Jason R. Bailey, W. Brent Lindquist, Svetlozar T. Rachev
Using data from 2000 through 2022, we analyze the predictive capability of the annual numbers of new home constructions and four available environmental, social, and governance factors on the average annual price of homes sold in eight major U.S. cities. We contrast the predictive capability of a P-spline generalized additive model (GAM) against a strictly linear version of the commonly used generalized linear model (GLM). As the data for the annual price and predictor variables constitute non-stationary time series, to avoid spurious correlations in the analysis we transform each time series appropriately to produce stationary series for use in the GAM and GLM models. While arithmetic returns or first differences are adequate transformations for the predictor variables, for the average price response variable we utilize the series of innovations obtained from AR(q)-ARCH(1) fits. Based on the GAM results, we find that the influence of ESG factors varies markedly by city, reflecting geographic diversity. Notably, the presence of air conditioning emerges as a strong factor. Despite limitations on the length of available time series, this study represents a pivotal step toward integrating ESG considerations into predictive real estate models.
{"title":"Hedonic Models Incorporating ESG Factors for Time Series of Average Annual Home Prices","authors":"Jason R. Bailey, W. Brent Lindquist, Svetlozar T. Rachev","doi":"arxiv-2404.07132","DOIUrl":"https://doi.org/arxiv-2404.07132","url":null,"abstract":"Using data from 2000 through 2022, we analyze the predictive capability of\u0000the annual numbers of new home constructions and four available environmental,\u0000social, and governance factors on the average annual price of homes sold in\u0000eight major U.S. cities. We contrast the predictive capability of a P-spline\u0000generalized additive model (GAM) against a strictly linear version of the\u0000commonly used generalized linear model (GLM). As the data for the annual price\u0000and predictor variables constitute non-stationary time series, to avoid\u0000spurious correlations in the analysis we transform each time series\u0000appropriately to produce stationary series for use in the GAM and GLM models.\u0000While arithmetic returns or first differences are adequate transformations for\u0000the predictor variables, for the average price response variable we utilize the\u0000series of innovations obtained from AR(q)-ARCH(1) fits. Based on the GAM\u0000results, we find that the influence of ESG factors varies markedly by city,\u0000reflecting geographic diversity. Notably, the presence of air conditioning\u0000emerges as a strong factor. Despite limitations on the length of available time\u0000series, this study represents a pivotal step toward integrating ESG\u0000considerations into predictive real estate models.","PeriodicalId":501294,"journal":{"name":"arXiv - QuantFin - Computational Finance","volume":"84 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140565598","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 research paper introduces innovative approaches for multivariate time series forecasting based on different variations of the combined regression strategy. We use specific data preprocessing techniques which makes a radical change in the behaviour of prediction. We compare the performance of the model based on two types of hyper-parameter tuning Bayesian optimisation (BO) and Usual Grid search. Our proposed methodologies outperform all state-of-the-art comparative models. We illustrate the methodologies through eight time series datasets from three categories: cryptocurrency, stock index, and short-term load forecasting.
{"title":"Some variation of COBRA in sequential learning setup","authors":"Aryan Bhambu, Arabin Kumar Dey","doi":"arxiv-2405.04539","DOIUrl":"https://doi.org/arxiv-2405.04539","url":null,"abstract":"This research paper introduces innovative approaches for multivariate time\u0000series forecasting based on different variations of the combined regression\u0000strategy. We use specific data preprocessing techniques which makes a radical\u0000change in the behaviour of prediction. We compare the performance of the model\u0000based on two types of hyper-parameter tuning Bayesian optimisation (BO) and\u0000Usual Grid search. Our proposed methodologies outperform all state-of-the-art\u0000comparative models. We illustrate the methodologies through eight time series\u0000datasets from three categories: cryptocurrency, stock index, and short-term\u0000load forecasting.","PeriodicalId":501294,"journal":{"name":"arXiv - QuantFin - Computational Finance","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-04-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140940996","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper we provide more general features for the variable annuity contract with LTC payouts and GLWB proposed by the state-of-the-art and we refine its pricing methods. In particular, as to product features, we allow dynamic withdrawal strategies, including the surrender option. Furthermore, we consider stochastic interest rate, described by a Cox-Ingersoll-Ross (CIR) process. As to the numerical methods, we solve the stochastic control problem involved by the selection of the optimal withdrawal strategy by means of a robust tree method. We use such a method to estimate the fair price of the product. Furthermore, our numerical results show how the optimal withdrawal strategy varies over time with the health status of the policyholder. Our proposed tree method, we name Tree-LTC, proves to be efficient and reliable, when tested against the Monte Carlo approach.
{"title":"The Life Care Annuity: enhancing product features and refining pricing methods","authors":"G. Apicella, A. Molent, M. Gaudenzi","doi":"arxiv-2404.02858","DOIUrl":"https://doi.org/arxiv-2404.02858","url":null,"abstract":"In this paper we provide more general features for the variable annuity\u0000contract with LTC payouts and GLWB proposed by the state-of-the-art and we\u0000refine its pricing methods. In particular, as to product features, we allow\u0000dynamic withdrawal strategies, including the surrender option. Furthermore, we\u0000consider stochastic interest rate, described by a Cox-Ingersoll-Ross (CIR)\u0000process. As to the numerical methods, we solve the stochastic control problem\u0000involved by the selection of the optimal withdrawal strategy by means of a\u0000robust tree method. We use such a method to estimate the fair price of the\u0000product. Furthermore, our numerical results show how the optimal withdrawal\u0000strategy varies over time with the health status of the policyholder. Our\u0000proposed tree method, we name Tree-LTC, proves to be efficient and reliable,\u0000when tested against the Monte Carlo approach.","PeriodicalId":501294,"journal":{"name":"arXiv - QuantFin - Computational Finance","volume":"26 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140565702","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In recent decades, financial quantification has emerged and matured rapidly. For financial institutions such as funds, investment institutions are increasingly dissatisfied with the situation of passively constructing investment portfolios with average market returns, and are paying more and more attention to active quantitative strategy investment portfolios. This requires the introduction of active stock investment fund management models. Currently, in my country's stock fund investment market, there are many active quantitative investment strategies, and the algorithms used vary widely, such as SVM, random forest, RNN recurrent memory network, etc. This article focuses on this trend, using the emerging LSTM-GRU gate-controlled long short-term memory network model in the field of financial stock investment as a basis to build a set of active investment stock strategies, and combining it with SVM, which has been widely used in the field of quantitative stock investment. Comparing models such as RNN, theoretically speaking, compared to SVM that simply relies on kernel functions for high-order mapping and classification of data, neural network algorithms such as RNN and LSTM-GRU have better principles and are more suitable for processing financial stock data. Then, through multiple By comparison, it was finally found that the LSTM- GRU gate-controlled long short-term memory network has a better accuracy. By selecting the LSTM-GRU algorithm to construct a trading strategy based on the Shanghai and Shenzhen 300 Index constituent stocks, the parameters were adjusted and the neural layer connection was adjusted. Finally, It has significantly outperformed the benchmark index CSI 300 over the long term. The conclusion of this article is that the research results can provide certain quantitative strategy references for financial institutions to construct active stock investment portfolios.
{"title":"Intelligent Optimization of Mine Environmental Damage Assessment and Repair Strategies Based on Deep Learning","authors":"Qishuo Cheng","doi":"arxiv-2404.01624","DOIUrl":"https://doi.org/arxiv-2404.01624","url":null,"abstract":"In recent decades, financial quantification has emerged and matured rapidly.\u0000For financial institutions such as funds, investment institutions are\u0000increasingly dissatisfied with the situation of passively constructing\u0000investment portfolios with average market returns, and are paying more and more\u0000attention to active quantitative strategy investment portfolios. This requires\u0000the introduction of active stock investment fund management models. Currently,\u0000in my country's stock fund investment market, there are many active\u0000quantitative investment strategies, and the algorithms used vary widely, such\u0000as SVM, random forest, RNN recurrent memory network, etc. This article focuses\u0000on this trend, using the emerging LSTM-GRU gate-controlled long short-term\u0000memory network model in the field of financial stock investment as a basis to\u0000build a set of active investment stock strategies, and combining it with SVM,\u0000which has been widely used in the field of quantitative stock investment.\u0000Comparing models such as RNN, theoretically speaking, compared to SVM that\u0000simply relies on kernel functions for high-order mapping and classification of\u0000data, neural network algorithms such as RNN and LSTM-GRU have better principles\u0000and are more suitable for processing financial stock data. Then, through\u0000multiple By comparison, it was finally found that the LSTM- GRU gate-controlled\u0000long short-term memory network has a better accuracy. By selecting the LSTM-GRU\u0000algorithm to construct a trading strategy based on the Shanghai and Shenzhen\u0000300 Index constituent stocks, the parameters were adjusted and the neural layer\u0000connection was adjusted. Finally, It has significantly outperformed the\u0000benchmark index CSI 300 over the long term. The conclusion of this article is\u0000that the research results can provide certain quantitative strategy references\u0000for financial institutions to construct active stock investment portfolios.","PeriodicalId":501294,"journal":{"name":"arXiv - QuantFin - Computational Finance","volume":"4 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-04-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140566129","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}
With the recent development of large language models (LLMs), models that focus on certain domains and languages have been discussed for their necessity. There is also a growing need for benchmarks to evaluate the performance of current LLMs in each domain. Therefore, in this study, we constructed a benchmark comprising multiple tasks specific to the Japanese and financial domains and performed benchmark measurements on some models. Consequently, we confirmed that GPT-4 is currently outstanding, and that the constructed benchmarks function effectively. According to our analysis, our benchmark can differentiate benchmark scores among models in all performance ranges by combining tasks with different difficulties.
{"title":"Construction of a Japanese Financial Benchmark for Large Language Models","authors":"Masanori Hirano","doi":"arxiv-2403.15062","DOIUrl":"https://doi.org/arxiv-2403.15062","url":null,"abstract":"With the recent development of large language models (LLMs), models that\u0000focus on certain domains and languages have been discussed for their necessity.\u0000There is also a growing need for benchmarks to evaluate the performance of\u0000current LLMs in each domain. Therefore, in this study, we constructed a\u0000benchmark comprising multiple tasks specific to the Japanese and financial\u0000domains and performed benchmark measurements on some models. Consequently, we\u0000confirmed that GPT-4 is currently outstanding, and that the constructed\u0000benchmarks function effectively. According to our analysis, our benchmark can\u0000differentiate benchmark scores among models in all performance ranges by\u0000combining tasks with different difficulties.","PeriodicalId":501294,"journal":{"name":"arXiv - QuantFin - Computational Finance","volume":"29 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140298270","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}
Francesco Zola, Lander Segurola, Erin King, Martin Mullins, Raul Orduna
Tools for fighting cyber-criminal activities using new technologies are promoted and deployed every day. However, too often, they are unnecessarily complex and hard to use, requiring deep domain and technical knowledge. These characteristics often limit the engagement of law enforcement and end-users in these technologies that, despite their potential, remain misunderstood. For this reason, in this study, we describe our experience in combining learning and training methods and the potential benefits of gamification to enhance technology transfer and increase adult learning. In fact, in this case, participants are experienced practitioners in professions/industries that are exposed to terrorism financing (such as Law Enforcement Officers, Financial Investigation Officers, private investigators, etc.) We define training activities on different levels for increasing the exchange of information about new trends and criminal modus operandi among and within law enforcement agencies, intensifying cross-border cooperation and supporting efforts to combat and prevent terrorism funding activities. On the other hand, a game (hackathon) is designed to address realistic challenges related to the dark net, crypto assets, new payment systems and dark web marketplaces that could be used for terrorist activities. The entire methodology was evaluated using quizzes, contest results, and engagement metrics. In particular, training events show about 60% of participants complete the 11-week training course, while the Hackathon results, gathered in two pilot studies (Madrid and The Hague), show increasing expertise among the participants (progression in the achieved points on average). At the same time, more than 70% of participants positively evaluate the use of the gamification approach, and more than 85% of them consider the implemented Use Cases suitable for their investigations.
{"title":"Enhancing Law Enforcement Training: A Gamified Approach to Detecting Terrorism Financing","authors":"Francesco Zola, Lander Segurola, Erin King, Martin Mullins, Raul Orduna","doi":"arxiv-2403.13625","DOIUrl":"https://doi.org/arxiv-2403.13625","url":null,"abstract":"Tools for fighting cyber-criminal activities using new technologies are\u0000promoted and deployed every day. However, too often, they are unnecessarily\u0000complex and hard to use, requiring deep domain and technical knowledge. These\u0000characteristics often limit the engagement of law enforcement and end-users in\u0000these technologies that, despite their potential, remain misunderstood. For\u0000this reason, in this study, we describe our experience in combining learning\u0000and training methods and the potential benefits of gamification to enhance\u0000technology transfer and increase adult learning. In fact, in this case,\u0000participants are experienced practitioners in professions/industries that are\u0000exposed to terrorism financing (such as Law Enforcement Officers, Financial\u0000Investigation Officers, private investigators, etc.) We define training\u0000activities on different levels for increasing the exchange of information about\u0000new trends and criminal modus operandi among and within law enforcement\u0000agencies, intensifying cross-border cooperation and supporting efforts to\u0000combat and prevent terrorism funding activities. On the other hand, a game\u0000(hackathon) is designed to address realistic challenges related to the dark\u0000net, crypto assets, new payment systems and dark web marketplaces that could be\u0000used for terrorist activities. The entire methodology was evaluated using\u0000quizzes, contest results, and engagement metrics. In particular, training\u0000events show about 60% of participants complete the 11-week training course,\u0000while the Hackathon results, gathered in two pilot studies (Madrid and The\u0000Hague), show increasing expertise among the participants (progression in the\u0000achieved points on average). At the same time, more than 70% of participants\u0000positively evaluate the use of the gamification approach, and more than 85% of\u0000them consider the implemented Use Cases suitable for their investigations.","PeriodicalId":501294,"journal":{"name":"arXiv - QuantFin - Computational Finance","volume":"159 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140204349","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 develop a provably convergent kernel-based solver for path-dependent PDEs (PPDEs). Our numerical scheme leverages signature kernels, a recently introduced class of kernels on path-space. Specifically, we solve an optimal recovery problem by approximating the solution of a PPDE with an element of minimal norm in the signature reproducing kernel Hilbert space (RKHS) constrained to satisfy the PPDE at a finite collection of collocation paths. In the linear case, we show that the optimisation has a unique closed-form solution expressed in terms of signature kernel evaluations at the collocation paths. We prove consistency of the proposed scheme, guaranteeing convergence to the PPDE solution as the number of collocation points increases. Finally, several numerical examples are presented, in particular in the context of option pricing under rough volatility. Our numerical scheme constitutes a valid alternative to the ubiquitous Monte Carlo methods.
{"title":"A path-dependent PDE solver based on signature kernels","authors":"Alexandre Pannier, Cristopher Salvi","doi":"arxiv-2403.11738","DOIUrl":"https://doi.org/arxiv-2403.11738","url":null,"abstract":"We develop a provably convergent kernel-based solver for path-dependent PDEs\u0000(PPDEs). Our numerical scheme leverages signature kernels, a recently\u0000introduced class of kernels on path-space. Specifically, we solve an optimal\u0000recovery problem by approximating the solution of a PPDE with an element of\u0000minimal norm in the signature reproducing kernel Hilbert space (RKHS)\u0000constrained to satisfy the PPDE at a finite collection of collocation paths. In\u0000the linear case, we show that the optimisation has a unique closed-form\u0000solution expressed in terms of signature kernel evaluations at the collocation\u0000paths. We prove consistency of the proposed scheme, guaranteeing convergence to\u0000the PPDE solution as the number of collocation points increases. Finally,\u0000several numerical examples are presented, in particular in the context of\u0000option pricing under rough volatility. Our numerical scheme constitutes a valid\u0000alternative to the ubiquitous Monte Carlo methods.","PeriodicalId":501294,"journal":{"name":"arXiv - QuantFin - Computational Finance","volume":"16 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140171822","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 develop several innovations designed to bring the best practices of traditional investment funds to the blockchain landscape. Our innovations combine the superior mechanisms of mutual funds and hedge funds. Specifically, we illustrate how fund prices can be updated regularly like mutual funds and performance fees can be charged like hedge funds. We show how mutually hedged blockchain investment funds can operate with investor protection schemes - high water marks - and measures to offset trading slippage when redemptions happen. We provide detailed steps - including mathematical formulations and instructive pointers - to implement these ideas as blockchain smart contracts. We discuss how our designs overcome several blockchain bottlenecks and how we can make smart contracts smarter. We provide numerical illustrations of several scenarios related to the mechanisms we have tailored for blockchain implementation. The concepts we have developed for blockchain implementation can also be useful in traditional financial funds to calculate performance fees in a simplified manner. We highlight two main issues with the operation of mutual funds and hedge funds and show how blockchain technology can alleviate those concerns. The ideas developed here illustrate on one hand, how blockchain can solve many issues faced by the traditional world and on the other hand, how many innovations from traditional finance can benefit decentralized finance and speed its adoption. This becomes an example of symbiosis between decentralized and traditional finance - bringing these two realms closer and breaking down barriers between such artificial distinctions - wherein the future will be about providing better risk adjusted wealth appreciation opportunities to end customers through secure, reliable, accessible and transparent services - without getting too caught up about how such services are being rendered.
{"title":"The Democratization of Wealth Management: Hedged Mutual Fund Blockchain Protocol","authors":"Ravi Kashyap","doi":"arxiv-2405.02302","DOIUrl":"https://doi.org/arxiv-2405.02302","url":null,"abstract":"We develop several innovations designed to bring the best practices of\u0000traditional investment funds to the blockchain landscape. Our innovations\u0000combine the superior mechanisms of mutual funds and hedge funds. Specifically,\u0000we illustrate how fund prices can be updated regularly like mutual funds and\u0000performance fees can be charged like hedge funds. We show how mutually hedged\u0000blockchain investment funds can operate with investor protection schemes - high\u0000water marks - and measures to offset trading slippage when redemptions happen.\u0000We provide detailed steps - including mathematical formulations and instructive\u0000pointers - to implement these ideas as blockchain smart contracts. We discuss\u0000how our designs overcome several blockchain bottlenecks and how we can make\u0000smart contracts smarter. We provide numerical illustrations of several\u0000scenarios related to the mechanisms we have tailored for blockchain\u0000implementation. The concepts we have developed for blockchain implementation can also be\u0000useful in traditional financial funds to calculate performance fees in a\u0000simplified manner. We highlight two main issues with the operation of mutual\u0000funds and hedge funds and show how blockchain technology can alleviate those\u0000concerns. The ideas developed here illustrate on one hand, how blockchain can\u0000solve many issues faced by the traditional world and on the other hand, how\u0000many innovations from traditional finance can benefit decentralized finance and\u0000speed its adoption. This becomes an example of symbiosis between decentralized\u0000and traditional finance - bringing these two realms closer and breaking down\u0000barriers between such artificial distinctions - wherein the future will be\u0000about providing better risk adjusted wealth appreciation opportunities to end\u0000customers through secure, reliable, accessible and transparent services -\u0000without getting too caught up about how such services are being rendered.","PeriodicalId":501294,"journal":{"name":"arXiv - QuantFin - Computational Finance","volume":"13 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140882105","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 presents an innovative approach for predicting cryptocurrency time series, specifically focusing on Bitcoin, Ethereum, and Litecoin. The methodology integrates the use of technical indicators, a Performer neural network, and BiLSTM (Bidirectional Long Short-Term Memory) to capture temporal dynamics and extract significant features from raw cryptocurrency data. The application of technical indicators, such facilitates the extraction of intricate patterns, momentum, volatility, and trends. The Performer neural network, employing Fast Attention Via positive Orthogonal Random features (FAVOR+), has demonstrated superior computational efficiency and scalability compared to the traditional Multi-head attention mechanism in Transformer models. Additionally, the integration of BiLSTM in the feedforward network enhances the model's capacity to capture temporal dynamics in the data, processing it in both forward and backward directions. This is particularly advantageous for time series data where past and future data points can influence the current state. The proposed method has been applied to the hourly and daily timeframes of the major cryptocurrencies and its performance has been benchmarked against other methods documented in the literature. The results underscore the potential of the proposed method to outperform existing models, marking a significant progression in the field of cryptocurrency price prediction.
{"title":"Enhancing Price Prediction in Cryptocurrency Using Transformer Neural Network and Technical Indicators","authors":"Mohammad Ali Labbaf Khaniki, Mohammad Manthouri","doi":"arxiv-2403.03606","DOIUrl":"https://doi.org/arxiv-2403.03606","url":null,"abstract":"This study presents an innovative approach for predicting cryptocurrency time\u0000series, specifically focusing on Bitcoin, Ethereum, and Litecoin. The\u0000methodology integrates the use of technical indicators, a Performer neural\u0000network, and BiLSTM (Bidirectional Long Short-Term Memory) to capture temporal\u0000dynamics and extract significant features from raw cryptocurrency data. The\u0000application of technical indicators, such facilitates the extraction of\u0000intricate patterns, momentum, volatility, and trends. The Performer neural\u0000network, employing Fast Attention Via positive Orthogonal Random features\u0000(FAVOR+), has demonstrated superior computational efficiency and scalability\u0000compared to the traditional Multi-head attention mechanism in Transformer\u0000models. Additionally, the integration of BiLSTM in the feedforward network\u0000enhances the model's capacity to capture temporal dynamics in the data,\u0000processing it in both forward and backward directions. This is particularly\u0000advantageous for time series data where past and future data points can\u0000influence the current state. The proposed method has been applied to the hourly\u0000and daily timeframes of the major cryptocurrencies and its performance has been\u0000benchmarked against other methods documented in the literature. The results\u0000underscore the potential of the proposed method to outperform existing models,\u0000marking a significant progression in the field of cryptocurrency price\u0000prediction.","PeriodicalId":501294,"journal":{"name":"arXiv - QuantFin - Computational Finance","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140056532","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}
Christian Bayer, Chiheb Ben Hammouda, Antonis Papapantoleon, Michael Samet, Raúl Tempone
Efficiently pricing multi-asset options poses a significant challenge in quantitative finance. The Monte Carlo (MC) method remains the prevalent choice for pricing engines; however, its slow convergence rate impedes its practical application. Fourier methods leverage the knowledge of the characteristic function to accurately and rapidly value options with up to two assets. Nevertheless, they face hurdles in the high-dimensional settings due to the tensor product (TP) structure of commonly employed quadrature techniques. This work advocates using the randomized quasi-MC (RQMC) quadrature to improve the scalability of Fourier methods with high dimensions. The RQMC technique benefits from the smoothness of the integrand and alleviates the curse of dimensionality while providing practical error estimates. Nonetheless, the applicability of RQMC on the unbounded domain, $mathbb{R}^d$, requires a domain transformation to $[0,1]^d$, which may result in singularities of the transformed integrand at the corners of the hypercube, and deteriorate the rate of convergence of RQMC. To circumvent this difficulty, we design an efficient domain transformation procedure based on the derived boundary growth conditions of the integrand. This transformation preserves the sufficient regularity of the integrand and hence improves the rate of convergence of RQMC. To validate this analysis, we demonstrate the efficiency of employing RQMC with an appropriate transformation to evaluate options in the Fourier space for various pricing models, payoffs, and dimensions. Finally, we highlight the computational advantage of applying RQMC over MC or TP in the Fourier domain, and over MC in the physical domain for options with up to 15 assets.
{"title":"Quasi-Monte Carlo for Efficient Fourier Pricing of Multi-Asset Options","authors":"Christian Bayer, Chiheb Ben Hammouda, Antonis Papapantoleon, Michael Samet, Raúl Tempone","doi":"arxiv-2403.02832","DOIUrl":"https://doi.org/arxiv-2403.02832","url":null,"abstract":"Efficiently pricing multi-asset options poses a significant challenge in\u0000quantitative finance. The Monte Carlo (MC) method remains the prevalent choice\u0000for pricing engines; however, its slow convergence rate impedes its practical\u0000application. Fourier methods leverage the knowledge of the characteristic\u0000function to accurately and rapidly value options with up to two assets.\u0000Nevertheless, they face hurdles in the high-dimensional settings due to the\u0000tensor product (TP) structure of commonly employed quadrature techniques. This\u0000work advocates using the randomized quasi-MC (RQMC) quadrature to improve the\u0000scalability of Fourier methods with high dimensions. The RQMC technique\u0000benefits from the smoothness of the integrand and alleviates the curse of\u0000dimensionality while providing practical error estimates. Nonetheless, the\u0000applicability of RQMC on the unbounded domain, $mathbb{R}^d$, requires a\u0000domain transformation to $[0,1]^d$, which may result in singularities of the\u0000transformed integrand at the corners of the hypercube, and deteriorate the rate\u0000of convergence of RQMC. To circumvent this difficulty, we design an efficient\u0000domain transformation procedure based on the derived boundary growth conditions\u0000of the integrand. This transformation preserves the sufficient regularity of\u0000the integrand and hence improves the rate of convergence of RQMC. To validate\u0000this analysis, we demonstrate the efficiency of employing RQMC with an\u0000appropriate transformation to evaluate options in the Fourier space for various\u0000pricing models, payoffs, and dimensions. Finally, we highlight the\u0000computational advantage of applying RQMC over MC or TP in the Fourier domain,\u0000and over MC in the physical domain for options with up to 15 assets.","PeriodicalId":501294,"journal":{"name":"arXiv - QuantFin - Computational Finance","volume":"23 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140045473","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：