|
on Computational Economics |
| By: | Hongyang Yang; Xiao-Yang Liu; Shan Zhong; Anwar Walid |
| Abstract: | Stock trading strategies play a critical role in investment. However, it is challenging to design a profitable strategy in a complex and dynamic stock market. In this paper, we propose an ensemble strategy that employs deep reinforcement schemes to learn a stock trading strategy by maximizing investment return. We train a deep reinforcement learning agent and obtain an ensemble trading strategy using three actor-critic based algorithms: Proximal Policy Optimization (PPO), Advantage Actor Critic (A2C), and Deep Deterministic Policy Gradient (DDPG). The ensemble strategy inherits and integrates the best features of the three algorithms, thereby robustly adjusting to different market situations. In order to avoid the large memory consumption in training networks with continuous action space, we employ a load-on-demand technique for processing very large data. We test our algorithms on the 30 Dow Jones stocks that have adequate liquidity. The performance of the trading agent with different reinforcement learning algorithms is evaluated and compared with both the Dow Jones Industrial Average index and the traditional min-variance portfolio allocation strategy. The proposed deep ensemble strategy is shown to outperform the three individual algorithms and two baselines in terms of the risk-adjusted return measured by the Sharpe ratio. This work is fully open-sourced at \href{https://github.com/AI4Finance-Foun dation/Deep-Reinforcement-Learning-for-A utomated-Stock-Trading-Ensemble-Strategy -ICAIF-2020}{GitHub}. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.12120 |
| By: | Daniel Moulton; Larry Santucci; Robyn Smith |
| Abstract: | This paper examines methodological approaches for extracting structured data from large-scale historical document archives, comparing “hyperspecialized” versus “adaptive modular” strategies. Using 56 years of Philadelphia property deeds as a case study, we show the benefits of the adaptive modular approach leveraging optical character recognition (OCR), full-text search, and frontier large language models (LLMs) to identify deeds containing specific restrictive use language— achieving 98% precision and 90–98% recall. Our adaptive modular methodology enables analysis of historically important economic phenomena including re strictive property covenants, their precise geographic locations, and the localized neighborhood effects of these restrictions. This approach should be easily adapt able to other research involving deeds and similar document. |
| Keywords: | large language models (LLMs); artificial intelligence (AI); machine learning (ML); restrictive covenants; deeds; property; real estate; housing; John Coltrane; digitization |
| JEL: | C81 N32 R31 R38 |
| Date: | 2025–10–25 |
| URL: | https://d.repec.org/n?u=RePEc:fip:fedpdp:102114 |
| By: | Julian Lütticke; Lukas Lautenschlaeger; Wolfgang Schäfers |
| Abstract: | The role of investor sentiment in real estate investment trust (REIT) markets is well-documented. However, traditional sentiment indicators often fail to fully capture real-time market dynamics. This study explores the potential of GenAI-extracted sentiment in forecasting U.S. REIT returns by leveraging large language models (LLMs) to analyze textual data from news media sources. The hypothesis underpinning this study is that LLMs can process textual data in a manner analogous to that of humans. The novel sentiment score is integrated into a machine learning model to predict REIT returns. The analysis differentiates between overall index returns and sector-specific REIT performance, thereby providing a more granular view of sentiment-driven market behavior. In addition to traditional statistical metrics the model performance is assessed by evaluating an active trading strategy based on sentiment signals. This strategy is benchmarked against a buy-and-hold approach to determine whether sentiment-based predictions can systematically outperform the market. The findings contribute to the growing field of AI-driven financial forecasting and offer valuable insights for investors and policymakers in the indirect real estate sector. |
| Keywords: | Generative AI; Large Language Model; News Sentiment; REIT |
| JEL: | R3 |
| Date: | 2025–01–01 |
| URL: | https://d.repec.org/n?u=RePEc:arz:wpaper:eres2025_242 |
| By: | Qian'ang Mao; Yuxuan Zhang; Jiaman Chen; Wenjun Zhou; Jiaqi Yan |
| Abstract: | As Decentralized Finance (DeFi) develops, understanding user intent behind DeFi transactions is crucial yet challenging due to complex smart contract interactions, multifaceted on-/off-chain factors, and opaque hex logs. Existing methods lack deep semantic insight. To address this, we propose the Transaction Intent Mining (TIM) framework. TIM leverages a DeFi intent taxonomy built on grounded theory and a multi-agent Large Language Model (LLM) system to robustly infer user intents. A Meta-Level Planner dynamically coordinates domain experts to decompose multiple perspective-specific intent analyses into solvable subtasks. Question Solvers handle the tasks with multi-modal on/off-chain data. While a Cognitive Evaluator mitigates LLM hallucinations and ensures verifiability. Experiments show that TIM significantly outperforms machine learning models, single LLMs, and single Agent baselines. We also analyze core challenges in intent inference. This work helps provide a more reliable understanding of user motivations in DeFi, offering context-aware explanations for complex blockchain activity. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.15456 |
| By: | Emmanuel Lwele; Sabuni Emmanuel; Sitali Gabriel Sitali |
| Abstract: | This paper presents a deep reinforcement learning (DRL) framework for dynamic portfolio optimization under market uncertainty and risk. The proposed model integrates a Sharpe ratio-based reward function with direct risk control mechanisms, including maximum drawdown and volatility constraints. Proximal Policy Optimization (PPO) is employed to learn adaptive asset allocation strategies over historical financial time series. Model performance is benchmarked against mean-variance and equal-weight portfolio strategies using backtesting on high-performing equities. Results indicate that the DRL agent stabilizes volatility successfully but suffers from degraded risk-adjusted returns due to over-conservative policy convergence, highlighting the challenge of balancing exploration, return maximization, and risk mitigation. The study underscores the need for improved reward shaping and hybrid risk-aware strategies to enhance the practical deployment of DRL-based portfolio allocation models. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.11481 |
| By: | Iñaki Aldasoro; Ajit Desai |
| Abstract: | Using prompt-based experiments with ChatGPT's reasoning model, we evaluate whether a generative artificial intelligence (AI) agent can perform high-level intraday liquidity management in a wholesale payment system. We simulate payment scenarios with liquidity shocks and competing priorities to test the agent's ability to maintain precautionary liquidity buffers, dynamically prioritize payments under tight constraints, and optimize the trade-off between settlement speed and liquidity usage. Our results show that even without domain-specific training, the AI agent closely replicates key prudential cash-management practices, issuing calibrated recommendations that preserve liquidity while minimizing delays. These findings suggest that routine cash-management tasks could be automated using general-purpose large language models, potentially reducing operational costs and improving intraday liquidity efficiency. We conclude with a discussion of the regulatory and policy safeguards that central banks and supervisors may need to consider in an era of AI-driven payment operations. |
| Keywords: | generative AI, agentic AI, LLM, payments systems, liquidity management |
| JEL: | A12 C7 D83 E42 E58 |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:bis:biswps:1310 |
| By: | Luca Corazzini; Elisa Deriu; Marco Guerzoni |
| Abstract: | Large language models (LLMs) increasingly mediate economic and organisational processes, from automated customer support and recruitment to investment advice and policy analysis. These systems are often assumed to embody rational decision making free from human error; yet they are trained on human language corpora that may embed cognitive and social biases. This study investigates whether advanced LLMs behave as rational agents or whether they reproduce human behavioural tendencies when faced with classic decision problems. Using two canonical experiments in behavioural economics, the ultimatum game and a gambling game, we elicit decisions from two state of the art models, Google Gemma7B and Qwen, under neutral and gender conditioned prompts. We estimate parameters of inequity aversion and loss-aversion and compare them with human benchmarks. The models display attenuated but persistent deviations from rationality, including moderate fairness concerns, mild loss aversion, and subtle gender conditioned differences. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.12319 |
| By: | Fang, Qixiang; Garcia-Bernardo, Javier; van Kesteren, Erik-Jan |
| Abstract: | Large language models (LLMs) have become an essential tool for social scientists and humanities (SSH) researchers who work with textual data. One particularly valuable use case is automating text annotation, traditionally a time-consuming step in preparing data for empirical analysis. Yet, many SSH researchers face two challenges: getting started with LLMs, and understanding how to evaluate and correct for their limitations. The rapid pace of model development can make LLMs appear inaccessible or intimidating, while even experienced users may overlook how annotation errors can bias results from downstream analyses (e.g., regression estimates, $p$-values), even when accuracy appears high. This tutorial provides a step-by-step, hands-on guide to using LLMs for text annotation in SSH research for both Python and R users. We cover (1) how to choose and access LLM APIs, (2) how to design and run annotation tasks programmatically, (3) how to evaluate annotation quality and iterate on prompts, (4) how to integrate annotations into statistical workflows while accounting for uncertainty, and (5) how to manage cost, efficiency, and reproducibility. Throughout, we provide concrete examples, code snippets, and best-practice checklists to help researchers confidently and transparently incorporate LLM-based annotation into their workflows. |
| Date: | 2025–11–13 |
| URL: | https://d.repec.org/n?u=RePEc:osf:socarx:v4eq6_v1 |
| By: | Jianfei Cao; Michael P. Leung |
| Abstract: | This paper studies double/debiased machine learning (DML) methods applied to weakly dependent data. We allow observations to be situated in a general metric space that accommodates spatial and network data. Existing work implements cross-fitting by excluding from the training fold observations sufficiently close to the evaluation fold. We find in simulations that this can result in exceedingly small training fold sizes, particularly with network data. We therefore seek to establish the validity of DML without cross-fitting, building on recent work by Chen et al. (2022). They study i.i.d. data and require the machine learner to satisfy a natural stability condition requiring insensitivity to data perturbations that resample a single observation. We extend these results to dependent data by strengthening stability to "neighborhood stability, " which requires insensitivity to resampling observations in any slowly growing neighborhood. We show that existing results on the stability of various machine learners can be adapted to verify neighborhood stability. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.10995 |
| By: | Hongyang Yang; Xiao-Yang Liu; Qingwei Wu |
| Abstract: | Stock recommendation is vital to investment companies and investors. However, no single stock selection strategy will always win while analysts may not have enough time to check all S&P 500 stocks (the Standard & Poor's 500). In this paper, we propose a practical scheme that recommends stocks from S&P 500 using machine learning. Our basic idea is to buy and hold the top 20% stocks dynamically. First, we select representative stock indicators with good explanatory power. Secondly, we take five frequently used machine learning methods, including linear regression, ridge regression, stepwise regression, random forest and generalized boosted regression, to model stock indicators and quarterly log-return in a rolling window. Thirdly, we choose the model with the lowest Mean Square Error in each period to rank stocks. Finally, we test the selected stocks by conducting portfolio allocation methods such as equally weighted, mean-variance, and minimum-variance. Our empirical results show that the proposed scheme outperforms the long-only strategy on the S&P 500 index in terms of Sharpe ratio and cumulative returns. This work is fully open-sourced at \href{https://github.com/AI4Finance-Foun dation/Dynamic-Stock-Recommendation-Mach ine_Learning-Published-Paper-IEEE}{GitHu b}. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.12129 |
| By: | Heyang Ma; Qirui Mi; Qipeng Yang; Zijun Fan; Bo Li; Haifeng Zhang |
| Abstract: | Economic decision-making depends not only on structured signals such as prices and taxes, but also on unstructured language, including peer dialogue and media narratives. While multi-agent reinforcement learning (MARL) has shown promise in optimizing economic decisions, it struggles with the semantic ambiguity and contextual richness of language. We propose LAMP (Language-Augmented Multi-Agent Policy), a framework that integrates language into economic decision-making and narrows the gap to real-world settings. LAMP follows a Think-Speak-Decide pipeline: (1) Think interprets numerical observations to extract short-term shocks and long-term trends, caching high-value reasoning trajectories; (2) Speak crafts and exchanges strategic messages based on reasoning, updating beliefs by parsing peer communications; and (3) Decide fuses numerical data, reasoning, and reflections into a MARL policy to optimize language-augmented decision-making. Experiments in economic simulation show that LAMP outperforms both MARL and LLM-only baselines in cumulative return (+63.5%, +34.0%), robustness (+18.8%, +59.4%), and interpretability. These results demonstrate the potential of language-augmented policies to deliver more effective and robust economic strategies. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.12876 |
| By: | Priyanka Verma; Balagopal Unnikrishnan |
| Abstract: | Fair resource division algorithms, like those implemented in Spliddit platform, have traditionally been considered difficult for the end users to manipulate due to its complexities. This paper demonstrates how Large Language Models (LLMs) can dismantle these protective barriers by democratizing access to strategic expertise. Through empirical analysis of rent division scenarios on Spliddit algorithms, we show that users can obtain actionable manipulation strategies via simple conversational queries to AI assistants. We present four distinct manipulation scenarios: exclusionary collusion where majorities exploit minorities, defensive counterstrategies that backfire, benevolent subsidization of specific participants, and cost minimization coalitions. Our experiments reveal that LLMs can explain algorithmic mechanics, identify profitable deviations, and generate specific numerical inputs for coordinated preference misreporting--capabilities previously requiring deep technical knowledge. These findings extend algorithmic collective action theory from classification contexts to resource allocation scenarios, where coordinated preference manipulation replaces feature manipulation. The implications reach beyond rent division to any domain using algorithmic fairness mechanisms for resource division. While AI-enabled manipulation poses risks to system integrity, it also creates opportunities for preferential treatment of equity deserving groups. We argue that effective responses must combine algorithmic robustness, participatory design, and equitable access to AI capabilities, acknowledging that strategic sophistication is no longer a scarce resource. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.14722 |
| By: | Sofiia Dolgikh; Bodan Potanin |
| Abstract: | We propose plug-in (PI) and double machine learning (DML) estimators of average treatment effect (ATE), average treatment effect on the treated (ATET) and local average treatment effect (LATE) in the multivariate sample selection model with ordinal selection equations. Our DML estimators are doubly-robust and based on the efficient influence functions. Finite sample properties of the proposed estimators are studied and compared on simulated data. Specifically, the results of the analysis suggest that without addressing multivariate sample selection, the estimates of the causal parameters may be highly biased. However, the proposed estimators allow us to avoid these biases. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.12640 |
| By: | Ghulam Mustafa (Pakistan Institute of Development Economics); Muhammad Hamza Amjad |
| Abstract: | Blockchain technology has the potential to transform Pakistan`s industrial sector, particularly the real estate market, by enhancing transparency, reducing transaction costs, and minimising market frictions. While blockchain adoption is increasing globally to improve market efficiency, Pakistan remains behind in integrating this technology. This study aims to bridge this gap by simulating the impact of blockchain adoption on real estate market efficiency, focusing on Islamabad. Using an Agent-Based Modeling (ABM) approach, the study tests four key hypotheses: (i) the effect of blockchain on transparency and fraudulent cases, (ii) its impact on average sale time and liquidity, (iii) the role of price discovery in the absence of traditional dealers, and (iv) changes in transaction costs due to blockchain implementation. The simulation results reveal that blockchain adoption significantly enhances market efficiency by reducing asymmetric information, increasing transparency, improving liquidity, and drastically lowering transaction costs. Through tokenisation, smart contracts, and decentralised ledgers, blockchain disrupts the role of intermediaries, leading to a more efficient and transparent market. By eliminating dealercentric liquidity and introducing decentralised mechanisms, Islamabad`s real estate sector can achieve true price discovery based on supply-demand dynamics rather than speculation. However, the Capital Development Authority (CDA) plays a crucial role as both regulator and innovator. Balancing blockchain`s disruptive potential with existing legal and institutional frameworks is essential for its successful implementation in Pakistan`s real estate sector. |
| Keywords: | Agent-Based Model, Blockchain Technology, REAL ESTATE |
| Date: | 2025 |
| URL: | https://d.repec.org/n?u=RePEc:pid:wpaper:2025:12 |
| By: | Paul C. Behler (University of Bonn); Laurenz Guenther (Bocconi University, Toulouse School of Economics) |
| Abstract: | While the recent rise of populism has led many scholars to study populism in the modern era, its long-run evolution remains underexplored. This paper analyzes German parliamentary speeches to study populism over the long run, covering the Weimar Republic (1918–1933) and the united Federal Republic (1991–today). We employ a tailored and validated machine learning model to measure populism and dissect it into anti-elitism and people-centrism. We find that in both republics, populism is similarly common, similarly distributed across the ideological spectrum, and increases over time. Moreover, in both states, left-wing parties were initially the most populist group but were eventually overtaken by right-wing parties. However, we find a difference in the form of populism: in the Weimar Republic, the increase in populism is driven by a surge in the anti-elitism of right-wing parties, while in the Federal Republic, it is due to a general rise in people-centrism. |
| Keywords: | Populism, Nazi, Weimar, Radical, Democracy, Right-wing, Far-right, Machine learning, BERT, Text analysis, Rhetoric |
| JEL: | P16 N40 C89 |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:ajk:ajkdps:381 |