Paper: arXiv 2509.23533

Authors: Gabriele Casto

Abstract

We introduce the Historical and Dynamic Volatility Ratios (HVR/DVR) and show that equity and index volatilities are cointegrated at intraday and daily horizons. This allows us to construct a VECM to forecast portfolio volatility by exploiting volatility cointegration. On S&P 500 data, HVR is generally stationary and cointegration with the index is frequent; the VECM implementation yields substantially lower mean absolute percentage error (MAPE) than covariance-based forecasts at short- to medium-term horizons across portfolio sizes. The approach is interpretable and readily implementable, factorizing covariance into market volatility, relative-volatility ratios, and correlations.

Complexity vs Empirical Score

  • Math Complexity: 8.0/10
  • Empirical Rigor: 7.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced econometric methods including cointegration analysis and VECM modeling, requiring significant mathematical sophistication. Empirical validation is demonstrated on S&P 500 data with quantitative metrics like MAPE, indicating strong practical implementation.

Research Flowchart

  flowchart TD
  A["Research Goal<br>Forecast portfolio volatility<br>exploiting cointegrated dynamics"] --> B["Data & Inputs<br>S&P 500 intraday & daily data<br>Equity & index volatilities"]
  B --> C["Methodology: Historical Volatility Ratio<br>Compute relative variance ratios<br>Check for stationarity & cointegration"]
  C --> D["Methodology: Dynamic Volatility Ratio<br>VECM estimation<br>Exploit cointegration for forecasts"]
  D --> E["Key Findings/Outcomes"]
  subgraph E ["Outcomes"]
      F1["HVR is typically stationary<br>Volatilities are cointegrated"]
      F2["VECM forecasts reduce MAPE<br>vs. covariance benchmarks<br>at short-medium horizons"]
      F3["Interpretable decomposition<br>Market Volatility +<br>Relative Ratios + Correlations"]
  end

  %% Styling
  classDef goal fill:#e1f5fe,stroke:#01579b,stroke-width:2px
  classDef data fill:#fff3e0,stroke:#e65100,stroke-width:2px
  classDef method fill:#f3e5f5,stroke:#4a148c,stroke-width:2px
  classDef outcome fill:#e8f5e9,stroke:#1b5e20,stroke-width:2px

  class A goal
  class B data
  class C,D method
  class E,F1,F2,F3 outcome

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 Historical and Dynamic Volatility Ratios (HVR/DVR) and a VECM framework that exploits cointegration among volatilities to forecast portfolio risk.

Source sentence
We introduce the Historical and Dynamic Volatility Ratios (HVR/DVR) and show that equity and index volatilities are cointegrated at intraday and daily horizons.

Assumes 3

  • Asset volatilities are cointegrated with the benchmark index. statistical
    Source sentence
    First, the cointegration assumption of asset volatilities, as experimentally justified in Section 7 of this work, may not be universally applicable over horizons, industries, or during periods of structural reforms.
  • The VECM framework relies on the assumption of cointegration among volatility series. model
    Source sentence
    we formulate a theoretical framework in terms of which portfolio volatility can be expressed and forecast by a Vector Error Correction Model (VECM) given the assumption of cointegration among volatility series
  • HVR series are stationary (mean‑reverting). statistical
    Source sentence
    The ADF test supports the hypothesis of mean reversion of the HVR and cointegration among volatilities for a sizable fraction of stocks at short horizons

Data

  • S&P 500 constituents (daily and intraday) market-prices unknown
    2018/09/01‑2025/09/01 (daily); 2024/09/0 · All S&P 500 constituents
    Source sentence
    Using a comprehensive dataset of all S&P 500 constituents at both daily and intraday frequencies, we validate the ability of the Vector Error Correction Model (VECM) to forecast portfolio volatility

Costs

Not applicable The paper reports no trading or execution result.

Finds 4

  • The VECM implementation yields substantially lower mean absolute percentage error (MAPE) than covariance‑based forecasts at short‑ to medium‑term horizons. predictive-accuracy
    Source sentence
    the VECM implementation yields substantially lower mean absolute percentage error (MAPE) than covariance-based forecasts at short- to medium-term horizons across portfolio sizes
  • For intraday (5–30 min) and short‑to‑medium daily horizons (5–30 days), VECM systematically outperforms the traditional covariance approach on MAPE. performance
    Source sentence
    For forecasting, the VECM framework systematically outperforms the traditional covariance approach on MAPE at intraday (5–30 minutes) and short-to-medium daily horizons (5–30 days)
  • Stationarity of HVR is observed for 100 % of series at horizons up to 10 days, decreasing at longer horizons. statistical
    Source sentence
    The share of series classified as stationary is 100.00% (5 minutes)
  • Cointegration between individual stock volatilities and the index is found for 100 % of series at short horizons, weakening at longer horizons. statistical
    Source sentence
    The share of series classified as cointegrated with the index volatility is 100.00% (5 minutes)

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 36 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 (16)

  1. Portfolio selection (1952)
  2. Generalized autoregressive conditional heteroskedasticity (1986)
  3. Dynamic conditional correlation: A simple class of multivariate GARCH models (2002)
  4. Co-integration and error correction: Representation, estimation, and testing (1987)
  5. Estimation and hypothesis testing of cointegration vectors in Gaussian vector autoregressive models (1991)
  6. Improved estimation of the covariance matrix of stock returns with an application to portfolio selection (2003)
  7. A well-conditioned estimator for large-dimensional covariance matrices (2004)
  8. The Markowitz optimization enigma: Is ``optimized’’ optimal? (1989)
and 8 more
  1. The distribution of realized exchange rate volatility (2001)
  2. Distribution of the estimators for autoregressive time series with a unit root (1979)
  3. Likelihood ratio statistics for autoregressive time series with a unit root (1981)
  4. Testing the null hypothesis of stationarity against the alternative of a unit root (1992)
  5. Testing for a unit root in time series regression (1988)
  6. Lag length selection and the construction of unit root tests with good size and power (2001)
  7. A simple, positive semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix (1987)
  8. Automatic time series forecasting: The forecast package for R (2008)

0 of 16 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.