-
Mean-Field Limits for Nearly Unstable Hawkes Processes
Authors:
Grégoire Szymanski,
Wei Xu
Abstract:
In this paper, we establish general scaling limits for nearly unstable Hawkes processes in a mean-field regime by extending the method introduced by Jaisson and Rosenbaum. Under a mild asymptotic criticality condition on the self-exciting kernels $\{φ^n\}$, specifically $\|φ^n\|_{L^1} \to 1$, we first show that the scaling limits of these Hawkes processes are necessarily stochastic Volterra diffus…
▽ More
In this paper, we establish general scaling limits for nearly unstable Hawkes processes in a mean-field regime by extending the method introduced by Jaisson and Rosenbaum. Under a mild asymptotic criticality condition on the self-exciting kernels $\{φ^n\}$, specifically $\|φ^n\|_{L^1} \to 1$, we first show that the scaling limits of these Hawkes processes are necessarily stochastic Volterra diffusions of affine type. Moreover, we establish a propagation of chaos result for Hawkes systems with mean-field interactions, highlighting three distinct regimes for the limiting processes, which depend on the asymptotics of $n(1-\|φ^n\|_{L^1})^2$. These results provide a significant generalization of the findings by Delattre, Fournier and Hoffmann.
△ Less
Submitted 20 January, 2025;
originally announced January 2025.
-
Path-dependent Fractional Volterra Equations and the Microstructure of Rough Volatility Models driven by Poisson Random Measures
Authors:
Ulrich Horst,
Wei Xu,
Rouyi Zhang
Abstract:
We consider a microstructure foundation for rough volatility models driven by Poisson random measures. In our model the volatility is driven by self-exciting arrivals of market orders as well as self-exciting arrivals of limit orders and cancellations. The impact of market order on future order arrivals is captured by a Hawkes kernel with power law decay, and is hence persistent. The impact of lim…
▽ More
We consider a microstructure foundation for rough volatility models driven by Poisson random measures. In our model the volatility is driven by self-exciting arrivals of market orders as well as self-exciting arrivals of limit orders and cancellations. The impact of market order on future order arrivals is captured by a Hawkes kernel with power law decay, and is hence persistent. The impact of limit orders on future order arrivals is temporary, yet possibly long-lived. After suitable scaling the volatility process converges to a fractional Heston model driven by an additional Poisson random measure. The random measure generates occasional spikes and clusters of spikes in the volatility process. Our results are based on novel existence and uniqueness of solutions results for stochastic path-dependent Volterra equations driven by Poisson random measures.
△ Less
Submitted 20 December, 2024;
originally announced December 2024.
-
MOT: A Mixture of Actors Reinforcement Learning Method by Optimal Transport for Algorithmic Trading
Authors:
Xi Cheng,
Jinghao Zhang,
Yunan Zeng,
Wenfang Xue
Abstract:
Algorithmic trading refers to executing buy and sell orders for specific assets based on automatically identified trading opportunities. Strategies based on reinforcement learning (RL) have demonstrated remarkable capabilities in addressing algorithmic trading problems. However, the trading patterns differ among market conditions due to shifted distribution data. Ignoring multiple patterns in the…
▽ More
Algorithmic trading refers to executing buy and sell orders for specific assets based on automatically identified trading opportunities. Strategies based on reinforcement learning (RL) have demonstrated remarkable capabilities in addressing algorithmic trading problems. However, the trading patterns differ among market conditions due to shifted distribution data. Ignoring multiple patterns in the data will undermine the performance of RL. In this paper, we propose MOT,which designs multiple actors with disentangled representation learning to model the different patterns of the market. Furthermore, we incorporate the Optimal Transport (OT) algorithm to allocate samples to the appropriate actor by introducing a regularization loss term. Additionally, we propose Pretrain Module to facilitate imitation learning by aligning the outputs of actors with expert strategy and better balance the exploration and exploitation of RL. Experimental results on real futures market data demonstrate that MOT exhibits excellent profit capabilities while balancing risks. Ablation studies validate the effectiveness of the components of MOT.
△ Less
Submitted 2 June, 2024;
originally announced July 2024.
-
Functional Limit Theorems for Hawkes Processes
Authors:
Ulrich Horst,
Wei Xu
Abstract:
We prove that the long-run behavior of Hawkes processes is fully determined by the average number and the dispersion of child events. For subcritical processes we provide FLLNs and FCLTs under minimal conditions on the kernel of the process with the precise form of the limit theorems depending strongly on the dispersion of child events. For a critical Hawkes process with weakly dispersed child eve…
▽ More
We prove that the long-run behavior of Hawkes processes is fully determined by the average number and the dispersion of child events. For subcritical processes we provide FLLNs and FCLTs under minimal conditions on the kernel of the process with the precise form of the limit theorems depending strongly on the dispersion of child events. For a critical Hawkes process with weakly dispersed child events, functional central limit theorems do not hold. Instead, we prove that the rescaled intensity processes and rescaled Hawkes processes behave like CIR-processes without mean-reversion, respectively integrated CIR-processes. We provide the rate of convergence by establishing an upper bound on the Wasserstein distance between the distributions of rescaled Hawkes process and the corresponding limit process. By contrast, critical Hawkes process with heavily dispersed child events share many properties of subcritical ones. In particular, functional limit theorems hold. However, unlike subcritical processes critical ones with heavily dispersed child events display long-range dependencies.
△ Less
Submitted 29 December, 2024; v1 submitted 21 January, 2024;
originally announced January 2024.
-
Convergence of Heavy-Tailed Hawkes Processes and the Microstructure of Rough Volatility
Authors:
Ulrich Horst,
Wei Xu,
Rouyi Zhang
Abstract:
We establish the weak convergence of the intensity of a nearly-unstable Hawkes process with heavy-tailed kernel. Our result is used to derive a scaling limit for a financial market model where orders to buy or sell an asset arrive according to a Hawkes process with power-law kernel. After suitable rescaling the price-volatility process converges weakly to a rough Heston model. Our convergence resu…
▽ More
We establish the weak convergence of the intensity of a nearly-unstable Hawkes process with heavy-tailed kernel. Our result is used to derive a scaling limit for a financial market model where orders to buy or sell an asset arrive according to a Hawkes process with power-law kernel. After suitable rescaling the price-volatility process converges weakly to a rough Heston model. Our convergence result is much stronger than previously established ones that have either focused on light-tailed kernels or the convergence of integrated volatility process. The key is to establish the tightness of the family of rescaled volatility processes. This is achieved by introducing a new methods to establish the $C$-tightness of càdlàg processes based on the classical Kolmogorov-Chentsov tightness criterion for continuous processes.
△ Less
Submitted 10 November, 2024; v1 submitted 14 December, 2023;
originally announced December 2023.
-
DeepScalper: A Risk-Aware Reinforcement Learning Framework to Capture Fleeting Intraday Trading Opportunities
Authors:
Shuo Sun,
Wanqi Xue,
Rundong Wang,
Xu He,
Junlei Zhu,
Jian Li,
Bo An
Abstract:
Reinforcement learning (RL) techniques have shown great success in many challenging quantitative trading tasks, such as portfolio management and algorithmic trading. Especially, intraday trading is one of the most profitable and risky tasks because of the intraday behaviors of the financial market that reflect billions of rapidly fluctuating capitals. However, a vast majority of existing RL method…
▽ More
Reinforcement learning (RL) techniques have shown great success in many challenging quantitative trading tasks, such as portfolio management and algorithmic trading. Especially, intraday trading is one of the most profitable and risky tasks because of the intraday behaviors of the financial market that reflect billions of rapidly fluctuating capitals. However, a vast majority of existing RL methods focus on the relatively low frequency trading scenarios (e.g., day-level) and fail to capture the fleeting intraday investment opportunities due to two major challenges: 1) how to effectively train profitable RL agents for intraday investment decision-making, which involves high-dimensional fine-grained action space; 2) how to learn meaningful multi-modality market representation to understand the intraday behaviors of the financial market at tick-level. Motivated by the efficient workflow of professional human intraday traders, we propose DeepScalper, a deep reinforcement learning framework for intraday trading to tackle the above challenges. Specifically, DeepScalper includes four components: 1) a dueling Q-network with action branching to deal with the large action space of intraday trading for efficient RL optimization; 2) a novel reward function with a hindsight bonus to encourage RL agents making trading decisions with a long-term horizon of the entire trading day; 3) an encoder-decoder architecture to learn multi-modality temporal market embedding, which incorporates both macro-level and micro-level market information; 4) a risk-aware auxiliary task to maintain a striking balance between maximizing profit and minimizing risk. Through extensive experiments on real-world market data spanning over three years on six financial futures, we demonstrate that DeepScalper significantly outperforms many state-of-the-art baselines in terms of four financial criteria.
△ Less
Submitted 21 August, 2022; v1 submitted 15 December, 2021;
originally announced January 2022.
-
Risk and return prediction for pricing portfolios of non-performing consumer credit
Authors:
Siyi Wang,
Xing Yan,
Bangqi Zheng,
Hu Wang,
Wangli Xu,
Nanbo Peng,
Qi Wu
Abstract:
We design a system for risk-analyzing and pricing portfolios of non-performing consumer credit loans. The rapid development of credit lending business for consumers heightens the need for trading portfolios formed by overdue loans as a manner of risk transferring. However, the problem is nontrivial technically and related research is absent. We tackle the challenge by building a bottom-up architec…
▽ More
We design a system for risk-analyzing and pricing portfolios of non-performing consumer credit loans. The rapid development of credit lending business for consumers heightens the need for trading portfolios formed by overdue loans as a manner of risk transferring. However, the problem is nontrivial technically and related research is absent. We tackle the challenge by building a bottom-up architecture, in which we model the distribution of every single loan's repayment rate, followed by modeling the distribution of the portfolio's overall repayment rate. To address the technical issues encountered, we adopt the approaches of simultaneous quantile regression, R-copula, and Gaussian one-factor copula model. To our best knowledge, this is the first study that successfully adopts a bottom-up system for analyzing credit portfolio risks of consumer loans. We conduct experiments on a vast amount of data and prove that our methodology can be applied successfully in real business tasks.
△ Less
Submitted 28 October, 2021;
originally announced October 2021.
-
HIST: A Graph-based Framework for Stock Trend Forecasting via Mining Concept-Oriented Shared Information
Authors:
Wentao Xu,
Weiqing Liu,
Lewen Wang,
Yingce Xia,
Jiang Bian,
Jian Yin,
Tie-Yan Liu
Abstract:
Stock trend forecasting, which forecasts stock prices' future trends, plays an essential role in investment. The stocks in a market can share information so that their stock prices are highly correlated. Several methods were recently proposed to mine the shared information through stock concepts (e.g., technology, Internet Retail) extracted from the Web to improve the forecasting results. However,…
▽ More
Stock trend forecasting, which forecasts stock prices' future trends, plays an essential role in investment. The stocks in a market can share information so that their stock prices are highly correlated. Several methods were recently proposed to mine the shared information through stock concepts (e.g., technology, Internet Retail) extracted from the Web to improve the forecasting results. However, previous work assumes the connections between stocks and concepts are stationary, and neglects the dynamic relevance between stocks and concepts, limiting the forecasting results. Moreover, existing methods overlook the invaluable shared information carried by hidden concepts, which measure stocks' commonness beyond the manually defined stock concepts. To overcome the shortcomings of previous work, we proposed a novel stock trend forecasting framework that can adequately mine the concept-oriented shared information from predefined concepts and hidden concepts. The proposed framework simultaneously utilize the stock's shared information and individual information to improve the stock trend forecasting performance. Experimental results on the real-world tasks demonstrate the efficiency of our framework on stock trend forecasting. The investment simulation shows that our framework can achieve a higher investment return than the baselines.
△ Less
Submitted 20 January, 2022; v1 submitted 26 October, 2021;
originally announced October 2021.
-
REST: Relational Event-driven Stock Trend Forecasting
Authors:
Wentao Xu,
Weiqing Liu,
Chang Xu,
Jiang Bian,
Jian Yin,
Tie-Yan Liu
Abstract:
Stock trend forecasting, aiming at predicting the stock future trends, is crucial for investors to seek maximized profits from the stock market. Many event-driven methods utilized the events extracted from news, social media, and discussion board to forecast the stock trend in recent years. However, existing event-driven methods have two main shortcomings: 1) overlooking the influence of event inf…
▽ More
Stock trend forecasting, aiming at predicting the stock future trends, is crucial for investors to seek maximized profits from the stock market. Many event-driven methods utilized the events extracted from news, social media, and discussion board to forecast the stock trend in recent years. However, existing event-driven methods have two main shortcomings: 1) overlooking the influence of event information differentiated by the stock-dependent properties; 2) neglecting the effect of event information from other related stocks. In this paper, we propose a relational event-driven stock trend forecasting (REST) framework, which can address the shortcoming of existing methods. To remedy the first shortcoming, we propose to model the stock context and learn the effect of event information on the stocks under different contexts. To address the second shortcoming, we construct a stock graph and design a new propagation layer to propagate the effect of event information from related stocks. The experimental studies on the real-world data demonstrate the efficiency of our REST framework. The results of investment simulation show that our framework can achieve a higher return of investment than baselines.
△ Less
Submitted 19 February, 2021; v1 submitted 15 February, 2021;
originally announced February 2021.
-
Every Corporation Owns Its Image: Corporate Credit Ratings via Convolutional Neural Networks
Authors:
Bojing Feng,
Wenfang Xue,
Bindang Xue,
Zeyu Liu
Abstract:
Credit rating is an analysis of the credit risks associated with a corporation, which reflect the level of the riskiness and reliability in investing. There have emerged many studies that implement machine learning techniques to deal with corporate credit rating. However, the ability of these models is limited by enormous amounts of data from financial statement reports. In this work, we analyze t…
▽ More
Credit rating is an analysis of the credit risks associated with a corporation, which reflect the level of the riskiness and reliability in investing. There have emerged many studies that implement machine learning techniques to deal with corporate credit rating. However, the ability of these models is limited by enormous amounts of data from financial statement reports. In this work, we analyze the performance of traditional machine learning models in predicting corporate credit rating. For utilizing the powerful convolutional neural networks and enormous financial data, we propose a novel end-to-end method, Corporate Credit Ratings via Convolutional Neural Networks, CCR-CNN for brevity. In the proposed model, each corporation is transformed into an image. Based on this image, CNN can capture complex feature interactions of data, which are difficult to be revealed by previous machine learning models. Extensive experiments conducted on the Chinese public-listed corporate rating dataset which we build, prove that CCR-CNN outperforms the state-of-the-art methods consistently.
△ Less
Submitted 2 December, 2020;
originally announced December 2020.
-
Evolutionary dynamics in financial markets with heterogeneities in strategies and risk tolerance
Authors:
Wen-Juan Xu,
Chen-Yang Zhong,
Fei Ren,
Tian Qiu,
Rong-Da Chen,
Yun-Xin He,
Li-Xin Zhong
Abstract:
In nature and human societies, the effects of homogeneous and heterogeneous characteristics on the evolution of collective behaviors are quite different from each other. It is of great importance to understand the underlying mechanisms of the occurrence of such differences. By incorporating pair pattern strategies and reference point strategies into an agent-based model, we have investigated the c…
▽ More
In nature and human societies, the effects of homogeneous and heterogeneous characteristics on the evolution of collective behaviors are quite different from each other. It is of great importance to understand the underlying mechanisms of the occurrence of such differences. By incorporating pair pattern strategies and reference point strategies into an agent-based model, we have investigated the coupled effects of heterogeneous investment strategies and heterogeneous risk tolerance on price fluctuations. In the market flooded with the investors with homogeneous investment strategies or homogeneous risk tolerance, large price fluctuations are easy to occur. In the market flooded with the investors with heterogeneous investment strategies or heterogeneous risk tolerance, the price fluctuations are suppressed. For a heterogeneous population, the coexistence of investors with pair pattern strategies and reference point strategies causes the price to have a slow fluctuation around a typical equilibrium point and both a large price fluctuation and a no-trading state are avoided, in which the pair pattern strategies push the system far away from the equilibrium while the reference point strategies pull the system back to the equilibrium. A theoretical analysis indicates that the evolutionary dynamics in the present model is governed by the competition between different strategies. The strategy that causes large price fluctuations loses more while the strategy that pulls the system back to the equilibrium gains more. Overfrequent trading does harm to one's pursuit for more wealth.
△ Less
Submitted 18 October, 2020;
originally announced October 2020.
-
The Microstructure of Stochastic Volatility Models with Self-Exciting Jump Dynamics
Authors:
Ulrich Horst,
Wei Xu
Abstract:
We provide a general probabilistic framework within which we establish scaling limits for a class of continuous-time stochastic volatility models with self-exciting jump dynamics. In the scaling limit, the joint dynamics of asset returns and volatility is driven by independent Gaussian white noises and two independent Poisson random measures that capture the arrival of exogenous shocks and the arr…
▽ More
We provide a general probabilistic framework within which we establish scaling limits for a class of continuous-time stochastic volatility models with self-exciting jump dynamics. In the scaling limit, the joint dynamics of asset returns and volatility is driven by independent Gaussian white noises and two independent Poisson random measures that capture the arrival of exogenous shocks and the arrival of self-excited shocks, respectively. Various well-studied stochastic volatility models with and without self-exciting price/volatility co-jumps are obtained as special cases under different scaling regimes. We analyze the impact of external shocks on the market dynamics, especially their impact on jump cascades and show in a mathematically rigorous manner that many small external shocks may tigger endogenous jump cascades in asset returns and stock price volatility.
△ Less
Submitted 29 November, 2019;
originally announced November 2019.
-
A Scaling Limit for Limit Order Books Driven by Hawkes Processes
Authors:
Ulrich Horst,
Wei Xu
Abstract:
In this paper we derive a scaling limit for an infinite dimensional limit order book model driven by Hawkes random measures. The dynamics of the incoming order flow is allowed to depend on the current market price as well as on a volume indicator. With our choice of scaling the dynamics converges to a coupled SDE-ODE system where limiting best bid and ask price processes follows a diffusion dynami…
▽ More
In this paper we derive a scaling limit for an infinite dimensional limit order book model driven by Hawkes random measures. The dynamics of the incoming order flow is allowed to depend on the current market price as well as on a volume indicator. With our choice of scaling the dynamics converges to a coupled SDE-ODE system where limiting best bid and ask price processes follows a diffusion dynamics, the limiting volume density functions follows an ODE in a Hilbert space and the limiting order arrival and cancellation intensities follow a Volterra-Fredholm integral equation.
△ Less
Submitted 8 August, 2018; v1 submitted 5 September, 2017;
originally announced September 2017.
-
A generalized public goods game with coupling of individual ability and project benefit
Authors:
Li-Xin Zhong,
Wen-Juan Xu,
Yun-Xin He,
Chen-Yang Zhong,
Rong-Da Chen,
Tian Qiu,
Yong-Dong Shi,
Fei Ren
Abstract:
Facing a heavy task, any single person can only make a limited contribution and team cooperation is needed. As one enjoys the benefit of the public goods, the potential benefits of the project are not always maximized and may be partly wasted. By incorporating individual ability and project benefit into the original public goods game, we study the coupling effect of the four parameters, the upper…
▽ More
Facing a heavy task, any single person can only make a limited contribution and team cooperation is needed. As one enjoys the benefit of the public goods, the potential benefits of the project are not always maximized and may be partly wasted. By incorporating individual ability and project benefit into the original public goods game, we study the coupling effect of the four parameters, the upper limit of individual contribution, the upper limit of individual benefit, the needed project cost and the upper limit of project benefit on the evolution of cooperation. Coevolving with the individual-level group size preferences, an increase in the upper limit of individual benefit promotes cooperation while an increase in the upper limit of individual contribution inhibits cooperation. The coupling of the upper limit of individual contribution and the needed project cost determines the critical point of the upper limit of project benefit, where the equilibrium frequency of cooperators reaches its highest level. Above the critical point, an increase in the upper limit of project benefit inhibits cooperation. The evolution of cooperation is closely related to the preferred group-size distribution. A functional relation between the frequency of cooperators and the dominant group size is found.
△ Less
Submitted 21 May, 2017; v1 submitted 23 February, 2017;
originally announced February 2017.
-
Self-organization and phase transition in financial markets with multiple choices
Authors:
Li-Xin Zhong,
Wen-Juan Xu,
Ping Huang,
Chen-Yang Zhong,
Tian Qiu
Abstract:
Market confidence is essential for successful investing. By incorporating multi-market into the evolutionary minority game, we investigate the effects of investor beliefs on the evolution of collective behaviors and asset prices. When there exists another investment opportunity, market confidence, including overconfidence and under-confidence, is not always good or bad for investment. The roles of…
▽ More
Market confidence is essential for successful investing. By incorporating multi-market into the evolutionary minority game, we investigate the effects of investor beliefs on the evolution of collective behaviors and asset prices. When there exists another investment opportunity, market confidence, including overconfidence and under-confidence, is not always good or bad for investment. The roles of market confidence is closely related to market impact. For low market impact, overconfidence in a particular asset makes an investor become insensitive to losses and a delayed strategy adjustment leads to a decline in wealth, and thereafter, one's runaway from the market. For high market impact, under-confidence in a particular asset makes an investor over-sensitive to losses and one's too frequent strategy adjustment leads to a large fluctuation in asset prices, and thereafter, a decrease in the number of agents. At an intermediate market impact, the phase transition occurs. No matter what the market impact is, an equilibrium between different markets exists, which is reflected in the occurrence of similar price fluctuations in different markets. A theoretical analysis indicates that such an equilibrium results from the coupled effects of strategy updating and shift in investment. The runaway of the agents trading a specific asset will lead to a decline in the asset price volatility and such a decline will be inhibited by the clustering of the strategies. A uniform strategy distribution will lead to a large fluctuation in asset prices and such a fluctuation will be suppressed by the decrease in the number of agents in the market. A functional relationship between the price fluctuations and the numbers of agents is found.
△ Less
Submitted 1 June, 2014; v1 submitted 2 December, 2013;
originally announced December 2013.
-
Coupled effects of market impact and asymmetric sensitivity in financial markets
Authors:
Li-Xin Zhong,
Wen-Juan Xu,
Fei Ren,
Yong-Dong Shi
Abstract:
By incorporating market impact and asymmetric sensitivity into the evolutionary minority game, we study the coevolutionary dynamics of stock prices and investment strategies in financial markets. Both the stock price movement and the investors' global behavior are found to be closely related to the phase region they fall into. Within the region where the market impact is small, investors' asymmetr…
▽ More
By incorporating market impact and asymmetric sensitivity into the evolutionary minority game, we study the coevolutionary dynamics of stock prices and investment strategies in financial markets. Both the stock price movement and the investors' global behavior are found to be closely related to the phase region they fall into. Within the region where the market impact is small, investors' asymmetric response to gains and losses leads to the occurrence of herd behavior, when all the investors are prone to behave similarly in an extreme way and large price fluctuations occur. A linear relation between the standard deviation of stock price changes and the mean value of strategies is found. With full market impact, the investors tend to self-segregate into opposing groups and the introduction of asymmetric sensitivity leads to the disappearance of dominant strategies. Compared with the situations in the stock market with little market impact, the stock price fluctuations are suppressed and an efficient market occurs. Theoretical analyses indicate that the mechanism of phase transition from clustering to self-segregation in the present model is similar to that in the majority-minority game and the occurrence and disappearance of efficient markets are related to the competition between the trend-following and the trend-aversion forces. The clustering of the strategies in the present model results from the majority-wins effect and the wealth-driven mechanism makes the market become predictable.
△ Less
Submitted 13 January, 2013; v1 submitted 15 September, 2012;
originally announced September 2012.
