Dynamic AMM pools, as found in Temporal Function Market Making, rebalance their holdings to a new desired ratio (e.g. moving from being 50-50 between two assets to being 90-10 in favour of one of them) by introducing an arbitrage opportunity that disappears when their holdings are in line with their target. Structuring this arbitrage opportunity reduces to the problem of choosing the sequence of portfolio weights the pool exposes to the market via its trading function. Linear interpolation from start weights to end weights has been used to reduce the cost paid by pools to arbitrageurs to rebalance. Here we obtain the $textit{optimal}$ interpolation in the limit of small weight changes (which has the downside of requiring a call to a transcendental function) and then obtain a cheap-to-compute approximation to that optimal approach that gives almost the same performance improvement. We then demonstrate this method on a range of market backtests, including simulating pool performance when trading fees are present, finding that the new approximately-optimal method of changing weights gives robust increases in pool performance. For a BTC-ETH-DAI pool from July 2022 to June 2023, the increases of pool P&L from approximately-optimal weight changes is $sim25%$ for a range of different strategies and trading fees.
{"title":"Optimal Rebalancing in Dynamic AMMs","authors":"Matthew Willetts, Christian Harrington","doi":"arxiv-2403.18737","DOIUrl":"https://doi.org/arxiv-2403.18737","url":null,"abstract":"Dynamic AMM pools, as found in Temporal Function Market Making, rebalance\u0000their holdings to a new desired ratio (e.g. moving from being 50-50 between two\u0000assets to being 90-10 in favour of one of them) by introducing an arbitrage\u0000opportunity that disappears when their holdings are in line with their target.\u0000Structuring this arbitrage opportunity reduces to the problem of choosing the\u0000sequence of portfolio weights the pool exposes to the market via its trading\u0000function. Linear interpolation from start weights to end weights has been used\u0000to reduce the cost paid by pools to arbitrageurs to rebalance. Here we obtain\u0000the $textit{optimal}$ interpolation in the limit of small weight changes\u0000(which has the downside of requiring a call to a transcendental function) and\u0000then obtain a cheap-to-compute approximation to that optimal approach that\u0000gives almost the same performance improvement. We then demonstrate this method\u0000on a range of market backtests, including simulating pool performance when\u0000trading fees are present, finding that the new approximately-optimal method of\u0000changing weights gives robust increases in pool performance. For a BTC-ETH-DAI\u0000pool from July 2022 to June 2023, the increases of pool P&L from\u0000approximately-optimal weight changes is $sim25%$ for a range of different\u0000strategies and trading fees.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140311683","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We reassess Boehmer et al. (2021, BJZZ)'s seminal work on the predictive power of retail order imbalance (ROI) for future stock returns. First, we replicate their 2010-2015 analysis in the more recent 2016-2021 period. We find that the ROI's predictive power weakens significantly. Specifically, past ROI can no longer predict weekly returns on large-cap stocks, and the long-short strategy based on past ROI is no longer profitable. Second, we analyze the effect of using the alternative quote midpoint (QMP) method to identify and sign retail trades on their main conclusions. While the results based on the QMP method align with BJZZ's findings in 2010-2015, the two methods provide different conclusions in 2016-2021. Our study shows that BJZZ's original findings are sensitive to the sample period and the approach to identify ROIs.
{"title":"Revisiting Boehmer et al. (2021): Recent Period, Alternative Method, Different Conclusions","authors":"David Ardia, Clément Aymard, Tolga Cenesizoglu","doi":"arxiv-2403.17095","DOIUrl":"https://doi.org/arxiv-2403.17095","url":null,"abstract":"We reassess Boehmer et al. (2021, BJZZ)'s seminal work on the predictive\u0000power of retail order imbalance (ROI) for future stock returns. First, we\u0000replicate their 2010-2015 analysis in the more recent 2016-2021 period. We find\u0000that the ROI's predictive power weakens significantly. Specifically, past ROI\u0000can no longer predict weekly returns on large-cap stocks, and the long-short\u0000strategy based on past ROI is no longer profitable. Second, we analyze the\u0000effect of using the alternative quote midpoint (QMP) method to identify and\u0000sign retail trades on their main conclusions. While the results based on the\u0000QMP method align with BJZZ's findings in 2010-2015, the two methods provide\u0000different conclusions in 2016-2021. Our study shows that BJZZ's original\u0000findings are sensitive to the sample period and the approach to identify ROIs.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140316775","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
National football teams increasingly issue tradeable blockchain-based fan tokens to strategically enhance fan engagement. This study investigates the impact of 2022 World Cup matches on the dynamic performance of each team's fan token. The event study uncovers fan token returns surged six months before the World Cup, driven by positive anticipation effects. However, intraday analysis reveals a reversal of fan token returns consistently declining and trading volumes rising as matches unfold. To explain findings, we uncover asymmetries whereby defeats in high-stake matches caused a plunge in fan token returns, compared to low-stake matches, intensifying in magnitude for knockout matches. Contrarily, victories enhance trading volumes, reflecting increased market activity without a corresponding positive effect on returns. We align findings with the classic market adage "buy the rumor, sell the news," unveiling cognitive biases and nuances in investor sentiment, cautioning the dichotomy of pre-event optimism and subsequent performance declines.
{"title":"Anticipatory Gains and Event-Driven Losses in Blockchain-Based Fan Tokens: Evidence from the FIFA World Cup","authors":"Aman Saggu, Lennart Ante, Ender Demir","doi":"arxiv-2403.15810","DOIUrl":"https://doi.org/arxiv-2403.15810","url":null,"abstract":"National football teams increasingly issue tradeable blockchain-based fan\u0000tokens to strategically enhance fan engagement. This study investigates the\u0000impact of 2022 World Cup matches on the dynamic performance of each team's fan\u0000token. The event study uncovers fan token returns surged six months before the\u0000World Cup, driven by positive anticipation effects. However, intraday analysis\u0000reveals a reversal of fan token returns consistently declining and trading\u0000volumes rising as matches unfold. To explain findings, we uncover asymmetries\u0000whereby defeats in high-stake matches caused a plunge in fan token returns,\u0000compared to low-stake matches, intensifying in magnitude for knockout matches.\u0000Contrarily, victories enhance trading volumes, reflecting increased market\u0000activity without a corresponding positive effect on returns. We align findings\u0000with the classic market adage \"buy the rumor, sell the news,\" unveiling\u0000cognitive biases and nuances in investor sentiment, cautioning the dichotomy of\u0000pre-event optimism and subsequent performance declines.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140302526","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Thanos Konstantinidis, Giorgos Iacovides, Mingxue Xu, Tony G. Constantinides, Danilo Mandic
There are multiple sources of financial news online which influence market movements and trader's decisions. This highlights the need for accurate sentiment analysis, in addition to having appropriate algorithmic trading techniques, to arrive at better informed trading decisions. Standard lexicon based sentiment approaches have demonstrated their power in aiding financial decisions. However, they are known to suffer from issues related to context sensitivity and word ordering. Large Language Models (LLMs) can also be used in this context, but they are not finance-specific and tend to require significant computational resources. To facilitate a finance specific LLM framework, we introduce a novel approach based on the Llama 2 7B foundational model, in order to benefit from its generative nature and comprehensive language manipulation. This is achieved by fine-tuning the Llama2 7B model on a small portion of supervised financial sentiment analysis data, so as to jointly handle the complexities of financial lexicon and context, and further equipping it with a neural network based decision mechanism. Such a generator-classifier scheme, referred to as FinLlama, is trained not only to classify the sentiment valence but also quantify its strength, thus offering traders a nuanced insight into financial news articles. Complementing this, the implementation of parameter-efficient fine-tuning through LoRA optimises trainable parameters, thus minimising computational and memory requirements, without sacrificing accuracy. Simulation results demonstrate the ability of the proposed FinLlama to provide a framework for enhanced portfolio management decisions and increased market returns. These results underpin the ability of FinLlama to construct high-return portfolios which exhibit enhanced resilience, even during volatile periods and unpredictable market events.
{"title":"FinLlama: Financial Sentiment Classification for Algorithmic Trading Applications","authors":"Thanos Konstantinidis, Giorgos Iacovides, Mingxue Xu, Tony G. Constantinides, Danilo Mandic","doi":"arxiv-2403.12285","DOIUrl":"https://doi.org/arxiv-2403.12285","url":null,"abstract":"There are multiple sources of financial news online which influence market\u0000movements and trader's decisions. This highlights the need for accurate\u0000sentiment analysis, in addition to having appropriate algorithmic trading\u0000techniques, to arrive at better informed trading decisions. Standard lexicon\u0000based sentiment approaches have demonstrated their power in aiding financial\u0000decisions. However, they are known to suffer from issues related to context\u0000sensitivity and word ordering. Large Language Models (LLMs) can also be used in\u0000this context, but they are not finance-specific and tend to require significant\u0000computational resources. To facilitate a finance specific LLM framework, we\u0000introduce a novel approach based on the Llama 2 7B foundational model, in order\u0000to benefit from its generative nature and comprehensive language manipulation.\u0000This is achieved by fine-tuning the Llama2 7B model on a small portion of\u0000supervised financial sentiment analysis data, so as to jointly handle the\u0000complexities of financial lexicon and context, and further equipping it with a\u0000neural network based decision mechanism. Such a generator-classifier scheme,\u0000referred to as FinLlama, is trained not only to classify the sentiment valence\u0000but also quantify its strength, thus offering traders a nuanced insight into\u0000financial news articles. Complementing this, the implementation of\u0000parameter-efficient fine-tuning through LoRA optimises trainable parameters,\u0000thus minimising computational and memory requirements, without sacrificing\u0000accuracy. Simulation results demonstrate the ability of the proposed FinLlama\u0000to provide a framework for enhanced portfolio management decisions and\u0000increased market returns. These results underpin the ability of FinLlama to\u0000construct high-return portfolios which exhibit enhanced resilience, even during\u0000volatile periods and unpredictable market events.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140170702","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Eduardo Abi Jaber, Eyal Neuman, Sturmius Tuschmann
We consider a class of optimal portfolio choice problems in continuous time where the agent's transactions create both transient cross-impact driven by a matrix-valued Volterra propagator, as well as temporary price impact. We formulate this problem as the maximization of a revenue-risk functional, where the agent also exploits available information on a progressively measurable price predicting signal. We solve the maximization problem explicitly in terms of operator resolvents, by reducing the corresponding first order condition to a coupled system of stochastic Fredholm equations of the second kind and deriving its solution. We then give sufficient conditions on the matrix-valued propagator so that the model does not permit price manipulation. We also provide an implementation of the solutions to the optimal portfolio choice problem and to the associated optimal execution problem. Our solutions yield financial insights on the influence of cross-impact on the optimal strategies and its interplay with alpha decays.
{"title":"Optimal Portfolio Choice with Cross-Impact Propagators","authors":"Eduardo Abi Jaber, Eyal Neuman, Sturmius Tuschmann","doi":"arxiv-2403.10273","DOIUrl":"https://doi.org/arxiv-2403.10273","url":null,"abstract":"We consider a class of optimal portfolio choice problems in continuous time\u0000where the agent's transactions create both transient cross-impact driven by a\u0000matrix-valued Volterra propagator, as well as temporary price impact. We\u0000formulate this problem as the maximization of a revenue-risk functional, where\u0000the agent also exploits available information on a progressively measurable\u0000price predicting signal. We solve the maximization problem explicitly in terms\u0000of operator resolvents, by reducing the corresponding first order condition to\u0000a coupled system of stochastic Fredholm equations of the second kind and\u0000deriving its solution. We then give sufficient conditions on the matrix-valued\u0000propagator so that the model does not permit price manipulation. We also\u0000provide an implementation of the solutions to the optimal portfolio choice\u0000problem and to the associated optimal execution problem. Our solutions yield\u0000financial insights on the influence of cross-impact on the optimal strategies\u0000and its interplay with alpha decays.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140151679","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper studies the market structure impact of cheaper and faster chains on the Uniswap v3 Protocol. The Uniswap Protocol is the largest decentralized application on Ethereum by both gas and blockspace used, and user behaviors of the protocol are very sensitive to fluctuations in gas prices and market structure due to the economic factors of the Protocol. We focus on the chains where Uniswap v3 has the most activity, giving us the best comparison to Ethereum mainnet. Because of cheaper gas and lower block times, we find evidence that the majority of swaps get better gas-adjusted execution on these chains, liquidity providers are more capital efficient, and liquidity providers have increased fee returns from more arbitrage. We also present evidence that two second block times may be too long for optimal liquidity provider returns, compared to first come, first served. We argue that many of the current drawbacks with AMMs may be due to chain dynamics and are vastly improved with cheaper and faster transactions
{"title":"Layer 2 be or Layer not 2 be: Scaling on Uniswap v3","authors":"Austin Adams","doi":"arxiv-2403.09494","DOIUrl":"https://doi.org/arxiv-2403.09494","url":null,"abstract":"This paper studies the market structure impact of cheaper and faster chains\u0000on the Uniswap v3 Protocol. The Uniswap Protocol is the largest decentralized\u0000application on Ethereum by both gas and blockspace used, and user behaviors of\u0000the protocol are very sensitive to fluctuations in gas prices and market\u0000structure due to the economic factors of the Protocol. We focus on the chains\u0000where Uniswap v3 has the most activity, giving us the best comparison to\u0000Ethereum mainnet. Because of cheaper gas and lower block times, we find\u0000evidence that the majority of swaps get better gas-adjusted execution on these\u0000chains, liquidity providers are more capital efficient, and liquidity providers\u0000have increased fee returns from more arbitrage. We also present evidence that\u0000two second block times may be too long for optimal liquidity provider returns,\u0000compared to first come, first served. We argue that many of the current\u0000drawbacks with AMMs may be due to chain dynamics and are vastly improved with\u0000cheaper and faster transactions","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140151436","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We investigate a market with a normal-speed informed trader (IT) who may employ mixed strategy and multiple anticipatory high-frequency traders (HFTs) who are under different inventory pressures, in a three-period Kyle's model. The pure- and mixed-strategy equilibria are considered and the results provide recommendations for IT's randomization strategy with different numbers of HFTs. Some surprising results about investors' profits arise: the improvement of anticipatory traders' speed or a more precise prediction may harm themselves but help IT.
{"title":"Trading Large Orders in the Presence of Multiple High-Frequency Anticipatory Traders","authors":"Ziyi Xu, Xue Cheng","doi":"arxiv-2403.08202","DOIUrl":"https://doi.org/arxiv-2403.08202","url":null,"abstract":"We investigate a market with a normal-speed informed trader (IT) who may\u0000employ mixed strategy and multiple anticipatory high-frequency traders (HFTs)\u0000who are under different inventory pressures, in a three-period Kyle's model.\u0000The pure- and mixed-strategy equilibria are considered and the results provide\u0000recommendations for IT's randomization strategy with different numbers of HFTs.\u0000Some surprising results about investors' profits arise: the improvement of\u0000anticipatory traders' speed or a more precise prediction may harm themselves\u0000but help IT.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140127632","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper focuses on computing the fill probabilities for limit orders positioned at various price levels within the limit order book, which play a crucial role in optimizing executions. We adopt a generic stochastic model to capture the dynamics of the order book as a series of queueing systems. This generic model is state-dependent and also incorporates stylized factors. We subsequently derive semi-analytical expressions to compute the relevant probabilities within the context of state-dependent stochastic order flows. These probabilities cover various scenarios, including the probability of a change in the mid-price, the fill probabilities of orders posted at the best quotes, and those posted at a price level deeper than the best quotes in the book, before the opposite best quote moves. These expressions can be further generalized to accommodate orders posted even deeper in the order book, although the associated probabilities are typically very small in such cases. Lastly, we conduct extensive numerical experiments using real order book data from the foreign exchange spot market. Our findings suggest that the model is tractable and possesses the capability to effectively capture the dynamics of the limit order book. Moreover, the derived formulas and numerical methods demonstrate reasonably good accuracy in estimating the fill probabilities.
{"title":"Fill Probabilities in a Limit Order Book with State-Dependent Stochastic Order Flows","authors":"Felix Lokin, Fenghui Yu","doi":"arxiv-2403.02572","DOIUrl":"https://doi.org/arxiv-2403.02572","url":null,"abstract":"This paper focuses on computing the fill probabilities for limit orders\u0000positioned at various price levels within the limit order book, which play a\u0000crucial role in optimizing executions. We adopt a generic stochastic model to\u0000capture the dynamics of the order book as a series of queueing systems. This\u0000generic model is state-dependent and also incorporates stylized factors. We\u0000subsequently derive semi-analytical expressions to compute the relevant\u0000probabilities within the context of state-dependent stochastic order flows.\u0000These probabilities cover various scenarios, including the probability of a\u0000change in the mid-price, the fill probabilities of orders posted at the best\u0000quotes, and those posted at a price level deeper than the best quotes in the\u0000book, before the opposite best quote moves. These expressions can be further\u0000generalized to accommodate orders posted even deeper in the order book,\u0000although the associated probabilities are typically very small in such cases.\u0000Lastly, we conduct extensive numerical experiments using real order book data\u0000from the foreign exchange spot market. Our findings suggest that the model is\u0000tractable and possesses the capability to effectively capture the dynamics of\u0000the limit order book. Moreover, the derived formulas and numerical methods\u0000demonstrate reasonably good accuracy in estimating the fill probabilities.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140045531","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In recent years, there has been quick developments of derivative markets in China and standardized derivative trading have reached considerable volumes. In this research, we collect all the daily data of ETF options traded at Shanghai Stock Exchange and start with a simple short-volatility strategy. The strategy delivers nice performance before 2018, providing significant excess return over the buy and hold benchmark. However, after 2018, this strategy starts to deteriorate and no obvious risk-adjusted return is shown. Based on the discussion of relationship between the strategy's performance and market's volatility, we improve the model by adjusting positions and exposure according to volatility forecasts using methods such as volatility momentum and GARCH. The new models have improved performance in different ways, where larger upside capture and smaller drawbacks can be achieved in market fluctuation. This research has shown potentials of volatility-based trading on Chinese equity index options, and with further improvement and implementation considerations, real-world practical trading strategies can be formed.
{"title":"Volatility-based strategy on Chinese equity index ETF options","authors":"Peng Yifeng","doi":"arxiv-2403.00474","DOIUrl":"https://doi.org/arxiv-2403.00474","url":null,"abstract":"In recent years, there has been quick developments of derivative markets in\u0000China and standardized derivative trading have reached considerable volumes. In\u0000this research, we collect all the daily data of ETF options traded at Shanghai\u0000Stock Exchange and start with a simple short-volatility strategy. The strategy\u0000delivers nice performance before 2018, providing significant excess return over\u0000the buy and hold benchmark. However, after 2018, this strategy starts to\u0000deteriorate and no obvious risk-adjusted return is shown. Based on the\u0000discussion of relationship between the strategy's performance and market's\u0000volatility, we improve the model by adjusting positions and exposure according\u0000to volatility forecasts using methods such as volatility momentum and GARCH.\u0000The new models have improved performance in different ways, where larger upside\u0000capture and smaller drawbacks can be achieved in market fluctuation. This\u0000research has shown potentials of volatility-based trading on Chinese equity\u0000index options, and with further improvement and implementation considerations,\u0000real-world practical trading strategies can be formed.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140026301","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Financial trading is a crucial component of the markets, informed by a multimodal information landscape encompassing news, prices, and Kline charts, and encompasses diverse tasks such as quantitative trading and high-frequency trading with various assets. While advanced AI techniques like deep learning and reinforcement learning are extensively utilized in finance, their application in financial trading tasks often faces challenges due to inadequate handling of multimodal data and limited generalizability across various tasks. To address these challenges, we present FinAgent, a multimodal foundational agent with tool augmentation for financial trading. FinAgent's market intelligence module processes a diverse range of data-numerical, textual, and visual-to accurately analyze the financial market. Its unique dual-level reflection module not only enables rapid adaptation to market dynamics but also incorporates a diversified memory retrieval system, enhancing the agent's ability to learn from historical data and improve decision-making processes. The agent's emphasis on reasoning for actions fosters trust in its financial decisions. Moreover, FinAgent integrates established trading strategies and expert insights, ensuring that its trading approaches are both data-driven and rooted in sound financial principles. With comprehensive experiments on 6 financial datasets, including stocks and Crypto, FinAgent significantly outperforms 9 state-of-the-art baselines in terms of 6 financial metrics with over 36% average improvement on profit. Specifically, a 92.27% return (a 84.39% relative improvement) is achieved on one dataset. Notably, FinAgent is the first advanced multimodal foundation agent designed for financial trading tasks.
{"title":"FinAgent: A Multimodal Foundation Agent for Financial Trading: Tool-Augmented, Diversified, and Generalist","authors":"Wentao Zhang, Lingxuan Zhao, Haochong Xia, Shuo Sun, Jiaze Sun, Molei Qin, Xinyi Li, Yuqing Zhao, Yilei Zhao, Xinyu Cai, Longtao Zheng, Xinrun Wang, Bo An","doi":"arxiv-2402.18485","DOIUrl":"https://doi.org/arxiv-2402.18485","url":null,"abstract":"Financial trading is a crucial component of the markets, informed by a\u0000multimodal information landscape encompassing news, prices, and Kline charts,\u0000and encompasses diverse tasks such as quantitative trading and high-frequency\u0000trading with various assets. While advanced AI techniques like deep learning\u0000and reinforcement learning are extensively utilized in finance, their\u0000application in financial trading tasks often faces challenges due to inadequate\u0000handling of multimodal data and limited generalizability across various tasks.\u0000To address these challenges, we present FinAgent, a multimodal foundational\u0000agent with tool augmentation for financial trading. FinAgent's market\u0000intelligence module processes a diverse range of data-numerical, textual, and\u0000visual-to accurately analyze the financial market. Its unique dual-level\u0000reflection module not only enables rapid adaptation to market dynamics but also\u0000incorporates a diversified memory retrieval system, enhancing the agent's\u0000ability to learn from historical data and improve decision-making processes.\u0000The agent's emphasis on reasoning for actions fosters trust in its financial\u0000decisions. Moreover, FinAgent integrates established trading strategies and\u0000expert insights, ensuring that its trading approaches are both data-driven and\u0000rooted in sound financial principles. With comprehensive experiments on 6\u0000financial datasets, including stocks and Crypto, FinAgent significantly\u0000outperforms 9 state-of-the-art baselines in terms of 6 financial metrics with\u0000over 36% average improvement on profit. Specifically, a 92.27% return (a 84.39%\u0000relative improvement) is achieved on one dataset. Notably, FinAgent is the\u0000first advanced multimodal foundation agent designed for financial trading\u0000tasks.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140009650","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：