|
on Big Data |
| By: | Gang Li; Dandan Qiao; Mingxuan Zheng |
| Abstract: | We find that event features extracted by large language models (LLMs) are effective for text-based stock return prediction. Using a pre-trained LLM to extract event features from news articles, we propose a novel deep learning model based on structured event representation (SER) and attention mechanisms to predict stock returns in the cross-section. Our SER-based model provides superior performance compared with other existing text-driven models to forecast stock returns out of sample and offers highly interpretable feature structures to examine the mechanisms underlying the stock return predictability. We further provide various implications based on SER and highlight the crucial benefit of structured model inputs in stock return predictability. |
| Date: | 2025–12 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2512.19484 |
| By: | Uddin, Md Azhar |
| Abstract: | Identifying farms at risk of exiting the dairy industry remains a major challenge, particularly due to data scarcity and the limitations of traditional econometric models such as logit and probit. This study applies machine learning (ML) techniques to predict dairy farm exit intentions in Wisconsin using data from the 2024 DATCP Dairy Producer Survey. Using a broad set of accessible survey based variables, including farm demographics, operations, environmental practices, and perceived challenges, we compare the performance of Lasso, Ridge, Random Forest, and Extreme Gradient Boosting (XGBoost) models. XGBoost outperforms all others in both overall accuracy and sensitivity, effectively identifying farms at risk of exit while maintaining strong performance in predicting continuation. Furthermore, SHAP (SHapley Additive exPlanations) analysis highlights succession planning, operators age, investment behavior, labor constraints, and conservation practices as key predictors. These findings demonstrate the practical utility of ML models for early risk detection and offer actionable insights for policymakers, industry stakeholders, and extension services aiming to sustain Wisconsin’s dairy sector. |
| Keywords: | Teaching/Communication/Extension/Profession |
| Date: | 2025 |
| URL: | https://d.repec.org/n?u=RePEc:ags:aaea25:362687 |
| By: | Uluc Aysun (University of Central Florida, Orlando, FL); Melanie Guldi (University of Central Florida, Orlando, FL) |
| Abstract: | We revisit the exchange-rate predictability puzzle by asking whether standard, widely used machine-learning (ML) algorithms convincingly improve exchange rate forecasting once evaluation is disciplined and implementation is made robust. Using monthly data from January 1986 to February 2025, we study US dollar to British pound as the baseline case (in both levels and monthly percent changes). We compare five ML methods -- random forests, neural networks, LASSO, gradient boosting, and linear support-vector classification -- against canonical benchmarks (random walk and ARIMA) in a rolling one-step-ahead out-of-sample forecasting design. To mitigate sensitivity to stochastic estimation, we average forecasts across multiple random seeds and assess performance using RMSE and Diebold-Mariano tests. We find that ML does not improve level forecasts and typically underperforms ARIMA. For exchange-rate changes, ML methods consistently outperform the random-walk benchmark, but only neural networks -- under a specific design -- reliably beat ARIMA. A theory-based UIP/PPP filtering approach improves accuracy for both ML and univariate methods, yet does not change the overall ranking. Extensive robustness checks across windows, currencies, frequencies, and tuning choices confirm that ML’s advantages are limited and fragile relative to conventional univariate benchmarks. |
| Keywords: | Machine learning, exchange rates, forecasting, theoretical filtering, random walk, ARIMA. |
| JEL: | C53 F31 F37 G17 |
| Date: | 2026–01 |
| URL: | https://d.repec.org/n?u=RePEc:cfl:wpaper:2026-01ua |
| By: | Molitor, Cullen; Cohen, Juliet; Lewin, Grace; Cognac, Steven; Hadunka, Protensia; Proctor, Jonathan; Carleton, Tamma |
| Abstract: | Recent innovations in task-agnostic imagery featurization have lowered the computational costs of using machine learning to predict ground conditions from satellite imagery. These methods hold particular promise for the development of imagery-based monitoring systems in low-income regions, where data and computational resources can be limited. However, these relatively simple prediction pipelines have not been evaluated in developing-country contexts over time, limiting our understanding of their performance in practice. Here, we compute task-agnostic random convolutional features from satellite imagery and use linear ridge regression models to predict maize yields over space and time in Zambia, a country prone to severe droughts and crop failure. Leveraging Landsat and Sentinel 2 satellite constellations, in combination with district-level yield data, our model explains 83% of the out-of-sample maize yield variation from 2016 to 2021, slightly outperforming a model trained on Normalized Difference Vegetation Index (NDVI) features, a common remote sensing approach used by practitioners to monitor crop health. Our approach maintains an R2 score of 0.74 when predicting temporal variation alone, while the performance of the NDVI-based approach drops to an R2 of 0.39. Our findings imply that this task-agnostic featurization can be used to predict spatial and temporal variation in agricultural outcomes, even in contexts with limited ground truth data. More broadly, these results point to imagery-based monitoring as a promising tool for assisting agricultural planning and food security, even in contexts where computationally expensive methodologies remain out of reach. |
| Keywords: | 37 Earth Sciences (for-2020), 3704 Geoinformatics (for-2020), Machine Learning and Artificial Intelligence (rcdc), Networking and Information Technology R&D (NITRD) (rcdc), Generic health relevance (hrcs-hc), 2 Zero Hunger (sdg), maize, yield prediction, Landsat, Sentinel, MOSAIKS, Zambia, 0203 Classical Physics (for), 0406 Physical Geography and Environmental Geoscience (for), 0909 Geomatic Engineering (for), 3701 Atmospheric sciences (for-2020), 3709 Physical geography and environmental geoscience (for-2020), 4013 Geomatic engineering (for-2020) |
| Date: | 2025–01–01 |
| URL: | https://d.repec.org/n?u=RePEc:cdl:agrebk:qt9c92m41k |
| By: | Ludovic Goudenège (Université Paris-Saclay); Andrea Molent (Università degli Studi di Udine - University of Udine [Italie]); Antonino Zanette (MATHRISK - Mathematical Risk Handling - UPEM - Université Paris-Est Marne-la-Vallée - Centre Inria de Paris - Inria - Institut National de Recherche en Informatique et en Automatique - ENPC - École nationale des ponts et chaussées - IP Paris - Institut Polytechnique de Paris) |
| Abstract: | Total value adjustment (XVA) is the change in value to be added to the price of a derivative to account for the bilateral default risk and the funding costs. In this paper, we compute such a premium for American basket derivatives whose payoff depends on multiple underlyings. In particular, in our model, those underlyings are supposed to follow the multidimensional Black-Scholes stochastic model. In order to determine the XVA, we follow the approach introduced by (Burgard and Kjaer in SSRN Electronic J 7:1–19, 2010) and afterward applied by (Arregui et al. in Appl Math Comput 308:31–53, 2017), (Arregui et al. in Int J Comput Math 96:2157–2176, 2019) for the one-dimensional American derivatives. The evaluation of the XVA for basket derivatives is particularly challenging as the presence of several underlings leads to a high-dimensional control problem. We tackle such an obstacle by resorting to Gaussian Process Regression, a machine learning technique that allows one to address the curse of dimensionality effectively. Moreover, the use of numerical techniques, such as control variates, turns out to be a powerful tool to improve the accuracy of the proposed methods. The paper includes the results of several numerical experiments that confirm the goodness of the proposed methodologies. |
| Keywords: | Control variates, Basket option, Gaussian process regression, XVA, American options, Transaction costs, Greeks, Hedging |
| Date: | 2025–08–08 |
| URL: | https://d.repec.org/n?u=RePEc:hal:journl:hal-05421581 |
| By: | Niclas Griesshaber; Jochen Streb |
| Abstract: | We leverage multimodal large language models (LLMs) to construct a dataset of 306, 070 German patents (1877-1918) from 9, 562 archival image scans using our LLM-based pipeline powered by Gemini-2.5-Pro and Gemini-2.5-Flash-Lite. Our benchmarking exercise provides tentative evidence that multimodal LLMs can create higher quality datasets than our research assistants, while also being more than 795 times faster and 205 times cheaper in constructing the patent dataset from our image corpus. About 20 to 50 patent entries are embedded on each page, arranged in a double-column format and printed in Gothic and Roman fonts. The font and layout complexity of our primary source material suggests to us that multimodal LLMs are a paradigm shift in how datasets are constructed in economic history. We open-source our benchmarking and patent datasets as well as our LLM-based data pipeline, which can be easily adapted to other image corpora using LLM-assisted coding tools, lowering the barriers for less technical researchers. Finally, we explain the economics of deploying LLMs for historical dataset construction and conclude by speculating on the potential implications for the field of economic history. |
| Date: | 2025–12 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2512.19675 |
| By: | Petr Koklev |
| Abstract: | Financial institutions face a trade-off between predictive accuracy and interpretability when deploying machine learning models for credit risk. Monotonicity constraints align model behavior with domain knowledge, but their performance cost - the price of monotonicity - is not well quantified. This paper benchmarks monotone-constrained versus unconstrained gradient boosting models for credit probability of default across five public datasets and three libraries. We define the Price of Monotonicity (PoM) as the relative change in standard performance metrics when moving from unconstrained to constrained models, estimated via paired comparisons with bootstrap uncertainty. In our experiments, PoM in AUC ranges from essentially zero to about 2.9 percent: constraints are almost costless on large datasets (typically less than 0.2 percent, often indistinguishable from zero) and most costly on smaller datasets with extensive constraint coverage (around 2-3 percent). Thus, appropriately specified monotonicity constraints can often deliver interpretability with small accuracy losses, particularly in large-scale credit portfolios. |
| Date: | 2025–12 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2512.17945 |
| By: | Mousavi, Ebrahim; Zare, Hassan; Moula, Ahmad |
| Abstract: | Sentiment analysis in tourism platforms plays a vital role in understanding customer feedback, enhancing service quality, and supporting strategic economic decision-making across tourism markets. Challenges such as imbalanced sentiment classes, domain-specific language, and noisy data reduce the economic efficiency and analytical value of conventional approaches. This paper introduces a novel hybrid framework that combines lexicon-based sentiment and emotion analysis with an economically optimized weighted kNearest Neighbors (kNN) classifier. The framework incorporates advanced data augmentation techniques and comprehensive feature engineering, including n-gram TF-IDF extraction and metric learning—to improve minority sentiment class recognition and increase the economic robustness of predictive analytics. A modified co-optimization layer jointly tunes augmentation parameters, feature extraction methods, and classifier hyperparameters to maximize minority-class F1-scores while minimizing computational and economic costs. Experimental evaluations on real-world tourism review datasets demonstrate significant improvements in classification performance compared to baseline models such as SVM, Random Forest, and CNN, highlighting the framework’s economic value in large-scale tourism data processing. Additionally, a real-time business intelligence dashboard is developed for economic monitoring and dynamic visualization of sentiment trends and minorityclass heatmaps, enabling tourism stakeholders to make informed economic and managerial decisions and strategically respond to customer sentiments. The findings confirm a predominance of positive sentiments across tourism services while identifying economically critical areas requiring improvement. Future work will explore multilingual sentiment analysis and aspect-based models to enhance granularity, scalability, and economic impact. This research contributes an effective, interpretable, and economically oriented solution for advanced sentiment analysis in tourism platforms. |
| Keywords: | Sentiment Analysis; Tourism Economic Platforms; Data Augmentation; kNN; Optimization |
| JEL: | A1 A10 O1 O10 R1 R10 R15 |
| Date: | 2025–06–13 |
| URL: | https://d.repec.org/n?u=RePEc:pra:mprapa:127063 |
| By: | Benjamin Born; Nora Lamersdorf; Jana-Lynn Schuster; Sascha Steffen |
| Abstract: | Using modern natural language processing, we construct a high-frequency inflation expectations index from German-language tweets. This index closely tracks realized inflation and aligns even more closely with household survey expectations. It also improves short-run forecasts relative to standard benchmarks. In response to monetary policy tightening, the index declines within about a week, with the effects concentrated in tweets by private individuals and during the recent period of elevated inflation. Using 117 million online transactions from German retailers, we show that higher inflation expectations are followed by lower household spending on discretionary goods. By linking these shifts in demand to stock returns, we find that, during periods of elevated inflation, firms operating in discretionary sectors experience significantly lower stock returns when inflation expectations rise. Thus, our Twitter-based index provides market participants and policymakers with a timely tool to monitor inflation sentiment and its economic consequences. |
| Keywords: | Inflation expectations, social media (Twitter/X), large language models (LLMs), NLP, household consumption, stock returns, monetary policy |
| JEL: | E31 D84 E58 C45 C81 |
| Date: | 2026–01 |
| URL: | https://d.repec.org/n?u=RePEc:bon:boncrc:crctr224_2025_724 |
| By: | Hayo, Bernd; Zahner, Johannes |
| Abstract: | We investigate the relationship between parliamentary debates and public expenditure by mapping legislative speeches to fiscally relevant topics and examining their connection in both long-term trends and short-term adjustments. Our analysis draws on transcripts of federal legislative discussions and federal government spending data in Germany (1950-2020), classified into nine policy functions (e.g. Social Security, National Defence and Education). We apply a state-of-the-art natural language processing technique - a structural topic model - to match identified debate topics to corresponding spending functions. Using cointegration analysis and error-correction models, we find (i) significant long-term equilibria between parliamentary debates and corresponding fiscal expenditure and (ii) that in cases of short-term disequilibrium, adjustments occur through government expenditure; that is, parliamentary debates are weakly exogenous. |
| Keywords: | Fiscal expenditure, parliamentary debate, Bundestag, text analysis, structural topic model, error-correction model |
| JEL: | E62 C32 D78 |
| Date: | 2025 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:imfswp:334478 |
| By: | Zare, Hassan; Mousavi, Ebrahim |
| Abstract: | This paper presents a novel deep learning framework for video analysis focused on automated key object detection and tactical action recognition within economic activity contexts. The proposed system integrates enhanced motion estimation for robust tracking of functional objects and state-of-the art 3D pose estimation to extract participant postures relevant to economic decision-making behavior. A deep semantic tactical ontology is employed to model the complex relationships between individuals, objects, and their actions, enabling interpretable and rule-based tactical insight extraction for economic interaction patterns beyond conventional classification. Evaluations conducted on benchmark datasets demonstrate high accuracy with approximately 91% in object detection and 96% in action recognition, highlighting the framework’s applicability to dynamic economic environments involving multi-agent interactions. Comparative analysis against baseline methods shows the effectiveness of the framework in handling complex scenarios with occlusions and rapidly changing economic behaviors. Future work will focus on enhancing preprocessing techniques, automating ontology rule learning, and extending the approach to a wider range of economically oriented domains. This research contributes to advancing intelligent analytics by bridging deep learning with semantic reasoning, fostering improved real-time tactical feedback and decision support in economic environments. |
| Keywords: | Economic Video Analysis; Tactical Action Recognition; Deep Learning; Semantic Ontology; 3D Pose Estimation |
| JEL: | C0 C01 L0 L00 P0 R1 R11 R13 |
| Date: | 2025–03–08 |
| URL: | https://d.repec.org/n?u=RePEc:pra:mprapa:127062 |
| By: | Maryam Arif; Soban Saeed |
| Abstract: | Our study investigates the interplay between young women's empowerment and Pakistan's economic growth, focusing on how social media use enhances their businesses and drives economic advancement. We utilize a mixed-methods research design, integrating both online and offline random sampling, for our survey of 51 respondents. We also utilized existing datasets consisting of both social media usage (n = 1000) and entrepreneurship (n = 1092). Our analysis identifies distinct social media engagement patterns via unsupervised learning and applies supervised models for entrepreneurship prediction, with logistic regression outperforming all other algorithms in terms of predictive accuracy and stability. In social media use, the cluster analysis reveals that at K=2, users form tightly packed, well-separated engagement groups. The results indicate that 39.4 percent of respondents believe social media positively impacts the economy by enabling businesses to generate increased revenue. However, only 14 percent of respondents participate in entrepreneurship, highlighting a substantial gap between digital engagement and business adoption. The analysis indicates that daily social media consumption is widespread with YouTube (66.7 percent) and WhatsApp (62.7 percent) being the most frequently used platforms. Key barriers identified are online harassment, limited digital literacy, and cultural constraints in a patriarchal society such as Pakistan. Additionally, 52.9 percent of respondents are unaware of government initiatives supporting women entrepreneurs, indicating limited policy outreach. |
| Date: | 2025–12 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2512.12685 |
| By: | Siddhartha Chib; Fei Tan |
| Abstract: | We show how state-of-the-art large language models (LLMs), seemingly inapplicable to the small samples typical of macroeconomics, can be trained to learn the language of macroeconomy. We estimate a large-scale dynamic stochastic general equilibrium (DSGE) model on an initial segment of the data and obtain a posterior distribution over structural parameters. We sample from this posterior to generate millions of theory-consistent synthetic panels that, when mixed with actual macroeconomic data, form the training corpus for a time-series transformer with attention. The trained model is then used to forecast out-of-sample through 2025. The results show that this hybrid forecaster, which combines the theoretical coherence of DSGE models with the representational power of modern LLMs, successfully learns the macroeconomic language. |
| Date: | 2025–12 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2512.21031 |
| By: | Foltas, Alexander |
| Abstract: | This paper contributes to previous research on the efficient integration of forecasters' narratives into business cycle forecasts. Using a Bidirectional Encoder Representations from Transformers (BERT) model, I quantify 19, 300 paragraphs from German business cycle reports (1998-2021) and use them to predict the direction of consumption forecast errors. By testing the model on an evaluation sample, I find a highly significant correlation of modest strength between predicted and actual sign of the forecast error. The correlation coefficient is substantially higher for 12.8% of paragraphs with a predicted class probability of 85% or higher. By qualitatively reviewing 150 of such high-probability paragraphs, I find recurring narratives correlated with consumption forecast errors. Underestimations of consumption growth often mention rising employment, increasing wages and transfer payments, low inflation, decreasing taxes, crisis-related fiscal support, and reduced relevance of marginal employment. Conversely, overestimated consumption forecasts present opposing narratives. Forecasters appear to particularly underestimate these factors when they disproportionately affect low-income households. |
| Abstract: | Diese Studie leistet einen Beitrag zur bisherigen Forschung hinsichtlich der effizienten Einbindung von Prognosenarrativen in Konjunkturprognosen. Unter Verwendung eines BERT-Modells (Bidirectional Encoder Representations from Transformers) quantifiziere ich 19.300 Absätze aus deutschen Konjunkturberichten (1998-2021) und nutze diese um die Richtung von Konsumprognosefehlern vorherzusagen. Durch die Überprüfung des Modells anhand einer Evaluationsstichprobe stelle ich eine hochsignifikante Korrelation moderater Stärke zwischen dem vorhergesagten und dem tatsächlichen Vorzeichen des Prognosefehlers fest. Der Korrelationskoeffizient ist für jene 12, 8 % der Absätze wesentlich höher, die eine vorhergesagte Klassenwahrscheinlichkeit von 85 % oder mehr aufweisen. Eine qualitative Untersuchung von 150 dieser Absätze mit hoher Wahrscheinlichkeit zeigt wiederkehrende Narrative, die mit Fehlern in der Konsumprognose korrelieren. Unterschätzungen des Konsumwachstums erwähnen häufig steigende Beschäftigung, zunehmende Löhne und Transferzahlungen, niedrige Inflation, sinkende Steuern, krisenbedingte fiskalische Unterstützung sowie eine abnehmende Bedeutung geringfügiger Beschäftigung. Umgekehrt weisen überschätzte Konsumprognosen gegensätzliche Narrative auf. Prognostiker scheinen diese Faktoren insbesondere dann zu unterschätzen, wenn sie einkommensschwache Haushalte überproportional betreffen. |
| Keywords: | Macroeconomic forecasting, Evaluating forecasts, Business cycles, Consumption forecasting, Natural language processing, Language Modeling, Machine learning, Judgmental forecasting |
| Date: | 2025 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:hwwiwp:334496 |
| By: | Ming Gu; David Hirshleifer; Siew Hong Teoh; Shijia Wu |
| Abstract: | We study dynamic visual representations as a proxy for investor sentiment about the stock market. Our sentiment index, GIFsentiment, is constructed from millions of posts in the Graphics Interchange Format (GIF) on a leading investment social media platform. GIFsentiment correlates with seasonal mood variations and the severity of COVID lockdowns. It is positively associated with contemporaneous market returns and negatively predicts returns for up to four weeks, even after controlling for other sentiment and attention measures. These effects are stronger among portfolios that are more susceptible to mispricing. GIFsentiment positively predicts trading volume, market volatility, and flows toward equity funds and away from debt funds. Our evidence suggests that GIFsentiment is a proxy for misperceptions that are later corrected. |
| Date: | 2025–12 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2512.20027 |
| By: | Drin, Svitlana (Örebro University School of Business); Zhuravlova, Anastasiia (National University of Kyiv-Mohyla Academy) |
| Abstract: | imely assessment of regional economic activity in Ukraine is severely constrained by institutional and data-related limitations. Official regional gross regional product (GRP) statistics are available only at low frequency, are published with substantial delays, and, in the post-2022 period, are further affected by disruptions to statistical production caused by martial law. At the same time, a growing set of potentially informative regional indicators derived from administrative records and official short-term statistics is available at higher frequencies but only over short and heterogeneous time spans. These features make the direct application of standard regional nowcasting models infeasible. This paper develops a mixed-frequency factor-augmented vector autoregressive framework tailored to the Ukrainian data environment and designed to incorporate short and incomplete regional indicators into the nowcasting of regional GDP. The model explicitly exploits the hierarchical structure of Ukrainian regional statistics by combining information from quarterly and annual measures of economic activity and by linking regional dynamics to national output developments. Short regional indicators are summarised through latent regional factors extracted using missing-data factor estimation techniques that are robust to ragged edges at both the beginning and the end of the sample. The proposed framework is implemented using Ukrainian macro-regional aggregates constructed from official data published by the State Statistics Service of Ukraine. Particular attention is paid to the treatment of labour market indicators, housing price dynamics, and other short-term variables that exhibit discontinuities or limited availability. A pseudo-real-time nowcasting exercise shows that conditioning regional GDP nowcasts on factor information derived from short regional data improves predictive performance when contemporaneous national GDP estimates are not yet available. Once national aggregates are released, the marginal informational contribution of regional short-term indicators diminishes. Overall, the results demonstrate that mixed-frequency factor-augmented VAR models provide a coherent and empirically viable framework for regional GDP nowcasting in Ukraine. The approach is particularly well suited to data environments 1 characterised by short samples, publication delays, and institutional disruptions, and thus offers a valuable tool for real-time regional economic monitoring in periods of heightened uncertainty. |
| Keywords: | MF-FAVAR; FAVAR; Nowcasting; EMPCA; GRP; Google Trends |
| JEL: | C53 E37 |
| Date: | 2026–01–02 |
| URL: | https://d.repec.org/n?u=RePEc:hhs:oruesi:2026_001 |
| By: | Lu, Qinan; Karwowski, Nicole; Liu, Pengfei; Wu, Karin |
| Abstract: | As renewable energy development accelerates, wind turbines are increasingly being installed on agricultural land, raising questions about their effects on crop production. This paper investigates the impact of wind turbine installations on agricultural productivity using a high-resolution dataset that combines parcel-level corn yield data with detailed information on wind turbine locations and weather characteristics. Using Difference-in-differences approach to address potential endogeneity, we find that parcels within an 8-kilometer radius of wind turbines experienced, on average, a 1% increase in corn yield after installation. These results suggest that localized microclimatic changes induced by turbines may improve growing conditions. Our findings highlight an overlooked positive externality of renewable energy infrastructure and underscore the importance of incorporating land-use interactions into energy policy and planning. |
| Keywords: | Agricultural and Food Policy |
| Date: | 2025 |
| URL: | https://d.repec.org/n?u=RePEc:ags:aaea25:360609 |