Black–Litterman portfolio optimization based on GARCH–EVT–Copula and LSTM models
The paper integrates investors' opinions into portfolio optimization utilizing Black-Litterman, GARCH-EVT Copula, and LSTM deep models.
What it examines
This paper integrates investor views into portfolio optimization using the Black-Litterman framework. It employs a deep learning LSTM model to estimate views, while GARCH-EVT-Copula models capture asset return dependencies. The study aims to enhance portfolio performance compared to traditional max-Sharpe and standard Black-Litterman approaches.
What it concludes
Results indicate that the Black-Litterman approach with GARCH-EVT-Copula and LSTM outperforms traditional models. The research offers potential applications in improved asset allocation, risk management, and dynamic portfolio adjustment, which can guide better investment decisions and motivate further study in advanced financial risk analysis.
Evidence objects
The study integrates the Black-Litterman model with advanced techniques such as GARCH, EVT, Copula models, and LSTM learning, offering a mix of statistical rigor and artificial intelligence for portfolio optimization.
key_findings bullet 1 · key_findings · validation V0
Empirical results demonstrate that portfolios constructed via these hybrid methods outperform traditional max-Sharpe and original Black-Litterman portfolios, with deep learning estimating investor views and GARCH-EVT-Copula capturing stock return dependencies effectively.
key_findings bullet 2 · key_findings · validation V0
The authors merge statistical analysis with machine learning, incorporating investor sentiment via deep learning while validating the approach through historical data, simulations, and back-testing, though noting computational challenges for scalability.
key_findings bullet 3 · key_findings · validation V0
Integrating advanced risk models ($\text{GARCH}$, $\text{EVT}$, $\text{Copula}$) with deep learning ($\text{LSTM}$), this paper revitalizes the $\text{Black--Litterman}$ framework. Its innovative, multidisciplinary approach fuses traditional econometrics with modern AI, delivering novel investor views for portfolio optimization and market prediction, making it a compelling, ever-relevant read for finance and quantitative research enthusiasts remarkably.
key_findings bullet 4 · key_findings · validation V0
Raw abstract and provenance
- … structure of financial returns, which is a critical aspect of financial risk modeling and … For the details of rigorous mathematical introduction and treatments, we refer to …
Source row: 325 · abstract type: snippet