-
Interpretability in deep learning for finance: a case study for the Heston model
Authors:
Damiano Brigo,
Xiaoshan Huang,
Andrea Pallavicini,
Haitz Saez de Ocariz Borde
Abstract:
Deep learning is a powerful tool whose applications in quantitative finance are growing every day. Yet, artificial neural networks behave as black boxes and this hinders validation and accountability processes. Being able to interpret the inner functioning and the input-output relationship of these networks has become key for the acceptance of such tools. In this paper we focus on the calibration…
▽ More
Deep learning is a powerful tool whose applications in quantitative finance are growing every day. Yet, artificial neural networks behave as black boxes and this hinders validation and accountability processes. Being able to interpret the inner functioning and the input-output relationship of these networks has become key for the acceptance of such tools. In this paper we focus on the calibration process of a stochastic volatility model, a subject recently tackled by deep learning algorithms. We analyze the Heston model in particular, as this model's properties are well known, resulting in an ideal benchmark case. We investigate the capability of local strategies and global strategies coming from cooperative game theory to explain the trained neural networks, and we find that global strategies such as Shapley values can be effectively used in practice. Our analysis also highlights that Shapley values may help choose the network architecture, as we find that fully-connected neural networks perform better than convolutional neural networks in predicting and interpreting the Heston model prices to parameters relationship.
△ Less
Submitted 19 April, 2021;
originally announced April 2021.
-
Scenario-decomposition Solution Framework for Nonseparable Stochastic Control Problems
Authors:
Xin Huang,
Duan Li,
Daniel Zhuoyu Long
Abstract:
When stochastic control problems do not possess separability and/or monotonicity, the dynamic programming pioneered by Bellman in 1950s fails to work as a time-decomposition solution method. Such cases have posted a great challenge to the control society in both theoretical foundation and solution methodologies for many years. With the help of the progressive hedging algorithm proposed by Rockafel…
▽ More
When stochastic control problems do not possess separability and/or monotonicity, the dynamic programming pioneered by Bellman in 1950s fails to work as a time-decomposition solution method. Such cases have posted a great challenge to the control society in both theoretical foundation and solution methodologies for many years. With the help of the progressive hedging algorithm proposed by Rockafellar and Wets in 1991, we develop a novel scenario-decomposition solution framework for stochastic control problems which could be nonseparable and/or non-monotonic, thus extending the reach of stochastic optimal control. We discuss then some of its promising applications, including online quadratic programming problems and dynamic portfolio selection problems with smoothing properties.
△ Less
Submitted 18 October, 2020;
originally announced October 2020.
-
BERT-based Financial Sentiment Index and LSTM-based Stock Return Predictability
Authors:
Joshua Zoen Git Hiew,
Xin Huang,
Hao Mou,
Duan Li,
Qi Wu,
Yabo Xu
Abstract:
Traditional sentiment construction in finance relies heavily on the dictionary-based approach, with a few exceptions using simple machine learning techniques such as Naive Bayes classifier. While the current literature has not yet invoked the rapid advancement in the natural language processing, we construct in this research a textual-based sentiment index using a well-known pre-trained model BERT…
▽ More
Traditional sentiment construction in finance relies heavily on the dictionary-based approach, with a few exceptions using simple machine learning techniques such as Naive Bayes classifier. While the current literature has not yet invoked the rapid advancement in the natural language processing, we construct in this research a textual-based sentiment index using a well-known pre-trained model BERT developed by Google, especially for three actively trading individual stocks in Hong Kong market with at the same time the hot discussion on Weibo.com. On the one hand, we demonstrate a significant enhancement of applying BERT in financial sentiment analysis when compared with the existing models. On the other hand, by combining with the other two commonly-used methods when it comes to building the sentiment index in the financial literature, i.e., the option-implied and the market-implied approaches, we propose a more general and comprehensive framework for the financial sentiment analysis, and further provide convincing outcomes for the predictability of individual stock return by combining LSTM (with a feature of a nonlinear map**). It is significantly distinct with the dominating econometric methods in sentiment influence analysis which are all of a nature of linear regression.
△ Less
Submitted 7 July, 2022; v1 submitted 21 June, 2019;
originally announced June 2019.
-
Enhancing countries' fitness with recommender systems on the international trade network
Authors:
Hao Liao,
Xiao-Min Huang,
Xing-Tong Wu,
Ming-Kai Liu,
Alexandre Vidmer,
Mingyang Zhou,
Yi-Cheng Zhang
Abstract:
Prediction is one of the major challenges in complex systems. The prediction methods have shown to be effective predictors of the evolution of networks. These methods can help policy makers to solve practical problems successfully and make better strategy for the future. In this work, we focus on exporting countries' data of the international trading network. A recommendation system is then used t…
▽ More
Prediction is one of the major challenges in complex systems. The prediction methods have shown to be effective predictors of the evolution of networks. These methods can help policy makers to solve practical problems successfully and make better strategy for the future. In this work, we focus on exporting countries' data of the international trading network. A recommendation system is then used to identify the products corresponding to the production capacity of each individual country, but are somehow overlook by the country. Then, we simulate the evolution of the country's fitness if it would have followed the recommendations. The result of this work is the combination combine these two methods to provide insights to countries on how to enhance the diversification of their exported products in a scientific way and improve national competitiveness significantly, especially for develo** countries.
△ Less
Submitted 4 April, 2019;
originally announced April 2019.
-
Systemic risk in dynamical networks with stochastic failure criterion
Authors:
B. Podobnik,
D. Horvatic,
M. Bertella,
L. Feng,
X. Huang,
B. Li
Abstract:
Complex non-linear interactions between banks and assets we model by two time-dependent Erdős Renyi network models where each node, representing bank, can invest either to a single asset (model I) or multiple assets (model II). We use dynamical network approach to evaluate the collective financial failure---systemic risk---quantified by the fraction of active nodes. The systemic risk can be calcul…
▽ More
Complex non-linear interactions between banks and assets we model by two time-dependent Erdős Renyi network models where each node, representing bank, can invest either to a single asset (model I) or multiple assets (model II). We use dynamical network approach to evaluate the collective financial failure---systemic risk---quantified by the fraction of active nodes. The systemic risk can be calculated over any future time period, divided on sub-periods, where within each sub-period banks may contiguously fail due to links to either (i) assets or (ii) other banks, controlled by two parameters, probability of internal failure $p$ and threshold $T_h$ ("solvency" parameter). The systemic risk non-linearly increases with $p$ and decreases with average network degree faster when all assets are equally distributed across banks than if assets are randomly distributed. The more inactive banks each bank can sustain (smaller $T_h$), the smaller the systemic risk---for some $T_h$ values in I we report a discontinuity in systemic risk. When contiguous spreading becomes stochastic (ii) controlled by probability $p_2$---a condition for the bank to be solvent (active) is stochastic---the systemic risk decreases with decreasing $p_2$. We analyse asset allocation for the U.S. banks.
△ Less
Submitted 10 April, 2014; v1 submitted 22 March, 2014;
originally announced March 2014.
-
Partial correlation analysis: Applications for financial markets
Authors:
Dror Y. Kenett,
Xuqing Huang,
Irena Vodenska,
Shlomo Havlin,
H. Eugene Stanley
Abstract:
The presence of significant cross-correlations between the synchronous time evolution of a pair of equity returns is a well-known empirical fact. The Pearson correlation is commonly used to indicate the level of similarity in the price changes for a given pair of stocks, but it does not measure whether other stocks influence the relationship between them. To explore the influence of a third stock…
▽ More
The presence of significant cross-correlations between the synchronous time evolution of a pair of equity returns is a well-known empirical fact. The Pearson correlation is commonly used to indicate the level of similarity in the price changes for a given pair of stocks, but it does not measure whether other stocks influence the relationship between them. To explore the influence of a third stock on the relationship between two stocks, we use a partial correlation measurement to determine the underlying relationships between financial assets. Building on previous work, we present a statistically robust approach to extract the underlying relationships between stocks from four different financial markets: the United States, the United Kingdom, Japan, and India. This methodology provides new insights into financial market dynamics and uncovers implicit influences in play between stocks. To demonstrate the capabilities of this methodology, we (i) quantify the influence of different companies and, by studying market similarity across time, present new insights into market structure and market stability, and (ii) we present a practical application, which provides information on the how a company is influenced by different economic sectors, and how the sectors interact with each other. These examples demonstrate the effectiveness of this methodology in uncovering information valuable for a range of individuals, including not only investors and traders but also regulators and policy makers.
△ Less
Submitted 6 February, 2014;
originally announced February 2014.
-
Cascading Failures in Bi-partite Graphs: Model for Systemic Risk Propagation
Authors:
Xuqing Huang,
Irena Vodenska,
Shlomo Havlin,
H. Eugene Stanley
Abstract:
As economic entities become increasingly interconnected, a shock in a financial network can provoke significant cascading failures throughout the system. To study the systemic risk of financial systems, we create a bi-partite banking network model composed of banks and bank assets and propose a cascading failure model to describe the risk propagation process during crises. We empirically test the…
▽ More
As economic entities become increasingly interconnected, a shock in a financial network can provoke significant cascading failures throughout the system. To study the systemic risk of financial systems, we create a bi-partite banking network model composed of banks and bank assets and propose a cascading failure model to describe the risk propagation process during crises. We empirically test the model with 2007 US commercial banks balance sheet data and compare the model prediction of the failed banks with the real failed banks after 2007. We find that our model efficiently identifies a significant portion of the actual failed banks reported by Federal Deposit Insurance Corporation. The results suggest that this model could be useful for systemic risk stress testing for financial systems. The model also identifies that commercial rather than residential real estate assets are major culprits for the failure of over 350 US commercial banks during 2008-2011.
△ Less
Submitted 11 January, 2013; v1 submitted 17 October, 2012;
originally announced October 2012.
