Paper on market state-based multi-asset tactical asset allocation optimization released
Get up to speed
Supriadi and Bambang Siswoyo released a new paper proposing a multi-asset tactical asset allocation framework that integrates Gaussian Hidden Markov Model (HMM) market state identification and Nested Clustered Optimization (NCO). This framework uses only information available before the decision point to identify three market states: bull, sideways, and bear. It introduces state-conditional parameters in covariance estimation, hierarchical clustering, and weight optimization, replacing fixed calendar rebalancing with adaptive rebalancing triggered by state switches or significant changes in optimal weights. This framework was validated through rolling walk-forward backtesting on a monthly multi-asset portfolio including Indonesian government bonds, the LQ45 stock index, IDR-denominated gold, CNY/IDR exchange rate, and BTC/IDR. The paper claims that this framework can limit tail risk, with a maximum drawdown of 4.44%.
AI-generated from reports · 2 hr agoUpdated
Coverage timeline
Follow the coverage to understand different sides of the event.
- arXiv q-finMarket Regime-Aware Multi-Asset Tactical Asset Allocation Framework Based on Nested Clustered Optimization
This arXiv paper proposes a multi-asset tactical asset allocation framework that integrates market regime identification with Nested Clustered Optimization (NCO). It uses a Gaussian Hidden Markov Model that only uses information available up to the decision time to identify three market regimes: bull, sideways, and bear markets. State-dependent parameters are introduced in covariance estimation, hierarchical clustering, and weight optimization, replacing fixed calendar rebalancing with adaptive rebalancing triggered by regime switches or significant changes in optimal weights.
Trend of this event
Not enough consecutive observations yet; trend is not drawn.