While research of reinforcement learning applied to financial markets predominantly concentrates on finding optimal behaviours, it is worth to realize that the reinforcement learning returns $G_t$ and state value functions themselves are of interest and play a pivotal role in the evaluation of assets. Instead of focussing on the more complex task of finding optimal decision rules, this paper studies and applies the power of distributional state value functions in the context of financial market valuation and machine learning based trading algorithms. Accurate and trustworthy estimates of the distributions of $G_t$ provide a competitive edge leading to better informed decisions and more optimal behaviour. Herein, ideas from predictive knowledge and deep reinforcement learning are combined to introduce a novel family of models called CDG-Model, resulting in a highly flexible framework and intuitive approach with minimal assumptions regarding underlying distributions. The models allow seamless integration of typical financial modelling pitfalls like transaction costs, slippage and other possible costs or benefits into the model calculation. They can be applied to any kind of trading strategy or asset class. The frameworks introduced provide concrete business value through their potential in market valuation of single assets and portfolios, in the comparison of strategies as well as in the improvement of market timing. They can positively impact the performance and enhance the learning process of existing or new trading algorithms. They are of interest from a scientific point-of-view and open up multiple areas of future research. Initial implementations and tests were performed on real market data. While the results are promising, applying a robust statistical framework to evaluate the models in general remains a challenge and further investigations are needed.
{"title":"Exploiting Distributional Value Functions for Financial Market Valuation, Enhanced Feature Creation and Improvement of Trading Algorithms","authors":"Colin D. Grab","doi":"arxiv-2405.11686","DOIUrl":"https://doi.org/arxiv-2405.11686","url":null,"abstract":"While research of reinforcement learning applied to financial markets\u0000predominantly concentrates on finding optimal behaviours, it is worth to\u0000realize that the reinforcement learning returns $G_t$ and state value functions\u0000themselves are of interest and play a pivotal role in the evaluation of assets.\u0000Instead of focussing on the more complex task of finding optimal decision\u0000rules, this paper studies and applies the power of distributional state value\u0000functions in the context of financial market valuation and machine learning\u0000based trading algorithms. Accurate and trustworthy estimates of the\u0000distributions of $G_t$ provide a competitive edge leading to better informed\u0000decisions and more optimal behaviour. Herein, ideas from predictive knowledge\u0000and deep reinforcement learning are combined to introduce a novel family of\u0000models called CDG-Model, resulting in a highly flexible framework and intuitive\u0000approach with minimal assumptions regarding underlying distributions. The\u0000models allow seamless integration of typical financial modelling pitfalls like\u0000transaction costs, slippage and other possible costs or benefits into the model\u0000calculation. They can be applied to any kind of trading strategy or asset\u0000class. The frameworks introduced provide concrete business value through their\u0000potential in market valuation of single assets and portfolios, in the\u0000comparison of strategies as well as in the improvement of market timing. They\u0000can positively impact the performance and enhance the learning process of\u0000existing or new trading algorithms. They are of interest from a scientific\u0000point-of-view and open up multiple areas of future research. Initial\u0000implementations and tests were performed on real market data. While the results\u0000are promising, applying a robust statistical framework to evaluate the models\u0000in general remains a challenge and further investigations are needed.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"14 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141149255","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}
Salma Elomari-Kessab, Guillaume Maitrier, Julius Bonart, Jean-Philippe Bouchaud
Understanding the micro-dynamics of asset prices in modern electronic order books is crucial for investors and regulators. In this paper, we use an order by order Eurostoxx database spanning over 3 years to analyze the joint dynamics of prices and order flow. In order to alleviate various problems caused by high-frequency noise, we propose a double coarse-graining procedure that allows us to extract meaningful information at the minute time scale. We use Principal Component Analysis to construct "microstructure modes" that describe the most common flow/return patterns and allow one to separate them into bid-ask symmetric and bid-ask anti-symmetric. We define and calibrate a Vector Auto-Regressive (VAR) model that encodes the dynamical evolution of these modes. The parameters of the VAR model are found to be extremely stable in time, and lead to relatively high $R^2$ prediction scores, especially for symmetric liquidity modes. The VAR model becomes marginally unstable as more lags are included, reflecting the long-memory nature of flows and giving some further credence to the possibility of "endogenous liquidity crises". Although very satisfactory on several counts, we show that our VAR framework does not account for the well known square-root law of price impact.
{"title":"\"Microstructure Modes\" -- Disentangling the Joint Dynamics of Prices & Order Flow","authors":"Salma Elomari-Kessab, Guillaume Maitrier, Julius Bonart, Jean-Philippe Bouchaud","doi":"arxiv-2405.10654","DOIUrl":"https://doi.org/arxiv-2405.10654","url":null,"abstract":"Understanding the micro-dynamics of asset prices in modern electronic order\u0000books is crucial for investors and regulators. In this paper, we use an order\u0000by order Eurostoxx database spanning over 3 years to analyze the joint dynamics\u0000of prices and order flow. In order to alleviate various problems caused by\u0000high-frequency noise, we propose a double coarse-graining procedure that allows\u0000us to extract meaningful information at the minute time scale. We use Principal\u0000Component Analysis to construct \"microstructure modes\" that describe the most\u0000common flow/return patterns and allow one to separate them into bid-ask\u0000symmetric and bid-ask anti-symmetric. We define and calibrate a Vector\u0000Auto-Regressive (VAR) model that encodes the dynamical evolution of these\u0000modes. The parameters of the VAR model are found to be extremely stable in\u0000time, and lead to relatively high $R^2$ prediction scores, especially for\u0000symmetric liquidity modes. The VAR model becomes marginally unstable as more\u0000lags are included, reflecting the long-memory nature of flows and giving some\u0000further credence to the possibility of \"endogenous liquidity crises\". Although\u0000very satisfactory on several counts, we show that our VAR framework does not\u0000account for the well known square-root law of price impact.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"218 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141149307","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}
Prediction of stock prices has been a crucial and challenging task, especially in the case of highly volatile digital currencies such as Bitcoin. This research examineS the potential of using neural network models, namely LSTMs and GRUs, to forecast Bitcoin's price movements. We employ five-fold cross-validation to enhance generalization and utilize L2 regularization to reduce overfitting and noise. Our study demonstrates that the GRUs models offer better accuracy than LSTMs model for predicting Bitcoin's price. Specifically, the GRU model has an MSE of 4.67, while the LSTM model has an MSE of 6.25 when compared to the actual prices in the test set data. This finding indicates that GRU models are better equipped to process sequential data with long-term dependencies, a characteristic of financial time series data such as Bitcoin prices. In summary, our results provide valuable insights into the potential of neural network models for accurate Bitcoin price prediction and emphasize the importance of employing appropriate regularization techniques to enhance model performance.
{"title":"Comparative Study of Bitcoin Price Prediction","authors":"Ali Mohammadjafari","doi":"arxiv-2405.08089","DOIUrl":"https://doi.org/arxiv-2405.08089","url":null,"abstract":"Prediction of stock prices has been a crucial and challenging task,\u0000especially in the case of highly volatile digital currencies such as Bitcoin.\u0000This research examineS the potential of using neural network models, namely\u0000LSTMs and GRUs, to forecast Bitcoin's price movements. We employ five-fold\u0000cross-validation to enhance generalization and utilize L2 regularization to\u0000reduce overfitting and noise. Our study demonstrates that the GRUs models offer\u0000better accuracy than LSTMs model for predicting Bitcoin's price. Specifically,\u0000the GRU model has an MSE of 4.67, while the LSTM model has an MSE of 6.25 when\u0000compared to the actual prices in the test set data. This finding indicates that\u0000GRU models are better equipped to process sequential data with long-term\u0000dependencies, a characteristic of financial time series data such as Bitcoin\u0000prices. In summary, our results provide valuable insights into the potential of\u0000neural network models for accurate Bitcoin price prediction and emphasize the\u0000importance of employing appropriate regularization techniques to enhance model\u0000performance.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"30 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141060709","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}
Statistical analysis of high-frequency stock market order transaction data is conducted to understand order transition dynamics. We employ a first-order time-homogeneous discrete-time Markov chain model to the sequence of orders of stocks belonging to six different sectors during the USA-China trade war of 2018. The Markov property of the order sequence is validated by the Chi-square test. We estimate the transition probability matrix of the sequence using maximum likelihood estimation. From the heat-map of these matrices, we found the presence of active participation by different types of traders during high volatility days. On such days, these traders place limit orders primarily with the intention of deleting the majority of them to influence the market. These findings are supported by high stationary distribution and low mean recurrence values of add and delete orders. Further, we found similar spectral gap and entropy rate values, which indicates that similar trading strategies are employed on both high and low volatility days during the trade war. Among all the sectors considered in this study, we observe that there is a recurring pattern of full execution orders in Finance & Banking sector. This shows that the banking stocks are resilient during the trade war. Hence, this study may be useful in understanding stock market order dynamics and devise trading strategies accordingly on high and low volatility days during extreme macroeconomic events.
{"title":"High-Frequency Stock Market Order Transitions during the US-China Trade War 2018: A Discrete-Time Markov Chain Analysis","authors":"Salam Rabindrajit Luwang, Anish Rai, Md. Nurujjaman, Om Prakash, Chittaranjan Hens","doi":"arxiv-2405.05634","DOIUrl":"https://doi.org/arxiv-2405.05634","url":null,"abstract":"Statistical analysis of high-frequency stock market order transaction data is\u0000conducted to understand order transition dynamics. We employ a first-order\u0000time-homogeneous discrete-time Markov chain model to the sequence of orders of\u0000stocks belonging to six different sectors during the USA-China trade war of\u00002018. The Markov property of the order sequence is validated by the Chi-square\u0000test. We estimate the transition probability matrix of the sequence using\u0000maximum likelihood estimation. From the heat-map of these matrices, we found\u0000the presence of active participation by different types of traders during high\u0000volatility days. On such days, these traders place limit orders primarily with\u0000the intention of deleting the majority of them to influence the market. These\u0000findings are supported by high stationary distribution and low mean recurrence\u0000values of add and delete orders. Further, we found similar spectral gap and\u0000entropy rate values, which indicates that similar trading strategies are\u0000employed on both high and low volatility days during the trade war. Among all\u0000the sectors considered in this study, we observe that there is a recurring\u0000pattern of full execution orders in Finance & Banking sector. This shows that\u0000the banking stocks are resilient during the trade war. Hence, this study may be\u0000useful in understanding stock market order dynamics and devise trading\u0000strategies accordingly on high and low volatility days during extreme\u0000macroeconomic events.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"76 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140929882","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}
Kundan Mukhia, Anish Rai, SR Luwang, Md Nurujjaman, Sushovan Majhi, Chittaranjan Hens
This paper identifies the cryptocurrency market crashes and analyses its dynamics using the complex network. We identify three distinct crashes during 2017-20, and the analysis is carried out by dividing the time series into pre-crash, crash, and post-crash periods. Partial correlation based complex network analysis is carried out to study the crashes. Degree density ($rho_D$), average path length ($bar{l}$), and average clustering coefficient ($overline{cc}$) are estimated from these networks. We find that both $rho_D$ and $overline{cc}$ are smallest during the pre-crash period, and spike during the crash suggesting the network is dense during a crash. Although $rho_D$ and $overline{cc}$ decrease in the post-crash period, they remain higher than pre-crash levels for the 2017-18 and 2018-19 crashes suggesting a market attempt to return to normalcy. We get $bar{l}$ is minimal during the crash period, suggesting a rapid flow of information. A dense network and rapid information flow suggest that during a crash uninformed synchronized panic sell-off happens. However, during the 2019-20 crash, the values of $rho_D$, $overline{cc}$, and $bar{l}$ did not vary significantly, indicating minimal change in dynamics compared to other crashes. The findings of this study may guide investors in making decisions during market crashes.
{"title":"Complex network analysis of cryptocurrency market during crashes","authors":"Kundan Mukhia, Anish Rai, SR Luwang, Md Nurujjaman, Sushovan Majhi, Chittaranjan Hens","doi":"arxiv-2405.05642","DOIUrl":"https://doi.org/arxiv-2405.05642","url":null,"abstract":"This paper identifies the cryptocurrency market crashes and analyses its\u0000dynamics using the complex network. We identify three distinct crashes during\u00002017-20, and the analysis is carried out by dividing the time series into\u0000pre-crash, crash, and post-crash periods. Partial correlation based complex\u0000network analysis is carried out to study the crashes. Degree density\u0000($rho_D$), average path length ($bar{l}$), and average clustering coefficient\u0000($overline{cc}$) are estimated from these networks. We find that both $rho_D$\u0000and $overline{cc}$ are smallest during the pre-crash period, and spike during\u0000the crash suggesting the network is dense during a crash. Although $rho_D$ and\u0000$overline{cc}$ decrease in the post-crash period, they remain higher than\u0000pre-crash levels for the 2017-18 and 2018-19 crashes suggesting a market\u0000attempt to return to normalcy. We get $bar{l}$ is minimal during the crash\u0000period, suggesting a rapid flow of information. A dense network and rapid\u0000information flow suggest that during a crash uninformed synchronized panic\u0000sell-off happens. However, during the 2019-20 crash, the values of $rho_D$,\u0000$overline{cc}$, and $bar{l}$ did not vary significantly, indicating minimal\u0000change in dynamics compared to other crashes. The findings of this study may\u0000guide investors in making decisions during market crashes.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"128 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140929827","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}
A Wright function based framework is proposed to combine and extend several distribution families. The $alpha$-stable distribution is generalized by adding the degree of freedom parameter. The PDF of this two-sided super distribution family subsumes those of the original $alpha$-stable, Student's t distributions, as well as the exponential power distribution and the modified Bessel function of the second kind. Its CDF leads to a fractional extension of the Gauss hypergeometric function. The degree of freedom makes possible for valid variance, skewness, and kurtosis, just like Student's t. The original $alpha$-stable distribution is viewed as having one degree of freedom, that explains why it lacks most of the moments. A skew-Gaussian kernel is derived from the characteristic function of the $alpha$-stable law, which maximally preserves the law in the new framework. To facilitate such framework, the stable count distribution is generalized as the fractional extension of the generalized gamma distribution. It provides rich subordination capabilities, one of which is the fractional $chi$ distribution that supplies the needed 'degree of freedom' parameter. Hence, the "new" $alpha$-stable distribution is a "ratio distribution" of the skew-Gaussian kernel and the fractional $chi$ distribution. Mathematically, it is a new form of higher transcendental function under the Wright function family. Last, the new univariate symmetric distribution is extended to the multivariate elliptical distribution successfully.
本文提出了一个基于赖特函数的框架来组合和扩展几个分布族。通过添加自由度参数，对 $alpha$ 稳定分布进行了泛化。这个双面超分布族的 PDF 包含了原始的 $alpha$-稳定分布、Student's t 分布、指数幂分布和修正的第二类贝塞尔函数的 PDF。它的 CDF 导致高斯超几何函数的分数扩展。自由度使得方差、偏斜度和峰度成为可能，就像 Student's t 分布一样。从 $alpha$ 稳定规律的特征函数中导出了一个偏高斯核，它在新框架中最大限度地保留了该规律。为了促进这种框架，稳定计数分布被概括为广义伽马分布的分数扩展。它提供了丰富的从属能力，其中之一就是分数 $chi$ 分布，它提供了所需的 "自由度 "参数。因此，"新的"$α$稳定分布是偏高斯核与分数$chi$分布的 "比率分布"。在数学上，它是赖特函数族下的一种新的高超越函数形式。最后，新的单变量对称分布成功地扩展到了多变量椭圆分布。
{"title":"Generalization of the Alpha-Stable Distribution with the Degree of Freedom","authors":"Stephen H. Lihn","doi":"arxiv-2405.04693","DOIUrl":"https://doi.org/arxiv-2405.04693","url":null,"abstract":"A Wright function based framework is proposed to combine and extend several\u0000distribution families. The $alpha$-stable distribution is generalized by\u0000adding the degree of freedom parameter. The PDF of this two-sided super\u0000distribution family subsumes those of the original $alpha$-stable, Student's t\u0000distributions, as well as the exponential power distribution and the modified\u0000Bessel function of the second kind. Its CDF leads to a fractional extension of\u0000the Gauss hypergeometric function. The degree of freedom makes possible for\u0000valid variance, skewness, and kurtosis, just like Student's t. The original\u0000$alpha$-stable distribution is viewed as having one degree of freedom, that\u0000explains why it lacks most of the moments. A skew-Gaussian kernel is derived\u0000from the characteristic function of the $alpha$-stable law, which maximally\u0000preserves the law in the new framework. To facilitate such framework, the\u0000stable count distribution is generalized as the fractional extension of the\u0000generalized gamma distribution. It provides rich subordination capabilities,\u0000one of which is the fractional $chi$ distribution that supplies the needed\u0000'degree of freedom' parameter. Hence, the \"new\" $alpha$-stable distribution is\u0000a \"ratio distribution\" of the skew-Gaussian kernel and the fractional $chi$\u0000distribution. Mathematically, it is a new form of higher transcendental\u0000function under the Wright function family. Last, the new univariate symmetric\u0000distribution is extended to the multivariate elliptical distribution\u0000successfully.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"12 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140929881","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 highlights the significance of mesoscale structures, particularly the core-periphery structure, in financial networks for portfolio optimization. We build portfolios of stocks belonging to the periphery part of the Planar maximally filtered subgraphs of the underlying network of stocks created from Pearson correlations between pairs of stocks and compare its performance with some well-known strategies of Pozzi et. al. hinging around the local indices of centrality in terms of the Sharpe ratio, returns and standard deviation. Our findings reveal that these portfolios consistently outperform traditional strategies and further the core-periphery profile obtained is statistically significant across time periods. These empirical findings substantiate the efficacy of using the core-periphery profile of the stock market network for both inter-day and intraday trading and provide valuable insights for investors seeking better returns.
{"title":"A novel portfolio construction strategy based on the core-periphery profile of stocks","authors":"Imran Ansari, Charu Sharma, Akshay Agrawal, Niteesh Sahni","doi":"arxiv-2405.12993","DOIUrl":"https://doi.org/arxiv-2405.12993","url":null,"abstract":"This paper highlights the significance of mesoscale structures, particularly\u0000the core-periphery structure, in financial networks for portfolio optimization.\u0000We build portfolios of stocks belonging to the periphery part of the Planar\u0000maximally filtered subgraphs of the underlying network of stocks created from\u0000Pearson correlations between pairs of stocks and compare its performance with\u0000some well-known strategies of Pozzi et. al. hinging around the local indices of\u0000centrality in terms of the Sharpe ratio, returns and standard deviation. Our\u0000findings reveal that these portfolios consistently outperform traditional\u0000strategies and further the core-periphery profile obtained is statistically\u0000significant across time periods. These empirical findings substantiate the\u0000efficacy of using the core-periphery profile of the stock market network for\u0000both inter-day and intraday trading and provide valuable insights for investors\u0000seeking better returns.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"173 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141149265","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}
Computing cost of equity for private corporations and performing comparable company analysis (comps) for both public and private corporations is an integral but tedious and time-consuming task, with important applications spanning the finance world, from valuations to internal planning. Performing comps traditionally often times include high ambiguity and subjectivity, leading to unreliability and inconsistency. In this paper, I will present a systematic and faster approach to compute cost of equity for private corporations and perform comps for both public and private corporations using spectral and agglomerative clustering. This leads to a reduction in the time required to perform comps by orders of magnitude and entire process being more consistent and reliable.
{"title":"Systematic Comparable Company Analysis and Computation of Cost of Equity using Clustering","authors":"Mohammed Perves","doi":"arxiv-2405.12991","DOIUrl":"https://doi.org/arxiv-2405.12991","url":null,"abstract":"Computing cost of equity for private corporations and performing comparable\u0000company analysis (comps) for both public and private corporations is an\u0000integral but tedious and time-consuming task, with important applications\u0000spanning the finance world, from valuations to internal planning. Performing\u0000comps traditionally often times include high ambiguity and subjectivity,\u0000leading to unreliability and inconsistency. In this paper, I will present a\u0000systematic and faster approach to compute cost of equity for private\u0000corporations and perform comps for both public and private corporations using\u0000spectral and agglomerative clustering. This leads to a reduction in the time\u0000required to perform comps by orders of magnitude and entire process being more\u0000consistent and reliable.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"59 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141149478","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 study, we utilize the Kalman-Filter analysis to assess market efficiency in major stock markets. The Kalman-Filter operates in two stages, assuming that the data contains a consistent trendline representing the true market value prior to being affected by noise. Unlike traditional methods, it can forecast stock price movements effectively. Our findings reveal significant portfolio returns in emerging markets such as Korea, Vietnam, and Malaysia, as well as positive returns in developed markets like the UK, Europe, Japan, and Hong Kong. This suggests that the Kalman-Filter-based price reversal indicator yields promising results across various market types.
{"title":"Analysis of market efficiency in main stock markets: using Karman-Filter as an approach","authors":"Beier Liu, Haiyun Zhu","doi":"arxiv-2404.16449","DOIUrl":"https://doi.org/arxiv-2404.16449","url":null,"abstract":"In this study, we utilize the Kalman-Filter analysis to assess market\u0000efficiency in major stock markets. The Kalman-Filter operates in two stages,\u0000assuming that the data contains a consistent trendline representing the true\u0000market value prior to being affected by noise. Unlike traditional methods, it\u0000can forecast stock price movements effectively. Our findings reveal significant\u0000portfolio returns in emerging markets such as Korea, Vietnam, and Malaysia, as\u0000well as positive returns in developed markets like the UK, Europe, Japan, and\u0000Hong Kong. This suggests that the Kalman-Filter-based price reversal indicator\u0000yields promising results across various market types.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"39 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140798055","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 paper benchmarks several Transformer models [4], to show how these models can judge sentiment from a news event. This signal can then be used for downstream modelling and signal identification for commodity trading. We find that fine-tuned BERT models outperform fine-tuned or vanilla GPT models on this task. Transformer models have revolutionized the field of natural language processing (NLP) in recent years, achieving state-of-the-art results on various tasks such as machine translation, text summarization, question answering, and natural language generation. Among the most prominent transformer models are Bidirectional Encoder Representations from Transformers (BERT) and Generative Pre-trained Transformer (GPT), which differ in their architectures and objectives. A CopBERT model training data and process overview is provided. The CopBERT model outperforms similar domain specific BERT trained models such as FinBERT. The below confusion matrices show the performance on CopBERT & CopGPT respectively. We see a ~10 percent increase in f1_score when compare CopBERT vs GPT4 and 16 percent increase vs CopGPT. Whilst GPT4 is dominant It highlights the importance of considering alternatives to GPT models for financial engineering tasks, given risks of hallucinations, and challenges with interpretability. We unsurprisingly see the larger LLMs outperform the BERT models, with predictive power. In summary BERT is partially the new XGboost, what it lacks in predictive power it provides with higher levels of interpretability. Concluding that BERT models might not be the next XGboost [2], but represent an interesting alternative for financial engineering tasks, that require a blend of interpretability and accuracy.
{"title":"BERT vs GPT for financial engineering","authors":"Edward Sharkey, Philip Treleaven","doi":"arxiv-2405.12990","DOIUrl":"https://doi.org/arxiv-2405.12990","url":null,"abstract":"The paper benchmarks several Transformer models [4], to show how these models\u0000can judge sentiment from a news event. This signal can then be used for\u0000downstream modelling and signal identification for commodity trading. We find\u0000that fine-tuned BERT models outperform fine-tuned or vanilla GPT models on this\u0000task. Transformer models have revolutionized the field of natural language\u0000processing (NLP) in recent years, achieving state-of-the-art results on various\u0000tasks such as machine translation, text summarization, question answering, and\u0000natural language generation. Among the most prominent transformer models are\u0000Bidirectional Encoder Representations from Transformers (BERT) and Generative\u0000Pre-trained Transformer (GPT), which differ in their architectures and\u0000objectives. A CopBERT model training data and process overview is provided. The CopBERT\u0000model outperforms similar domain specific BERT trained models such as FinBERT.\u0000The below confusion matrices show the performance on CopBERT & CopGPT\u0000respectively. We see a ~10 percent increase in f1_score when compare CopBERT vs\u0000GPT4 and 16 percent increase vs CopGPT. Whilst GPT4 is dominant It highlights\u0000the importance of considering alternatives to GPT models for financial\u0000engineering tasks, given risks of hallucinations, and challenges with\u0000interpretability. We unsurprisingly see the larger LLMs outperform the BERT\u0000models, with predictive power. In summary BERT is partially the new XGboost,\u0000what it lacks in predictive power it provides with higher levels of\u0000interpretability. Concluding that BERT models might not be the next XGboost\u0000[2], but represent an interesting alternative for financial engineering tasks,\u0000that require a blend of interpretability and accuracy.","PeriodicalId":501139,"journal":{"name":"arXiv - QuantFin - Statistical Finance","volume":"19 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141149266","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：