Identification of market abuse is an extremely complicated activity that requires the analysis of large and complex datasets. We propose an unsupervised machine learning method for contextual anomaly detection, which allows to support market surveillance aimed at identifying potential insider trading activities. This method lies in the reconstruction-based paradigm and employs principal component analysis and autoencoders as dimensionality reduction techniques. The only input of this method is the trading position of each investor active on the asset for which we have a price sensitive event (PSE). After determining reconstruction errors related to the trading profiles, several conditions are imposed in order to identify investors whose behavior could be suspicious of insider trading related to the PSE. As a case study, we apply our method to investor resolved data of Italian stocks around takeover bids.
识别市场滥用是一项极其复杂的工作，需要对大量复杂的数据集进行分析。我们提出了一种用于上下文异常检测的无监督机器学习方法，可用于支持旨在识别潜在内部交易活动的市场监控。该方法属于基于重构的范例，采用了主成分分析和自动编码器作为降维技术。在确定了与交易概况相关的重构误差后，我们施加了几个条件，以识别其行为可能涉嫌与 PSE 相关的内幕交易的投资者。作为一项案例研究，我们将我们的方法应用于意大利股票收购要约前后的投资者解决数据。
{"title":"Dimensionality reduction techniques to support insider trading detection","authors":"Adele Ravagnani, Fabrizio Lillo, Paola Deriu, Piero Mazzarisi, Francesca Medda, Antonio Russo","doi":"arxiv-2403.00707","DOIUrl":"https://doi.org/arxiv-2403.00707","url":null,"abstract":"Identification of market abuse is an extremely complicated activity that\u0000requires the analysis of large and complex datasets. We propose an unsupervised\u0000machine learning method for contextual anomaly detection, which allows to\u0000support market surveillance aimed at identifying potential insider trading\u0000activities. This method lies in the reconstruction-based paradigm and employs\u0000principal component analysis and autoencoders as dimensionality reduction\u0000techniques. The only input of this method is the trading position of each\u0000investor active on the asset for which we have a price sensitive event (PSE).\u0000After determining reconstruction errors related to the trading profiles,\u0000several conditions are imposed in order to identify investors whose behavior\u0000could be suspicious of insider trading related to the PSE. As a case study, we\u0000apply our method to investor resolved data of Italian stocks around takeover\u0000bids.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"108 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140034841","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 stock market plays a pivotal role in economic development, yet its intricate volatility poses challenges for investors. Consequently, research and accurate predictions of stock price movements are crucial for mitigating risks. Traditional time series models fall short in capturing nonlinearity, leading to unsatisfactory stock predictions. This limitation has spurred the widespread adoption of neural networks for stock prediction, owing to their robust nonlinear generalization capabilities. Recently, Mamba, a structured state space sequence model with a selection mechanism and scan module (S6), has emerged as a powerful tool in sequence modeling tasks. Leveraging this framework, this paper proposes a novel Mamba-based model for stock price prediction, named MambaStock. The proposed MambaStock model effectively mines historical stock market data to predict future stock prices without handcrafted features or extensive preprocessing procedures. Empirical studies on several stocks indicate that the MambaStock model outperforms previous methods, delivering highly accurate predictions. This enhanced accuracy can assist investors and institutions in making informed decisions, aiming to maximize returns while minimizing risks. This work underscores the value of Mamba in time-series forecasting. Source code is available at https://github.com/zshicode/MambaStock.
{"title":"MambaStock: Selective state space model for stock prediction","authors":"Zhuangwei Shi","doi":"arxiv-2402.18959","DOIUrl":"https://doi.org/arxiv-2402.18959","url":null,"abstract":"The stock market plays a pivotal role in economic development, yet its\u0000intricate volatility poses challenges for investors. Consequently, research and\u0000accurate predictions of stock price movements are crucial for mitigating risks.\u0000Traditional time series models fall short in capturing nonlinearity, leading to\u0000unsatisfactory stock predictions. This limitation has spurred the widespread\u0000adoption of neural networks for stock prediction, owing to their robust\u0000nonlinear generalization capabilities. Recently, Mamba, a structured state\u0000space sequence model with a selection mechanism and scan module (S6), has\u0000emerged as a powerful tool in sequence modeling tasks. Leveraging this\u0000framework, this paper proposes a novel Mamba-based model for stock price\u0000prediction, named MambaStock. The proposed MambaStock model effectively mines\u0000historical stock market data to predict future stock prices without handcrafted\u0000features or extensive preprocessing procedures. Empirical studies on several\u0000stocks indicate that the MambaStock model outperforms previous methods,\u0000delivering highly accurate predictions. This enhanced accuracy can assist\u0000investors and institutions in making informed decisions, aiming to maximize\u0000returns while minimizing risks. This work underscores the value of Mamba in\u0000time-series forecasting. Source code is available at\u0000https://github.com/zshicode/MambaStock.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"8 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140010546","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 analyzes the statistical properties of constrained portfolio formation in a high dimensional portfolio with a large number of assets. Namely, we consider portfolios with tracking error constraints, portfolios with tracking error jointly with weight (equality or inequality) restrictions, and portfolios with only weight restrictions. Tracking error is the portfolio's performance measured against a benchmark (an index usually), {color{black}{and weight constraints refers to specific allocation of assets within the portfolio, which often come in the form of regulatory requirement or fund prospectus.}} We show how these portfolios can be estimated consistently in large dimensions, even when the number of assets is larger than the time span of the portfolio. We also provide rate of convergence results for weights of the constrained portfolio, risk of the constrained portfolio and the Sharpe Ratio of the constrained portfolio. To achieve those results we use a new machine learning technique that merges factor models with nodewise regression in statistics. Simulation results and empirics show very good performance of our method.
{"title":"Navigating Complexity: Constrained Portfolio Analysis in High Dimensions with Tracking Error and Weight Constraints","authors":"Mehmet Caner, Qingliang Fan, Yingying Li","doi":"arxiv-2402.17523","DOIUrl":"https://doi.org/arxiv-2402.17523","url":null,"abstract":"This paper analyzes the statistical properties of constrained portfolio\u0000formation in a high dimensional portfolio with a large number of assets.\u0000Namely, we consider portfolios with tracking error constraints, portfolios with\u0000tracking error jointly with weight (equality or inequality) restrictions, and\u0000portfolios with only weight restrictions. Tracking error is the portfolio's\u0000performance measured against a benchmark (an index usually), {color{black}{and\u0000weight constraints refers to specific allocation of assets within the\u0000portfolio, which often come in the form of regulatory requirement or fund\u0000prospectus.}} We show how these portfolios can be estimated consistently in\u0000large dimensions, even when the number of assets is larger than the time span\u0000of the portfolio. We also provide rate of convergence results for weights of\u0000the constrained portfolio, risk of the constrained portfolio and the Sharpe\u0000Ratio of the constrained portfolio. To achieve those results we use a new\u0000machine learning technique that merges factor models with nodewise regression\u0000in statistics. Simulation results and empirics show very good performance of\u0000our method.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"53 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140011231","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}
Yifan Duan, Guibin Zhang, Shilong Wang, Xiaojiang Peng, Wang Ziqi, Junyuan Mao, Hao Wu, Xinke Jiang, Kun Wang
Credit card fraud poses a significant threat to the economy. While Graph Neural Network (GNN)-based fraud detection methods perform well, they often overlook the causal effect of a node's local structure on predictions. This paper introduces a novel method for credit card fraud detection, the textbf{underline{Ca}}usal textbf{underline{T}}emporal textbf{underline{G}}raph textbf{underline{N}}eural textbf{N}etwork (CaT-GNN), which leverages causal invariant learning to reveal inherent correlations within transaction data. By decomposing the problem into discovery and intervention phases, CaT-GNN identifies causal nodes within the transaction graph and applies a causal mixup strategy to enhance the model's robustness and interpretability. CaT-GNN consists of two key components: Causal-Inspector and Causal-Intervener. The Causal-Inspector utilizes attention weights in the temporal attention mechanism to identify causal and environment nodes without introducing additional parameters. Subsequently, the Causal-Intervener performs a causal mixup enhancement on environment nodes based on the set of nodes. Evaluated on three datasets, including a private financial dataset and two public datasets, CaT-GNN demonstrates superior performance over existing state-of-the-art methods. Our findings highlight the potential of integrating causal reasoning with graph neural networks to improve fraud detection capabilities in financial transactions.
{"title":"CaT-GNN: Enhancing Credit Card Fraud Detection via Causal Temporal Graph Neural Networks","authors":"Yifan Duan, Guibin Zhang, Shilong Wang, Xiaojiang Peng, Wang Ziqi, Junyuan Mao, Hao Wu, Xinke Jiang, Kun Wang","doi":"arxiv-2402.14708","DOIUrl":"https://doi.org/arxiv-2402.14708","url":null,"abstract":"Credit card fraud poses a significant threat to the economy. While Graph\u0000Neural Network (GNN)-based fraud detection methods perform well, they often\u0000overlook the causal effect of a node's local structure on predictions. This\u0000paper introduces a novel method for credit card fraud detection, the\u0000textbf{underline{Ca}}usal textbf{underline{T}}emporal\u0000textbf{underline{G}}raph textbf{underline{N}}eural textbf{N}etwork\u0000(CaT-GNN), which leverages causal invariant learning to reveal inherent\u0000correlations within transaction data. By decomposing the problem into discovery\u0000and intervention phases, CaT-GNN identifies causal nodes within the transaction\u0000graph and applies a causal mixup strategy to enhance the model's robustness and\u0000interpretability. CaT-GNN consists of two key components: Causal-Inspector and\u0000Causal-Intervener. The Causal-Inspector utilizes attention weights in the\u0000temporal attention mechanism to identify causal and environment nodes without\u0000introducing additional parameters. Subsequently, the Causal-Intervener performs\u0000a causal mixup enhancement on environment nodes based on the set of nodes.\u0000Evaluated on three datasets, including a private financial dataset and two\u0000public datasets, CaT-GNN demonstrates superior performance over existing\u0000state-of-the-art methods. Our findings highlight the potential of integrating\u0000causal reasoning with graph neural networks to improve fraud detection\u0000capabilities in financial transactions.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"41 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139949644","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}
Steven Y. K. Wong, Jennifer S. K. Chan, Lamiae Azizi
Time-series with time-varying variance pose a unique challenge to uncertainty quantification (UQ) methods. Time-varying variance, such as volatility clustering as seen in financial time-series, can lead to large mismatch between predicted uncertainty and forecast error. Building on recent advances in neural network UQ literature, we extend and simplify Deep Evidential Regression and Deep Ensembles into a unified framework to deal with UQ under the presence of volatility clustering. We show that a Scale Mixture Distribution is a simpler alternative to the Normal-Inverse-Gamma prior that provides favorable complexity-accuracy trade-off. To illustrate the performance of our proposed approach, we apply it to two sets of financial time-series exhibiting volatility clustering: cryptocurrencies and U.S. equities.
{"title":"Quantifying neural network uncertainty under volatility clustering","authors":"Steven Y. K. Wong, Jennifer S. K. Chan, Lamiae Azizi","doi":"arxiv-2402.14476","DOIUrl":"https://doi.org/arxiv-2402.14476","url":null,"abstract":"Time-series with time-varying variance pose a unique challenge to uncertainty\u0000quantification (UQ) methods. Time-varying variance, such as volatility\u0000clustering as seen in financial time-series, can lead to large mismatch between\u0000predicted uncertainty and forecast error. Building on recent advances in neural\u0000network UQ literature, we extend and simplify Deep Evidential Regression and\u0000Deep Ensembles into a unified framework to deal with UQ under the presence of\u0000volatility clustering. We show that a Scale Mixture Distribution is a simpler\u0000alternative to the Normal-Inverse-Gamma prior that provides favorable\u0000complexity-accuracy trade-off. To illustrate the performance of our proposed\u0000approach, we apply it to two sets of financial time-series exhibiting\u0000volatility clustering: cryptocurrencies and U.S. equities.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"67 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139949666","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 objective of this research is to examine how sentiment analysis can be employed to generate trading signals for the Foreign Exchange (Forex) market. The author assessed sentiment in social media posts and news articles pertaining to the United States Dollar (USD) using a combination of methods: lexicon-based analysis and the Naive Bayes machine learning algorithm. The findings indicate that sentiment analysis proves valuable in forecasting market movements and devising trading signals. Notably, its effectiveness is consistent across different market conditions. The author concludes that by analyzing sentiment expressed in news and social media, traders can glean insights into prevailing market sentiments towards the USD and other pertinent countries, thereby aiding trading decision-making. This study underscores the importance of weaving sentiment analysis into trading strategies as a pivotal tool for predicting market dynamics.
{"title":"Applying News and Media Sentiment Analysis for Generating Forex Trading Signals","authors":"Oluwafemi F Olaiyapo","doi":"arxiv-2403.00785","DOIUrl":"https://doi.org/arxiv-2403.00785","url":null,"abstract":"The objective of this research is to examine how sentiment analysis can be\u0000employed to generate trading signals for the Foreign Exchange (Forex) market.\u0000The author assessed sentiment in social media posts and news articles\u0000pertaining to the United States Dollar (USD) using a combination of methods:\u0000lexicon-based analysis and the Naive Bayes machine learning algorithm. The\u0000findings indicate that sentiment analysis proves valuable in forecasting market\u0000movements and devising trading signals. Notably, its effectiveness is\u0000consistent across different market conditions. The author concludes that by\u0000analyzing sentiment expressed in news and social media, traders can glean\u0000insights into prevailing market sentiments towards the USD and other pertinent\u0000countries, thereby aiding trading decision-making. This study underscores the\u0000importance of weaving sentiment analysis into trading strategies as a pivotal\u0000tool for predicting market dynamics.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"52 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140034646","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}
Hanshuang Tong, Jun Li, Ning Wu, Ming Gong, Dongmei Zhang, Qi Zhang
Recent advancements in large language models (LLMs) have opened new pathways for many domains. However, the full potential of LLMs in financial investments remains largely untapped. There are two main challenges for typical deep learning-based methods for quantitative finance. First, they struggle to fuse textual and numerical information flexibly for stock movement prediction. Second, traditional methods lack clarity and interpretability, which impedes their application in scenarios where the justification for predictions is essential. To solve the above challenges, we propose Ploutos, a novel financial LLM framework that consists of PloutosGen and PloutosGPT. The PloutosGen contains multiple primary experts that can analyze different modal data, such as text and numbers, and provide quantitative strategies from different perspectives. Then PloutosGPT combines their insights and predictions and generates interpretable rationales. To generate accurate and faithful rationales, the training strategy of PloutosGPT leverage rearview-mirror prompting mechanism to guide GPT-4 to generate rationales, and a dynamic token weighting mechanism to finetune LLM by increasing key tokens weight. Extensive experiments show our framework outperforms the state-of-the-art methods on both prediction accuracy and interpretability.
{"title":"Ploutos: Towards interpretable stock movement prediction with financial large language model","authors":"Hanshuang Tong, Jun Li, Ning Wu, Ming Gong, Dongmei Zhang, Qi Zhang","doi":"arxiv-2403.00782","DOIUrl":"https://doi.org/arxiv-2403.00782","url":null,"abstract":"Recent advancements in large language models (LLMs) have opened new pathways\u0000for many domains. However, the full potential of LLMs in financial investments\u0000remains largely untapped. There are two main challenges for typical deep\u0000learning-based methods for quantitative finance. First, they struggle to fuse\u0000textual and numerical information flexibly for stock movement prediction.\u0000Second, traditional methods lack clarity and interpretability, which impedes\u0000their application in scenarios where the justification for predictions is\u0000essential. To solve the above challenges, we propose Ploutos, a novel financial\u0000LLM framework that consists of PloutosGen and PloutosGPT. The PloutosGen\u0000contains multiple primary experts that can analyze different modal data, such\u0000as text and numbers, and provide quantitative strategies from different\u0000perspectives. Then PloutosGPT combines their insights and predictions and\u0000generates interpretable rationales. To generate accurate and faithful\u0000rationales, the training strategy of PloutosGPT leverage rearview-mirror\u0000prompting mechanism to guide GPT-4 to generate rationales, and a dynamic token\u0000weighting mechanism to finetune LLM by increasing key tokens weight. Extensive\u0000experiments show our framework outperforms the state-of-the-art methods on both\u0000prediction accuracy and interpretability.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"13 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140034385","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}
Efforts to predict stock market outcomes have yielded limited success due to the inherently stochastic nature of the market, influenced by numerous unpredictable factors. Many existing prediction approaches focus on single-point predictions, lacking the depth needed for effective decision-making and often overlooking market risk. To bridge this gap, we propose a novel model, RAGIC, which introduces sequence generation for stock interval prediction to quantify uncertainty more effectively. Our approach leverages a Generative Adversarial Network (GAN) to produce future price sequences infused with randomness inherent in financial markets. RAGIC's generator includes a risk module, capturing the risk perception of informed investors, and a temporal module, accounting for historical price trends and seasonality. This multi-faceted generator informs the creation of risk-sensitive intervals through statistical inference, incorporating horizon-wise insights. The interval's width is carefully adjusted to reflect market volatility. Importantly, our approach relies solely on publicly available data and incurs only low computational overhead. RAGIC's evaluation across globally recognized broad-based indices demonstrates its balanced performance, offering both accuracy and informativeness. Achieving a consistent 95% coverage, RAGIC maintains a narrow interval width. This promising outcome suggests that our approach effectively addresses the challenges of stock market prediction while incorporating vital risk considerations.
{"title":"RAGIC: Risk-Aware Generative Adversarial Model for Stock Interval Construction","authors":"Jingyi Gu, Wenlu Du, Guiling Wang","doi":"arxiv-2402.10760","DOIUrl":"https://doi.org/arxiv-2402.10760","url":null,"abstract":"Efforts to predict stock market outcomes have yielded limited success due to\u0000the inherently stochastic nature of the market, influenced by numerous\u0000unpredictable factors. Many existing prediction approaches focus on\u0000single-point predictions, lacking the depth needed for effective\u0000decision-making and often overlooking market risk. To bridge this gap, we\u0000propose a novel model, RAGIC, which introduces sequence generation for stock\u0000interval prediction to quantify uncertainty more effectively. Our approach\u0000leverages a Generative Adversarial Network (GAN) to produce future price\u0000sequences infused with randomness inherent in financial markets. RAGIC's\u0000generator includes a risk module, capturing the risk perception of informed\u0000investors, and a temporal module, accounting for historical price trends and\u0000seasonality. This multi-faceted generator informs the creation of\u0000risk-sensitive intervals through statistical inference, incorporating\u0000horizon-wise insights. The interval's width is carefully adjusted to reflect\u0000market volatility. Importantly, our approach relies solely on publicly\u0000available data and incurs only low computational overhead. RAGIC's evaluation\u0000across globally recognized broad-based indices demonstrates its balanced\u0000performance, offering both accuracy and informativeness. Achieving a consistent\u000095% coverage, RAGIC maintains a narrow interval width. This promising outcome\u0000suggests that our approach effectively addresses the challenges of stock market\u0000prediction while incorporating vital risk considerations.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"23 6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139902514","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 inclusion ensures that individuals have access to financial products and services that meet their needs. As a key contributing factor to economic growth and investment opportunity, financial inclusion increases consumer spending and consequently business development. It has been shown that institutions are more profitable when they provide marginalised social groups access to financial services. Customer segmentation based on consumer transaction data is a well-known strategy used to promote financial inclusion. While the required data is available to modern institutions, the challenge remains that segment annotations are usually difficult and/or expensive to obtain. This prevents the usage of time series classification models for customer segmentation based on domain expert knowledge. As a result, clustering is an attractive alternative to partition customers into homogeneous groups based on the spending behaviour encoded within their transaction data. In this paper, we present a solution to one of the key challenges preventing modern financial institutions from providing financially inclusive credit, savings and insurance products: the inability to understand consumer financial behaviour, and hence risk, without the introduction of restrictive conventional credit scoring techniques. We present a novel time series clustering algorithm that allows institutions to understand the financial behaviour of their customers. This enables unique product offerings to be provided based on the needs of the customer, without reliance on restrictive credit practices.
{"title":"Towards Financially Inclusive Credit Products Through Financial Time Series Clustering","authors":"Tristan Bester, Benjamin Rosman","doi":"arxiv-2402.11066","DOIUrl":"https://doi.org/arxiv-2402.11066","url":null,"abstract":"Financial inclusion ensures that individuals have access to financial\u0000products and services that meet their needs. As a key contributing factor to\u0000economic growth and investment opportunity, financial inclusion increases\u0000consumer spending and consequently business development. It has been shown that\u0000institutions are more profitable when they provide marginalised social groups\u0000access to financial services. Customer segmentation based on consumer\u0000transaction data is a well-known strategy used to promote financial inclusion.\u0000While the required data is available to modern institutions, the challenge\u0000remains that segment annotations are usually difficult and/or expensive to\u0000obtain. This prevents the usage of time series classification models for\u0000customer segmentation based on domain expert knowledge. As a result, clustering\u0000is an attractive alternative to partition customers into homogeneous groups\u0000based on the spending behaviour encoded within their transaction data. In this\u0000paper, we present a solution to one of the key challenges preventing modern\u0000financial institutions from providing financially inclusive credit, savings and\u0000insurance products: the inability to understand consumer financial behaviour,\u0000and hence risk, without the introduction of restrictive conventional credit\u0000scoring techniques. We present a novel time series clustering algorithm that\u0000allows institutions to understand the financial behaviour of their customers.\u0000This enables unique product offerings to be provided based on the needs of the\u0000customer, without reliance on restrictive credit practices.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"140 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139925578","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 the burgeoning realm of cryptocurrency, social media platforms like Twitter have become pivotal in influencing market trends and investor sentiments. In our study, we leverage GPT-4 and a fine-tuned transformer-based BERT model for a multimodal sentiment analysis, focusing on the impact of emoji sentiment on cryptocurrency markets. By translating emojis into quantifiable sentiment data, we correlate these insights with key market indicators like BTC Price and the VCRIX index. This approach may be fed into the development of trading strategies aimed at utilizing social media elements to identify and forecast market trends. Crucially, our findings suggest that strategies based on emoji sentiment can facilitate the avoidance of significant market downturns and contribute to the stabilization of returns. This research underscores the practical benefits of integrating advanced AI-driven analyses into financial strategies, offering a nuanced perspective on the interplay between digital communication and market dynamics in an academic context.
{"title":"Emoji Driven Crypto Assets Market Reactions","authors":"Xiaorui Zuo, Yao-Tsung Chen, Wolfgang Karl Härdle","doi":"arxiv-2402.10481","DOIUrl":"https://doi.org/arxiv-2402.10481","url":null,"abstract":"In the burgeoning realm of cryptocurrency, social media platforms like\u0000Twitter have become pivotal in influencing market trends and investor\u0000sentiments. In our study, we leverage GPT-4 and a fine-tuned transformer-based\u0000BERT model for a multimodal sentiment analysis, focusing on the impact of emoji\u0000sentiment on cryptocurrency markets. By translating emojis into quantifiable\u0000sentiment data, we correlate these insights with key market indicators like BTC\u0000Price and the VCRIX index. This approach may be fed into the development of\u0000trading strategies aimed at utilizing social media elements to identify and\u0000forecast market trends. Crucially, our findings suggest that strategies based\u0000on emoji sentiment can facilitate the avoidance of significant market downturns\u0000and contribute to the stabilization of returns. This research underscores the\u0000practical benefits of integrating advanced AI-driven analyses into financial\u0000strategies, offering a nuanced perspective on the interplay between digital\u0000communication and market dynamics in an academic context.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"147 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139904116","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：