Enhancing portfolio optimization with multi-LLM sentiment aggregation: A Black-Litterman integration approach

The rise of robo-advisor wealth management services, which constitute a key element of fintech revolution, unveils the question whether they can dominate human-based advice, namely how to address the client’s behavioral biases in an automated way. One approach to it would be the application of machine learning tools during client profiling. However, trained neural network is often considered as a black box, which may raise concerns from the customers and regulators in terms of model validity, transparency, and related risks. In order to address these issues and shed more light on how neurons work, especially to figure out how they perform computation at intermediate layers, this paper visualizes and estimates the neurons’ sensitivity to different input parameters. Before it, the comprehensive review of the most popular optimization algorithms is presented and based on them respective data set is generated to train convolutional neural network. It was found that selected hidden units to some extent are not only specializing in the reaction to such features as, for example, risk, return or risk-aversion level but also they are learning more complex concepts like Sharpe ratio. These findings should help to understand robo-advisor mechanics deeper, which finally will provide more room to improve and significantly innovate the automated wealth management process and make it more transparent.

This paper focuses on classical portfolio strategies applied to five countries, which are Brazil, Russia, India, China and South Africa. These five countries form the so-called BRICS group. In particular, the authors investigate their corporate and sovereign bond market and evaluate whether these markets can represent a profitable investment for non-satiable and risk-averse investors. Two-step optimization is proposed to control price risk and market risk. For price risk management, classical immunization strategies and are obtained funds of bond are obtained that share the same risk measure. For market risk control, the previously found funds are used and a performance measure optimization commonly used in stock markets is applied to define the best portfolio of funds. Therefore, the resulting optimal portfolio controls the price risk and jointly maximizes a desired performance measure that includes the market risk. Finally, the authors propose an empirical analysis to evaluate the profitability of the suggested two-step optimization for the five BRICS countries and compare the ex-post sample paths of the obtained portfolios for testing the stochastic dominance relations.

Several contributions in the literature argue that a significant in-sample risk reduction can be obtained by investing in a relatively small number of assets in an investment universe. Furthermore, selecting small portfolios seems to yield good out-of-sample performances in practice. This analysis provides further evidence that an appropriate preselection of the assets in a market can lead to an improvement in portfolio performance. For preselection, this paper investigates the effectiveness of a minimum variance approach and that of an innovative index (the new Altman Z-score) based on the creditworthiness of the companies. Different classes of portfolio models are examined on real-world data by applying both the minimum variance and the Z-score preselection methods. Preliminary results indicate that the new Altman Z-score preselection provides encouraging out-of-sample performances with respect to those obtained with the minimum variance approach.