Paper: arXiv 2601.03927
Authors: Vrinda Dhingra, Amita Sharma, Anubha Goel
Abstract
Index tracking, also known as passive investing, has gained significant traction in financial markets due to its cost-effective and efficient approach to replicating the performance of a specific market index. This review paper provides a comprehensive overview of the various modeling approaches and strategies developed for index tracking, highlighting the strengths and limitations of each approach. We categorize the index tracking models into three broad frameworks: optimization-based models, statistical-based models and machine learning based data-driven approach. A comprehensive empirical study conducted on the S&P 500 dataset demonstrates that the tracking error volatility model under the optimization-based framework delivers the most precise index tracking, the convex co-integration model, under the statistical-based framework achieves the strongest return-risk balance, and the deep neural network with fixed noise model within the data-driven framework provides a competitive performance with notably low turnover and high computational efficiency. By combining a critical review of the existing literature with comparative empirical analysis, this paper aims to provide insights into the evolving landscape of index tracking and its practical implications for investors and fund managers.
Complexity vs Empirical Score
- Math Complexity: 6.5/10
- Empirical Rigor: 7.0/10
- Quadrant: Holy Grail — high math complexity, high empirical rigor
Why this score: The paper involves optimization-based models with quadratic programming and combinatorial constraints (NP-hard problems) alongside statistical and machine learning frameworks, indicating significant mathematical density. It also presents a comprehensive empirical study on the S&P 500 with specific model comparisons and performance metrics, demonstrating strong data and implementation focus.
Research Flowchart
flowchart TD
A["Research Goal: Review Index Tracking Models<br>and Evaluate Performance"] --> B["Methodology: Categorize Models<br>into 3 Frameworks"]
B --> C{"Data Input: S&P 500 Dataset"}
C --> D["Computational Process 1:<br>Optimization-based Models"]
C --> E["Computational Process 2:<br>Statistical-based Models"]
C --> F["Computational Process 3:<br>Machine Learning/Deep Learning"]
D --> G["Outcome 1: Tracking Error Volatility<br>Model achieves highest precision"]
E --> H["Outcome 2: Convex Co-integration<br>Model offers best risk-return balance"]
F --> I["Outcome 3: DNN with Fixed Noise<br>Model provides low turnover & high efficiency"]
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 a comprehensive review of index‑tracking models across three paradigms and an empirical comparison on the S&P 500, highlighting that the tracking‑error variance model is most accurate, the convex co‑integration model has the best risk‑adjusted performance, and a deep neural networ
Source sentence
This review paper provides a comprehensive overview of the various modeling approaches and strategies developed for index tracking, highlighting the strengths and limitations of each approach.
Assumes 4
- Both the index and constituent asset price series are integrated of order one (I(1)). statistical
Source sentence
The approach assumes that both the index and the constituent asset price series are integrated of order one, i.e., I(1), and that their logarithms can be linearly related through a co‑integrating vector.
- Residuals from the co‑integration regression are stationary when the null of a unit root is rejected at the 5% level. statistical
Source sentence
Subsets for which the null of a unit root is rejected at the 5% level (p < 0.05) are retained as candidate co‑integrated portfolios.
- The MSE tracking‑error function is convex and quadratic, leading to a convex quadratic programming problem. model
Source sentence
This objective function is convex and quadratic, resulting in the following convex quadratic programming problem: (MSE) Min
- All underlying formulations are convex programs, so solutions are globally optimal when convergence occurs. model
Source sentence
It is worth noting that, since all underlying formulations are convex programs, the solutions reported are globally optimal when convergence occurs.
Data
- S&P 500 dataset benchmark commercial
Source sentence
We characterize the S&P 500 index and its constituents over the study period (12 October 2012 to 8 November 2022) using three descriptive statistics...
Costs
Partial
Turnover is used as a proxy for transaction costs; the paper reports turnover ratios but does not embed explicit cost numbers in the performance results.
Source sentence
Turnover, a proxy for transaction costs, further differentiates the paradigms.
Finds 5
- The tracking‑error variance (TEV) model achieves the lowest tracking error and highest correlation with the index among optimization‑based formulations. performance
Source sentence
TEV model delivers the most accurate replication, achieving the lowest tracking error and the highest correlation with the index.
- The convex co‑integration (Cvx CoInt) model attains the strongest risk‑adjusted performance, with superior Sharpe, Sortino, and information ratios. performance
Source sentence
The convex cointegration (Cvx CoInt) model attains the strongest risk‑adjusted performance, as reflected in superior Sharpe, Sortino, and information ratios.
- The deep neural network with fixed noise (DNNF) provides competitive replication accuracy while delivering substantially lower turnover and faster computation. performance
Source sentence
The deep neural network with fixed noise (DNNF) achieves competitive replication accuracy while offering substantial advantages in terms of turnover and computational efficiency.
- Optimization‑based portfolios have higher turnover (median ~1.20) compared to statistical (median ~1.18) and data‑driven models (median ~0.87). performance
Source sentence
Turnover, a proxy for transaction costs, further differentiates the paradigms. Optimization‑based portfolios fall in the range 1.14–1.91 (median = 1.20)
- Statistical models, especially Cvx CoInt, are computationally intensive (≈18 hours), whereas regression‑based and penalized regression models solve in seconds; data‑driven models are generally the most efficient. performance
Source sentence
All optimization models (except SES) are computationally intensive, requiring 17–18 hours on average per evaluation window
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 28 related papers). Coverage is partial, so this is neither replication nor refutation.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.
Cited References (98)
- An evolutionary heuristic for the index tracking problem (2003) — link
- Mixed-integer programming approaches for index tracking and enhanced indexation (2009) — link
- Combining random forest and multicollinearity modeling for index tracking (2022) — link
- Active versus passive investing: Evidence from the 2009-2017 market (2018)
- The Shift from Active to Passive Investing: Risks to Financial Stability? (2020) — link
- Passive Investment Strategies and Efficient Markets (2003) — link
- A particle swarm optimization heuristic for the index tacking problem (2010) — link
- Optimal index tracking under transaction costs and impulse control (1998)
and 90 more
- Empirical insights on indexing: how capitalization, stratification and weighting can affect tracking error (1998)
- Portfolio optimization and index tracking for the shipping stock and freight markets using evolutionary algorithms (2013) — link
- Performance replication of the Spot Energy Index with optimal equity portfolio selection: Evidence from the UK, US and Brazilian markets (2014) — link
- A linear model for tracking error minimization (1999) — link
- Kernel Search: An application to the index tracking problem (2012) — link
- Clustering of financial time series with application to index and enhanced index tracking portfolio (2005) — link
- Tracking S&P 500 Index Funds (2001) — link
- Optimal portfolio selection and dynamic benchmark tracking (2005) — link
- Bond portfolio optimization problems and their applications to index tracking: A partial optimization approach (1996) — link
- A hybrid optimization approach to index tracking (2009) — link
- Factor based index tracking (2006) — link
- Nonnegative-lasso and application in index tracking (2014) — link
- Robust portfolio selection for index tracking (2012) — link
- Differential evolution and combinatorial search for constrained index-tracking (2009) — link
- Simultaneous pursuit of out-of-sample performance and sparsity in index tracking portfolios (2013) — link
- A methodology for index tracking based on time-series clustering (2004) — link
- A constrained cluster-based approach for tracking the S&P 500 index (2017) — link
- Index tracking and enhanced indexing using cointegration and correlation with endogenous portfolio selection (2017) — link
- On the index tracking and the statistical arbitrage choosing the stocks by means of cointegration: the role of stock picking (2015) — link
- Solving the index tracking problem based on a convex reformulation for cointegration (2020) — link
- Index tracking based on deep neural network (2019) — link
- Index tracking through deep latent representation learning (2020) — link
- Index tracking with cardinality constraints: A stochastic neural networks approach (2020)
- Neural network with fixed noise for index-tracking portfolio optimization (2021) — link
- Factor-based robust index tracking (2017) — link
- Optimal construction and rebalancing of index-tracking portfolios (2018) — link
- Cardinality versus q-norm constraints for index tracking (2014) — link
- Sparse Portfolios for High-Dimensional Financial Index Tracking (2018) — link
- Convex Optimization (2004)
- Support vector regression as conditional value-at-risk minimization with application to financial time-series analysis (2010) — link
- Conditional value-at-risk for general loss distributions (2002) — link
- A mean/variance analysis of tracking error (1992)
- Co-integration and error correction: Representation, estimation, and testing (1987) — link
- Some properties of time series data and their use in econometric model specification (1981) — link
- Portfolio selection (1952)
- Index tracking with controlled number of assets using a hybrid heuristic combining genetic algorithm and non-linear programming (2017) — link
- Meta-heuristic based decision support for portfolio optimization with a case study on tracking error minimization in passive portfolio management (2003) — link
- Nonnegative-lasso and application in index tracking (2014) — link
- Nonnegative Elastic Net and application in index tracking (2014) — link
- Lasso-based index tracking and statistical arbitrage long-short strategies (2020) — link
- Quantile regression for index tracking and enhanced indexation (2013) — link
- Efficient sparse portfolios based on composite quantile regression for high-dimensional index tracking (2020) — link
- Creating better tracking portfolios with quantiles (2022) — link
- Sparse Portfolio Selection via Topological Data Analysis based Clustering — → 2401.16920
- Sparse Index Tracking via Topological Learning — → 2310.09578
- Network-based index tracking using asset dependency structures (2026) — link
- Exact and heuristic approaches for the index tracking problem with UCITS constraints (2013) — link
- An efficient optimization approach for a cardinality-constrained index tracking problem (2016) — link
- An index tracking model with stratified sampling and optimal allocation (2018) — link
- Minimizing the tracking error of cardinality constrained portfolios (2018) — link
- Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk (1964)
- Efficient Capital Markets: A Review of Theory and Empirical Work (1970)
- Index tracking and enhanced indexing using mixed conditional value-at-risk (2018) — link
- A systematic literature review on solution approaches for the index tracking problem (2024) — link
- Artificial intelligence and machine learning in finance: Identifying foundations, themes, and research clusters from bibliometric analysis (2021) — link
- Optimal hedging using cointegration (1999) — link
- Stock-Index Tracking Optimization Using Auto-Encoders (2020) — link
- Risk-allocation-based index tracking (2023) — link
- Index optimization replication algorithm by using the soft subspace clustering method (2014) — link
- Index Tracking Based on Dynamic Time Warping and Constrained k-medoids Clustering (2021) — link
- Sparse Portfolio Selection via Topological Data Analysis based Clustering — → 2401.16920
- Accurate solution of the Index Tracking problem with a hybrid simulated annealing algorithm — → 2303.13282
- A fuzzy index tracking portfolio selection model (2005) — link
- Robust and sparse portfolio model for index tracking (2019) — link
- Lasso-based index tracking and statistical arbitrage long-short strategies (2020) — link
- The Performance of Mutual Funds in the Period 1945–1964 (1968)
- Support Vector Regression for Time Series Analysis (2011) — link
- Splitting criteria for ordinal decision trees: an experimental study (2024)
- Regression-Based Index Tracking Versus Clustering-Based Index Tracking: An Empirical Study (2025) — link
- Risk reduced sparse index tracking portfolio: A topological data analysis approach (2026) — link
- A robust possibilistic programming approach for selecting a clustering-based index tracking portfolio (2025) — link
- A benchmark-asset principal component factorization for index tracking on large investment universes (2025) — link
- Index Tracking Based on Norm-Constraints and Regularization (2025) — link
- Enhancing Sparse Index-Tracking Portfolios Using Deep Learning Models (2025) — link
- Harmony Search Based Metaheuristic for the Index Tracking Problem (2024) — link
- Optimizing index tracking: A Random Matrix Theory approach to portfolio selection (2025) — link
- Risk measure index tracking model (2022) — link
- A mixed 0-1 LP for index tracking problem with CVaR risk constraints (2012) — link
- Index tracking by using sparse support vector regression (2017) — link
- Combining random forest and multicollinearity modeling for index tracking (2022) — link
- Reinforcement Learning for Financial Index Tracking — → 2308.02820
- An integrative extraction approach for index-tracking portfolio construction and forecasting under a deep learning framework (2024) — link
- Fast and stable portfolios through Huber’s criterion for constrained index tracking (2025) — link
- How the index fund was born (2011)
- SPDR S&P 500 ETF Trust (SPY) (1993) — link
- Deep learning for enhanced index tracking (2024)
- Reinforcement Learning for Financial Index Tracking — → 2308.02820
- Index tracking with cardinality constraints: A stochastic neural networks approach (2020)
- Index tracking using shapley additive explanations and one-dimensional pointwise convolutional autoencoders (2024)
- State-dependent stock selection in index tracking: a machine learning approach (2022)
6 of 98 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.