Paper: arXiv 2401.16920
Authors: Anubha Goel, Damir Filipović, Puneet Pasricha
Abstract
This paper uses topological data analysis (TDA) tools and introduces a data-driven clustering-based stock selection strategy tailored for sparse portfolio construction. Our asset selection strategy exploits the topological features of stock price movements to select a subset of topologically similar (different) assets for a sparse index tracking (Markowitz) portfolio. We introduce new distance measures, which serve as an input to the clustering algorithm, on the space of persistence diagrams and landscapes that consider the time component of a time series. We conduct an empirical analysis on the S&P index from 2009 to 2022, including a study on the COVID-19 data to validate the robustness of our methodology. Our strategy to integrate TDA with the clustering algorithm significantly enhanced the performance of sparse portfolios across various performance measures in diverse market scenarios.
Complexity vs Empirical Score
- Math Complexity: 7.0/10
- Empirical Rigor: 7.5/10
- Quadrant: Holy Grail — high math complexity, high empirical rigor
Why this score: The paper employs advanced topological data analysis (TDA) concepts like persistence diagrams and landscapes, which are mathematically dense, and backs this with a specific empirical study on S&P data from 2009–2022 including COVID-19 validation, demonstrating data-heavy implementation.
Research Flowchart
flowchart TD A["Research Goal"] --> B["Data: S&P Stocks 2009-2022"] B --> C["Topological Data Analysis<br>Persistence Diagrams & Landscapes"] C --> D["New Topological Distance Measures<br>incorporating time"] D --> E["Clustering & Asset Selection<br>Topologically Similar/Different Assets"] E --> F["Portfolio Construction<br>Sparse Index Tracking / Markowitz"] F --> G["Key Findings"] G --> G1["Enhanced Performance<br>across diverse markets"] G --> G2["Robustness<br>validated via COVID-19 study"]
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 new time‑aware distance measures on persistence diagrams and landscapes and a TDA‑based clustering stock selection strategy for sparse portfolio construction.
Source sentence
We introduce new distance measures, which serve as an input to the clustering algorithm, on the space of persistence diagrams and landscapes that consider the time component of a time series.
Assumes 2
- The analysis uses daily closing prices of S&P 500 constituents. data
Source sentence
We consider the daily closing prices of constituents of the S&P 500 index, obtained from the Thomson Reuters datastream spanning a period from 1st Dec 2009 to 31st August 2022.
- Portfolios are re‑balanced without explicit modeling of market impact or slippage. execution
Source sentence
Fourth, we calculate the portfolio turnover ratio (TR) as a proxy for the transaction cost.
Data
- S&P 500 constituents daily closing prices market-prices commercial
Source sentence
We consider the daily closing prices of constituents of the S&P 500 index, obtained from the Thomson Reuters datastream spanning a period from 1st Dec 2009 to 31st August 2022.
Costs
Partial
Turnover ratio is used as a proxy for transaction costs; no explicit cost model is applied.
Source sentence
Fourth, we calculate the portfolio turnover ratio (TR) as a proxy for the transaction cost.
Finds 5
- The sparse TDA‑based portfolio MV2 achieves the highest out‑of‑sample mean, Sharpe ratio and certainty‑equivalent return, and the lowest turnover, outperforming benchmark and other TDA portfolios. performance
Source sentence
MV 2 achieves the maximum values for mean, SR, and CEQ, and the minimum value of TR, thus outperforming the other portfolios, including TDA‑based portfolios and the benchmark portfolios.
- Cluster C2 delivers the highest information ratio and best trade‑off between excess return and tracking error among TDA‑based index‑tracking portfolios. performance
Source sentence
C2 excels C4 as it attains the highest information ratio and maintains a trade‑off between the return over and above the benchmark index and the tracking error.
- Affinity propagation clustering (APC) provides high correlation with the benchmark while keeping turnover low, balancing tracking accuracy and transaction costs. performance
Source sentence
APC achieves this alignment without excessively high turnover, striking a balance between tracking the benchmark and maintaining low trading costs.
- During the COVID‑19 volatile period, the sparse TDA‑based portfolio MV2 continues to outperform the benchmark, MV A, and the naive portfolio. performance
Source sentence
the sparse TDA‑based portfolio MV 2 performs really well even in this highly volatile period
- All clustering‑based portfolios show less out‑of‑sample performance degradation than non‑clustering benchmarks, indicating robustness of the TDA filtering strategy. performance
Source sentence
the difference between the in‑sample and out‑of‑sample average values is less than the MV A and GV A portfolios, which suggests that TDA‑based filtering strategy is more robust to the changing market dynamics.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 23 related papers). Coverage is partial, so this is neither replication nor refutation.
3 papers in the archive
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.