Investment portfolio allocation focuses on the allocation of a given capital investment among different assets in the presence of risk, return and liquidity. The literature review discusses the application of Linear Programming (LP) in the above decision making problem ranging from Markowitz’s quadratic model to linear models such as mean absolute deviation (MAD). This includes the LP stock selection of Mbah and Onwukwe (2016) that evaluates five Nigerian stocks based on dividend, risk and liquidity. The published model is reviewed and solved using an LP solver. It reproduces the same choices for Zenith Bank, Fidelity Bank and Diamond Bank but the audit indicates that the reported “amounts invested” and optimal solutions are results of the normalisation in the model, that every LP allocates the entire capital to a single stock, and that the initial choice varies when liquidity is in ₦ billion instead of ₦ trillion. Two budget-constrained LPs are therefore built on the same data: one with a maximum weight per stock, and one that adds a risk ceiling and a liquidity floor. A 40% cap gives an allocation of 40-40-20 to the same three banks in a single LP. The addition of the liquidity floor moves about 7% of capital into the most liquid stock, which also carries the highest risk. The selected set does not change in all 10,000 random tests of up to ±30%, but the ranking of stocks is highly dependent on the defining system of the score. A MAD-based LP could be a proposed extension but it would require return time-series data and has not been solved here. The outcome or results show that LP can be a transparent baseline. Its usefulness depends on constraints, it does not by itself prove investment performance.