Recent advances in large language models (LLMs) have opened new possibilities for artificial intelligence applications in finance. In this paper, we provide a practical survey focused on two key aspects of utilizing LLMs for financial tasks: existing solutions and guidance for adoption. First, we review current approaches employing LLMs in finance, including leveraging pretrained models via zero-shot or few-shot learning, fine-tuning on domain-specific data, and training custom LLMs from scratch. We summarize key models and evaluate their performance improvements on financial natural language processing tasks. Second, we propose a decision framework to guide financial professionals in selecting the appropriate LLM solution based on their use case constraints around data, compute, and performance needs. The framework provides a pathway from lightweight experimentation to heavy investment in customized LLMs. Lastly, we discuss limitations and challenges around leveraging LLMs in financial applications. Overall, this survey aims to synthesize the state-of-the-art and provide a roadmap for responsibly applying LLMs to advance financial AI.
{"title":"Large Language Models in Finance: A Survey","authors":"Yinheng Li, Shaofei Wang, Han Ding, Hang Chen","doi":"arxiv-2311.10723","DOIUrl":"https://doi.org/arxiv-2311.10723","url":null,"abstract":"Recent advances in large language models (LLMs) have opened new possibilities\u0000for artificial intelligence applications in finance. In this paper, we provide\u0000a practical survey focused on two key aspects of utilizing LLMs for financial\u0000tasks: existing solutions and guidance for adoption. First, we review current approaches employing LLMs in finance, including\u0000leveraging pretrained models via zero-shot or few-shot learning, fine-tuning on\u0000domain-specific data, and training custom LLMs from scratch. We summarize key\u0000models and evaluate their performance improvements on financial natural\u0000language processing tasks. Second, we propose a decision framework to guide financial professionals in\u0000selecting the appropriate LLM solution based on their use case constraints\u0000around data, compute, and performance needs. The framework provides a pathway\u0000from lightweight experimentation to heavy investment in customized LLMs. Lastly, we discuss limitations and challenges around leveraging LLMs in\u0000financial applications. Overall, this survey aims to synthesize the\u0000state-of-the-art and provide a roadmap for responsibly applying LLMs to advance\u0000financial AI.","PeriodicalId":501372,"journal":{"name":"arXiv - QuantFin - General Finance","volume":"138 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138523135","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 performance management is at the core of business management and has historically relied on financial ratio analysis using Balance Sheet and Income Statement data to assess company performance as compared with competitors. Little progress has been made in predicting how a company will perform or in assessing the risks (probabilities) of financial underperformance. In this study I introduce a new set of financial and macroeconomic ratios that supplement standard ratios of Balance Sheet and Income Statement. I also provide a set of supervised learning models (ML Regressors and Neural Networks) and Bayesian models to predict company performance. I conclude that the new proposed variables improve model accuracy when used in tandem with standard industry ratios. I also conclude that Feedforward Neural Networks (FNN) are simpler to implement and perform best across 6 predictive tasks (ROA, ROE, Net Margin, Op Margin, Cash Ratio and Op Cash Generation); although Bayesian Networks (BN) can outperform FNN under very specific conditions. BNs have the additional benefit of providing a probability density function in addition to the predicted (expected) value. The study findings have significant potential helping CFOs and CEOs assess risks of financial underperformance to steer companies in more profitable directions; supporting lenders in better assessing the condition of a company and providing investors with tools to dissect financial statements of public companies more accurately.
{"title":"Predictive AI for SME and Large Enterprise Financial Performance Management","authors":"Ricardo Cuervo","doi":"arxiv-2311.05840","DOIUrl":"https://doi.org/arxiv-2311.05840","url":null,"abstract":"Financial performance management is at the core of business management and\u0000has historically relied on financial ratio analysis using Balance Sheet and\u0000Income Statement data to assess company performance as compared with\u0000competitors. Little progress has been made in predicting how a company will\u0000perform or in assessing the risks (probabilities) of financial\u0000underperformance. In this study I introduce a new set of financial and\u0000macroeconomic ratios that supplement standard ratios of Balance Sheet and\u0000Income Statement. I also provide a set of supervised learning models (ML\u0000Regressors and Neural Networks) and Bayesian models to predict company\u0000performance. I conclude that the new proposed variables improve model accuracy\u0000when used in tandem with standard industry ratios. I also conclude that\u0000Feedforward Neural Networks (FNN) are simpler to implement and perform best\u0000across 6 predictive tasks (ROA, ROE, Net Margin, Op Margin, Cash Ratio and Op\u0000Cash Generation); although Bayesian Networks (BN) can outperform FNN under very\u0000specific conditions. BNs have the additional benefit of providing a probability\u0000density function in addition to the predicted (expected) value. The study\u0000findings have significant potential helping CFOs and CEOs assess risks of\u0000financial underperformance to steer companies in more profitable directions;\u0000supporting lenders in better assessing the condition of a company and providing\u0000investors with tools to dissect financial statements of public companies more\u0000accurately.","PeriodicalId":501372,"journal":{"name":"arXiv - QuantFin - General Finance","volume":"88 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138522772","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 investigates the issue of an adequate loss function in the optimization of machine learning models used in the forecasting of financial time series for the purpose of algorithmic investment strategies (AIS) construction. We propose the Mean Absolute Directional Loss (MADL) function, solving important problems of classical forecast error functions in extracting information from forecasts to create efficient buy/sell signals in algorithmic investment strategies. Finally, based on the data from two different asset classes (cryptocurrencies: Bitcoin and commodities: Crude Oil), we show that the new loss function enables us to select better hyperparameters for the LSTM model and obtain more efficient investment strategies, with regard to risk-adjusted return metrics on the out-of-sample data.
{"title":"Mean Absolute Directional Loss as a New Loss Function for Machine Learning Problems in Algorithmic Investment Strategies","authors":"Jakub Michańków, Paweł Sakowski, Robert Ślepaczuk","doi":"arxiv-2309.10546","DOIUrl":"https://doi.org/arxiv-2309.10546","url":null,"abstract":"This paper investigates the issue of an adequate loss function in the\u0000optimization of machine learning models used in the forecasting of financial\u0000time series for the purpose of algorithmic investment strategies (AIS)\u0000construction. We propose the Mean Absolute Directional Loss (MADL) function,\u0000solving important problems of classical forecast error functions in extracting\u0000information from forecasts to create efficient buy/sell signals in algorithmic\u0000investment strategies. Finally, based on the data from two different asset\u0000classes (cryptocurrencies: Bitcoin and commodities: Crude Oil), we show that\u0000the new loss function enables us to select better hyperparameters for the LSTM\u0000model and obtain more efficient investment strategies, with regard to\u0000risk-adjusted return metrics on the out-of-sample data.","PeriodicalId":501372,"journal":{"name":"arXiv - QuantFin - General Finance","volume":"203 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138522832","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 this paper, we explore the aftermath of the Silicon Valley Bank (SVB) collapse, with a particular focus on its impact on crypto markets. We conduct a multi-dimensional investigation, which includes a factual summary, analysis of user sentiment, and examination of market performance. Based on such efforts, we uncover a somewhat counterintuitive finding: the SVB collapse did not lead to the destruction of cryptocurrencies; instead, they displayed resilience.
{"title":"Cryptocurrency in the Aftermath: Unveiling the Impact of the SVB Collapse","authors":"Qin Wang, Guangsheng Yu, Shiping Chen","doi":"arxiv-2311.10720","DOIUrl":"https://doi.org/arxiv-2311.10720","url":null,"abstract":"In this paper, we explore the aftermath of the Silicon Valley Bank (SVB)\u0000collapse, with a particular focus on its impact on crypto markets. We conduct a\u0000multi-dimensional investigation, which includes a factual summary, analysis of\u0000user sentiment, and examination of market performance. Based on such efforts,\u0000we uncover a somewhat counterintuitive finding: the SVB collapse did not lead\u0000to the destruction of cryptocurrencies; instead, they displayed resilience.","PeriodicalId":501372,"journal":{"name":"arXiv - QuantFin - General Finance","volume":"52 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138522782","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 present a new financial domain large language model, InvestLM, tuned on LLaMA-65B (Touvron et al., 2023), using a carefully curated instruction dataset related to financial investment. Inspired by less-is-more-for-alignment (Zhou et al., 2023), we manually curate a small yet diverse instruction dataset, covering a wide range of financial related topics, from Chartered Financial Analyst (CFA) exam questions to SEC filings to Stackexchange quantitative finance discussions. InvestLM shows strong capabilities in understanding financial text and provides helpful responses to investment related questions. Financial experts, including hedge fund managers and research analysts, rate InvestLM's response as comparable to those of state-of-the-art commercial models (GPT-3.5, GPT-4 and Claude-2). Zero-shot evaluation on a set of financial NLP benchmarks demonstrates strong generalizability. From a research perspective, this work suggests that a high-quality domain specific LLM can be tuned using a small set of carefully curated instructions on a well-trained foundation model, which is consistent with the Superficial Alignment Hypothesis (Zhou et al., 2023). From a practical perspective, this work develops a state-of-the-art financial domain LLM with superior capability in understanding financial texts and providing helpful investment advice, potentially enhancing the work efficiency of financial professionals. We release the model parameters to the research community.
我们提出了一个新的金融领域大型语言模型，InvestLM，在llama - 65b上进行了调整(Touvron等人，2023)，使用了精心策划的与金融投资相关的指令数据集。受“少即是多”(Zhouet al.， 2023)的启发，我们手动策划了一个小而多样的指令数据集，涵盖了广泛的金融相关主题，从特许金融分析师(CFA)考试问题到SEC文件，再到Stackexchange定量金融讨论。InvestLM在理解金融文本方面表现出很强的能力，并对投资相关问题提供了有益的回答。包括对冲基金经理和研究分析师在内的金融专家认为，ateinvestlm的反应与最先进的商业模型(GPT-3.5、GPT-4和Claude-2)相当。在一组金融NLP基准上的零射击评估显示出很强的泛化性。从研究的角度来看，这项工作表明，一个高质量的特定领域的法学硕士可以在一个训练有素的基础模型上使用一组精心策划的指令，这与表面对齐假设是一致的(Zhou et al.， 2023)。从实际应用的角度来看，本研究开发的金融领域法学硕士具有卓越的金融文本理解能力和提供有用的投资建议，有可能提高金融专业人士的工作效率。我们向研究界发布模型参数。
{"title":"InvestLM: A Large Language Model for Investment using Financial Domain Instruction Tuning","authors":"Yi Yang, Yixuan Tang, Kar Yan Tam","doi":"arxiv-2309.13064","DOIUrl":"https://doi.org/arxiv-2309.13064","url":null,"abstract":"We present a new financial domain large language model, InvestLM, tuned on\u0000LLaMA-65B (Touvron et al., 2023), using a carefully curated instruction dataset\u0000related to financial investment. Inspired by less-is-more-for-alignment (Zhou\u0000et al., 2023), we manually curate a small yet diverse instruction dataset,\u0000covering a wide range of financial related topics, from Chartered Financial\u0000Analyst (CFA) exam questions to SEC filings to Stackexchange quantitative\u0000finance discussions. InvestLM shows strong capabilities in understanding\u0000financial text and provides helpful responses to investment related questions.\u0000Financial experts, including hedge fund managers and research analysts, rate\u0000InvestLM's response as comparable to those of state-of-the-art commercial\u0000models (GPT-3.5, GPT-4 and Claude-2). Zero-shot evaluation on a set of\u0000financial NLP benchmarks demonstrates strong generalizability. From a research\u0000perspective, this work suggests that a high-quality domain specific LLM can be\u0000tuned using a small set of carefully curated instructions on a well-trained\u0000foundation model, which is consistent with the Superficial Alignment Hypothesis\u0000(Zhou et al., 2023). From a practical perspective, this work develops a\u0000state-of-the-art financial domain LLM with superior capability in understanding\u0000financial texts and providing helpful investment advice, potentially enhancing\u0000the work efficiency of financial professionals. We release the model parameters\u0000to the research community.","PeriodicalId":501372,"journal":{"name":"arXiv - QuantFin - General Finance","volume":"65 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138542560","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 consider the problem of being a cross-chain wealth management platform with deposits, redemptions and investment assets across multiple networks. We discuss the need for blockchain bridges to facilitates fund flows across platforms. We point out several issues with existing bridges. We develop an algorithm - tailored to overcome current constraints - that dynamically changes the utilization of bridge capacities and hence the amounts to be transferred across networks. We illustrate several scenarios using numerical simulations.
{"title":"Arguably Adequate Aqueduct Algorithm: Crossing A Bridge-Less Block-Chain Chasm","authors":"Ravi Kashyap","doi":"arxiv-2311.10717","DOIUrl":"https://doi.org/arxiv-2311.10717","url":null,"abstract":"We consider the problem of being a cross-chain wealth management platform\u0000with deposits, redemptions and investment assets across multiple networks. We\u0000discuss the need for blockchain bridges to facilitates fund flows across\u0000platforms. We point out several issues with existing bridges. We develop an\u0000algorithm - tailored to overcome current constraints - that dynamically changes\u0000the utilization of bridge capacities and hence the amounts to be transferred\u0000across networks. We illustrate several scenarios using numerical simulations.","PeriodicalId":501372,"journal":{"name":"arXiv - QuantFin - General Finance","volume":"9 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138523137","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}
The National Basketball Association (NBA) imposes a player salary cap. It is therefore useful to develop tools to measure the relative realized return of a player's salary given their on court performance. Very few such studies exist, however. We thus present the first known framework to estimate a return on investment (ROI) for NBA player contracts. The framework operates in five parts: (1) decide on a measurement time horizon, such as the standard 82-game NBA regular season; (2) calculate the novel game contribution percentage (GCP) measure we propose, which is a single game summary statistic that sums to unity for each competing team and is comprised of traditional, playtype, hustle, box outs, defensive, tracking, and rebounding per game NBA statistics; (3) estimate the single game value (SGV) of each regular season NBA game using a standard currency conversion calculation; (4) multiply the SGV by the vector of realized GCPs to obtain a series of realized per-player single season cash flows; and (5) use the player salary as an initial investment to perform the traditional ROI calculation. We illustrate our framework by compiling a novel, sharable dataset of per game GCP statistics and salaries for the 2022-2023 NBA regular season. A scatter plot of ROI by salary for all players is presented, including the top and bottom 50 performers. Notably, missed games are treated as defaults because GCP is a per game metric. This allows for break-even calculations between high-performing players with frequent missed games and average performers with few missed games, which we demonstrate with a comparison of the 2023 NBA regular seasons of Anthony Davis and Brook Lopez. We conclude by suggesting uses of our framework, discussing its flexibility through customization, and outlining potential future improvements.
{"title":"A New Framework to Estimate Return on Investment for Player Salaries in the National Basketball Association","authors":"Jackson P. Lautier","doi":"arxiv-2309.05783","DOIUrl":"https://doi.org/arxiv-2309.05783","url":null,"abstract":"The National Basketball Association (NBA) imposes a player salary cap. It is\u0000therefore useful to develop tools to measure the relative realized return of a\u0000player's salary given their on court performance. Very few such studies exist,\u0000however. We thus present the first known framework to estimate a return on\u0000investment (ROI) for NBA player contracts. The framework operates in five\u0000parts: (1) decide on a measurement time horizon, such as the standard 82-game\u0000NBA regular season; (2) calculate the novel game contribution percentage (GCP)\u0000measure we propose, which is a single game summary statistic that sums to unity\u0000for each competing team and is comprised of traditional, playtype, hustle, box\u0000outs, defensive, tracking, and rebounding per game NBA statistics; (3) estimate\u0000the single game value (SGV) of each regular season NBA game using a standard\u0000currency conversion calculation; (4) multiply the SGV by the vector of realized\u0000GCPs to obtain a series of realized per-player single season cash flows; and\u0000(5) use the player salary as an initial investment to perform the traditional\u0000ROI calculation. We illustrate our framework by compiling a novel, sharable\u0000dataset of per game GCP statistics and salaries for the 2022-2023 NBA regular\u0000season. A scatter plot of ROI by salary for all players is presented, including\u0000the top and bottom 50 performers. Notably, missed games are treated as defaults\u0000because GCP is a per game metric. This allows for break-even calculations\u0000between high-performing players with frequent missed games and average\u0000performers with few missed games, which we demonstrate with a comparison of the\u00002023 NBA regular seasons of Anthony Davis and Brook Lopez. We conclude by\u0000suggesting uses of our framework, discussing its flexibility through\u0000customization, and outlining potential future improvements.","PeriodicalId":501372,"journal":{"name":"arXiv - QuantFin - General Finance","volume":"65 3","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138522843","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}
An increase in the novelty of news predicts negative stock market returns and negative macroeconomic outcomes over the next year. We quantify news novelty - changes in the distribution of news text - through an entropy measure, calculated using a recurrent neural network applied to a large news corpus. Entropy is a better out-of-sample predictor of market returns than a collection of standard measures. Cross-sectional entropy exposure carries a negative risk premium, suggesting that assets that positively covary with entropy hedge the aggregate risk associated with shifting news language. Entropy risk cannot be explained by existing long-short factors.
{"title":"New News is Bad News","authors":"Paul Glasserman, Harry Mamaysky, Jimmy Qin","doi":"arxiv-2309.05560","DOIUrl":"https://doi.org/arxiv-2309.05560","url":null,"abstract":"An increase in the novelty of news predicts negative stock market returns and\u0000negative macroeconomic outcomes over the next year. We quantify news novelty -\u0000changes in the distribution of news text - through an entropy measure,\u0000calculated using a recurrent neural network applied to a large news corpus.\u0000Entropy is a better out-of-sample predictor of market returns than a collection\u0000of standard measures. Cross-sectional entropy exposure carries a negative risk\u0000premium, suggesting that assets that positively covary with entropy hedge the\u0000aggregate risk associated with shifting news language. Entropy risk cannot be\u0000explained by existing long-short factors.","PeriodicalId":501372,"journal":{"name":"arXiv - QuantFin - General Finance","volume":"87 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138522779","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 volatility obeys two fascinating empirical regularities that apply to various assets, on various markets, and on various time scales: it is fat-tailed (more precisely power-law distributed) and it tends to be clustered in time. Many interesting models have been proposed to account for these regularities, notably agent-based models, which mimic the two empirical laws through a complex mix of nonlinear mechanisms such as traders' switching between trading strategies in highly nonlinear way. This paper explains the two regularities simply in terms of traders' attitudes towards news, an explanation that follows almost by definition of the traditional dichotomy of financial market participants, investors versus speculators, whose behaviors are reduced to their simplest forms. Long-run investors' valuations of an asset are assumed to follow a news-driven random walk, thus capturing the investors' persistent, long memory of fundamental news. Short-term speculators' anticipated returns, on the other hand, are assumed to follow a news-driven autoregressive process, capturing their shorter memory of fundamental news, and, by the same token, the feedback intrinsic to the short-sighted, trend-following (or herding) mindset of speculators. These simple, linear, models of traders' expectations, it is shown, explain the two financial regularities in a generic and robust way. Rational expectations, the dominant model of traders' expectations, is not assumed here, owing to the famous no-speculation, no-trade results
{"title":"News-driven Expectations and Volatility Clustering","authors":"Sabiou Inoua","doi":"arxiv-2309.04876","DOIUrl":"https://doi.org/arxiv-2309.04876","url":null,"abstract":"Financial volatility obeys two fascinating empirical regularities that apply\u0000to various assets, on various markets, and on various time scales: it is\u0000fat-tailed (more precisely power-law distributed) and it tends to be clustered\u0000in time. Many interesting models have been proposed to account for these\u0000regularities, notably agent-based models, which mimic the two empirical laws\u0000through a complex mix of nonlinear mechanisms such as traders' switching\u0000between trading strategies in highly nonlinear way. This paper explains the two\u0000regularities simply in terms of traders' attitudes towards news, an explanation\u0000that follows almost by definition of the traditional dichotomy of financial\u0000market participants, investors versus speculators, whose behaviors are reduced\u0000to their simplest forms. Long-run investors' valuations of an asset are assumed\u0000to follow a news-driven random walk, thus capturing the investors' persistent,\u0000long memory of fundamental news. Short-term speculators' anticipated returns,\u0000on the other hand, are assumed to follow a news-driven autoregressive process,\u0000capturing their shorter memory of fundamental news, and, by the same token, the\u0000feedback intrinsic to the short-sighted, trend-following (or herding) mindset\u0000of speculators. These simple, linear, models of traders' expectations, it is\u0000shown, explain the two financial regularities in a generic and robust way.\u0000Rational expectations, the dominant model of traders' expectations, is not\u0000assumed here, owing to the famous no-speculation, no-trade results","PeriodicalId":501372,"journal":{"name":"arXiv - QuantFin - General Finance","volume":"30 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138542559","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 present a formal framework for the aggregation of financial markets mediated by arbitrage. Our main tool is to characterize markets via utility functions and to employ a one-to-one correspondence to limit order book states. Inspired by the theory of thermodynamics we argue that the arbitrage-mediated aggregation mechanism gives rise to a market-dynamical entropy, which quantifies the loss of liquidity caused by aggregation. We also discuss future directions of research in this emerging theory of market dynamics.
{"title":"Aggregation of financial markets","authors":"Georg Menz, Moritz Voß","doi":"arxiv-2309.04116","DOIUrl":"https://doi.org/arxiv-2309.04116","url":null,"abstract":"We present a formal framework for the aggregation of financial markets\u0000mediated by arbitrage. Our main tool is to characterize markets via utility\u0000functions and to employ a one-to-one correspondence to limit order book states.\u0000Inspired by the theory of thermodynamics we argue that the arbitrage-mediated\u0000aggregation mechanism gives rise to a market-dynamical entropy, which\u0000quantifies the loss of liquidity caused by aggregation. We also discuss future\u0000directions of research in this emerging theory of market dynamics.","PeriodicalId":501372,"journal":{"name":"arXiv - QuantFin - General Finance","volume":"52 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138522776","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：