Abstract
Physical cash has, and continues to, play an important role in payment history for several reasons. It continues to facilitate financial transactions, closely linked to economic activity, symbolises sovereignty, and broadly recognised as a medium of exchange, provides people with liquidity, is used as a crucial tool for financial inclusion, and is used for emergency preparedness in cases where electronic payments may not be available (natural disasters, power outages, etc.). It also offers privacy and anonymity and is used as a store of value. Cash has remained a viable, resilient and essential mode of payment globally, even in the face of considerable improvements in the digital era.
The importance of cash to the economies of the world necessitates a close understanding of the determinants of its demand, especially to key agents like Central Banks which are at the core of the cash supply chain. The study aimed to providing insights into cash demand that will assist central banks in making informed decisions about their monetary supply and currency management strategies.
This study compared the findings with existing empirical research and extended the model specification and analysis within the context of South African market conditions. To attain the study's objectives, the Vector Error Correction Model (VECM), Impulse Response Function (IRF), and additional econometric techniques were employed. Furthermore, various forecasting methods were utilised to estimate cash demand effectively.
The results indicated the acute impact of payment technology on cash demand in South Africa. The results demonstrate a distinct shift towards alternative payment methods, with mobile penetration playing a significant role in reducing cash demand. This underscores the rising prominence of mobile banking in financial transactions. Traditional cash channels, automated teller machines (ATMs) and bank branches were seen to marginally increase cash demand. Moreover, the impact of electronic fund transfers (EFT’s) has become highly significant, indicating that digital transactions are increasingly supplanting traditional cash usage. While gross domestic product (GDP), interest, and tax ratios continue to affect cash demand, the emphasis is increasingly moving towards digital and mobile financial platforms.
The estimation analysis of cash demand revealed significant shifts in the ranking of various forecasting techniques when comparing short- and long-term horizons, highlighting the evolving effectiveness of these methods over time. Neural network models, particularly Neural
v
Network Time Series Forecast (NNETAR), emerged as the best performers, demonstrating their ability to capture long-term trends and nonlinear relationships – crucial in complex data environments. In contrast, models like the VECM and Exponential Smoothing underperformed, likely due to their reliance on linear relationships, making them less adaptable to structural shifts over extended periods.
The empirical findings hold important implications for the South African Reserve Bank’s (SARB) monetary policy and currency management strategies. The findings revealed complex relationships between cash demand and key macroeconomic indicators such as GDP growth, inflation, and the rise of digital financial services. In response, SARB must adapt its monetary frameworks to account for these evolving dynamics, particularly the reduced need for physical currency as digital transactions continue to grow.
This shift suggests the need for SARB to reconsider its approach to currency production and distribution. The ability to accurately forecast cash demand is crucial for managing currency fluctuations and maintaining confidence in the South African financial system. By identifying long-term drivers of cash demand, this research provides valuable insights that could inform and enhance SARB's currency management policies moving forward.