We propose a novel approach to marked Hawkes kernel inference which we name the moment-based neural Hawkes estimation method. Hawkes processes are fully characterized by their first and second order statistics through a Fredholm integral equation of the second kind. Using recent advances in solving partial differential equations with physics-informed neural networks, we provide a numerical procedure to solve this integral equation in high dimension. Together with an adapted training pipeline, we give a generic set of hyperparameters that produces robust results across a wide range of kernel shapes. We conduct an extensive numerical validation on simulated data. We finally propose two applications of the method to the analysis of the microstructure of cryptocurrency markets. In a first application we extract the influence of volume on the arrival rate of BTC-USD trades and in a second application we analyze the causality relationships and their directions amongst a universe of 15 cryptocurrency pairs in a centralized exchange.
{"title":"Neural Hawkes: Non-Parametric Estimation in High Dimension and Causality Analysis in Cryptocurrency Markets","authors":"Timothée Fabre, Ioane Muni Toke","doi":"arxiv-2401.09361","DOIUrl":"https://doi.org/arxiv-2401.09361","url":null,"abstract":"We propose a novel approach to marked Hawkes kernel inference which we name\u0000the moment-based neural Hawkes estimation method. Hawkes processes are fully\u0000characterized by their first and second order statistics through a Fredholm\u0000integral equation of the second kind. Using recent advances in solving partial\u0000differential equations with physics-informed neural networks, we provide a\u0000numerical procedure to solve this integral equation in high dimension. Together\u0000with an adapted training pipeline, we give a generic set of hyperparameters\u0000that produces robust results across a wide range of kernel shapes. We conduct\u0000an extensive numerical validation on simulated data. We finally propose two\u0000applications of the method to the analysis of the microstructure of\u0000cryptocurrency markets. In a first application we extract the influence of\u0000volume on the arrival rate of BTC-USD trades and in a second application we\u0000analyze the causality relationships and their directions amongst a universe of\u000015 cryptocurrency pairs in a centralized exchange.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":"143 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139495926","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 study a new "laminated" queueing model for orders on batched trading venues such as decentralised exchanges. The model aims to capture and generalise transaction queueing infrastructure that has arisen to organise MEV activity on public blockchains such as Ethereum, providing convenient channels for sophisticated agents to extract value by acting on end-user order flow by performing arbitrage and related HFT activities. In our model, market orders are interspersed with orders created by arbitrageurs that under idealised conditions reset the marginal price to a global equilibrium between each trade, improving predictability of execution for liquidity traders. If an arbitrageur has a chance to land multiple opportunities in a row, he may attempt to manipulate the execution price of the intervening market order by a probabilistic blind sandwiching strategy. To study how bad this manipulation can get, we introduce and bound a price manipulation coefficient that measures the deviation from global equilibrium of local pricing quoted by a rational arbitrageur. We exhibit cases in which this coefficient is well approximated by a "zeta value' with interpretable and empirically measurable parameters.
{"title":"Do backrun auctions protect traders?","authors":"Andrew W. Macpherson","doi":"arxiv-2401.08302","DOIUrl":"https://doi.org/arxiv-2401.08302","url":null,"abstract":"We study a new \"laminated\" queueing model for orders on batched trading\u0000venues such as decentralised exchanges. The model aims to capture and\u0000generalise transaction queueing infrastructure that has arisen to organise MEV\u0000activity on public blockchains such as Ethereum, providing convenient channels\u0000for sophisticated agents to extract value by acting on end-user order flow by\u0000performing arbitrage and related HFT activities. In our model, market orders\u0000are interspersed with orders created by arbitrageurs that under idealised\u0000conditions reset the marginal price to a global equilibrium between each trade,\u0000improving predictability of execution for liquidity traders. If an arbitrageur has a chance to land multiple opportunities in a row, he\u0000may attempt to manipulate the execution price of the intervening market order\u0000by a probabilistic blind sandwiching strategy. To study how bad this\u0000manipulation can get, we introduce and bound a price manipulation coefficient\u0000that measures the deviation from global equilibrium of local pricing quoted by\u0000a rational arbitrageur. We exhibit cases in which this coefficient is well\u0000approximated by a \"zeta value' with interpretable and empirically measurable\u0000parameters.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":"84 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139482763","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}
Equity auctions display several distinctive characteristics in contrast to continuous trading. As the auction time approaches, the rate of events accelerates causing a substantial liquidity buildup around the indicative price. This, in turn, results in a reduced price impact and decreased volatility of the indicative price. In this study, we adapt the latent/revealed order book framework to the specifics of equity auctions. We provide precise measurements of the model parameters, including order submissions, cancellations, and diffusion rates. Our setup allows us to describe the full dynamics of the average order book during closing auctions in Euronext Paris. These findings support the relevance of the latent liquidity framework in describing limit order book dynamics. Lastly, we analyze the factors contributing to a sub-diffusive indicative price and demonstrate the absence of indicative price predictability.
{"title":"Equity auction dynamics: latent liquidity models with activity acceleration","authors":"Mohammed Salek, Damien Challet, Ioane Muni Toke","doi":"arxiv-2401.06724","DOIUrl":"https://doi.org/arxiv-2401.06724","url":null,"abstract":"Equity auctions display several distinctive characteristics in contrast to\u0000continuous trading. As the auction time approaches, the rate of events\u0000accelerates causing a substantial liquidity buildup around the indicative\u0000price. This, in turn, results in a reduced price impact and decreased\u0000volatility of the indicative price. In this study, we adapt the latent/revealed\u0000order book framework to the specifics of equity auctions. We provide precise\u0000measurements of the model parameters, including order submissions,\u0000cancellations, and diffusion rates. Our setup allows us to describe the full\u0000dynamics of the average order book during closing auctions in Euronext Paris.\u0000These findings support the relevance of the latent liquidity framework in\u0000describing limit order book dynamics. Lastly, we analyze the factors\u0000contributing to a sub-diffusive indicative price and demonstrate the absence of\u0000indicative price predictability.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":"24 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139470909","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 find closed-form solutions to the stochastic game between a broker and a mean-field of informed traders. In the finite player game, the informed traders observe a common signal and a private signal. The broker, on the other hand, observes the trading speed of each of his clients and provides liquidity to the informed traders. Each player in the game optimises wealth adjusted by inventory penalties. In the mean field version of the game, using a G^ateaux derivative approach, we characterise the solution to the game with a system of forward-backward stochastic differential equations that we solve explicitly. We find that the optimal trading strategy of the broker is linear on his own inventory, on the average inventory among informed traders, and on the common signal or the average trading speed of the informed traders. The Nash equilibrium we find helps informed traders decide how to use private information, and helps brokers decide how much of the order flow they should externalise or internalise when facing a large number of clients.
{"title":"A Mean Field Game between Informed Traders and a Broker","authors":"Philippe Bergault, Leandro Sánchez-Betancourt","doi":"arxiv-2401.05257","DOIUrl":"https://doi.org/arxiv-2401.05257","url":null,"abstract":"We find closed-form solutions to the stochastic game between a broker and a\u0000mean-field of informed traders. In the finite player game, the informed traders\u0000observe a common signal and a private signal. The broker, on the other hand,\u0000observes the trading speed of each of his clients and provides liquidity to the\u0000informed traders. Each player in the game optimises wealth adjusted by\u0000inventory penalties. In the mean field version of the game, using a G^ateaux\u0000derivative approach, we characterise the solution to the game with a system of\u0000forward-backward stochastic differential equations that we solve explicitly. We\u0000find that the optimal trading strategy of the broker is linear on his own\u0000inventory, on the average inventory among informed traders, and on the common\u0000signal or the average trading speed of the informed traders. The Nash\u0000equilibrium we find helps informed traders decide how to use private\u0000information, and helps brokers decide how much of the order flow they should\u0000externalise or internalise when facing a large number of clients.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":"127 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139420996","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 study the temporal evolution of the holding-time distribution of bitcoins and find that the average distribution of holding-time is a heavy-tailed power law extending from one day to over at least $200$ weeks with an exponent approximately equal to $0.9$, indicating very long memory effects. We also report significant sample-to-sample variations of the distribution of holding times, which can be best characterized as multiscaling, with power-law exponents varying between $0.3$ and $2.5$ depending on bitcoin price regimes. We document significant differences between the distributions of book-to-market and of realized returns, showing that traders obtain far from optimal performance. We also report strong direct qualitative and quantitative evidence of the disposition effect in the Bitcoin Blockchain data. Defining age-dependent transaction flows as the fraction of bitcoins that are traded at a given time and that were born (last traded) at some specific earlier time, we document that the time-averaged transaction flow fraction has a power law dependence as a function of age, with an exponent close to $-1.5$, a value compatible with priority queuing theory. We document the existence of multifractality on the measure defined as the normalized number of bitcoins exchanged at a given time.
{"title":"Scaling Laws And Statistical Properties of The Transaction Flows And Holding Times of Bitcoin","authors":"Didier Sornette, Yu Zhang","doi":"arxiv-2401.04702","DOIUrl":"https://doi.org/arxiv-2401.04702","url":null,"abstract":"We study the temporal evolution of the holding-time distribution of bitcoins\u0000and find that the average distribution of holding-time is a heavy-tailed power\u0000law extending from one day to over at least $200$ weeks with an exponent\u0000approximately equal to $0.9$, indicating very long memory effects. We also\u0000report significant sample-to-sample variations of the distribution of holding\u0000times, which can be best characterized as multiscaling, with power-law\u0000exponents varying between $0.3$ and $2.5$ depending on bitcoin price regimes.\u0000We document significant differences between the distributions of book-to-market\u0000and of realized returns, showing that traders obtain far from optimal\u0000performance. We also report strong direct qualitative and quantitative evidence\u0000of the disposition effect in the Bitcoin Blockchain data. Defining\u0000age-dependent transaction flows as the fraction of bitcoins that are traded at\u0000a given time and that were born (last traded) at some specific earlier time, we\u0000document that the time-averaged transaction flow fraction has a power law\u0000dependence as a function of age, with an exponent close to $-1.5$, a value\u0000compatible with priority queuing theory. We document the existence of\u0000multifractality on the measure defined as the normalized number of bitcoins\u0000exchanged at a given time.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":"33 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139410456","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}
Kenan Wood, Maurice Herlihy, Hammurabi Mendes, Jonad Pulaj
An automated market maker (AMM) is a state machine that manages pools of assets, allowing parties to buy and sell those assets according to a fixed mathematical formula. AMMs are typically implemented as smart contracts on blockchains, and its prices are kept in line with the overall market price by arbitrage: if the AMM undervalues an asset with respect to the market, an "arbitrageur" can make a risk-free profit by buying just enough of that asset to bring the AMM's price back in line with the market. AMMs, however, are not designed for assets that expire: that is, assets that cannot be produced or resold after a specified date. As assets approach expiration, arbitrage may not be able to reconcile supply and demand, and the liquidity providers that funded the AMM may have excessive exposure to risk due to rapid price variations. This paper formally describes the design of a decentralized exchange (DEX) for assets that expire, combining aspects of AMMs and limit-order books. We ensure liveness and market clearance, providing mechanisms for liquidity providers to control their exposure to risk and adjust prices dynamically in response to situations where arbitrage may fail.
{"title":"Expiring Assets in Automated Market Makers","authors":"Kenan Wood, Maurice Herlihy, Hammurabi Mendes, Jonad Pulaj","doi":"arxiv-2401.04289","DOIUrl":"https://doi.org/arxiv-2401.04289","url":null,"abstract":"An automated market maker (AMM) is a state machine that manages pools of\u0000assets, allowing parties to buy and sell those assets according to a fixed\u0000mathematical formula. AMMs are typically implemented as smart contracts on\u0000blockchains, and its prices are kept in line with the overall market price by\u0000arbitrage: if the AMM undervalues an asset with respect to the market, an\u0000\"arbitrageur\" can make a risk-free profit by buying just enough of that asset\u0000to bring the AMM's price back in line with the market. AMMs, however, are not designed for assets that expire: that is, assets that\u0000cannot be produced or resold after a specified date. As assets approach\u0000expiration, arbitrage may not be able to reconcile supply and demand, and the\u0000liquidity providers that funded the AMM may have excessive exposure to risk due\u0000to rapid price variations. This paper formally describes the design of a decentralized exchange (DEX)\u0000for assets that expire, combining aspects of AMMs and limit-order books. We\u0000ensure liveness and market clearance, providing mechanisms for liquidity\u0000providers to control their exposure to risk and adjust prices dynamically in\u0000response to situations where arbitrage may fail.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":"83 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139409635","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 addresses the problem of meta order execution from a broker-dealer's point of view in Almgren-Chriss model under order fill uncertainty. A broker-dealer agency is authorized to execute an order of trading on client's behalf. The strategies that the agent is allowed to deploy is subject to a benchmark, referred to as the reservation strategy, regulated by the client. We formulate the broker's problem as a utility maximization problem in which the broker seeks to maximize his utility of excess profit-and-loss at the execution horizon. Optimal strategy in feedback form is obtained in closed form. In the absence of execution risk, the optimal strategies subject to reservation strategies are deterministic. We establish an affine structure among the trading trajectories under optimal strategies subject to general reservation strategies using implementation shortfall and target close orders as basis. We conclude the paper with numerical experiments illustrating the trading trajectories as well as histograms of terminal wealth and utility at investment horizon under optimal strategies versus those under TWAP strategies.
{"title":"Optimal Order Execution subject to Reservation Strategies under Execution Risk","authors":"Xue Cheng, Peng Guo, Tai-ho Wang","doi":"arxiv-2401.03305","DOIUrl":"https://doi.org/arxiv-2401.03305","url":null,"abstract":"The paper addresses the problem of meta order execution from a\u0000broker-dealer's point of view in Almgren-Chriss model under order fill\u0000uncertainty. A broker-dealer agency is authorized to execute an order of\u0000trading on client's behalf. The strategies that the agent is allowed to deploy\u0000is subject to a benchmark, referred to as the reservation strategy, regulated\u0000by the client. We formulate the broker's problem as a utility maximization\u0000problem in which the broker seeks to maximize his utility of excess\u0000profit-and-loss at the execution horizon. Optimal strategy in feedback form is\u0000obtained in closed form. In the absence of execution risk, the optimal\u0000strategies subject to reservation strategies are deterministic. We establish an\u0000affine structure among the trading trajectories under optimal strategies\u0000subject to general reservation strategies using implementation shortfall and\u0000target close orders as basis. We conclude the paper with numerical experiments\u0000illustrating the trading trajectories as well as histograms of terminal wealth\u0000and utility at investment horizon under optimal strategies versus those under\u0000TWAP strategies.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":"30 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139409761","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 conducts an extensive analysis of Bitcoin return series, with a primary focus on three volatility metrics: historical volatility (calculated as the sample standard deviation), forecasted volatility (derived from GARCH-type models), and implied volatility (computed from the emerging Bitcoin options market). These measures of volatility serve as indicators of market expectations for conditional volatility and are compared to elucidate their differences and similarities. The central finding of this study underscores a notably high expected level of volatility, both on a daily and annual basis, across all the methodologies employed. However, it's crucial to emphasize the potential challenges stemming from suboptimal liquidity in the Bitcoin options market. These liquidity constraints may lead to discrepancies in the computed values of implied volatility, particularly in scenarios involving extreme moneyness or maturity. This analysis provides valuable insights into Bitcoin's volatility landscape, shedding light on the unique characteristics and dynamics of this cryptocurrency within the context of financial markets.
{"title":"Forecasting Bitcoin Volatility: A Comparative Analysis of Volatility Approaches","authors":"Cristina Chinazzo, Vahidin Jeleskovic","doi":"arxiv-2401.02049","DOIUrl":"https://doi.org/arxiv-2401.02049","url":null,"abstract":"This paper conducts an extensive analysis of Bitcoin return series, with a\u0000primary focus on three volatility metrics: historical volatility (calculated as\u0000the sample standard deviation), forecasted volatility (derived from GARCH-type\u0000models), and implied volatility (computed from the emerging Bitcoin options\u0000market). These measures of volatility serve as indicators of market\u0000expectations for conditional volatility and are compared to elucidate their\u0000differences and similarities. The central finding of this study underscores a\u0000notably high expected level of volatility, both on a daily and annual basis,\u0000across all the methodologies employed. However, it's crucial to emphasize the\u0000potential challenges stemming from suboptimal liquidity in the Bitcoin options\u0000market. These liquidity constraints may lead to discrepancies in the computed\u0000values of implied volatility, particularly in scenarios involving extreme\u0000moneyness or maturity. This analysis provides valuable insights into Bitcoin's\u0000volatility landscape, shedding light on the unique characteristics and dynamics\u0000of this cryptocurrency within the context of financial markets.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":"80 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139105462","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}
Célestin Coquidé, José Lages, Dima L. Shepelyansky
We extend the opinion formation approach to probe the world influence of economical organizations. Our opinion formation model mimics a battle between currencies within the international trade network. Based on the United Nations Comtrade database, we construct the world trade network for the years of the last decade from 2010 to 2020. We consider different core groups constituted by countries preferring to trade in a specific currency. We will consider principally two core groups, namely, 5 Anglo-Saxon countries which prefer to trade in US dollar and the 11 BRICS+ which prefer to trade in a hypothetical currency, hereafter called BRI, pegged to their economies. We determine the trade currency preference of the other countries via a Monte Carlo process depending on the direct transactions between the countries. The results obtained in the frame of this mathematical model show that starting from year 2014 the majority of the world countries would have preferred to trade in BRI than USD. The Monte Carlo process reaches a steady state with 3 distinct groups: two groups of countries preferring, whatever is the initial distribution of the trade currency preferences, to trade, one in BRI and the other in USD, and a third group of countries swinging as a whole between USD and BRI depending on the initial distribution of the trade currency preferences. We also analyze the battle between USD, EUR and BRI, and present the reduced Google matrix description of the trade relations between the Anglo-Saxon countries and the BRICS+.
{"title":"Opinion formation in the world trade network","authors":"Célestin Coquidé, José Lages, Dima L. Shepelyansky","doi":"arxiv-2401.02378","DOIUrl":"https://doi.org/arxiv-2401.02378","url":null,"abstract":"We extend the opinion formation approach to probe the world influence of\u0000economical organizations. Our opinion formation model mimics a battle between\u0000currencies within the international trade network. Based on the United Nations\u0000Comtrade database, we construct the world trade network for the years of the\u0000last decade from 2010 to 2020. We consider different core groups constituted by\u0000countries preferring to trade in a specific currency. We will consider\u0000principally two core groups, namely, 5 Anglo-Saxon countries which prefer to\u0000trade in US dollar and the 11 BRICS+ which prefer to trade in a hypothetical\u0000currency, hereafter called BRI, pegged to their economies. We determine the\u0000trade currency preference of the other countries via a Monte Carlo process\u0000depending on the direct transactions between the countries. The results\u0000obtained in the frame of this mathematical model show that starting from year\u00002014 the majority of the world countries would have preferred to trade in BRI\u0000than USD. The Monte Carlo process reaches a steady state with 3 distinct\u0000groups: two groups of countries preferring, whatever is the initial\u0000distribution of the trade currency preferences, to trade, one in BRI and the\u0000other in USD, and a third group of countries swinging as a whole between USD\u0000and BRI depending on the initial distribution of the trade currency\u0000preferences. We also analyze the battle between USD, EUR and BRI, and present\u0000the reduced Google matrix description of the trade relations between the\u0000Anglo-Saxon countries and the BRICS+.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":"20 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139102949","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}
Maochun Xu, Zixun Lan, Zheng Tao, Jiawei Du, Zongao Ye
Artificial Intelligence (AI) and Machine Learning (ML) are transforming the domain of Quantitative Trading (QT) through the deployment of advanced algorithms capable of sifting through extensive financial datasets to pinpoint lucrative investment openings. AI-driven models, particularly those employing ML techniques such as deep learning and reinforcement learning, have shown great prowess in predicting market trends and executing trades at a speed and accuracy that far surpass human capabilities. Its capacity to automate critical tasks, such as discerning market conditions and executing trading strategies, has been pivotal. However, persistent challenges exist in current QT methods, especially in effectively handling noisy and high-frequency financial data. Striking a balance between exploration and exploitation poses another challenge for AI-driven trading agents. To surmount these hurdles, our proposed solution, QTNet, introduces an adaptive trading model that autonomously formulates QT strategies through an intelligent trading agent. Incorporating deep reinforcement learning (DRL) with imitative learning methodologies, we bolster the proficiency of our model. To tackle the challenges posed by volatile financial datasets, we conceptualize the QT mechanism within the framework of a Partially Observable Markov Decision Process (POMDP). Moreover, by embedding imitative learning, the model can capitalize on traditional trading tactics, nurturing a balanced synergy between discovery and utilization. For a more realistic simulation, our trading agent undergoes training using minute-frequency data sourced from the live financial market. Experimental findings underscore the model's proficiency in extracting robust market features and its adaptability to diverse market conditions.
{"title":"Deep Reinforcement Learning for Quantitative Trading","authors":"Maochun Xu, Zixun Lan, Zheng Tao, Jiawei Du, Zongao Ye","doi":"arxiv-2312.15730","DOIUrl":"https://doi.org/arxiv-2312.15730","url":null,"abstract":"Artificial Intelligence (AI) and Machine Learning (ML) are transforming the\u0000domain of Quantitative Trading (QT) through the deployment of advanced\u0000algorithms capable of sifting through extensive financial datasets to pinpoint\u0000lucrative investment openings. AI-driven models, particularly those employing\u0000ML techniques such as deep learning and reinforcement learning, have shown\u0000great prowess in predicting market trends and executing trades at a speed and\u0000accuracy that far surpass human capabilities. Its capacity to automate critical\u0000tasks, such as discerning market conditions and executing trading strategies,\u0000has been pivotal. However, persistent challenges exist in current QT methods,\u0000especially in effectively handling noisy and high-frequency financial data.\u0000Striking a balance between exploration and exploitation poses another challenge\u0000for AI-driven trading agents. To surmount these hurdles, our proposed solution,\u0000QTNet, introduces an adaptive trading model that autonomously formulates QT\u0000strategies through an intelligent trading agent. Incorporating deep\u0000reinforcement learning (DRL) with imitative learning methodologies, we bolster\u0000the proficiency of our model. To tackle the challenges posed by volatile\u0000financial datasets, we conceptualize the QT mechanism within the framework of a\u0000Partially Observable Markov Decision Process (POMDP). Moreover, by embedding\u0000imitative learning, the model can capitalize on traditional trading tactics,\u0000nurturing a balanced synergy between discovery and utilization. For a more\u0000realistic simulation, our trading agent undergoes training using\u0000minute-frequency data sourced from the live financial market. Experimental\u0000findings underscore the model's proficiency in extracting robust market\u0000features and its adaptability to diverse market conditions.","PeriodicalId":501478,"journal":{"name":"arXiv - QuantFin - Trading and Market Microstructure","volume":"573 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139054147","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：