Jul 2026· ZERO Jurnal Sains Matematika dan Terapan· 0 citations· 26 references
TL;DR
Findings indicate that combining machine-learning-based predictive modelling with adaptive, regime-driven allocation enhances portfolio stability, mitigates extreme losses, and improves risk-return efficiency under dynamic emerging-market conditions.
Abstract
This study proposes an adaptive portfolio optimization framework that integrates Random Forest(RF)-based return forecasting into a Mean-Variance-Forecast Error (MVF) model, augmented by a Hidden Markov Model (HMM) for market regime identification. Using weekly historical return data from 40 LQ45-listed stocks spanning January 2014 to January 2025, the framework dynamically adjusts portfolio allocations in response to bull and bear market conditions detected by a two-state HMM. The primary methodological contribution lies in addressing the static limitation of conventional MVF under shifting market regimes. Out-of-sample evaluation over a 138-week test period demonstrates that regime-switching MVF achieves Sharpe ratios above 1.30, substantially lower maximum drawdowns than the MVF-only portfolio, and cumulative returns of 291.96%. Bootstrap-validated 95% confidence intervals confirm the statistical robustness of these improvements. Nevertheless, portfolio turnover remains high during active reallocations. These findings indicate that combining machine-learning-based predictive modelling with adaptive, regime-driven allocation enhances portfolio stability, mitigates extreme losses, and improves risk-return efficiency under dynamic emerging-market conditions.
This paper explores the effectiveness of integrating Artificial Neural Network (ANN) based return forecasts into three portfolio optimization frameworks: Equal-Weight (EW), Mean-Variance (MV), and Black-Litterman (BL), under varying market regimes, using 67 stocks listed on Thailand’s Market for Alternative Investment (MAI) as a case study. Portfolios are constructed using ANN predictions and tested across pre-COVID stable conditions (2019) and a volatile period (2020–2024), with an additional evaluation of rebalanced portfolios on 2024 data. Results indicate that BL achieves the highest risk-adjusted returns under stable conditions, while EW outperforms significantly during high volatility, driven largely by outlier stock performance during the COVID-19 recovery. Rebalancing with updated ANN forecasts did not guarantee improved performance, highlighting both the promise and limitations of machine learning-enhanced optimization in emerging markets.
Pema Yangchen, Rujira Chaysiri· International Conference on...· 0 citations
Weekly macro-financial and financial market data, combined with machine learning methods, offer new possibilities for identifying latent economic states in real time, but the portfolio value of regime detection depends critically on how detected states are translated into allocation rules. This paper develops and evaluates a data-driven macro-financial framework that combines a four-state Gaussian Hidden Markov Model (HMM), estimated on eight weekly macro-financial features, with Conditional Value-at-Risk (CVaR) portfolio optimization across European multi-asset portfolios from January 2000 to April 2026. Using a strictly out-of-sample walk-forward design, we show that naive regime-conditional CVaR allocation generates excessive turnover (approximately 226% per year) that erodes net performance below a simple benchmark under any realistic transaction cost, whereas implementation-aware alternatives recover the gap substantially: regime-constrained weight bands attain a net Sharpe ratio within 0.009 of the static benchmark at roughly 29% annual turnover. Expanding the universe to include sovereign bonds improves drawdown control but introduces duration risk that materializes in rate-hiking episodes. These findings demonstrate that, in data-driven macro-financial systems, the bottleneck is not regime detection but transparent, stable, and cost-aware decision-rule design, with implications for next-generation, AI-assisted macro-financial monitoring and policy surveillance systems.
Jorge Grube Martín-Lunas, Ana Lazcano, Julio E. Sandubete· Economies· 0 citations
Portfolio construction aims to balance expected return and risk through effective asset allocation. This study proposes a portfolio formation framework that integrates machine learning-based return prediction with Markowitz mean–variance portfolio optimization. Random Forest, XGBoost, Multilayer Perceptron, and Support Vector Regression models are employed to predict the cross-sectional excess returns of stocks using financial indicators derived from technical and macroeconomic variables. These predictions are incorporated into the portfolio optimization process to determine portfolio weights. The resulting strategies are evaluated against benchmark portfolios including an equal-weighted portfolio and the BIST 100 index. Empirical results show that machine learning-based stock selection improves portfolio performance. In particular, the XGBoost-based portfolio achieves the best results with an annual return of 75.30% and a Sharpe ratio of 1.80. SHAP analysis further indicates that momentum and price-based technical indicators play a dominant role in model predictions.
Mustafa Etcil, Hüseyin Akkaş, Burak Kolukısa et al.· Signal Processing and Commun...· 0 citations
This paper proposes an Exponentially Weighted Moving Mean–Variance (EMMV) model that discounts older data with a forgetting factor. Unlike traditional asset-level smoothing, the EMMV aggregates portfolio-level moments across rolling windows, yielding stable, adaptive allocations. Hyperparameters are selected objectively via a genetic algorithm with walk-forward cross-validation. Using 30 U.S. stocks from 2010 to 2025, the model achieves a buy-and-hold Sharpe ratio of 1.49 — more than double that of the equally weighted portfolio — and a rebalanced Sharpe ratio of 0.69 after transaction costs, outperforming all benchmarks. Robustness analysis confirms consistently high performance over a wide parameter range, and bootstrap tests verify significant or near-significant improvements. The framework provides a practical, data-driven tool for dynamic portfolio selection in nonstationary markets.
Kumsong Jo, Sunhak Kim, Bongnam Ri· International Journal of Fin...· 0 citations
This study evaluates the out-of-sample forecasting ability of six model types: ARIMA, VAR, Random Forest, XGBoost, LSTM, and GRU, for monthly Canadian inflation from January 2012 to April 2026 (n = 172). The evaluation employs expanding-window walk-forward validation across 1-, 3-, 6-, and 12-month horizons. Results reveal a horizon-dependent shift: ARIMA significantly outperforms all machine learning and deep learning models at the one-month horizon (Diebold-Mariano p<0.05). However, Random Forest and XGBoost become notably superior at six and twelve months, reducing RMSE by approximately 30-75 percent compared to ARIMA and VAR. LSTM and GRU perform well only at the shortest horizon, likely due to overfitting given the limited data. Analyzing four macroeconomic sub-periods shows that no single model consistently dominates. SHAP analysis of the top-performing XGBoost model indicates that lagged inflation is more influential than unemployment, which only becomes significantly impactful during the pandemic tail. The findings clarify when machine learning methods can surpass traditional benchmarks.
The rapid growth of retail investors in Indonesia, from 2.48 million in 2019 to over 20 million by 2025, underscores an urgent need for empirically grounded portfolio optimization frameworks adoptable into practical tools such as robo-advisory systems. This study applies the Markowitz Mean-Variance model to construct and evaluate an optimal stock portfolio from the LQ45 index over January 2022 to December 2025, a post-pandemic period characterized by predominantly adverse risk-adjusted returns. Using daily closing price data for 27 consistently listed LQ45 stocks (958 trading days) from the Indonesia Stock Exchange (IDX), with sample consistency verified through official BEI constituent evaluation announcements, this study constitutes an ex-post empirical analysis designed to isolate the mathematical efficacy of the Mean-Variance model under adverse market conditions. The analysis encompasses individual return and risk profiling, construction of a 27x27 covariance matrix, efficient frontier derivation via constrained quadratic optimization using Microsoft Excel Solver (GRG Nonlinear), and Sharpe ratio-based performance evaluation against a risk-free rate of 5.3281% per annum (average BI Rate 2022-2025). The core finding is that 19 of 27 sample stocks (70.4%) generated negative Sharpe ratios, confirming the inadequacy of undiversified single-stock strategies in adverse markets. In contrast, the tangency portfolio achieved a Sharpe ratio that materially surpasses all 27 individual stocks, establishing that quantitative portfolio optimization delivers superior risk-adjusted outcomes precisely when markets are most challenging. These results validate the enduring relevance of Modern Portfolio Theory in Indonesia's capital market and provide a transparent, replicable framework for investors and practitioners.
Irfan Andi Pramudya, Intan Shaferi· The International Conference...· 0 citations
