Paper: arXiv 2410.01732

Authors: Jun Cai, Zhanyi Jiao, Tiantian Mao

Abstract

The expected regret and target semi-variance are two of the most important risk measures for downside risk. When the distribution of a loss is uncertain, and only partial information of the loss is known, their worst-case values play important roles in robust risk management for finance, insurance, and many other fields. Jagannathan (1977) derived the worst-case expected regrets when only the mean and variance of a loss are known and the loss is arbitrary, symmetric, or non-negative. While Chen et al. (2011) obtained the worst-case target semi-variances under similar conditions but focusing on arbitrary losses. In this paper, we first complement the study of Chen et al. (2011) on the worst-case target semi-variances and derive the closed-form expressions for the worst-case target semi-variance when only the mean and variance of a loss are known and the loss is symmetric or non-negative. Then, we investigate worst-case target semi-variances over uncertainty sets that represent undesirable scenarios faced by an investors. Our methods for deriving these worst-case values are different from those used in Jagannathan (1977) and Chen et al. (2011). As applications of the results derived in this paper, we propose robust portfolio selection methods that minimize the worst-case target semi-variance of a portfolio loss over different uncertainty sets. To explore the insights of our robust portfolio selection methods, we conduct numerical experiments with real financial data and compare our portfolio selection methods with several existing portfolio selection models related to the models proposed in this paper.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • Empirical Rigor: 6.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper presents highly advanced mathematical derivations for worst-case target semi-variances under distributional uncertainty, requiring sophisticated optimization and probability theory. It also includes numerical experiments with real financial data, demonstrating implementation relevance, though the empirical focus is on methodological validation rather than live backtesting.

Research Flowchart

  flowchart TD
  A["Research Goal<br>Derive worst-case target semi-variance expressions<br>for symmetric & non-negative losses"] --> B["Methodology<br>Mathematical derivations & robust optimization"]
  B --> C["Inputs<br>Mean & variance constraints<br>Uncertainty sets for distributions"]
  C --> D["Computation<br>Closed-form solutions &<br>Worst-case portfolio optimization"]
  D --> E{"Outcomes"}
  E --> F["Closed-form expressions<br>for worst-case target semi-variances"]
  E --> G["Robust portfolio selection models<br>minimizing worst-case risk"]
  E --> H["Numerical experiments with real data<br>Comparative performance analysis"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes robust portfolio selection methods that minimize the worst-case target semi-variance of a portfolio loss over different uncertainty sets.

Source sentence
As applications of the results derived in this paper, we propose robust portfolio selection methods that minimize the worst-case target semi-variance of a portfolio loss over different uncertainty sets.

Assumes 6

  • Only the mean and variance of a loss are known. statistical
    Source sentence
    when only the mean and variance of a loss are known and the loss is symmetric or non-negative
  • The loss distribution is symmetric or non‑negative. statistical
    Source sentence
    derive the closed-form expressions for the worst-case target semi-variance when only the mean and variance of a loss are known and the loss is symmetric or non-negative
  • The investor’s total initial wealth to be invested is 1 (w⊤e = 1). model
    Source sentence
    Without loss of generality, we assume that the investor’s total initial wealth to be invested is 1, which means that w⊤e = 1
  • Short selling may be allowed (set W) or prohibited (set W+). market
    Source sentence
    we use the set of portfolios W = {w ∈Rd : w⊤e = 1} to denote that an investor may short sell a stock, and use the set of portfolios W+ = {w ∈Rd : w⊤e = 1, w ⩾0} to indicate that an investor does not short sell any stock
Show 2 more
  • Mean vector µ and covariance matrix Σ are known (mean‑variance uncertainty set). statistical
    Source sentence
    In the classical mean-variance (M‑V) model, the mean vector µ and covariance matrix Σ of the loss vector X are assumed to be known
  • µ and e are not linearly dependent (µ ≠ c e for any c). model
    Source sentence
    we assume that µ and e are not linearly dependent, or equivalently, assume that for any c ∈R, µ̸ = c e

Data

Backtest sample: 2020-01 – 2023-01 · selected US stocks (AAPL, MSFT, GOOG, ...) · equities · daily

  • Real financial data (12 selected stocks) market-prices unknown
    January 2, 2019 to January 2, 2020 (in‑s · 12 selected stocks
    Source sentence
    The initial portfolio w∗0 is calculated on January 3, 2020 using the data from January 2, 2019 to January 2, 2020 as the in‑sample dataset (252 trading days).
  • S&P 500 index benchmark public
    Source sentence
    We also include the performance of the S&P 500 index in Figure 2 to compare it with the performance of these portfolio selection models (a)-(e).

Costs

Not modeled A trading or execution result is reported without costs.

Finds 3

  • All strategies except the TSV model outperform the passive S&P 500 investment. performance
    Source sentence
    It is evident that all the strategies, except for the TSV model, outperform the passive investment strategy of the S&P 500 index.
  • The EEP‑TSV‑S model (d) outperforms all other models, showing the benefit of the expected excess profit constraint. performance
    Source sentence
    We can also see from Figure 2 that the EEP‑TSV‑S model (d) outperforms all the other models listed above. The expected excess profit constraint enhances the capability of controlling the downside risk.
  • Incorporating both symmetry information and expected downside constraints significantly improves investment performance. performance
    Source sentence
    Therefore, incorporating both symmetric information and expected downside constraints into portfolio selection models can significantly improve investment performance when using the models proposed in this paper.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 15 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 1 paper in the archive (1 by independent authors): Pareto frontier of portfolio investment under volatility … (uses its method or data)

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.