Title 1 : Power of deep learning: Quantifying language to explain cross-sectional returns
Speaker: Weijun Xu ( Research Fellow of South China University of Technology )
Title 2 : Consistency of pairwise comparison matrices
Speaker: Prof. Fang Liu ( Guangxi University )
Title 3: Competitive online Portfolio Strategy Research based on Weak Integration Algorithm
Speaker: Prof. Yong Zhang ( Guangdong University of Technology )
Title 4: Competitive online Portfolio Strategy Research based on Weak Integration Algorithm
Speaker: Prof. Yong Ma ( Hunan University )
Title 5: Volatility-managed portfolio in Chinese market，Attention enhanced long short-term memory network with multi-source heterogeneous information fusion: An application to BGI Genomics
Speaker: Prof. Xili Zhang ( Zhejiang University )
Prof. Qun Zhang ( Guangdong University of Foreign Studies )
Time: Sat, Dec.5 2020, PM:14:30-18:00
Location: Tencent Conference
Meeting Number: 899 226 356
Password: 202012
Inviter: Dr. Li Gao
Abstract1:
When quantifying qualitative information from unstructured textual data, the traditional bag-of-words approach only captures semantic features of single words or phrases. The context, the sequence of words, and the relationship between words are ignored. This paper introduces deep neural networks (NNs) to encode and mimic human intelligence in processing natural language. Using the NN-based artificial intelligence, we construct a new measure of sentiment that is specific to performancediscussions and is adjusted for complex contextual negations. We find that this performance-specific sentiment explains cross-sectional returns and future operating performance better than the umbrella sentiment proxies used in the literature.
Abstract2:
The axiomatic properties are proposed to characterize the consistency of pairwise comparison matrices (PCMs).. The novel consistency indexes are given to quantify the inconsistency degree of PCMs. It is found that uncertain preference relations are inconsistent in nature. The concept of approxiamte consistency of interval-valued comparison matrices is proposed. The consistency index of interval-valued comparison matrices is constructed.
Abstract4:
This paper presents a valuation of VIX options employing a Hawkes jump-diffusion model that captures the clustering pattern of jumps observed extensively in the financial markets. In the consistent framework, the valuation problem of VIX options is solved efficiently via the Fourier cosine expansion (COS) method. The Monte Carlo (MC) simulations are carried out to demonstrate the reliability and efficiency of the COS method. Furthermore, a sensitivity analysis is performed to show how option prices response to different parameters associated with jump clustering. Finally, empirical studies are conducted to provide evidence to support our jump specification in matching the VIX option surface.
