The extreme level of volatilities and dynamic correlations between cryptocurrencies and traditional investments represent major problems that affect traditional portfolio allocation techniques. Attempting to measure the influence of volatility of cryptocurrencies on the optimal portfolios, this research is conducted by analyzing the daily prices of four cryptocurrencies (Bitcoin, Ethereum, Binance Coin, and Ripple), as well as five traditional investments (equities, bonds, gold, and oil) from 2018 to 2024. In order to determine the volatilities and conditional correlations, DCC-GARCH methodology was employed while the volatility spillovers were calculated using the Diebold-Yilmaz index. The performance of the traditional, hybrid, and unconstrained portfolios was assessed on the basis of returns, Sharpe ratio, and conditional value-at-risk measures. According to the findings, the correlation between Bitcoin and the S&P 500 surged from 0.08 pre-COVID-19 to 0.34 during the pandemic period and stayed at 0.27 in the post-COVID era. The total volatility spillover index was 34.2% during the peak of the turmoil. With the inclusion of cryptocurrencies, the Sharpe ratio increased from 0.67 (portfolio without constraints) to 1.02 (unconstrained portfolio); however, the maximum capital drawdown became even higher – from -22.5% to -38.5%. The optimal level of exposure to cryptocurrencies was found to be 15%. The Black-Letterman model adjusted the Bitcoin weighting to 8.2%. In conclusion, the empirical evidence confirms that while cryptocurrencies significantly enhance portfolio returns and risk-adjusted performance – with the unconstrained portfolio achieving an 18.2% annualized return compared to 7.8% for traditional assets, they simultaneously introduce substantially higher downside risks, as reflected in the increased maximum drawdown and CVaR measures. The dynamic correlation analysis reveals that cryptocurrencies have become increasingly integrated with traditional equity markets, particularly during crisis periods, thereby diminishing their diversification benefits precisely when they are most needed. The optimal cryptocurrency allocation of 15% represents a pragmatic balance between return enhancement and risk control, while the Black-Litterman adjustment to 8.2% for Bitcoin reflects the importance of incorporating forward-looking market views into portfolio construction. These findings provide critical quantitative evidence for policymakers, institutional investors, and portfolio managers in both developed and emerging economies, including Uzbekistan, highlighting the need for prudent position limits, robust risk management frameworks, and continuous monitoring of evolving cross-asset correlations in the era of digital asset integration.
Tax compliance continues to be an issue for transition economies, and it is evident that the conventional system of penalties has not been sufficient to address the problem. In order to examine the effectiveness of behavioral policies as low-cost solutions to increase tax compliance, a randomized field experiment was conducted in Uzbekistan. Eight thousand two hundred taxpayers who failed to submit their tax returns by 2024 were randomized into a control group and four different treatment groups namely, social norms treatment, simplification treatment, moral obligation treatment, and mild deterrent treatment. Each treatment group received a letter, which contained contents related to behavioral economics. Simplification of process was found to have the highest impact in terms of the percentage of increase, which was 8.5 points above that of the control group (22.5%). In addition, the other interventions included mild deterrence (7.4 percentage points), social norms (6.2 percentage points), and moral obligation (3.8 percentage points). Process simplification was also found to generate the greatest improvement in payment of taxes (27.1%) and the most efficient benefit-to-cost ratio (21.6). Low income earners and elderly taxpayers were found to be the least responsive to any form of reminder. The empirical results show that behavioral policies, especially process simplification, have a much greater impact on improving tax compliance in transition economies than conventional enforcement methods at a much smaller cost. The effect of the simplification treatment on tax compliance was an increase in filing rates by 8.5 percentage points and tax payments by 27.1%, with the benefit-cost ratio of 21.6 being considerably superior to those of moral suasion (7.2) and deterrence (18.7). Still, these policies can be quite ineffective for some specific groups of taxpayers: people with low income reacted only to simplification, while the elderly ones (aged 56 or above) were not sensitive to any interventions. This research gives valuable implications for the tax authorities in transition economies in terms of the design of policy programs that can increase revenues from taxes and do not impose additional enforcement expenses and damage to the taxpayers’ trust.
This article analyzes the interrelationship between tourism and creative industries within the structure of contemporary scientific discourse. The study is based on a dataset of 665 publications indexed in the scientometric database Web of Science Core Collection. The sample was formed by sequentially selecting publications according to a developed search strategy, verifying their content relevance, and subsequently normalizing the bibliographic data. The analysis was conducted using scientometric mapping tools, which allow us to trace not only the dynamics of research interest but also the internal configuration of thematic connections. The study examines changes in publication activity, the distribution of works across subject areas, the structure of key concepts, and the patterns of their co-occurrence in scientific literature. The analysis shows that research on tourism and the creative industries emerges at the intersection of several academic fields and does not constitute a fully consolidated research space. The most common categories retain a central position in the system of scientific connections but are characterized by a relatively low degree of internal connectivity. At the same time, certain specialized fields demonstrate a more stable conceptual structure, while remaining relatively isolated from the general research landscape. There is a gradual shift in research focus from general issues of the interaction between tourism and creative industries toward more specific topics related to the tourist experience, territorial development, cultural policy, creative cities, and destination management. The development of this field of study occurs primarily through the layering of new thematic directions onto an already established conceptual core, which leads not to the replacement of existing research approaches but to a more complex system of connections between them.
Introduction: The knowledge economy reframes tax literacy as a measurable component of human capital that conditions voluntary compliance, fiscal capacity and the efficiency of public revenue mobilisation. Kazakhstan presents an analytically informative case: near-universal digital access coexists with a comparatively low adult financial-literacy index, while the State Revenue Committee is undertaking an artificial-intelligence-driven transition from a control-based to a service-based model of tax administration. The paper examines how AI technologies can convert the country’s digital readiness into higher population tax literacy. Methods: A mixed design combines desk analysis of official Kazakh and international statistics for 2019-2025 with a systematic mapping of eight AI-enabled and digital systems of the State Revenue Committee to specific tax-literacy functions, and the construction of a composite Tax-Literacy Readiness Index from six cited national sub-indicators under an explicit, reproducible weighting scheme. Results: Adult financial literacy rose from 39.5 percent in 2021 to 42.0 percent in 2025, while internet use among adults aged 16-74 reached 96.2 percent, population digital literacy exceeded 92 percent, and 92 percent of public services moved online. Kazakhstan ranked 24th of 193 on the UN E-Government Development Index (EGDI 0.9009) and 78th of 133 on the Global Innovation Index. The composite Tax-Literacy Readiness Index equals 71.9 out of 100, and the knowledge dimension is the single binding constraint, contributing only 0.126 of the 0.719 total. Discussion: High infrastructure and digital-capability scores are not matched by knowledge: tax literacy in Kazakhstan is constrained on the cognitive rather than the access margin. AI-enabled instruments – pre-filled declarations, the Smart Data Finance analytics core, conversational assistants and proactive notifications – act on precisely this margin by relocating tax knowledge into the taxpayer’s own digital environment, in line with the OECD Tax Administration 3.0 vision. Scientific Novelty: The paper provides the first composite, reproducible Tax-Literacy Readiness Index for Kazakhstan that links knowledge-economy indicators, adult financial literacy and the AI-enabled tax-administration technology stack within a single analytical frame. Practical Implications: For the tax administration, the binding constraint is the knowledge margin, so AI investment yields the highest marginal return when targeted at automated taxpayer education for low-literacy segments rather than at enforcement alone. For policymakers, tax literacy should be tracked as a named sub-component of the national financial-literacy strategy and integrated into the eGov ecosystem.
The ability to forecast volatility precisely plays an important role in risk management in financial markets. In this paper, we performed a comparative analysis of forecasting accuracy of machine learning approaches and the traditional GARCH model in the case of the Amman Stock Exchange in Jordan. Historical daily index data from 2010 to March 2025 were used to compare the forecasting performance of the LSTM algorithm, support vector machine algorithm, and random forest algorithm for one-day out-of-sample horizon (January-March 2025) to GARCH(1,1)-t model. According to the MSE and MAE measures, the LSTM model demonstrated the lowest forecasting errors and performed significantly better by the Diebold-Mariano test. In economic terms, this finding implies that more accurate volatility forecasts directly translate into better risk-adjusted portfolio performance and more efficient capital allocation. The LSTM model provided up to 30% MSE reduction compared to GARCH during the periods of high volatility regimes. As such, from the economic analysis, the LSTM forecast portfolio is observed to have a Sharpe ratio of 0.89 with a maximum drawdown of -4.15%, which is a better performance compared to the GARCH forecast portfolio, proving the economic benefit of using machine learning models for forecasting volatility in emerging markets. In summary, it can be seen that the use of machine learning, especially the LSTM method, does not only improve forecast accuracy but also has strong economic benefits. As such, the problem of choosing the best volatility forecasting model is well solved in the Jordanian financial market.
