Financial Econometrics · Fiscal Dominance
The SDCCA–MTV framework with high-dimensional evidence from Brazil.
The paper studies whether fiscal stress changes not only the level of Brazilian equity volatility, but the geometry of the coupling itself. Instead of treating dimensionality reduction and volatility modeling as separate steps, the framework uses the same fiscal transition state to govern both dynamic Ridge-CCA regularization and long-run unconditional variance, creating a regime-contingent high-dimensional transmission model.
Research question
Does fiscal dominance alter the latent coupling structure of Brazilian equity volatility, and can that change be measured jointly with nonstationary long-run variance?
Method
SDCCA-MTV combines Airy-wavelet decomposition, rolling dynamic Ridge-CCA, and multivariate time-varying GARCH. The same fiscal transition variable drives both the time-varying canonical shrinkage parameter and the long-run variance component, while the Selic rate enters the short-run GARCH-X channel. The application uses 36 continuously listed IBOV constituents over 2,721 trading days from 2015 to 2026.
Main findings
Medium-band coupling
0.994
At the 15–45 day horizon, the dominant canonical correlation averages 0.994, with a minimum of 0.979 across the sample, indicating extremely stable coherence between constituent volatility and the fiscal-monetary signal block.
Fiscal amplification
8.67×
The dominant coupling direction's long-run variance amplifies 8.67-fold between the low- and high-stress ends of the estimated fiscal transition range, providing a direct volatility-scale measure of fiscal dominance.
Corrected LM evidence
0.049 / 0.012
After correcting for first-stage estimation, the LM test rejects constant unconditional variance for the first two canonical directions with p-values of 0.0486 and 0.0119, while the third direction does not reject.
Monte Carlo calibration
2.0% size
The corrected LM test is conservative in simulation, with 2.0% empirical size against a nominal 5% level and 17.1% power at the paper's empirical 8.67-fold variance-amplification calibration.
Why it matters
The framework makes fiscal dominance measurable as a structural volatility multiplier rather than only a macroeconomic narrative. More importantly, it makes the dimensionality reduction itself state-dependent: under fiscal stress, the canonical geometry changes at the same time that long-run variance changes. That joint identification is the methodological core of the paper.
Full research record
The external scholarly record contains the complete manuscript and citation metadata. This ShockBridge page is the research showroom summary, not a substitute for the full paper.
Open scholarly record ↗