Making a Markowitz portfolio with agricultural commodity futures

Abstract

This paper constructs a minimum-variance portfolio of six agricultural futures. We make a full sample analysis as well as a pre-COVID and COVID examination. Using Markowitz portfolio optimisation, we find that soybean futures have the highest share (31%) in the full sample portfolio because it has the lowest variance. Both soybean oil and rice futures have the second highest weight in the full sample portfolio, in an amount of 24%, because soybean oil has the second lowest variance, whereas rice has, by far, the lowest average correlation with other agricultural futures. Soybean oil has the highest share of 35% in the pre-COVID period, whereas rice follows with 27%. On the other hand, in the COVID period, soybean has a very high share in an amount of 47% due to the lowest risk, while rice takes second place with 19%. Based on the results, investors should invest the most in soybean oil and rice in tranquil periods, while the choice should be soybean and rice in crisis periods. Rice is the choice in both sub-periods because rice has a very low correlation with other agricultural commodities, which happens due to the price stabilisation of rice that is often conducted by Asian countries.

Scheduling vegetable sales to supermarkets in Europe: The tomato case

“…We used the classic method for production programming, the mean-variance quadratic equation (M-V) developed by Markowitz (1952). Some studies have applied this formula in order to find efficient crop planning (Gómez-Limón et al 2003;Živkov et al 2022). Other programming models utilized in the farming sector are minimization of total absolute deviation (MOTAD) and target MOTAD (Tauer 1983).…”