Cross-sectional Learning of Extremal Dependence among Financial Assets

Authors: Xing Yan, Qi Wu, Wen Zhang

NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We show its flexible tail dependence structure through simulation. Coupled with a GARCH model to eliminate serial dependence of each individual asset return series, we use this novel method to model and forecast multivariate conditional distribution of stock returns, and obtain notable performance improvements in multi-dimensional coverage tests.
Researcher Affiliation Collaboration Xing Yan School of Data Science City University of Hong Kong yanxing128@gmail.com Qi Wu School of Data Science City University of Hong Kong qiwu55@cityu.edu.hk Wen Zhang JD Digits zhangwen.jd@gmail.com
Pseudocode Yes Algorithm 1 Algorithm for learning parameters of our proposed tail dependence model with data.
Open Source Code No The paper does not provide any information about the availability of open-source code for the described methodology.
Open Datasets No The paper mentions using stock return data (Apple, IBM, Microsoft, JP Morgan, Walmart, Dow-Jones stocks, SP500) and specifies the date ranges and number of observations for each. However, it does not provide concrete access information (e.g., links, DOIs, specific repositories, or formal citations to publicly accessible versions) for the exact datasets used.
Dataset Splits No The paper specifies a 'testing set' but does not explicitly mention the use of a separate 'validation' set or its split details.
Hardware Specification No The paper does not provide specific details about the hardware (e.g., GPU/CPU models, memory) used to run the experiments.
Software Dependencies No The paper mentions various software components and models like 'GARCH model', 'AR(1)-GARCH(1,1)', 'quantile regression', 't-distribution innovation', 'multivariate normal', 'multivariate t', 'Clayton copula', and 'Gumbel copula'. However, it does not specify any version numbers for these software dependencies, which is necessary for reproducibility.
Experiment Setup Yes A is a positive constant satisfying A 3 (see [Wu and Yan, 2019]). We set A = 4 in this paper. Now we have totally n location parameters µ1, . . . , µn, (n2+n)/2 usual correlation/scale parameters σ11, σ21, σ22, . . . , σnn, (n2+n)/2 right-tail parameters u11, u21, u22, . . . , unn, and (n2 + n)/2 left-tail parameters v11, v21, v22, . . . , vnn. The total number of parameters is n + 3(n2 + n)/2. ... we set Ψ = {0.01, 0.02, . . . , 0.98, 0.99} with 99 probability levels. Other smaller set that covers the interval (0, 1) sufficiently is also acceptable, e.g., {0.01, 0.05, 0.1, . . . , 0.9, 0.95, 0.99} with 21 levels.