvent the inefficiencies of the traditional banking system, reduce exposure to corruption and access financial instruments that are independent of national authorities. Conversely, in stronger institutional environments, where financial markets are better regulated and banking services are more reliable, the marginal utility of cryptocurrencies as a risk management tool diminishes. Households already have effective mechanisms for saving, credit and insurance, making the relative advantages of digital assets less relevant or even counterproductive due to their high volatility.

These findings are consistent with the literature on digital finance, which highlights the substitutive nature of financial innovations in contexts marked by institutional failures (Levine, 2005). Overall, these findings highlight that the impact of cryptocurrencies on financial resilience is strongly conditioned by the quality of institutions, with particularly beneficial effects in countries with fragile institutional quality. In addition, Goodell et al. (2020) and Viñuela et al. (2020) showed that in countries with reliable institutions, the rise in cryptocurrencies can be seen as a source of instability or a means of circumventing and weakening traditional financial regulation tools.

5. Conclusions

The paper assesses the effect of cryptocurrency adoption on household financial resilience. Empirical results indicate that the integration of digital assets has, on average, a positive and significant impact on individuals’ ability to cope with financial shocks. In particular, cryptocurrencies are an alternative instrument for reducing exposure to the instability of traditional currencies and circumventing certain constraints of the conventional banking system, particularly in contexts marked by institutional failures. However, the analysis also highlights an important moderating mechanism linked to institutional quality. While the direct effect of cryptocurrencies is broadly beneficial, negative interaction terms suggest that this impact is attenuated or even reversed in countries with strong institutions. This result indicates that cryptocurrencies mainly serve as a substitute for failing financial institutions rather than a complement to well-functioning financial systems. In environments with effective institutional quality, adequate regulation and a well-established Rule of Law, households already have reliable formal mechanisms for managing risk, which reduces the marginal utility of digital assets in terms of financial resilience. Furthermore, the level of education appears to be a key factor in the relationship between cryptocurrencies and financial resilience. A more educated and potentially more financially literate population is better able to make optimal use of digital technologies, while limiting exposure to the risks of fraud and the high volatility of crypto-asset markets. Overall, these results suggest cryptocurrencies can be an effective lever for strengthening financial resilience, particularly in countries with fragile institutions and limited financial inclusion. However, their positive contribution depends heavily on the institutional framework and human capital.

While the empirical evidence points toward a significant relationship between digital asset adoption and financial resilience, these findings should be interpreted with caution as they represent predictive correlations rather than direct causal links. Nonetheless, these associations suggest a need for calibrated regulation that seeks to monitor the integration of cryptocurrencies within the household production function without prematurely stifling financial innovation.

Rather than viewing technology as a standalone solution, authorities might consider strengthening financial and digital literacy frameworks. Such programs could potentially empower households to navigate the volatility of new technologies while more effectively mitigating associated risks. Furthermore, the exploration of Central Bank Digital Currencies (CBDCs) could emerge as a structured intermediate path. By potentially combining the distributed ledger efficiencies of blockchain with the stability and institutional credibility of formal financial systems, CBDCs may offer a more resilient framework for financial inclusion in countries worldwide, provided they are supported by robust digital infrastructure and transparent governance.

In order to ensure the resilience of households whilst also creating an environment that fosters innovation, a tripartite regulatory framework is proposed. This framework is focused on institutional stability and user safety. This approach is predicated on the principle of consumer protection through the implementation of a “Proof-of-Reserve” transparency model for exchange platforms. It entails the enforcement of oversight mechanisms for stablecoins by ensuring a 1:1 asset-to-backing ratio to mitigate the risk of liquidity crises. Additionally, it integrates risk-based Anti-Money Laundering/Counter-Terrorist Financing (AML/CFT) protocols at fiat gateways. By establishing these three pillars, authorities can address systemic issues such as custodial failure and illicit capital flight without compromising the fundamental utility of digital assets in emerging markets.

There are two other folds of policy consequences: governments should first investigate a regulatory sandbox approach that would enable the tracking of P2P transaction flows while also offering legal safeguards for retail consumers against volatility and fraud. Second, since our results show that the utility of these assets decreases as formal institutional quality improves, the long-term policy goal should be the institutionalisation of their advantages rather than the displacement of crypto-assets. Policymakers may naturally move homes from volatile decentralised assets toward more reliable, integrated financial systems by lowering transaction costs and enhancing formal banking reliability, therefore utilising technology to close the ‘resilience gap’ without jeopardising macroeconomic stability.