Statistical analysis of gold production in South Africa using ARIMA, VAR and ARNN modelling techniques: Extrapolating future gold production, Resources–Reserves depletion, and Implication on South Africa's gold exploration

Litshedzani Mutele and Emmanuel John M. Carranza

Resources Policy, 2024, vol. 93, issue C

Abstract: South African gold mining is one of the main contributors to the country's economic growth and employment creation. Assessment of future gold production is pivotal for positioning the country's supply security and contribution to the projected global demand. In this study, we modelled and forecasted South African and global gold production trends (1990–2022) using the Autoregressive Integrated Moving Average (ARIMA), Vector Autoregressive (VAR), and Autoregressive Neural Network (ARNN) techniques. The study also evaluated the impact of annual productions on the collated reserves (2008–2021) and the contribution of resource–reserve conversion in sustaining gold reserves within the South African goldfields. The ARIMA (1,1,0) model, based on root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE), outperformed both ARNN (1,1) and VAR (2) methods. The forecast of the ARIMA model showed an increase in South Africa's gold production, with a compounded annual growth rate (CAGR) of ∼0.6%. The model also predicted an increase in global gold production, with CAGR of ∼2.2%. The analyses of reserves–production trends showed that most South African mining companies have been depleting their gold reserves due to production and were minimally converting resources to reserves, averaging at ∼2%/year from 2017 till present. The minimal conversion may also reflect the challenging mining conditions (such as mining depths and underground flooding) and the mining industry's socio-political dynamics. We demonstrated that, if the resource–reserve conversion factor is maintained (at ∼2%/year and above), the forecasted production may sustain the current reserves (at > 2000 t Au excluding tailings) but deplete ∼27 % of the resources to reserves by 2040. However, if the conversion factor continues to decline (to <1 %), the forecasted gold production will reduce the reserves to < 500 t Au with ∼3% of the resources converted to reserves.

Keywords: South Africa; ARIMA; Multivariate VAR; Autoregressive neural network; Gold production; Resource–reserve conversion rate (search for similar items in EconPapers)
Date: 2024
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0301420724004434
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:jrpoli:v:93:y:2024:i:c:s0301420724004434

DOI: 10.1016/j.resourpol.2024.105076

Access Statistics for this article

Resources Policy is currently edited by R. G. Eggert

More articles in Resources Policy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().