Abstract:
This paper presents Aeonaux Capital’s proprietary quantitative framework for equity investing in India, designed to respond adaptively to shifting market and macroeconomic regimes. Built from first principles, the approach integrates volatility regime detection, macroeconomic clustering, and high-integrity factor signal engineering to deliver consistent alpha with controlled drawdowns. The research introduces a live-tested, execution-aware model that is institutionally robust, SEBI-compliant, and suitable for PMS-level deployment.
- Context and Rationale
Indian equity markets have historically exhibited persistent volatility clustering, liquidity imbalances, and abrupt macro shifts conditions under which static factor models fail. Aeonaux Capital recognized this gap and engineered a system that dynamically realigns portfolio weights in response to changing market states. The result is a resilient architecture designed for real-world constraints and investor outcomes.
- Volatility Regime Modeling at Aeonaux
Aeonaux Capital segments the market into three volatility states compressed, trending, and distressed using a proprietary Hidden Markov regime detector calibrated on NIFTY 50 volatility. Each state triggers distinct portfolio rules, including leverage bounds, hedging overlays, and factor re-activation schedules.
- Macroeconomic Clustering and Signal Gating
Indian macro data is synthesized using PCA-reduced indicators across inflation, credit, and policy spreads. DBSCAN clustering isolates stable macro regimes, enabling Aeonaux to selectively activate signals suited to prevailing macroeconomic conditions. This regime-aware filtering minimizes noise and improves signal persistence.
- Factor Signal Framework
Signals used by Aeonaux are validated across three axes: statistical persistence, economic interpretability, and regime stability. Actively deployed signals include:
- Momentum: Residual-adjusted medium-term returns
- Quality: Accrual-filtered ROE volatility
- Earnings Stability: Dispersion of forward EPS revisions
- Liquidity Resilience: Bid-ask compressibility and ADV continuity
A stacking ensemble blends these signals with live monitoring via PSI (Population Stability Index). Auto-throttling is engaged when drift exceeds 1.5σ from training baseline.
- Portfolio Optimization Engine
Aeonaux constructs portfolios using a conditional Sortino frontier optimizer with shrinkage-adjusted regime-specific covariance matrices. Constraints include:
- Sector diversification (as per SEBI)
- Annual turnover < 1.1
- Execution-adjusted alpha using a proprietary TCA model
- Model Validation Protocol
- Walk-Forward Tests: 36-month train / 12-month test over 200+ rolling slices
- Drift Autopsies: KS-tests and live PSI logs
- Execution Simulation: Hybrid fill models adjusted for STT, bid-ask width, and fill risk
- Research Depth and Cross-Disciplinary Applications
The Aeonaux framework is not limited to traditional equity allocation. Its adaptive structure, driven by volatility clustering and macroeconomic segmentation, opens up applications in multi-asset rotation, risk parity overlays, and volatility targeting in currency pairs. The methodology’s modularity allows for the insertion of domain-specific signals—such as commodities-based inflation expectations or real estate-linked rate spreads—without retraining the core architecture.
By leveraging principles of market microstructure and execution-aware design, the Aeonaux system remains resilient across low-liquidity segments and macro regime bifurcations. This capacity to abstract market states into actionable clusters aligns with emerging trends in portfolio neuroscience and behavior-aware alpha modeling.
- What Makes Aeonaux Capital’s Research Differentiated
Aeonaux Capital is not a traditional PMS it is a research lab with capital attached. Each signal is tracked from inception to retirement, with drift and attribution logs made available to clients. Unlike industry practices that rely on fixed-factor blends, Aeonaux enforces live decay diagnostics, regime filters, and cost-aware rebalancing.
Our commitment is not only to alpha, but to integrity of alpha.
- Path Ahead
Future extensions include:
- Framework generalization to small- and mid-cap universes
- Machine learning enhancements for non-linear macro-signal interactions
- Cross-market transferability tests under macro stress conditions
Explore Further:
- Learn more about Aeonaux’s live PMS strategies and signal engine: https://www.aeonauxcapital.com/services
- Discover Aeonaux’s research philosophy and institutional insights: https://www.aeonauxcapital.com
Conclusion
Aeonaux Capital’s regime-adaptive framework reflects a new standard in Indian quant investing—one where research depth meets execution fidelity. In an environment rife with overfitted models and underperforming promises, Aeonaux delivers real-time robustness grounded in empirical discipline.
Author:
Tushar Joshi, Aeonaux Capital- Bengaluru
