Ranking Metrics: Extending Acceptability and Performance Indices

Battle Count: 55

The paper is highly relevant to quantitative trading in terms of portfolio performance evaluation, ranking of investment alternatives, and portfolio optimization. The introduction of bibliometric-based ranking metrics offers novel perspectives on portfolio construction beyond traditional mean-variance or risk-adjusted ratios. The Λ-quantile metrics provide tail-performance assessment useful for strategy evaluation. However, the paper is more theoretical/axiomatic than directly implementable as a trading strategy, and focuses on evaluation rather than signal generation.

The theoretical framework requires understanding of functional analysis, convex analysis, and risk measure theory. The empirical implementation involves: (1) computing traditional metrics (RAROC, GLR, Omega) which are straightforward; (2) implementing Λ-quantile ranking metrics requiring quantile estimation; (3) constructing bibliometric indices requiring sorting, rescaling, and curve comparison; (4) portfolio optimization via genetic algorithms in MATLAB. The mathematical proofs are complex but the computational implementation is moderate.

The paper provides detailed mathematical formulations and uses publicly available data (European Environment Agency) and Bloomberg data (NASDAQ, S&P 500). Portfolio optimization uses MATLAB's Global Optimization Toolbox with genetic algorithms. However, specific code is not provided, and Bloomberg data requires a subscription. The theoretical framework is fully self-contained with proofs.