Evaluation of financial models for portfolio optimization in Nigeria’s non-interest real estate sector: Integrating mean-variance optimization, the capital asset pricing model, and Value-at-Risk analysis
Keywords:
Financial modeling, Portfolio management, Real estate investment, Sukuk instruments, Risk analysisAbstract
This study evaluates how modern financial-modeling tools can improve investment decision-making in Nigeria's non-interest real estate sector. Secondary data covering 2018--2024 were obtained from the Central Bank of Nigeria (CBN), Nigerian Exchange Group (NGX), and Real Estate Investment Trusts (REITs). Three models were applied within an Islamic-finance framework: Mean-Variance Optimization (MVO), the Capital Asset Pricing Model (CAPM), and Value-at-Risk (VaR). The empirical results indicate that appropriately modeled non-interest portfolios outperformed conventional benchmarks. On average, optimized portfolios achieved a 17.6% improvement in expected return and a 12.9% reduction in risk. The inclusion of Sukuk-backed instruments significantly stabilized portfolio returns, contributing to an overall 12.9% reduction in portfolio risk and lower drawdown compared to conventional benchmarks. Regression and sensitivity analyses yielded R2 = 0.88 and p < 0.05, indicating strong explanatory power. The study concludes that structured financial modeling can support ethical investment and long-term wealth creation in Nigeria's real estate sector. It recommends wider use of data-driven investment tools, gradual adoption of advanced analytical methods such as machine learning, and stronger coordination among the CBN, Securities and Exchange Commission (SEC), and National Insurance Commission (NAICOM) to improve transparency, regulation, and portfolio management in Nigeria's expanding non-interest financial system.
Published
How to Cite
Issue
Section
Copyright (c) 2026 Jeremiah Adamu Rikaya, Hannatu Lucius Datukun, Pheobe Oine Rikaya, Job Olanipekun Israel (Author)

This work is licensed under a Creative Commons Attribution 4.0 International License.