|
on Market Microstructure |
| By: | Tomas Espana; Yadh Hafsi; Fabrizio Lillo; Edoardo Vittori |
| Abstract: | We investigate the use of Reinforcement Learning for the optimal execution of meta-orders, where the objective is to execute incrementally large orders while minimizing implementation shortfall and market impact over an extended period of time. Departing from traditional parametric approaches to price dynamics and impact modeling, we adopt a model-free, data-driven framework. Since policy optimization requires counterfactual feedback that historical data cannot provide, we employ the Queue-Reactive Model to generate realistic and tractable limit order book simulations that encompass transient price impact, and nonlinear and dynamic order flow responses. Methodologically, we train a Double Deep Q-Network agent on a state space comprising time, inventory, price, and depth variables, and evaluate its performance against established benchmarks. Numerical simulation results show that the agent learns a policy that is both strategic and tactical, adapting effectively to order book conditions and outperforming standard approaches across multiple training configurations. These findings provide strong evidence that model-free Reinforcement Learning can yield adaptive and robust solutions to the optimal execution problem. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.15262 |
| By: | Xuzhu ZHENG (Graduate School of Economics, The University of Osaka); Masato UBUKATA (The Faculty of Economics, Meiji Gakuin University); Kosuke OYA (Graduate School of Economics, The University of Osaka, Center for Mathematical Modeling and Data Science, The University of Osaka) |
| Abstract: | This study examines the existence of rough volatility, which has recently attracted considerable attention and is characterized by volatility dynamics that cannot be fully captured by conventional volatility models. Specifically, we investigate whether the observed roughness in volatility is merely an artifact induced by microstructure noise inherent in high-frequency price data, or whether such rough behavior persists even after accounting for the effects of noise. The empirical analysis utilizes high-frequency data from the Nikkei 225 index as well as two representatives, actively traded individual stocks. Applying several representative volatility estimation methods, we first construct volatility series and then estimate their Hurst exponents using a nonparametric estimation procedure proposed in the literature. Our results show that, regardless of the presence of microstructure noise, the estimated Hurst exponents consistently take low values, suggesting that the volatility processes under study exhibit rough behavior. These findings provide supporting evidence for the necessity of incorporating roughness into volatility modeling to achieve a more refined understanding of volatility dynamics in financial markets. |
| Keywords: | High-frequency data, Volatility, Roughness |
| JEL: | C14 C55 C58 |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:osk:wpaper:2517 |
| By: | Babolmorad, N.; Massoud, N. |
| Abstract: | We build a framework to examine how the training regime-rather than model architecture-drives the performance of financial sentiment models. Using firm-level news and parsimonious classifiers, we compare three supervision regimes: human-only, hybrid, and market-only (fully automated). The framework opens the "black box" of sentiment modeling by tracing how supervision shapes each component of the classifier. Across extensive tests, the hybrid regime consistently outperforms fully automated training in explaining variation in stock returns and trading volume, enhancing interpretability and economic relevance. Human input improves sentiment inference, offering new insights into information processing and price formation in financial markets. |
| Keywords: | Sentiment Analysis, Financial Media News, Investor Sentiment, Stock Markets, Human versus Machine |
| JEL: | G02 G11 G12 G14 |
| Date: | 2025–10–31 |
| URL: | https://d.repec.org/n?u=RePEc:cam:camdae:2577 |
| By: | Jianhao Su; Yanliang Zhang |
| Abstract: | The information released to investors in financial markets has various forms. We refer to range information as information about the upper and lower bound which the payoff of a risky asset may reach in the future. This study develops rational expectation models to explore the market impacts of disclosure of range information. Our model shows that its disclosure can decrease the sensitivity of market price to private signal and increase market liquidity. The market impact of its disclosure depends on the position and precision of the disclosed range. When the linear combination of private signal and noise trading volume is distant from the disclosed range, the reaction of price to a variation in private signal will almost vanish, whereas a movement of the disclosed range can affect the price efficiently. If the midpoint of the disclosed range is higher (lower) than a criterion which is specified in this study, the disclosure will reduce (raise) asset premium. |
| Date: | 2025–11 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2511.11405 |
| By: | Ferreira Batista Martins, Igor (Örebro University School of Business); Virbickaitè, Audronè (CUNEF University, Madrid, Spain); Nguyen, Hoang (Linköping University); Freitas Lopes, Hedibert (nsper Institute of Education and Research, Sao Paulo, Brazil) |
| Abstract: | We propose a high-frequency stochastic volatility model that integrates persistent component, intraday periodicity, and volume-driven time-of-day effects. By allowing intraday volatility patterns to respond to lagged trading activity, the model captures economically and statistically relevant departures from traditional intraday seasonality effects. We find that the volumedriven component accounts for a substantial share of intraday volatility for futures data across equity indexes, currencies, and commodities. Out-of-sample, our forecasts achieve near-zero intercepts, unit slopes, and the highest R2 values in Mincer-Zarnowitz regressions, while horserace regressions indicate that competing forecasts add little information once our predictions are included. These statistical improvements translate into economically meaningful gains, as volatility-managed portfolio strategies based on our model consistently improve Sharpe ratios. Our results highlight the value of incorporating lagged trading activity into high-frequency volatility models. |
| Keywords: | Intraday volatility; high-frequency; volume; periodicity. |
| JEL: | C11 C22 C53 C58 |
| Date: | 2025–11–21 |
| URL: | https://d.repec.org/n?u=RePEc:hhs:oruesi:2025_014 |
| 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 |