Tracking the build-up of financial vulnerabilities is a key component of financial stability policy. Due to the complexity of the financial system, this task is daunting, and there have been several proposals on how to manage this goal. One way to do this is by the creation of indices that act as a signal for the policy maker. While factor modelling in finance and economics has a rich history, most of the applications tend to focus on stationary factors. Nevertheless, financial stress (and in particular tail events) can exhibit a high degree of inertia. This paper advocates moving away from the stationary paradigm and instead proposes non-stationary factor models as measures of financial stress. Key advantage of a non-stationary factor model is that while some popular measures of financial stress describe the variance-covariance structure of the financial stress indicators, the new index can capture the tails of the distribution. To showcase this, we use the obtained factors as variables in a growth-at-risk exercise. This paper offers an overview of how to construct non-stationary dynamic factors of financial stress using the UK financial market as an example.
{"title":"Non-stationary Financial Risk Factors and Macroeconomic Vulnerability for the UK","authors":"Katalin Varga, Tibor Szendrei","doi":"arxiv-2404.01451","DOIUrl":"https://doi.org/arxiv-2404.01451","url":null,"abstract":"Tracking the build-up of financial vulnerabilities is a key component of\u0000financial stability policy. Due to the complexity of the financial system, this\u0000task is daunting, and there have been several proposals on how to manage this\u0000goal. One way to do this is by the creation of indices that act as a signal for\u0000the policy maker. While factor modelling in finance and economics has a rich\u0000history, most of the applications tend to focus on stationary factors.\u0000Nevertheless, financial stress (and in particular tail events) can exhibit a\u0000high degree of inertia. This paper advocates moving away from the stationary\u0000paradigm and instead proposes non-stationary factor models as measures of\u0000financial stress. Key advantage of a non-stationary factor model is that while\u0000some popular measures of financial stress describe the variance-covariance\u0000structure of the financial stress indicators, the new index can capture the\u0000tails of the distribution. To showcase this, we use the obtained factors as\u0000variables in a growth-at-risk exercise. This paper offers an overview of how to\u0000construct non-stationary dynamic factors of financial stress using the UK\u0000financial market as an example.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"62 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140595994","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 study examines the effects of macroeconomic policies on financial markets using a novel approach that combines Machine Learning (ML) techniques and causal inference. It focuses on the effect of interest rate changes made by the US Federal Reserve System (FRS) on the returns of fixed income and equity funds between January 1986 and December 2021. The analysis makes a distinction between actively and passively managed funds, hypothesizing that the latter are less susceptible to changes in interest rates. The study contrasts gradient boosting and linear regression models using the Double Machine Learning (DML) framework, which supports a variety of statistical learning techniques. Results indicate that gradient boosting is a useful tool for predicting fund returns; for example, a 1% increase in interest rates causes an actively managed fund's return to decrease by -11.97%. This understanding of the relationship between interest rates and fund performance provides opportunities for additional research and insightful, data-driven advice for fund managers and investors
{"title":"Unveiling the Impact of Macroeconomic Policies: A Double Machine Learning Approach to Analyzing Interest Rate Effects on Financial Markets","authors":"Anoop Kumar, Suresh Dodda, Navin Kamuni, Rajeev Kumar Arora","doi":"arxiv-2404.07225","DOIUrl":"https://doi.org/arxiv-2404.07225","url":null,"abstract":"This study examines the effects of macroeconomic policies on financial\u0000markets using a novel approach that combines Machine Learning (ML) techniques\u0000and causal inference. It focuses on the effect of interest rate changes made by\u0000the US Federal Reserve System (FRS) on the returns of fixed income and equity\u0000funds between January 1986 and December 2021. The analysis makes a distinction\u0000between actively and passively managed funds, hypothesizing that the latter are\u0000less susceptible to changes in interest rates. The study contrasts gradient\u0000boosting and linear regression models using the Double Machine Learning (DML)\u0000framework, which supports a variety of statistical learning techniques. Results\u0000indicate that gradient boosting is a useful tool for predicting fund returns;\u0000for example, a 1% increase in interest rates causes an actively managed fund's\u0000return to decrease by -11.97%. This understanding of the relationship between\u0000interest rates and fund performance provides opportunities for additional\u0000research and insightful, data-driven advice for fund managers and investors","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"44 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140595815","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}
Silvia García-Méndez, Francisco de Arriba-Pérez, Ana Barros-Vila, Francisco J. González-Castaño
Finance-related news such as Bloomberg News, CNN Business and Forbes are valuable sources of real data for market screening systems. In news, an expert shares opinions beyond plain technical analyses that include context such as political, sociological and cultural factors. In the same text, the expert often discusses the performance of different assets. Some key statements are mere descriptions of past events while others are predictions. Therefore, understanding the temporality of the key statements in a text is essential to separate context information from valuable predictions. We propose a novel system to detect the temporality of finance-related news at discourse level that combines Natural Language Processing and Machine Learning techniques, and exploits sophisticated features such as syntactic and semantic dependencies. More specifically, we seek to extract the dominant tenses of the main statements, which may be either explicit or implicit. We have tested our system on a labelled dataset of finance-related news annotated by researchers with knowledge in the field. Experimental results reveal a high detection precision compared to an alternative rule-based baseline approach. Ultimately, this research contributes to the state-of-the-art of market screening by identifying predictive knowledge for financial decision making.
彭博新闻社、CNN Business 和《福布斯》等与金融相关的新闻是市场筛选系统宝贵的真实数据来源。在新闻中，专家分享的观点不仅仅是简单的技术分析，还包括政治、社会和文化因素等背景。在同一篇文章中，专家经常讨论不同资产的表现。有些关键言论只是对过去事件的描述，而有些则是预测。因此，理解文本中关键语句的时间性对于将上下文信息与有价值的预测区分开来至关重要。我们提出了一种在话语层面检测金融相关新闻时间性的新系统，该系统结合了自然语言处理和机器学习技术，并利用了句法和语义依赖性等复杂特征。我们在由该领域研究人员标注的金融相关新闻标签数据集上测试了我们的系统。实验结果表明，与其他基于规则的基线方法相比，我们的系统具有很高的检测精度。最终，这项研究通过识别金融决策的预测性知识，为市场筛选的先进水平做出了贡献。
{"title":"Detection of Temporality at Discourse Level on Financial News by Combining Natural Language Processing and Machine Learning","authors":"Silvia García-Méndez, Francisco de Arriba-Pérez, Ana Barros-Vila, Francisco J. González-Castaño","doi":"arxiv-2404.01337","DOIUrl":"https://doi.org/arxiv-2404.01337","url":null,"abstract":"Finance-related news such as Bloomberg News, CNN Business and Forbes are\u0000valuable sources of real data for market screening systems. In news, an expert\u0000shares opinions beyond plain technical analyses that include context such as\u0000political, sociological and cultural factors. In the same text, the expert\u0000often discusses the performance of different assets. Some key statements are\u0000mere descriptions of past events while others are predictions. Therefore,\u0000understanding the temporality of the key statements in a text is essential to\u0000separate context information from valuable predictions. We propose a novel\u0000system to detect the temporality of finance-related news at discourse level\u0000that combines Natural Language Processing and Machine Learning techniques, and\u0000exploits sophisticated features such as syntactic and semantic dependencies.\u0000More specifically, we seek to extract the dominant tenses of the main\u0000statements, which may be either explicit or implicit. We have tested our system\u0000on a labelled dataset of finance-related news annotated by researchers with\u0000knowledge in the field. Experimental results reveal a high detection precision\u0000compared to an alternative rule-based baseline approach. Ultimately, this\u0000research contributes to the state-of-the-art of market screening by identifying\u0000predictive knowledge for financial decision making.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"42 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140595877","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 model a portfolio of crypto assets that does not respond well to multivariate autoregressive models because of discontinuity in conditional covariance matrix and posterior covariance matrix caused by extreme liquidity. We adjust asset-level return and volatility with liquidity to reduce such discontinuity, and restore the effectiveness of a set of liquidity-adjusted VECM-DCC/ADCC-BL models at extreme liquidity. We establish two distinctive yet complementary portfolio liquidity measures: portfolio liquidity jump that quantifies the effect of liquidity adjustment in forecasting the conditional covariance matrix, and portfolio liquidity diffusion that quantifies the effect of liquidity adjustment in estimating the posterior covariance matrix.
{"title":"Liquidity Jump, Liquidity Diffusion, and Portfolio of Assets with Extreme Liquidity","authors":"Qi Deng, Zhong-guo Zhou","doi":"arxiv-2407.00813","DOIUrl":"https://doi.org/arxiv-2407.00813","url":null,"abstract":"We model a portfolio of crypto assets that does not respond well to\u0000multivariate autoregressive models because of discontinuity in conditional\u0000covariance matrix and posterior covariance matrix caused by extreme liquidity.\u0000We adjust asset-level return and volatility with liquidity to reduce such\u0000discontinuity, and restore the effectiveness of a set of liquidity-adjusted\u0000VECM-DCC/ADCC-BL models at extreme liquidity. We establish two distinctive yet\u0000complementary portfolio liquidity measures: portfolio liquidity jump that\u0000quantifies the effect of liquidity adjustment in forecasting the conditional\u0000covariance matrix, and portfolio liquidity diffusion that quantifies the effect\u0000of liquidity adjustment in estimating the posterior covariance matrix.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"11 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141531028","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}
Silvia García-Méndez, Francisco de Arriba-Pérez, Ana Barros-Vila, Francisco J. González-Castaño, Enrique Costa-Montenegro
Financial news items are unstructured sources of information that can be mined to extract knowledge for market screening applications. Manual extraction of relevant information from the continuous stream of finance-related news is cumbersome and beyond the skills of many investors, who, at most, can follow a few sources and authors. Accordingly, we focus on the analysis of financial news to identify relevant text and, within that text, forecasts and predictions. We propose a novel Natural Language Processing (NLP) system to assist investors in the detection of relevant financial events in unstructured textual sources by considering both relevance and temporality at the discursive level. Firstly, we segment the text to group together closely related text. Secondly, we apply co-reference resolution to discover internal dependencies within segments. Finally, we perform relevant topic modelling with Latent Dirichlet Allocation (LDA) to separate relevant from less relevant text and then analyse the relevant text using a Machine Learning-oriented temporal approach to identify predictions and speculative statements. We created an experimental data set composed of 2,158 financial news items that were manually labelled by NLP researchers to evaluate our solution. The ROUGE-L values for the identification of relevant text and predictions/forecasts were 0.662 and 0.982, respectively. To our knowledge, this is the first work to jointly consider relevance and temporality at the discursive level. It contributes to the transfer of human associative discourse capabilities to expert systems through the combination of multi-paragraph topic segmentation and co-reference resolution to separate author expression patterns, topic modelling with LDA to detect relevant text, and discursive temporality analysis to identify forecasts and predictions within this text.
{"title":"Automatic detection of relevant information, predictions and forecasts in financial news through topic modelling with Latent Dirichlet Allocation","authors":"Silvia García-Méndez, Francisco de Arriba-Pérez, Ana Barros-Vila, Francisco J. González-Castaño, Enrique Costa-Montenegro","doi":"arxiv-2404.01338","DOIUrl":"https://doi.org/arxiv-2404.01338","url":null,"abstract":"Financial news items are unstructured sources of information that can be\u0000mined to extract knowledge for market screening applications. Manual extraction\u0000of relevant information from the continuous stream of finance-related news is\u0000cumbersome and beyond the skills of many investors, who, at most, can follow a\u0000few sources and authors. Accordingly, we focus on the analysis of financial\u0000news to identify relevant text and, within that text, forecasts and\u0000predictions. We propose a novel Natural Language Processing (NLP) system to\u0000assist investors in the detection of relevant financial events in unstructured\u0000textual sources by considering both relevance and temporality at the discursive\u0000level. Firstly, we segment the text to group together closely related text.\u0000Secondly, we apply co-reference resolution to discover internal dependencies\u0000within segments. Finally, we perform relevant topic modelling with Latent\u0000Dirichlet Allocation (LDA) to separate relevant from less relevant text and\u0000then analyse the relevant text using a Machine Learning-oriented temporal\u0000approach to identify predictions and speculative statements. We created an\u0000experimental data set composed of 2,158 financial news items that were manually\u0000labelled by NLP researchers to evaluate our solution. The ROUGE-L values for\u0000the identification of relevant text and predictions/forecasts were 0.662 and\u00000.982, respectively. To our knowledge, this is the first work to jointly\u0000consider relevance and temporality at the discursive level. It contributes to\u0000the transfer of human associative discourse capabilities to expert systems\u0000through the combination of multi-paragraph topic segmentation and co-reference\u0000resolution to separate author expression patterns, topic modelling with LDA to\u0000detect relevant text, and discursive temporality analysis to identify forecasts\u0000and predictions within this text.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"32 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140595820","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}
Francisco de Arriba-Pérez, Silvia García-Méndez, José A. Regueiro-Janeiro, Francisco J. González-Castaño
Micro-blogging sources such as the Twitter social network provide valuable real-time data for market prediction models. Investors' opinions in this network follow the fluctuations of the stock markets and often include educated speculations on market opportunities that may have impact on the actions of other investors. In view of this, we propose a novel system to detect positive predictions in tweets, a type of financial emotions which we term "opportunities" that are akin to "anticipation" in Plutchik's theory. Specifically, we seek a high detection precision to present a financial operator a substantial amount of such tweets while differentiating them from the rest of financial emotions in our system. We achieve it with a three-layer stacked Machine Learning classification system with sophisticated features that result from applying Natural Language Processing techniques to extract valuable linguistic information. Experimental results on a dataset that has been manually annotated with financial emotion and ticker occurrence tags demonstrate that our system yields satisfactory and competitive performance in financial opportunity detection, with precision values up to 83%. This promising outcome endorses the usability of our system to support investors' decision making.
{"title":"Detection of financial opportunities in micro-blogging data with a stacked classification system","authors":"Francisco de Arriba-Pérez, Silvia García-Méndez, José A. Regueiro-Janeiro, Francisco J. González-Castaño","doi":"arxiv-2404.07224","DOIUrl":"https://doi.org/arxiv-2404.07224","url":null,"abstract":"Micro-blogging sources such as the Twitter social network provide valuable\u0000real-time data for market prediction models. Investors' opinions in this\u0000network follow the fluctuations of the stock markets and often include educated\u0000speculations on market opportunities that may have impact on the actions of\u0000other investors. In view of this, we propose a novel system to detect positive\u0000predictions in tweets, a type of financial emotions which we term\u0000\"opportunities\" that are akin to \"anticipation\" in Plutchik's theory.\u0000Specifically, we seek a high detection precision to present a financial\u0000operator a substantial amount of such tweets while differentiating them from\u0000the rest of financial emotions in our system. We achieve it with a three-layer\u0000stacked Machine Learning classification system with sophisticated features that\u0000result from applying Natural Language Processing techniques to extract valuable\u0000linguistic information. Experimental results on a dataset that has been\u0000manually annotated with financial emotion and ticker occurrence tags\u0000demonstrate that our system yields satisfactory and competitive performance in\u0000financial opportunity detection, with precision values up to 83%. This\u0000promising outcome endorses the usability of our system to support investors'\u0000decision making.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"61 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140595819","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}
Stijn De Backer, Luis E. C. Rocha, Jan Ryckebusch, Koen Schoors
Accurate modeling of the temporal evolution of asset prices is crucial for understanding financial markets. We explore the potential of discrete-time quantum walks to model the evolution of asset prices. Return distributions obtained from a model based on the quantum walk algorithm are compared with those obtained from classical methodologies. We focus on specific limitations of the classical models, and illustrate that the quantum walk model possesses great flexibility in overcoming these. This includes the potential to generate asymmetric return distributions with complex market tendencies and higher probabilities for extreme events than in some of the classical models. Furthermore, the temporal evolution in the quantum walk possesses the potential to provide asset price dynamics.
{"title":"On the potential of quantum walks for modeling financial return distributions","authors":"Stijn De Backer, Luis E. C. Rocha, Jan Ryckebusch, Koen Schoors","doi":"arxiv-2403.19502","DOIUrl":"https://doi.org/arxiv-2403.19502","url":null,"abstract":"Accurate modeling of the temporal evolution of asset prices is crucial for\u0000understanding financial markets. We explore the potential of discrete-time\u0000quantum walks to model the evolution of asset prices. Return distributions\u0000obtained from a model based on the quantum walk algorithm are compared with\u0000those obtained from classical methodologies. We focus on specific limitations\u0000of the classical models, and illustrate that the quantum walk model possesses\u0000great flexibility in overcoming these. This includes the potential to generate\u0000asymmetric return distributions with complex market tendencies and higher\u0000probabilities for extreme events than in some of the classical models.\u0000Furthermore, the temporal evolution in the quantum walk possesses the potential\u0000to provide asset price dynamics.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"45 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140324623","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 complex financial markets, recommender systems can play a crucial role in empowering individuals to make informed decisions. Existing studies predominantly focus on price prediction, but even the most sophisticated models cannot accurately predict stock prices. Also, many studies show that most individual investors do not follow established investment theories because they have their own preferences. Hence, the tricky point in stock recommendation is that recommendations should give good investment performance but also should not ignore individual preferences. To develop effective stock recommender systems, it is essential to consider three key aspects: 1) individual preferences, 2) portfolio diversification, and 3) temporal aspect of both stock features and individual preferences. In response, we develop the portfolio temporal graph network recommender PfoTGNRec, which can handle time-varying collaborative signals and incorporates diversification-enhancing contrastive learning. As a result, our model demonstrated superior performance compared to various baselines, including cutting-edge dynamic embedding models and existing stock recommendation models, in a sense that our model exhibited good investment performance while maintaining competitive in capturing individual preferences. The source code and data are available at https://anonymous.4open.science/r/IJCAI2024-12F4.
{"title":"Stock Recommendations for Individual Investors: A Temporal Graph Network Approach with Diversification-Enhancing Contrastive Learning","authors":"Youngbin Lee, Yejin Kim, Yongjae Lee","doi":"arxiv-2404.07223","DOIUrl":"https://doi.org/arxiv-2404.07223","url":null,"abstract":"In complex financial markets, recommender systems can play a crucial role in\u0000empowering individuals to make informed decisions. Existing studies\u0000predominantly focus on price prediction, but even the most sophisticated models\u0000cannot accurately predict stock prices. Also, many studies show that most\u0000individual investors do not follow established investment theories because they\u0000have their own preferences. Hence, the tricky point in stock recommendation is\u0000that recommendations should give good investment performance but also should\u0000not ignore individual preferences. To develop effective stock recommender\u0000systems, it is essential to consider three key aspects: 1) individual\u0000preferences, 2) portfolio diversification, and 3) temporal aspect of both stock\u0000features and individual preferences. In response, we develop the portfolio\u0000temporal graph network recommender PfoTGNRec, which can handle time-varying\u0000collaborative signals and incorporates diversification-enhancing contrastive\u0000learning. As a result, our model demonstrated superior performance compared to\u0000various baselines, including cutting-edge dynamic embedding models and existing\u0000stock recommendation models, in a sense that our model exhibited good\u0000investment performance while maintaining competitive in capturing individual\u0000preferences. The source code and data are available at\u0000https://anonymous.4open.science/r/IJCAI2024-12F4.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"68 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140595875","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}
Twenty five years ago, several authors proposed to model the forward interest rate curve (FRC) as an elastic string along which idiosyncratic shocks propagate, accounting for the peculiar structure of the return correlation across different maturities. In this paper, we revisit the specific "stiff'' elastic string field theory of Baaquie and Bouchaud (2004) in a way that makes its micro-foundation more transparent. Our model can be interpreted as capturing the effect of market forces that set the rates of nearby tenors in a self-referential fashion. The model is parsimonious and accurately reproduces the whole correlation structure of the FRC over the time period 1994-2023, with an error below 2%. We need only two parameters, the values of which being very stable except perhaps during the Quantitative Easing period 2009-2014. The dependence of correlation on time resolution (also called the Epps effect) is also faithfully reproduced within the model and leads to a cross-tenor information propagation time of 10 minutes. Finally, we confirm that the perceived time in interest rate markets is a strongly sub-linear function of real time, as surmised by Baaquie and Bouchaud (2004). In fact, our results are fully compatible with hyperbolic discounting, in line with the recent behavioural literature (Farmer and Geanakoplos, 2009).
{"title":"Revisiting Elastic String Models of Forward Interest Rates","authors":"Victor Le Coz, Jean-Philippe Bouchaud","doi":"arxiv-2403.18126","DOIUrl":"https://doi.org/arxiv-2403.18126","url":null,"abstract":"Twenty five years ago, several authors proposed to model the forward interest\u0000rate curve (FRC) as an elastic string along which idiosyncratic shocks\u0000propagate, accounting for the peculiar structure of the return correlation\u0000across different maturities. In this paper, we revisit the specific \"stiff''\u0000elastic string field theory of Baaquie and Bouchaud (2004) in a way that makes\u0000its micro-foundation more transparent. Our model can be interpreted as\u0000capturing the effect of market forces that set the rates of nearby tenors in a\u0000self-referential fashion. The model is parsimonious and accurately reproduces\u0000the whole correlation structure of the FRC over the time period 1994-2023, with\u0000an error below 2%. We need only two parameters, the values of which being very\u0000stable except perhaps during the Quantitative Easing period 2009-2014. The\u0000dependence of correlation on time resolution (also called the Epps effect) is\u0000also faithfully reproduced within the model and leads to a cross-tenor\u0000information propagation time of 10 minutes. Finally, we confirm that the\u0000perceived time in interest rate markets is a strongly sub-linear function of\u0000real time, as surmised by Baaquie and Bouchaud (2004). In fact, our results are\u0000fully compatible with hyperbolic discounting, in line with the recent\u0000behavioural literature (Farmer and Geanakoplos, 2009).","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"2010 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140314405","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}
As a branch of time series forecasting, stock movement forecasting is one of the challenging problems for investors and researchers. Since Transformer was introduced to analyze financial data, many researchers have dedicated themselves to forecasting stock movement using Transformer or attention mechanisms. However, existing research mostly focuses on individual stock information but ignores stock market information and high noise in stock data. In this paper, we propose a novel method using the attention mechanism in which both stock market information and individual stock information are considered. Meanwhile, we propose a novel EMD-based algorithm for reducing short-term noise in stock data. Two randomly selected exchange-traded funds (ETFs) spanning over ten years from US stock markets are used to demonstrate the superior performance of the proposed attention-based method. The experimental analysis demonstrates that the proposed attention-based method significantly outperforms other state-of-the-art baselines. Code is available at https://github.com/DurandalLee/ACEFormer.
{"title":"An End-to-End Structure with Novel Position Mechanism and Improved EMD for Stock Forecasting","authors":"Chufeng Li, Jianyong Chen","doi":"arxiv-2404.07969","DOIUrl":"https://doi.org/arxiv-2404.07969","url":null,"abstract":"As a branch of time series forecasting, stock movement forecasting is one of\u0000the challenging problems for investors and researchers. Since Transformer was\u0000introduced to analyze financial data, many researchers have dedicated\u0000themselves to forecasting stock movement using Transformer or attention\u0000mechanisms. However, existing research mostly focuses on individual stock\u0000information but ignores stock market information and high noise in stock data.\u0000In this paper, we propose a novel method using the attention mechanism in which\u0000both stock market information and individual stock information are considered.\u0000Meanwhile, we propose a novel EMD-based algorithm for reducing short-term noise\u0000in stock data. Two randomly selected exchange-traded funds (ETFs) spanning over\u0000ten years from US stock markets are used to demonstrate the superior\u0000performance of the proposed attention-based method. The experimental analysis\u0000demonstrates that the proposed attention-based method significantly outperforms\u0000other state-of-the-art baselines. Code is available at\u0000https://github.com/DurandalLee/ACEFormer.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"54 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140595816","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：