Search Results (215)

Short-term financial risk in intelligent trading systems is reflected not only in prices, trading volumes, and textual sentiment but also in infrastructure operating states, including server workload, device power consumption, network latency, and packet loss rate. We propose HSF-LLMNet, a hardware sensor and large language model semantic fusion network for jointly modeling external information shocks and infrastructure responses. A large language model extracts event category, sentiment polarity, risk intensity, and semantic uncertainty from financial texts. Reliability-aware temporal modeling handles sensor missingness, drift, and abnormal noise, while asynchronous soft alignment, bidirectional cross-attention, and reliability-aware gated fusion integrate irregular textual events with continuous hardware signals. The model jointly predicts market direction, realized volatility, and three-level risk over the subsequent 30 min. Experiments were conducted on eight Chinese A-share indices: the SSE Composite Index (000001.SH), SSE 50 Index (000016.SH), CSI 300 Index (000300.SH), STAR 50 Index (000688.SH), CSI 500 Index (000905.SH), CSI 1000 Index (000852.SH), Shenzhen Component Index (399001.SZ), and ChiNext Index (399006.SZ). The common observation period for market, textual, and hardware data extended from 1 March 2024 to 30 June 2025. After data cleaning, timestamp matching, and multimodal temporal alignment, 169,208 aligned asset–time prediction windows were retained for the 30 min forecasting task. Realized volatility was defined as the square root of the sum of squared one-minute log returns over the future 30 min interval. The three-level risk label was constructed from future realized volatility, absolute 30 min return, and liquidity stress, with all thresholds estimated exclusively from the training portion of each fold. A sample was labeled high risk when at least two of the three indicators exceeded their 85th-percentile thresholds or when any indicator exceeded its 95th-percentile threshold. It was labeled medium risk when, after excluding high-risk samples, at least two indicators exceeded their 60th-percentile thresholds or any indicator exceeded its 85th-percentile threshold; all remaining samples were labeled low risk. Results showed that HSF-LLMNet achieved an accuracy of 78.62%, a precision of 78.14%, a recall of 77.83%, an F1-score of 77.98%, an area under the receiver operating characteristic curve of 84.91%, and a Matthews correlation coefficient of 57.36% for directional prediction. For realized-volatility regression, the MAE, RMSE, MAPE, and $R^2$ were 0.0089, 0.0135, 9.21%, and 0.812, respectively. For high-risk-event warning, the mean effective warning time, defined as the interval between the first valid alarm and the corresponding event, was 15.37 min; the false-alarm rate and missed-alarm rate were 6.82% and 8.14%, respectively. Ablation experiments showed performance reductions after removing semantic encoding, sensor-reliability estimation, asynchronous alignment, bidirectional cross-attention, gated fusion, or multi-task learning. These results indicate that textual events and infrastructure operating states provide complementary information for quantitative risk analytics and fintech applications Full article

Sustainable finance increasingly treats green assets as instruments for climate-risk hedging; however, it remains unclear whether they protect investors during energy-market shocks and financial-market stress or merely transmit different transition, equity-market, and growth risks. This study asks whether green assets hedge better than brown assets, or whether the two asset classes hedge different risks across market states. Using daily data from 2010 to 2025 on exchange-traded clean-energy, fossil-fuel, green-bond, ESG, and market-risk instruments, we construct green and brown portfolios and analyze the green–brown return spread across normal conditions, high-volatility regimes, market-stress days, positive and negative oil-price shocks, the COVID-19 crisis, and the recent energy-crisis period. The empirical design combines performance and downside-risk metrics, rolling correlations and betas, Newey–West regressions, stress-state comparisons, quantile regressions, portfolio allocation tests, and supplementary machine-learning classification. The results reveal strong oil-shock asymmetry: brown assets outperform during positive oil-price shocks, while green assets perform relatively better when oil prices fall sharply. However, green assets do not function as broad safe havens during financial-market stress, reflecting persistent equity-market downside exposure and growth-factor repricing. Quantile regressions confirm nonlinear and state-dependent risk transmission, while portfolio tests show weak full-sample return-risk performance for clean-energy equity exposure alone. Shorter-sample suggests that green bonds and ESG assets display more defensive characteristics, whereas machine-learning models show limited short-horizon predictive power. Green assets are therefore not universal hedges but conditional transition-risk assets whose value depends on the shock source, market regime, and sustainable instrument. Full article

Although stablecoins occupy a segment of digital-asset markets in which price stability is central by design, their temporary departures from reference values may reveal important information about latent risk and market stress. In this paper, we examine whether bubble and crash signals extracted from traditional cryptocurrencies and stablecoins improve volatility, Value-at-Risk, and Expected Shortfall forecasting and, in connection with these forecasting gains, contribute to the assessment of cross-asset contagions. The analysis applies the Bubble Crash–GARCH models, in which extreme price phases are identified through the Phillips, Shi, and Yu real-time monitoring procedure and incorporated into the conditional mean of returns through event-based dummy variables. For stablecoins, extreme episodes are not inferred from price dynamics in isolation but from deviations between the observed price and the asset-specific reference value. The empirical investigation focuses on Bitcoin, Ethereum, Tether’s USD-pegged (USDT), and Tether Gold and evaluates asset-specific bubble–crash effects and bidirectional contagion channels between traditional cryptocurrencies and stablecoins, using Bitcoin and Tether as the leading representatives of the two market segments. The findings indicate that accounting for bubble and crash episodes leads to more accurate volatility forecasts than standard GARCH benchmarks. For Value-at-Risk and Expected Shortfall, the bubble–crash specifications can improve tail risk forecasting at several tail probability levels through more accurate coverage, lower quantile loss, and stronger ESR backtesting performance. The results also reveal different degrees of price exuberance across the two asset categories: while extreme price dynamics are more evident among traditional cryptocurrencies, deviations from fundamentals are rare for stablecoins. Among stablecoins, USDT exhibits limited but detectable exuberance, whereas Tether Gold does not display extreme price episodes. However, when such deviations occur, as in the case of USDT, they generate significant contagion effects on major cryptocurrencies. Notably, extreme episodes originating in USDT have a stronger impacts on Bitcoin and Ethereum than the reverse spillovers from traditional cryptocurrencies to USDT. Overall, the evidence suggests that stablecoins are not merely passive instruments within the digital-asset ecosystem. Even temporary deviations from their reference values contain valuable information for risk forecasting and contagion monitoring. Full article

This study examines behavior among mutual funds across five sectors, including energy, utilities, real estate, technology, and healthcare, over a 27-year period (1999–2025). For each year, using daily NAV returns, each fund is regressed against the S&P 500 to separate fund-specific performance from broader market movements. All the resulting residual vectors for each year and fund are clustered together. These annual cluster assignments are then linked across time to construct each fund’s trajectory, showing how its relative performance position shifts from year to year. These trajectories capture long-term behavioral divergence and provide a simple, intuitive visualization. To identify and characterize fund trajectory and distinctiveness, we apply methods based on residual magnitudes, quantile migration patterns, and Hamming distance measures of (cluster, time) fund trajectories. These trajectories reveal how certain funds consistently diverge from market behavior, thereby contributing to portfolio diversity in terms of trajectory separation. We introduce a portfolio Hamming diversification index that measures separation between trajectories. Using trajectories and Hamming distances, we examine how fund behavior changes during major market disruptions, including the dot-com crash (2001), the financial crisis (2009), and the COVID-19 pandemic (2020–2021), and identify sector-specific differences in how fund trajectories respond to these events. The proposed methodology is intended as an exploratory descriptive methodology for studying long-term behavioral trajectories rather than as a replacement for traditional asset-pricing or performance-evaluation models. Full article

Climate-related risks are now perceived as financially material, affecting banks, investors, firms, regulators, and central banks. This article provides a systematic literature review of the relationship between climate-related risks and financial decision-making. The review, which is based on Scopus and Web of Science and follows a PRISMA-based selection process, includes a final sample of 80 studies retrieved using database publication-year filters for 2015–2025, including three online-first records subsequently assigned to 2026 issues. A Quality Appraisal Matrix, descriptive analysis, keyword co-occurrence mapping with VOSviewer version 1.6.20, and cluster-based thematic synthesis are used. The results reveal four key streams in the literature: climate risk, Environmental, Social, and Governance (ESG), and financial modeling; climate change, sustainable finance, and systemic stability; transition risk, investment, and risk assessment; and banks, performance, and financial impact. The review underscores the role of climate risks in asset pricing, portfolio allocation, lending, credit-risk assessment, disclosure, stress testing, and financial stability supervision, and identifies key gaps with respect to emerging economies, data quality, and the practical incorporation of climate risk considerations into financial decisions. Full article

In South Korea, apartments dominate the residential housing market, accounting for 67.5% of total housing transactions in the fourth quarter of 2018. With this figure continuing to rise, apartments are the most significant asset for many families. Consequently, precise and timely valuations are crucial for stakeholders, including homeowners, buyers, and mortgage lenders. Traditionally, these stakeholders have relied on the qualitative judgments of certified real estate agents. Because of market opacity and low liquidity, agents often use a comparative approach, referencing the most recent transaction prices of nearby comparable apartments. However, this method is subjective, potentially biased, time-consuming, and costly. Our study seeks to offer a more objective and quantitative method for determining fair apartment prices in Korea, helping market participants make informed decisions. The prediction target is the monthly representative price of an apartment complex (the within-complex average of transaction prices), from which a complex-level price index is subsequently constructed; we distinguish this target from individual transaction prices throughout. By employing clustering methods to identify similar apartments and imputation techniques for missing values, our model demonstrates promising results, with a mean absolute percentage error as low as 5.38% in the worst-case (consecutive-mask) setting and 4.76% in the typical (random-mask) setting. Because the training (2006–2015) and test (2016–2022) periods are temporally disjoint, these figures reflect out-of-sample performance rather than in-sample fit. We further validate the resulting series against external references: it attains a 5.03% MAPE against actual transactions nationwide—outperforming the appraiser-based Kookmin Bank index (7.29%)—and, once aggregated, closely tracks the official KREB transaction-based index while becoming available earlier; complex-level Granger tests confirm that our series temporally leads the appraiser-based series about 1.5 times as often as the reverse. We also outline the missing-data assumptions under which the approach is valid. Full article

Inflation expectations are central to monetary policy transmission, yet relatively little is known about whether individuals’ own investment experiences are associated with how they form such expectations across different forecast horizons. This study examines the association between self-reported past investment performance and horizon-specific expected cumulative consumer price changes at the one-, three-, and five-year horizons using a large-scale online survey of 157,523 active Japanese investors. Because the survey asks respondents how consumer prices will change over each horizon, the three- and five-year responses are interpreted as expected cumulative price changes rather than annualized inflation rates. Ordered probit models are estimated while controlling for demographic, socioeconomic, and behavioral characteristics. The results show a horizon-dependent conditional association: self-reported investment performance is not significantly associated with one-year expectations in the full specification, whereas it is positively and significantly associated with three- and five-year expectations. Formal stacked OLS interaction tests indicate that the association differs significantly across horizons, and additional threshold-specific probit models show that the pattern is most evident for moderate inflation-expectation thresholds. The economic magnitudes are statistically precise but modest. Heterogeneity analyses further suggest that the association is weaker among respondents with higher financial literacy and higher assets, and stronger among respondents with a more myopic view of the future. Because the analysis relies on cross-sectional observational data and subjective performance measures, the findings should be interpreted as conditional associations rather than causal effects. Full article

The increasing penetration of variable renewable energy sources (RESs) poses significant challenges to power system flexibility and reliability, particularly in systems with high solar generation. At the same time, existing Concentrating Solar Power (CSP) plants in Europe face declining economic viability due to high capital costs and the expiration of incentivized tariff schemes. This study proposes and evaluates a novel approach to repurpose CSP plants as flexible energy assets through the integration of cold thermal energy storage (CTES) within a Thermally Integrated Power-to-Heat-to-Power Energy Storage (TI-PTES) framework. The proposed system combines an ice/water-based cold storage with a CO₂-based refrigeration cycle to enhance the efficiency of the CSP steam cycle by reducing condenser temperatures, while also enabling temporal shifting of electricity consumption. A techno-economic optimization model based on PyPSA is developed to determine the optimal sizing and operation of the storage and refrigeration system under realistic load and electricity price conditions representative of the Spanish market. Results show that the integration of cold storage significantly alters system operation, shifting the chiller from a continuous demand-following mode to an intermittent, high-intensity regime. This leads to a reduction in annual operating expenditures by approximately 32% and an increase in annual profit and net present value (NPV), despite higher capital investment. While hourly net revenue becomes more volatile, with negative values during charging periods, cumulative annual performance improves due to effective temporal optimization. However, the absence of strong electricity price arbitrage and negative price signals limits the revenue potential of the storage system, which primarily acts as a cost-reduction mechanism. The findings demonstrate that cold thermal storage can successfully reposition CSP plants as flexible, value-generating assets in modern electricity systems. The proposed concept offers a promising pathway for extending the operational lifetime of existing CSP infrastructure while supporting higher integration of renewable energy sources. Full article

This study develops a leakage-free hybrid Markov–Random Forest framework for cryptocurrency price forecasting and evaluates it on Bitcoin and Ethereum. Daily OHLCV features are lagged by one trading day to prevent look-ahead bias, while regime labels are assigned from observed price changes using a two-state Markov chain with increasing and decreasing states. Regime-specific Random Forest models are then tuned independently via time-series cross-validation, allowing the predictive structure to adapt to regime-specific market conditions. The empirical results exhibit clear asset dependence. For Ethereum, the hybrid model outperforms the standalone Random Forest on magnitude-based metrics, attaining lower MAE and RMSE while also delivering a modest improvement in directional accuracy. Regime-specific tuning further identifies distinct optimal hyperparameter configurations across the increasing and decreasing states, suggesting that Ethereum’s upward and downward dynamics are structurally heterogeneous and can be better captured through regime-aware learning. By contrast, for Bitcoin, the standalone Random Forest delivers superior magnitude forecasting performance, while the regime-specific models differ only in tree depth and share the remaining tuning parameters, indicating that regime conditioning adds limited incremental value in a more persistent market. Statistical tests reinforce these findings. For Ethereum, Diebold–Mariano tests show that the hybrid significantly outperforms the standalone Random Forest under squared loss, while the absolute-loss comparison is only marginal. Across both assets, directional accuracy remains close to random chance, confirming the limited predictability of next-day price direction from lagged OHLCV features. Overall, the hybrid framework is most valuable when regime-specific dynamics are sufficiently distinct, offering improved forecasting performance and greater interpretability than a single global model. Full article

This study examines the relationship between valuation multiples and investment performance in the U.S. stock market. Specifically, it tests whether portfolios constructed with high-multiple stocks consistently outperform portfolios with low-multiple stocks. The analysis spans the Technology, Consumer Staples, and Healthcare sectors from 2018 to 2022. A sector-based portfolio construction framework was employed using quarterly portfolio-return data. Quantitative financial modelling, including regression analysis and descriptive statistics, was applied to assess the correlation between portfolio returns and valuation multiples (P/E and EV/EBITDA), while interpreting results within the broader context of market volatility and the COVID-19 period. The results show no statistically significant relationship between valuation multiples and portfolio performance. Low-multiple portfolios demonstrated marginally higher average returns over the period, offering weak support for value-based investment strategies. Results further suggest limited standalone predictive power in high-multiple valuations. Drawing on the Efficient Market Hypothesis, Value Investing, Growth Investing, and the Fama-French Three-Factor Model, this paper empirically tests the impact of valuation multiples within a sector-based portfolio framework. Accordingly, the study adds to the asset pricing literature by offering a structured null-result framework, demonstrating that valuation multiples, when applied in isolation, may not provide sufficiently reliable standalone signals for portfolio performance. The COVID-19 period is interpreted as an economically meaningful contextual regime characterized by elevated volatility, liquidity intervention, and sectoral divergence, rather than as a formally estimated event-study framework. Full article

This study examines the impact of public attention on financial markets during the Israel–Hamas conflict, focusing on the TA35 stock index and the Israeli Shekel (ILS) exchange rate over the period October 2023 to April 2025. By distinguishing between global and domestic Google search activity, the analysis investigates whether the origin of attention differentially affects market performance and currency dynamics. Public attention is treated as a real-time proxy for investor sentiment and perceived risk. Methodologically, the study combines Google Trends data with EGARCH(1,1) models to capture both return effects and asymmetric volatility responses. To enhance robustness, Principal Component Analysis (PCA) is applied separately to global and domestic search datasets, generating latent indices that reflect conflict-related and humanitarian narratives. These indices are subsequently incorporated into the empirical models. The findings reveal that global search intensity related to conflict topics exerts a significant negative effect on stock returns and contributes to currency depreciation, reflecting heightened uncertainty and risk aversion. In contrast, domestic search activity is associated with stabilizing or positive effects, suggesting local resilience and confidence. PCA-based models improve explanatory power and confirm that the geographical origin of attention plays a crucial role in shaping financial outcomes. Additionally, the results indicate that attention-driven shocks influence volatility asymmetrically, amplifying downside risk during periods of intensified global concern. Overall, the study contributes to the literature by integrating behavioral indicators into financial risk modeling and providing a novel, real-time framework for assessing how digital attention transmits geopolitical risk into asset prices. Full article

The efficient market hypothesis, which holds that prices completely reflect available information, is commonly used in financial market analysis. However, emerging empirical evidence shows that market efficiency develops with time, as posited by the adaptive market hypothesis (AMH), with predictability varying across shifting economic and behavioural regimes. Despite the increasing use of deep learning in financial forecasting, there has been little systematic investigation into whether neural network topologies can successfully identify time-varying efficiency trends across diverse markets. Furthermore, the relevance of transfer learning in studying adaptive behaviour between foreign exchange markets and extremely volatile cryptocurrency markets has received little attention. Using these data, we investigate the AMH by comparing the forecasting performance of various deep learning architectures and determining whether knowledge transfer from a relatively stable fiat currency market, Botswana Pula/US Dollar (BWP/USD), improves the predictive accuracy in a highly volatile cryptocurrency market, Bitcoin/US Dollar (BTC/USD). We use daily data from 1 January 2015 to 11 January 2026 to develop deep neural networks (DNNs) and alpha-recurrent neural networks, and, for generalisation, we benchmark using a recurrent temporal neural network (RTNN), a domain-adversarial neural network (DANN), and KLIEP-based importance-weighted regression. A transfer learning technique is used, in which models are initially trained on BWP/USD and then re-estimated on BTC/USD without freezing any network layers, ensuring complete flexibility and enabling parameters to respond to changing market dynamics. Out-of-sample accuracy measures and rolling long-memory diagnostics are used to evaluate forecast performance in terms of time-varying efficiency. The findings reveal that the RTNN regularly outperforms other forecasting models across marketplaces. Predictive accuracy fluctuates with time, and rolling long-memory measurements show persistent departures from random walk behaviour, which supports the AMH. Transfer learning improves predictive stability in the cryptocurrency market by identifying the existence of transferable informational structures between fiat and digital asset markets. Overall, our results support the idea that market efficiency is dynamic rather than static, and they show that adaptive deep learning systems are an excellent way to test the AMH. The paper suggests that cross-market transfer mechanisms and adaptive modelling methodologies be investigated further in growing foreign exchange and cryptocurrency markets. Full article

The rapid growth of environmental, social, and governance (ESG) investing has intensified the debate regarding whether ESG-oriented investment strategies exhibit performance patterns that differ from those of conventional investments, particularly during periods of market disruption. This study examines the risk-adjusted performance of ESG-oriented and non-ESG exchange-traded funds (ETFs) across market regimes surrounding the COVID-19 shock. The analysis classifies 28 passively managed ETFs into four sustainability-based categories and evaluates their performance using factor-based asset pricing models derived from the Fama–French framework. Additional analyses assess benchmark-relative performance using the S&P 500 and MSCI World indices and consider alternative ETF classifications based on investment mandates. The study estimates regime-specific regressions for the pre-COVID, COVID, and post-COVID periods. The results indicate that performance patterns vary across market regimes and ETF categories. Non-ESG ETFs tend to underperform on a risk-adjusted basis during the pre-COVID period, although this effect disappears thereafter. ESG-oriented ETFs generally exhibit limited evidence of abnormal performance, while factor exposures vary across regimes, reflecting changes in sector composition and macro-financial conditions. The findings suggest that, in addition to ESG orientation, market regimes and sectoral exposures play an important role in explaining differences in ETF performance. Full article

The inclusion of the cement industry into China’s national carbon emissions trading system in 2025 has fundamentally altered the compliance environment for high-emission enterprises, transforming carbon allowances from passive regulatory instruments into dynamic assets whose management directly affects financial performance. We develop a multi-scenario carbon asset management decision model tailored to the intensity-based benchmarking mechanism adopted by the national market. The model centres on the quota surplus-deficit variable EA4, which is computed from enterprise-level emission intensity relative to the industry benchmark, and decomposes the management problem into sequential selling and buying subproblems linked by coupled decision boundaries. A systematic parameter framework is constructed, and the model is applied to two cement enterprises—Enterprise A, a leading producer with a clear allowance surplus, and Enterprise B, a mid-tier producer operating near the benchmark boundary—through historical backtesting over the 2024–2025 period. Three principal findings emerge. First, the intensity benchmarking mechanism creates a dual-leverage effect whereby a 1.4% improvement in emission intensity (from 0.8112 to 0.8000 t/t) increases the quota surplus by 27%, a nonlinearity not captured by conventional compliance-cost models. Second, the model-driven strategy outperforms traditional experience-based approaches by 36.8% (baseline scenario, +95.20 vs. +69.58 MRMB) and 37.3% (risk scenario, −44.55 vs. −71.08 MRMB), with the improvement rate remaining consistent across both enterprises, suggesting that trading timing outweighs instrument selection in determining compliance cost outcomes. Third, dynamic CEA–CCER allocation captures an incremental 2.33 MRMB through the exploitation of a transient price inversion, a gain invisible to single-instrument strategies. Sensitivity analysis confirms that the relative advantage is robust to carbon price variations (±30%) and CCER offset caps (2–10%), while emission intensity and carry-over allowances represent the most consequential parameters for strategy direction, with EA4 crossing zero near the industry benchmark (I ≈ 0.85). The framework provides actionable decision support for cement and other high-emission enterprises navigating the unified carbon market, and contributes a quantitative methodology to the emerging field of environmental management accounting. This study contributes to Sustainable Development Goal 13 (Climate Action), Goal 7 (Affordable and Clean Energy), and Goal 9 (Industry, Innovation, and Infrastructure) by providing operational tools for decarbonisation in carbon-intensive industries. Full article

Volatility is a fundamental indicator for assessing the risk of financial assets. By integrating unstructured data, such as earnings call transcripts, the limitations of traditional time series data can be transcended, enabling collaborative forecasting from multiple data sources, enhancing the robustness of volatility prediction, and improving the efficiency of risk management. Although current research has effectively utilized earnings call data to predict asset volatility, price trends, and stock correlations, it often overlooks the inherent challenges of integrating textual and time series data, as well as the self-exciting and clustering characteristics of financial events. While conventional Long Short-Term Memory (LSTM) networks excel in processing fused data, they lack the structural capacity to explicitly model event-driven temporal decay, often failing to differentiate the varying influence of historical shocks over time. To surmount this limitation, we have significantly enhanced the predictive model by focusing on extracting salient information and integrating temporal dependency modeling with dynamic state adjustment mechanisms. The core innovation is introducing the Hawkes process to explicitly capture the self-exciting effect of financial events, which is the key to modeling volatility clustering around earnings releases. The proposed Hawkes LSTM model introduces a decay gating module and a textual information knowledge enhancement module. The decay gating module is specifically designed to more effectively capture the temporal dependencies between events within an event sequence. This allows the model to focus more on recent significant events, with the influence of an event on subsequent events typically diminishing as the temporal interval between them increases. By integrating temporal dependency modeling, the model is enabled to utilize historical data in a more flexible manner. The dynamic state adjustment mechanism further enhances its capacity to capture dynamically changing characteristics. Together, these features provide a more robust and precise solution for volatility prediction. Experimental results on two real-world earnings call datasets show that this approach significantly outperforms existing benchmark models on most prediction horizons, achieving competitive and superior performance and verifying its effectiveness and robustness. Full article