Paper: arXiv 2601.08571

Authors: Salam Rabindrajit Luwang, Buddha Nath Sharma, Kundan Mukhia, Md. Nurujjaman, Anish Rai, Filippo Petroni, Luis E. C. Rocha

Abstract

Financial markets alternate between tranquil periods and episodes of stress, and return dynamics can change substantially across these regimes. We study regime-dependent dynamics in developed and developing equity indices using a data-driven Hilbert–Huang-based regime identification and profiling pipeline, followed by variable-length Markov modeling of categorized returns. Market regimes are identified using an Empirical Mode Decomposition-based Hilbert–Huang Transform, where instantaneous energy from the Hilbert spectrum separates Normal, High, and Extreme regimes. We then profile each regime using Holo–Hilbert Spectral Analysis, which jointly resolves carrier frequencies, amplitude-modulation frequencies, and amplitude-modulation energy (AME). AME, interpreted as volatility intensity, declines monotonically from Extreme to High to Normal regimes. This decline is markedly sharper in developed markets, while developing markets retain higher baseline volatility intensity even in Normal regimes. Building on these regime-specific volatility signatures, we discretize daily returns into five quintile states \(\mathtt{“R”}_1\) to \(\mathtt{“R”}_5\) and estimate Variable-Length Markov Chains via context trees within each regime. Unconditional state probabilities show tail states dominate in Extreme regimes and recede as regimes stabilize, alongside persistent downside asymmetry. Entropy peaks in High regimes, indicating maximum unpredictability during moderate-volatility periods. Conditional transition dynamics, evaluated over contexts of length up to three days from the context-tree estimates, indicate that developed markets normalize more effectively as stress subsides, whereas developing markets retain residual tail dependence and downside persistence even in Normal regimes, consistent with a coexistence of continuation and burst-like shifts.

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 uses advanced mathematical concepts like Hilbert-Huang Transform, Empirical Mode Decomposition, and Variable-Length Markov Chains, requiring significant mathematical sophistication, while also implementing a detailed empirical pipeline on real market data with regime-specific analysis and quantitative metrics.

Research Flowchart

  flowchart TD
  A["Research Goal<br>Regime-Dependent Return Dynamics"] --> B["Data<br>Developed & Developing Equity Indices"]
  B --> C["Methodology: HHT Pipeline<br>EMD + Hilbert-Huang Transform"]
  C --> D{"Regime Classification<br>Instantaneous Energy Thresholds"}
  D --> E["<b>Outcome 1: Volatility Signatures</b><br>AME declines sharply in Developed<br>Higher baseline in Developing"]
  C --> F["Methodology: Profiling<br>Holo-Hilbert Spectral Analysis"]
  F --> G["<b>Outcome 2: Regime Profiles</b><br>Carrier/Modulation Frequency Analysis"]
  E & G --> H["Methodology: VLMC<br>Context Trees & Categorized Returns R1-R5"]
  H --> I["<b>Outcome 3: Transition Dynamics</b><br>Entropy peaks in High Regimes<br>Developing markets retain persistence"]
  I --> J["<b>Key Insight</b><br>Developed normalize; Developing show coexistence of continuation/bursts"]

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 data‑driven pipeline that uses Hilbert–Huang Transform‑based instantaneous energy to identify market regimes, profiles them with Holo‑Hilbert Spectral Analysis, and models intra‑regime return dynamics with Variable‑Length Markov Chains.

Source sentence
We study regime-dependent dynamics in developed and developing equity indices using a data-driven Hilbert--Huang-based regime identification and profiling pipeline, followed by variable-length Markov modeling of categorized returns. Building on these regime-specific volatility signatures, we discretize daily returns into five quintile states $\mathtt{R}_1$ to $\mathtt{R}_5$ and estimate Variable-Length Markov Chains

Assumes 6

  • Index return series are weakly stationary while their volatility dynamics are highly time‑varying. model
    Source sentence
    While index return series are often weakly stationary [24], their volatility dynamics are highly time-varying and exhibit intermittent bursts, par-ticularly during periods of market stress [25, 26].
  • Regime thresholds are set using the sample mean and standard deviation of the normalized instantaneous energy, with cutoffs at µ+σ and µ+6σ. statistical
    Source sentence
    The regime thresholds are then computed using the sample mean µ and sample standard deviation σ of the normalized series {E(t)}, and regimes are identified using cutoffs at µ + σ and µ + 6σ.
  • Daily returns are discretized into five quintile‑based states using fixed full‑sample cutoffs, making state probabilities comparable across regimes. data
    Source sentence
    We use fixed full-sample quintile cutoffs for each index so that regime-specific state probabilities and transition behavior are comparable across regimes. 2.2.
  • Variable‑Length Markov Chains adequately capture transition dynamics within each identified regime. model
    Source sentence
    Building on these regime-specific volatility signatures, we discretize daily returns into five quintile states $\mathtt{R}_1$ to $\mathtt{R}_5$ and estimate Variable-Length Markov Chains via context trees within each regime.
Show 2 more
  • Developed and developing markets are correctly classified according to the UN World Economic Situation and Prospects 2025 report. data
    Source sentence
    Data We analyze daily closing prices of the standard stock market indices for twenty coun-tries—ten developed and ten developing economies, as classified by the World Economic Situation and Prospects 2025 (United Nations) report [38, 39].
  • The Hilbert–Huang Transform provides a reliable, data‑driven measure of instantaneous energy for regime identification. model
    Source sentence
    Let IAj(t) denote the instantaneous amplitude of the jth IMF from HHT. We compute the raw instantaneous energy as Eraw(t) = n X j=1 IAj(t)2, (10) and apply max-normalization, E(t) = Eraw(t) maxt[Eraw(t)].

Data

  • Daily closing prices of twenty stock market indices (ten developed, ten developing) · Yahoo Finance market-prices public
    January 2000 to April 2025 · Standard stock market indices for twenty countries
    Source sentence
    Data We analyze daily closing prices of the standard stock market indices for twenty coun-tries—ten developed and ten developing economies, as classified by the World Economic Situation and Prospects 2025 (United Nations) report [38, 39]. The daily data is analyzed from January 2000 to April 2025.

Recognised sources: Yahoo Finance

Costs

Not applicable The paper reports no trading or execution result.

Finds 5

  • Amplitude‑modulation energy (interpreted as volatility intensity) declines monotonically from Extreme to High to Normal regimes. statistical
    Source sentence
    AME, interpreted as volatility intensity, declines monotonically from Extreme to High to Normal regimes.
  • The decline in volatility intensity is markedly sharper in developed markets, while developing markets retain higher baseline intensity even in Normal regimes. statistical
    Source sentence
    This decline is markedly sharper in developed markets, while developing markets retain higher baseline volatility intensity even in Normal regimes.
  • Unconditional state probabilities show tail states dominate in Extreme regimes and recede as regimes stabilize, with persistent downside asymmetry. statistical
    Source sentence
    Unconditional state probabilities show tail states dominate in Extreme regimes and recede as regimes stabilize, alongside persistent downside asymmetry. To examine regime-dependent return dynamics, daily in-dex returns are categorized into discrete states and analyzed using variable-length Markov chains.
  • Shannon entropy peaks in High regimes, indicating maximum unpredictability during moderate‑volatility periods. statistical
    Source sentence
    Entropy peaks in High regimes, indicating maximum unpredictability during moderate-volatility periods.
  • Developed markets normalize more effectively as stress subsides, whereas developing markets retain residual tail dependence and downside persistence even in Normal regimes. statistical
    Source sentence
    Conditional transition dynamics, evaluated over contexts of length up to three days from the context-tree estimates, indicate that developed markets normalize more effectively as stress subsides, whereas developing markets retain residual tail dependence and downside persistence even in Normal regimes, consistent with a coexistence of continuation and burst-like shifts. Conditional transition dynamics indicate

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 27 related papers). Coverage is partial, so this is neither replication nor refutation.

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.

Cited References (113)

  1. Analysis of time series subject to changes in regime (1990)
  2. Regime changes and financial markets (2012)
  3. Risk and volatility: Econometric models and financial practice (2004)
  4. The asymptotic distribution of extreme stock market returns (1996)
  5. Detection of high and low states in stock market returns with MCMC method in a Markov switching model (2014)
  6. Dynamics of Trade-by-Trade Price Movements: Decomposition and Models (2003)
  7. The frequency of regime switching in financial market volatility (2015)
  8. The variation of certain speculative prices (1963)
and 105 more
  1. Why does stock market volatility change over time? (1989)
  2. Answering the skeptics: Yes, standard volatility models do provide accurate forecasts (1998)
  3. Nonstationarities in financial time series, the long-range dependence, and the IGARCH effects (2004)
  4. The volatility surface: a practitioner’s guide (2011)
  5. Regime shifts: Implications for dynamic strategies (corrected) (2012)
  6. Volatility clustering in financial markets: a microsimulation of interacting agents (2000)
  7. What good is a volatility model? (2007)
  8. Asset allocation under multivariate regime switching (2007)
  9. Regime-based versus static asset allocation: Letting the data speak (2015)
  10. Modeling precious metal returns through fractional jump-diffusion processes combined with Markov regime-switching stochastic volatility (2021)
  11. Markov-Regime switches in oil markets: The fear factor dynamics (2023)
  12. Regime shifts in the behaviour of international currency and equity markets: A Markov-switching analysis (2021)
  13. Weighted-indexed semi-Markov models for modeling financial returns (2012)
  14. Detection and forecasting of extreme events in stock price triggered by fundamental, technical, and external factors (2023)
  15. Characteristics of 2020 stock market crash: The COVID-19 induced extreme event (2021)
  16. A practical guide to wavelet analysis (1998)
  17. Time-frequency analysis (1995)
  18. Long short-term memory (1997)
  19. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis (1998)
  20. Application of Hilbert-Huang transform in financial time series analysis (2013)
  21. Hilbert-Huang transform-based volatility model (2008)
  22. Business confidence and stock returns in the USA: a time-varying Markov regime-switching model (2012)
  23. A Markov model of financial returns (2006)
  24. A semi-Markov model for price returns (2012)
  25. Re-examination of risk-return dynamics in international equity markets and the role of policy uncertainty, geopolitical risk and VIX: Evidence using Markov-switching copulas (2022)
  26. Volatility in stock returns for new EU member states: Markov regime switching model (2007)
  27. Predictability of stock returns: Robustness and economic significance (1995)
  28. Stock return predictability: Is it there? (2007)
  29. Macroeconomic factors do influence aggregate stock returns (2002)
  30. Predicting stock returns (2006)
  31. The economic value of predicting stock index returns and volatility (2004)
  32. Variable length Markov chain with exogenous covariates (2022)
  33. Nonlinear effects of oil shocks on stock returns: a Markov-switching approach (2010)
  34. Applying a Markov chain for the stock pricing of a novel forecasting model (2017)
  35. Ten lectures on wavelets (1992)
  36. Fourier analysis for stock price forecasting: Assumption and evidence (2016)
  37. Estimation of intraday stock market resiliency: Short-Time Fourier Transform approach (2019)
  38. Epicasting: an ensemble wavelet neural network for forecasting epidemics (2023)
  39. International comovement of stock market returns: A wavelet analysis (2009)
  40. PyDTMC (2024) — link
  41. Variable length Markov chains (1999)
  42. Stock returns and real activity: A century of evidence (1990)
  43. Do macroeconomic variables have regime-dependent effects on stock return dynamics? Evidence from the Markov regime switching model (2009)
  44. High-frequency stock market order transitions during the US–China trade war 2018: A discrete-time Markov chain analysis — → 2405.05634
  45. Demand-based option pricing (2008)
  46. Asset fire sales (and purchases) in equity markets (2007)
  47. The benefits of bagging for forecast models of realized volatility (2010)
  48. A test for independence based on the correlation dimension (1996)
  49. Chaos and nonlinear dynamics: application to financial markets (1991)
  50. Nonlinear dynamics, chaos, and instability: statistical theory and economic evidence (1991)
  51. On holo-Hilbert spectral analysis: a full informational spectral representation for nonlinear and non-stationary data (2016)
  52. Unraveling nonlinear electrophysiologic processes in the human visual system with full dimension spectral analysis (2019)
  53. On instantaneous frequency (2009)
  54. Empirical mode decomposition as a filter bank (2004)
  55. Frontoparietal beta amplitude modulation and its interareal cross-frequency coupling in visual working memory (2021)
  56. The uniqueness of the instantaneous frequency based on intrinsic mode function (2013)
  57. Detecting strange attractors in turbulence (2006)
  58. Measuring the strangeness of strange attractors (1983)
  59. Evaluating the different stages of Parkinson’s disease using electroencephalography with holo-Hilbert spectral analysis (2022)
  60. Liquidity of emerging markets (2005)
  61. A new approach to the economic analysis of nonstationary time series and the business cycle (1989)
  62. Modeling the conditional distribution of interest rates as a regime-switching process (1996)
  63. Autoregressive conditional heteroskedasticity and changes in regime (1994)
  64. Threshold heteroskedastic models (1994)
  65. A double-threshold GARCH model for the French Franc/Deutschmark exchange rate (2001)
  66. An inequality with applications to statistical estimation for probabilistic functions of Markov processes and to a model for ecology (1967)
  67. On a threshold model (1978)
  68. Threshold autoregression, limit cycles and cyclical data (1980)
  69. Specification, estimation, and evaluation of smooth transition autoregressive models (1994)
  70. Multi-regime nonlinear capital asset pricing models (2011)
  71. Maximum likelihood estimation of the Markov-switching GARCH model (2014)
  72. On a double-threshold autoregressive heteroscedastic time series model (1996)
  73. On a threshold heteroscedastic model (2006)
  74. Power-law relaxation in a complex system: Omori law after a financial market crash (2003)
  75. Early-warning signals for critical transitions (2009)
  76. Statistical properties of the aftershocks of stock market crashes revisited: Analysis based on the 1987 crash, financial-crisis-2008 and COVID-19 pandemic (2022)
  77. mixvlmc: Variable Length Markov Chains with Covariates (2025) — link
  78. World Economic Situation and Prospects 2025 (2025) — link
  79. Yahoo Finance (2025) — link
  80. The full informational spectral analysis for auditory steady-state responses in human brain using the combination of canonical correlation analysis and holo-Hilbert spectral analysis (2022)
  81. Multiscale three-dimensional Holo–Hilbert spectral entropy: A novel complexity-based early fault feature representation method for rotating machinery (2023)
  82. Intraday order transitions across market capitalizations through Markov analysis (2025)
  83. Change point dynamics for financial data: an indexed Markov chain approach (2019)
  84. Copula based multivariate semi-Markov models with applications in high-frequency finance (2018)
  85. A semi-Markov model with memory for price changes (2011)
  86. Order-frequency holo-Hilbert spectral analysis for machinery fault diagnosis under time-varying operating conditions (2024)
  87. Using holo-Hilbert spectral analysis to quantify the modulation of Dansgaard-Oeschger events by obliquity (2018)
  88. Adaptive fast iterative filter Holo-spectrum analysis and its applications to fault diagnosis of rolling bearing (2024)
  89. The World Bank’s classification of countries by income (2016)
  90. Markov chain model application on share price movement in stock market (2013)
  91. Are stock returns predictable? A test using Markov chains (1991)
  92. Stochastic analysis of stock price changes as markov chain in finite states (2022)
  93. Bid-ask bounce and speads in the foreign exchange futures market (1996)
  94. Stylized facts of daily return series and the hidden Markov model (1998)
  95. An autoregressive model with suddenly changing parameters and an application to stock market prices (1988)
  96. Systemic Risk and Default Cascades in Global Equity Markets: Extending the Gai-Kapadia Framework with Stochastic Simulations and Network Analysis — → 2504.01969
  97. The tail risk of emerging stock markets (2009)
  98. Do institutional investors stabilize equity markets in crisis periods? Evidence from COVID-19 (2025)
  99. Modelling extremal events: for insurance and finance (2013)
  100. Persistence in variance, structural change, and the GARCH model (1990)
  101. Deciphering the liquidity and credit crunch 2007–2008 (2009)
  102. Fire sales in finance and macroeconomics (2011)
  103. Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation (1982)
  104. Generalized autoregressive conditional heteroskedasticity (1986)
  105. Empirical properties of asset returns: stylized facts and statistical issues (2001)

2 of 113 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.