Portfolio Management Using the Black–Litterman Model and Incorporating Investor Views

Abstract

Portfolio management has consistently faced the fundamental challenge of uncertainty in estimating asset returns and risks, which has limited the practical effectiveness of classical optimization approaches, particularly the Markowitz mean–variance model. In this context, the Black–Litterman model, as a Bayesian framework, enables the derivation of more stable estimates that are consistent with market economic logic by combining market equilibrium returns with investor views. The primary objective of the present study is to propose a structured framework for portfolio management using the Black–Litterman model with the systematic incorporation of investor views grounded in fundamental analysis and asset pricing factors. From a methodological perspective, this research is applied–developmental and quantitative, and it is empirically conducted using daily stock data of manufacturing firms listed on the Tehran Stock Exchange over the period 2015 to 2023. Rather than relying on subjective judgments, investor views are extracted based on the results of regressions from the Fama–French three-factor model and are incorporated into the Black–Litterman framework in the form of relative views. After computing the implied equilibrium returns and incorporating the views, posterior Black–Litterman returns are derived, and optimal portfolio weights are determined through mean–variance optimization. The performance of the resulting portfolio is compared with that of the classical Markowitz portfolio and the baseline Black–Litterman model using return, risk, and risk-adjusted performance measures. The findings indicate that the implied equilibrium returns obtained from the Black–Litterman model are generally more conservative than historical averages and help prevent overfitting to past data. The empirical results show that incorporating Fama–French–based views leads to a significant increase in cumulative returns, improvements in the Sharpe and Sortino ratios, and reductions in the standard deviation and maximum drawdown of the portfolio compared with both the Markowitz approach and the Black–Litterman model without views. Moreover, sensitivity analysis with respect to the τ parameter demonstrates that selecting intermediate values of this parameter can establish an appropriate balance between market information and investor views. Overall, the results confirm that the structured integration of Fama–French factor analysis with the Bayesian Black–Litterman framework enhances the stability of asset allocation and improves the risk-adjusted performance of portfolios, and can therefore serve as a practical and reliable framework for investment managers in volatile and emerging markets.