Macroeconomic indicators and the Johannesburg Stock Exchange Financial 15 Index nexus in South Africa

Instability in stock markets has been linked to major economic crises and remains a central concern in contemporary macro-financial analysis (Demir, 2019). In developing economies, stock prices often react strongly to movements in key macroeconomic indicators such as exchange rates, gross domestic product and trade flows (Panta, 2020). South Africa is no exception. The Johannesburg Stock Exchange, founded in 1887, is the largest in Africa and ranks around seventeenth globally (JSE, n.d.). Within this market, the JSE Financial 15 Index tracks fifteen highly traded banking, insurance and broader financial services firms that form the backbone of financial intermediation. Over the past year, this index rose by about 26.31% and traded between roughly R16 974.57 and R27 806.87, illustrating both growth and volatility in a core segment of the financial system (Financial Times, n.d.). Because financial institutions are sensitive to inflation trends, interest rate movements, economic growth and currency fluctuations, understanding how these macroeconomic forces affect the JSE Financial 15 Index is relevant for financial stability and inclusive growth.

Even though the JSE Financial 15 Index is a specific services-related sector, how financial institutions perform is linked to the aggregate macroeconomic conditions. Therefore, macroeconomic variables such as inflation, interest rates, exchange rates and real GDP play a crucial role when it comes to the demand for credit, performance of loans, as well as investment behaviour and capital flows. These are important for financial firms, especially for profitability and company valuations.

The nexus between macroeconomic indicators and the JSE Financial 15 Index is closely linked to the Sustainable Development Goals because the index reflects the performance of major financial institutions that support intermediation, credit allocation, savings mobilisation, investment financing and risk management in South Africa (FSDO, 2024). A resilient financial sector contributes to SDG 8 by promoting inclusive growth, employment and decent work, as highlighted in the 2030 Agenda for Sustainable Development (Sayed, 2015; Frey, 2018). Financial sector development also improves access to payments, credit, savings and insurance, thereby supporting poverty reduction, private-sector expansion, job creation and economic resilience (World Bank Group, 2013, 2016). The nexus further relates to SDG 9 through the mobilisation of long-term finance for infrastructure, innovation and industrial development, and to SDG 16 through credible monetary policy, exchange rate stability, effective regulation and transparent capital-market governance, which strengthen institutional trust, investor confidence and macro-financial stability (United Nations Conference on Trade and Development (UNCTAD), 2017). Therefore, analyzing how inflation, interest rates, exchange rates and real GDP affect the JSE Financial 15 Index provides evidence on whether South Africa's financial sector can support sustainable development through stable capital markets, inclusive finance and resilient economic growth. Table 1 presents the macroeconomic indicators, the JSE Financial 15 Index, and the sustainable development channels as they relate to the present study.

Existing empirical work shows that exchange rates, output, inflation and interest rates influence stock prices, but results are often mixed and depend on country group, period and modelling choices. For exchange rates, Areli Bermudez Delgado et al. (2018) and Chang et al. (2019) report a negative and significant long-run impact on stock market indices, consistent with the idea that currency depreciation raises imported input costs and compresses profits (Bahmani Oskooee and Saha, 2015). At the same time, Megaravali and Sampagnaro (2018) and Chang et al. (2019) find a positive short-run relation, while Ogunsakin and Awe (2020) report an inverse and significant effect in the short run. For output, Humpe et al. (2025) and Ogunsakin and Awe (2020) show that real GDP is positively related to stock prices. These findings suggests that stronger economic growth is associated with higher stock prices.

Inflation and interest rate effects are similarly ambiguous. Humpe et al. (2025) find that consumer prices are negatively related to stock prices in Anglosphere economies but positively related in BRICS economies, where equities may hedge against inflation. Chang et al. (2019), Giri and Pooja (2017) and Keswani et al. (2024) also document that consumer prices tend to undermine stock performance in the short run, although some evidence points to a positive relation in the long run. For interest rates, the Fisher effect suggests that nominal rates and expected inflation move together, so higher rates are usually associated with lower stock prices. Chang et al. (2019), Tursoy (2019), Bhuiyan and Chowdhury (2020) and Keswani et al. (2024) find an inverse association between interest rates and stock prices, and Ramatlo (2019) shows that unexpected repo rate increases reduce JSE All Share Index returns, although Bhuiyan and Chowdhury (2020) also report cases with no long run relationship. Overall, the literature confirms that macroeconomic indicators matter for stock prices but does not provide a clear message on the magnitude or sign for financial sector indices in South Africa.

Most South African studies have focused on broad market indices. Hackland (2015) finds that higher GDP growth raises the JSE Financial 15 Index while prime lending rate shocks reduce it, but the sample ends in 2014 and does not capture the post COVID-19 policy cycle. More recent work by Mokoena (2023) and Maake et al. (2024) concentrates on the JSE All Share Index or Top 40 indices, leaving sector-specific dynamics largely unexplored. In addition, many studies do not explicitly allow for structural breaks associated with episodes such as the global financial crisis, load shedding-related growth shocks or the COVID period, even though such events can alter macro financial relationships over time. This creates a practical problem for investors, portfolio managers and policymakers who need reliable guidance for asset allocation, risk management and the design of monetary and exchange rate interventions.

Motivated by these gaps, this study examines the relationship between macroeconomic indicators and the JSE Financial 15 Index in South Africa using quarterly data from 2004 to 2024. The main research questions are, first, what is the causal nexus between interest rates, inflation, GDP growth and exchange rate movements and the JSE Financial 15 Index and, second, what are the short-run and long-run effects of these macroeconomic indicators when estimated through a Vector Error Correction Model (VECM)? Figure 1 documents that consumer prices, GDP, the exchange rate and the JSE Financial 15 Index all follow rising trends over the sample, with visible breaks around 2008 quarter four, 2020 quarter one and 2020 quarter three. These visual patterns motivate formal tests for structural breaks and the explicit incorporation of break dummies into the econometric framework.

The study contributes by focusing on a systemically important sector index rather than an aggregate market index, by extending the sample to 2024 to cover multiple policy rate cycles, load shedding episodes and the COVID shock, and by jointly modelling long-run cointegration and short-run adjustment dynamics with structural breaks. The evidence is expected to support more informed investment strategies for institutional and retail investors, while providing the South African Reserve Bank, National Treasury and other authorities with practical insight on how interest rate decisions and exchange rate developments are transmitted into the valuations of major financial institutions that underpin growth and progress toward the Sustainable Development Goals.

The purpose of this literature review is to examine existing research on the relationship between macroeconomic indicators and how the stock market performs by focusing on the JSE Financial 15 index, covering articles and empirical studies published between 2015 and 2025. Unlike the resources and industrial sectors, which are mainly affected by global commodity prices and overseas demand, the JSE15 is more sensitive to what is happening in the local economy, changes in interest rates and regulations. However, exchange rate fluctuations also play an important role, especially because they influence the cost of foreign borrowing, the value of investments abroad and capital flows.

All these factors give the financial sector a unique mix of risks and opportunities compared to other parts of the market. Macroeconomic variables affect the financial sector in several ways, through the monetary policy transmission mechanism, real economic activity, as well as global factors. These mechanisms can influence the profitability of firms, expectations of investors and future cash flow valuations, which can influence stock market movements. Monetary conditions can influence stock prices, particularly through interest rates and inflation. Interest rates and inflation can affect the cost of borrowing, discount rates and real returns. Stock price movements can also be affected by real economic activity through corporate earnings changes, while exchange rates reflect external shocks and capital flows.

Monetary variables, particularly, interest rates and inflation, are essential for explaining stock market movements. According to the Fisher's effect, nominal interest rates and expected inflation tend to move together. As a result, interest rates and stock prices are generally expected to be inversely related due to the impact of inflation. Empirical studies largely support this inverse relationship between interest rates and stock prices. For instance, a study conducted by Chang et al. (2019) has indicated that the nominal interest rate significantly affects stock prices and that they are negatively related in the long run. Similarly, Ramatlo (2019) found that an unexpected rise in the repo rate is estimated to reduce the JSE All Share Index returns. Tursoy (2019), Bhuiyan and Chowdhury (2020), Keswani et al. (2024) and Rizk and Challita (2026) also discovered that the interest rates and stock prices are inversely related. However, Rizk and Challita (2026) also discovered that interest rates and stock prices have a positive relationship in the short run.

Humpe et al. (2025) have found that CPI, which is a measure of inflation, is negatively related to stock prices for Anglosphere countries in the long run. However, they also found that CPI and stock prices have a positive relationship in BRICS countries as this may help protect investors' money from losing value over time in response to economic instability these countries may have faced in the past. Similarly, Rizk and Challita (2026) found that inflation has a positive but statistically insignificant effect on stock prices in both the short and long run. This highlights the difference between macroeconomic conditions in developed and developing countries. In contrast to Chang et al. (2019), Giri and Pooja (2017) as well as Keswani et al. (2024) who discovered that the performance of the stock market is negatively affected by CPI in the short run.

Exchange rates can influence stock prices both positively and negatively due to the reduction in import input costs, which enhances profits when the domestic currency appreciates, while a domestic depreciation results in the rise of import input costs, which reduces profits for domestic producers (Bahmani-Oskooee and Saha, 2015). For instance, Chang et al. (2019) and Areli Bermudez Delgado et al. (2018) results indicate that the exchange rate has an inverse and statistically significant impact on the stock market index in the long run using monthly nominal exchange rate and quarterly data, respectively. However, Megaravali and Sampagnaro (2018) and Chang et al. (2019) found that the exchange rate and stock prices are positively related in the long run and short run using monthly and quarterly data, respectively. While Ogunsakin and Awe (2020) findings indicate that the exchange rate has an inverse and statistically significant influence on the performance of the stock market in the short run.

Real economic activity represents another important channel that links macroeconomic factors to stock market performance. GDP can affect stock prices both positively and negatively. This is because an increase in GDP is normally due to an increase in consumption spending, capital formation, among other components, which can lead to investors expecting high corporate revenue and earnings in the future and therefore increasing stock prices. However, GDP can negatively affect stock prices because strong GDP growth can increase inflation due to demand-pull inflation. This high inflation results in lower stock prices due to higher uncertainty and expected discount rates (Humpe et al., 2025).

Humpe et al. (2025), when there is high economic growth, it usually means that companies are earning more and their value has increased. Humpe et al. (2025) found that real GDP significantly affects stock prices in Anglosphere countries and has a positive relationship using quarterly data from 1995 to 2023. Similar to Ogunsakin and Awe (2020), who found a direct relationship in both short-term and long-term periods. This shows that economic growth is important for improving how the stock market can perform. However, Rizk and Challita (2026) have found that there is an inverse relationship between economic growth and stock prices.

The nexus between stock prices and various macroeconomic variables has been explored by previous studies. However, some results contradict each other. Whether the variables are directly or inversely correlated depends on a number of factors, such as whether the data is for developed or developing economies, whether structural breaks are considered, as well as the short- or long-term periods. Some of these studies have looked at the overall indices like the JSE All Share Index, and limited studies have looked into sectoral indices. This research seeks to fill this gap by analyzing the relationship between macroeconomic indicators and the JSE15 index using a data set from 2001Q1 to 2024Q4. It considers structural breaks such as the COVID-19 shock. Table 2 presents the summary of empirical studies.

The purpose of this methodology section is to explore the data as well as the research methodology that this study utilizes to examine the relationship between the JSE Financial 15 index and the macroeconomic indicators.

The following quarterly data from 2004Q1 to 2024Q4 is utilized due to availability of data; the dependent variable, long-run JSE Financial 15 index (LJSE15) was accessed from Reuters, independent variables such as inflation using Consumer Price Index (CPI) as its proxy, Repo rate, a proxy for interest rate (INT) as well as the exchange rate (EXR) data for ZAR against USD are all accessed from EasyData by Quantec. Lastly, real Gross Domestic Product (GDP) data are accessed from SARB.

Previous studies such as those by Chang et al. (2019) as well as Giri and Pooja (2017) have found that inflation and stock prices have an inverse short-run relationship. Bhuiyan and Chowdhury (2020) and Tursoy (2019) found that, in the long run, the S&P 500 index, all sector indices and long-term interest rates are inversely related. Areli Bermudez Delgado et al. (2018) results indicate that the exchange rate is statistically significant and negatively related to the stock market index. Humpe et al. (2025) and Ogunsakin and Awe (2020) found that stock prices and GDP are positively related. We therefore expect CPI, INT and EXR to have a negative relationship with the LJSE15, while GDP to be positively related to the index.

The justification for including the independent variables in the model:

Inflation (CPI): Can affect the stock markets through interest rates, as they are used by the central bank, such as the South African Reserve Bank, to control inflation. A continuous increase in inflation can erode real returns and also negatively affect investor confidence in financial markets.

Economic growth (GDP): Is an essential macroeconomic indicator that shows the overall performance of economic activity. When GDP increases, it means that there is more consumer spending, an increase in demand for goods and services or an increase in capital formation, which can lead to investors expecting high corporate revenue and profit in the future. As a result, this tends to increase stock prices as stock prices rely on the current value of the expected future earnings.

Interest rate (repo rate): Is set by the Reserve Bank. When they increase the interest rate (repo rate), the prime rate also increases, affecting consumer spending and investment spending. Increases in interest rates tend to result in reductions in corporate earnings expectations and asset valuation, which can then affect stock prices.

Exchange rate (EXR) movements: Can influence trade balances, import and export prices as well as foreign investor behavior. For example, when the Rand depreciates against the US Dollar, it can reduce investment in domestic assets as well as increase exports as goods become cheaper for foreign buyers.

These variables were selected due to their theoretical and empirical relevance in explaining aggregate stock market movements. Firm-specific variables such as asset size or market share were not included as control variables because the main analysis of this study is a sectoral stock market index rather than individual institutions. The focus of this study is therefore on macroeconomic influences operating at the aggregate level rather than micro-level institutional influences.

Theoretically, financial markets are affected by macroeconomic fundamentals proposed by various asset pricing models such as the Arbitrage Pricing Theory. It provides a theoretical framework that explains the link between macroeconomic fundamentals and stock prices (Fromentin, 2022). According to Bhuiyan and Chowdhury (2020), some studies using Arbitrage Pricing Theory have found that even though stock returns and factors such as changes in the term structure, industrial production and shifts in the risk premium are positively related, anticipated and unexpected inflation showed a negative relationship with the returns of stocks mainly due to how they negatively affect expected dividends and the discount rate.

The appropriate model for analyzing the nexus among the variables depends on their order of integration and whether cointegration exists. VECM is used as all variables are integrated of order one and cointegration is confirmed using the Johansen cointegration test. VECM not only explores the short- and long-run relationships but it also corrects any deviations caused by shocks in the short run from the variables' state of long-run equilibrium state (Shao et al., 2021; Xuan, 2025). The following model is used:

Where:

1. $Yt$ is the level vector of endogenous variables:

• (2)

  $X=αβ′$⁠, where:

• •

  $β′$ contains the cointegration vectors.

• •

  $α$ contains the adjustment coefficients

• (3)

  $Ki$ are the short-run dynamics

• (4)

  $εt$ is the error term

A unit root test is performed in order to see if a time series is stationary or not (Fame et al., 2019). By conducting a unit root test, it can assist in making sure that the data can be used for forecasting and modelling. According to Moniz and Kantar (2024), data that is not stationary can lead to spurious regression results. There are different unit root tests that can be utilized, such as the Kwiatkowsi–Philips–Schmidt–Shin (KPSS) test, Phillips–Perron (PP) test as well as the Augmented Dickey-Fuller (ADF) test. This study uses the ADF test to test for stationarity and the PP and KPSS tests for robustness. The following ADF model is used:

Where:

1. $Δyt=yt−yt−1$⁠: First difference of the series

2. $∑i=1pβiΔyt−1$⁠: Lagged difference terms to control for autocorrelation

3. $γ$⁠: The main coefficient tested (unit root test)

4. $a0$⁠: Intercept (constant)

5. $a2t$⁠: The time trend

6. $εt$⁠: White noise error term

$H0$⁠: $γ=0$ (Variable has a unit root.)

$H1$⁠: $γ≠0$ (Variable does not have a unit root.)

This study makes use of the Johansen Cointegration test. This test tests whether there is a long-run equilibrium relationship between the variables (Xuan, 2025).

According to Parab and Reddy (2020), results may be deteriorated by structural breaks, as they are unexpected changes that can occur in time series data. Therefore, including structural breaks in time series is important in order to indicate changes in the behaviour of the data and ignoring them can lead to misleading conclusions as well as biased estimates. There are different tests that can be conducted for structural breaks, such as the Chow test, Quant Andrews and the Bai–Perron tests. According to Bhattachrjee and Das (2023), the Bai–Perron test can test for multiple breakpoints in the data. To account for potential structural changes over the sample period 2004–2024, the Bai-Perron multiple breakpoints test was applied. Once the breakpoints were identified, they were then incorporated into the VECM as dummy variables to improve model robustness and control regime shifts.

This study uses the pairwise Granger causality test in order to establish whether, in the short run there exists causality between the variables. This test is used to determine if one variable can be used to predict the other variable (Bhattacharjee and Das, 2023).

For robustness and reliability of the estimated VECM, a series of post-estimation diagnostic tests are conducted:

1. Serial correlation test: The Breusch-Godfrey LM test is utilized to detect whether there is autocorrelation in residuals.

2. Heteroscedasticity test: This test shows that the model is well-defined, suggesting that the performance variance of the dependent variable is well-defined (Keswani et al., 2024). The White test is applied.

3. Residual normality test: The Jarque–Bera test is conducted to assess whether the residuals are normally distributed.

The purpose of this results and discussion section is to explore and discuss the results of this study. To explore the relationship between the JSE Financial 15 index and the macroeconomic variables, an econometric model such as the VECM was used. The results and discussion of the results are based on the descriptive statistics, unit root tests, Structural break test, Johansen cointegration test, VECM, Granger Causality test, as well as the Diagnostics.

Descriptive statistics have been conducted to evaluate the properties of the variables. The results in Table 3 indicates that this data has 84 observations. The highest and lowest means are indicated by CPI (62.7726) and INT (6.7579), respectively. LJSE15 and LGDP are both negatively skewed, suggesting that the observations' mean is not greater than the median, while the other variables are positively skewed. The Jarque–Bera test indicates that LJSE15, CPI and LGDP are normally distributed, while CPI and INT are not normally distributed.

The ADF test in Table 4 indicates that, at levels, we fail to reject the null hypothesis that CPI has a unit root; however, at first difference we reject the null hypothesis as it has a p-value that is less than 0.05 and conclude that it is integrated of order 1. These results are the same for EXR, log of GDP (LGDP), INT, and the log of JSE15 (LJSE15), which indicates that their p-values are greater than 0.05 at levels and less than 0.05 at first difference, and concludes that they are all integrated of order 1. These results are consistent with the PP test, which also shows that all the variables are integrated of order 1. However, the KPSS test results indicate that only CPI, LGDP, and LJSE15 are integrated of order 1, while EXR and INT are integrated of order 0. Given that ADF and PP tests share the null hypothesis of a unit root, the variables are all treated as integrated of order 1. This decision is further supported by the stationarity of the first differences.

The multiple breakpoint tests in Table 5 Panel A indicate that the break dates are 2012Q4 and 2020Q1. The Chow test performed in Panel B indicated that the F-statistics for both dates are statistically significant; therefore, we cannot accept the null hypothesis that there are no breaks at these specified breakpoints. Dummy variables for the structural break dates were created and included in the VECM.

Before conducting the Johansen cointegration test, a lag length selection test was conducted and according to Table 6, AIC indicates that the optimal lag is 5. Table 7 results of the cointegration test indicate that the Trace test shows that there are about 3 cointegrating equations that exist in this study. The test fails to accept the hypothesis that there are no cointegrating equations with the p value that is less than 0.05. The test also did not accept the hypothesis that there are at most 1 and 2 cointegrating equations, indicating that there exist 3 cointegrating equations.

Table 8 panel A shows that in the long run, a CPI and INT's 1% increase will result in a 4% and 0% increase in LJSE15, respectively. This implies that there is a positive relation between CPI and INT and LJSE15, respectively. The results for CPI are consistent with those of Giri and Pooja (2017) and Chang et al. (2019). These results, however, contradict the Fisher's effect and the findings of Tursoy (2019), Bhuiyan and Chowdhury (2020) and Keswani et al. (2024) because, according to this effect, expected inflation and nominal interest rates tend to move together; as a result, interest rates are generally expected to have a negative correlation with stock prices due to the impact of inflation. On the other hand, a 1% increase in EXR and LGDP results in a 1% and 2% reduction in LJSE15, respectively, suggesting that the exchange rate and real GDP are inversely related to LJSE15. The results for EXR are consistent with Keswani et al. (2024) and Chang et al. (2019), while Ogunsakin and Awe (2020) and Keswani et al. (2024) results indicate a direct long-term relationship between economic growth and stock returns.

In panel B, the speed of adjustment or the error correction term (CointEq1) is negative and statistically significant. This means that, in the case there are disturbances in the short run, the disequilibrium caused by these disturbances will be corrected in the long run as the CointEq1 coefficient towards equilibrium in the long run is 20% with JSE15 as the dependent variable. This indicates a very slow correction to its equilibrium state. The previous quarter's change in LJSE15 has a statistically insignificant, small and negative impact on the current period (quarter). This suggests that a percentage change in LJSE15 itself will reduce LJSE15 by 6%. However, in other lags, changes in LJSE15 have a positive effect on LJSE15 itself.

In the short run, only previous quarters' CPI and INT at lags 1 and 2 change negatively and significantly affect LJSE15 in the current period, respectively. This suggests that a percentage change in inflation and interest rates can result in a decline in JSE Financial 15 index returns. This is because a continuous increase in inflation can erode real returns and also negatively affect investor confidence in financial markets. This result is consistent with Kaur and Chaudhary (2022). Tursory (2019) found similar results using the ARDL bound test, suggesting that a rise in interest rates will lower stock prices. This is because increased interest rates tend to result in a reduction in corporate earnings expectations and asset valuation, which can then affect stock prices. However, CPI at lags 2 and 3 indicate that there is a positive, insignificant relationship between CPI and LJSE15, suggesting that when inflation increases, it will increase LJSE15. This is consistent with the findings of Giri and Pooja (2017) as well as Chang et al. (2019).

Lags 1 and 3 of EXR show a negative, insignificant effect on JSE Financial 15. This suggests that when exchange rates increase, the LJSE15 decreases. Bahmani-Oskooee and Saha (2015) as well as Ogunsakin and Awe (2020), also found similar results using an ARDL model. However, Chang et al. (2019) found a positive short-term nexus between the exchange rate and stock prices. LGDP has positive coefficients for all lags and a statistically insignificant effect on the LJSE15. This indicates that there is a positive relation between economic growth and JSE Financial 15 returns, suggesting that when LGDP rises, LJSE15 can increase as a result. However, it also implies that LGDP does not significantly affect LJSE15. Ogunsakin and Awe (2020) found a similar positive short-run nexus. Suggesting that as economic growth rises, stock prices will increase as a result.

Lastly, the dummy variable DM1 for 2012Q4 results show a statistically insignificant impact in the short run. This implies that any shock or shift occurring in this quarter did not significantly affect the JSE Financial 15 index return in the short run. However, the dummy variable for DM2 for 2020Q1, representing the beginning of the COVID-19 pandemic, has a negative and statistically significant coefficient of −6% (−0.6449). This indicates that the pandemic had negative, significant impact on the JSE Financial 15 index return in the short run, which is consistent with the global market disruptions observed at the time.

The results in Table 9 show that D(INT) and D(EXR) have a probability value that is lower than 0.05. Therefore, we reject the null hypothesis that INT and EXR do not Granger-cause D(LJSE15) and conclude that INT and EXR Granger-cause LJSE15. This indicates that they can be included when forecasting LJSE15. D(CPI) and D(LGDP), on the other hand, have a p-value that is more than 0.05, failing to reject the null hypothesis that CPI and LGDP do not Granger-cause LJSE15. Keswani et al. (2024) also found similar results that INT and EXR do Granger-cause stock prices, implying that in the short run, the fluctuations in the stock prices are explained by these variables. However, their results are contradictory to this study as they have found that GDP and interest rates also do granger cause Indian stock prices. D(LJSE15) has a probability value that is higher than the 5% level; therefore, we cannot reject the null hypothesis that D(LJSE15) does not Granger-cause D(CPI) and D(INT). This implies that LJSE15 cannot be included when forecasting INT and CPI. However, it does granger cause D(EXR) and D(LGDP) because it has a probability value is not greater than 0.05.

In Table 10 panel A, the probability value is higher than the 5% level; therefore, accept the null hypothesis that there is no serial correlation at lag 5. In panel B, the joint components/variables have a p-value of 0.2361, which is higher than the 0.05 level; therefore, we can conclude that it is normally distributed. The heteroscedasticity test results in panel C indicate that, for the joint test, the p-value of 0.1096 is greater than 0.05; therefore, we can conclude that this model is not heteroskedastic.

This study's findings have shown that a long-run nexus exists between the chosen macroeconomic indicators and the JSE Financial 15 index. It also indicated that a short-run nexus between the macroeconomic indicators and the JSE Financial 15 index exists, although some variables indicated a statistically significant relation while others indicated a statistically insignificant relation between the variables. Some results were contradictory to previous studies due to data frequency, time periods as well as lags chosen. INT and EXR do Granger-cause LJSE15; therefore, when forecasting the JSE Financial 15 Index returns, INT and EXR can be included while CPI and LGDP can be excluded. The COVID-19 pandemic shock in the short run had a significant, negative impact on LJSE15 indicated by the DM2 results.

The fact that INT (Repo rate) Granger-causes LJSE15 means that monetary policy has a predictive and transmission effect on the performance of the financial sector. The South African Reserve Bank should therefore continue to use interest rate policies that are clear, transparent and predictable because, according to the findings of Ramatlo (2019), unexpected rate changes can significantly influence market movements. While the exchange rate Granger causality results show that the financial market is sensitive to currency fluctuations, authorities should aim for exchange rate stability because unexpected and sharp depreciations can have a negative impact on financial stocks and excessive appreciation could reduce competitiveness in exports. Investors and portfolio managers should also avoid relying only on domestic macroeconomic indicators when creating short-term investment strategies for the JSE Financial 15 index. They should also focus on firm-specific determinants and global financial trends which can affect the performance of the index.

While the model captures key macroeconomic variables, it omits certain financial variables that are commonly used in stock price modelling, such as money supply, credit growth and global financial indices. Future researchers should consider incorporating firm-specific variables as well as global financial indices to explore whether they add further explanatory value.

This article used secondary data and no qualitative data such as surveys were utilized for analysis.

This manuscript is an original work produced by the author(s). KM and CSS are aware of its content and approves its submission. It is also important to mention that the manuscript has not been published elsewhere in part or in entirety and is not under consideration by another journal. The author(s) has given consent for this article to be submitted for publication in Journal of Economic and Administrative Sciences.

KM conceived the key ideas for this research paper. She collected and analyzed the data. KM and CSS worked on the introduction, literature review, methodology, results and conclusion.

During the preparation of this work the author(s) used ChatGPT in order to proofread. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication. Thanks to the DSI/NRF Trilateral Chair in Transformative Innovation, the 4IR and Sustainable Development for its support.