Asset allocation

This paper investigates whether longer historical training periods improve the out-of-sample performance of a Hidden Markov Model (HMM) used for identifying financial market regimes. Four four-state Gaussian HMMs are compared, differing only in the starting point of their training data (1970, 1980, 1990, and 2000), while being evaluated on the same test period. The model uses five market-based features capturing volatility, drawdowns, credit conditions, and sector rotation to identify latent risk regimes.

The study highlights that validation methodology and evaluation criteria are as important as model design and data length when assessing regime-based investment models.

Diversification is a key principle in portfolio construction, yet equal-weight portfolios often fail to deliver true risk diversification. This study shows that capital-based allocation can mask strong concentration in a small number of underlying risk factors. We analyze a simple multi-asset portfolio of ten ETFs spanning equities, bonds, commodities, credit, private equity, and Bitcoin. Despite equal weights, risk is highly concentrated in a few volatile assets and amplified by strong cross-asset correlations, particularly within equity and credit markets. Risk parity reduces concentration by balancing risk contributions and improves risk-adjusted performance, though at the cost of lower returns. Further improvement is achieved through clustering-based allocation, which groups similar assets and allocates risk across more independent sources of return. The results demonstrate that effective diversification depends on the structure of risk factors rather than the number of assets or equal capital weights.

Commodities represent a vital but highly volatile asset class, characterized by pronounced cyclicality, lack of yield, and susceptibility to severe macroeconomic drawdowns. While cross-sectional (relative) momentum is a well-documented anomaly, its application in commodities often forces portfolios to hold the “least declining” assets during broad-based bear markets, resulting in unacceptable tail-risk. This study empirically evaluates the efficacy of a Dual Momentum framework—combining relative strength ranking with an absolute time-series trend filter—applied to a diversified suite of commodity sector ETFs (DBA, DBB, DBE, DBP) from 2007 to 2026. We demonstrate that while pure relative momentum exhibits high parameter sensitivity and inconsistent benchmark outperformance, the inclusion of an absolute momentum filter structurally mitigates drawdowns and universally outperforms a static, equally weighted benchmark across all tested parameter combinations. The findings suggest that Dual Momentum provides a robust, parameter-agnostic framework for portfolio managers seeking tactical commodity exposure with superior risk-adjusted return profiles.

Our study explores a weekly-rebalanced dual-momentum-based Global Tactical Asset Allocation (GTAA) strategy applied to a diversified set of ETFs. The strategy selects assets based on relative momentum and applies an absolute momentum filter to avoid declining investments. Ultimately, a single combined strategy was created by merging two sub-strategies, incorporating both shorter- and longer-term momentum signals. Backtesting over an extended period demonstrates that this approach delivers attractive risk-adjusted returns, achieving attractive Sharpe and Calmar ratios, while maintaining lower drawdowns compared to a simple equally weighted benchmark.

For U.S. equities, fixed income, and commodities, reconstructing long-term historical datasets is relatively straightforward, and we have already explored these challenges in several previous studies, including 100 Years of Multi-Asset Trend Following, Extending Historical Daily Bond Data to 100 Years, and Extending Historical Daily Commodities Data to 100 Years. Moreover, the broader methodology of reconstructing missing market histories shares many similarities with the techniques discussed in How to Replicate Any Portfolio. Emerging markets, however, represent a particularly interesting opportunity for historical reconstruction, as reliable long-term data is often unavailable for much of the 20th century despite the growing importance of these markets in modern portfolio construction and asset allocation. In this article, we present the framework we developed to extend emerging market histories in a consistent and economically meaningful way, enabling more robust long-term quantitative research and modelling.

Most systematic trading strategies are pro cyclical by nature. They perform best when markets trend higher and volatility remains contained. During broad market expansions, equity risk premia, momentum and trend following approaches tend to generate stable positive returns.

This analysis demonstrates how crisis hedge strategies can be identified, evaluated and integrated into a model portfolio using the Quantpedia Pro framework.