Skip to content

The impact of economic fundamentals on stock markets in southern Africa

Literature › Works

Keith R Jefferis; Charles C Okeahalam (2000). The impact of economic fundamentals on stock markets in southern Africa. Development Southern Africa 17(1). Development Bank of Southern Africa.
Type journal article
Scope Comparative and methods
Method econometric
Period studied 1985–1996
Regions South Africa, Zimbabwe, Botswana
Themes macro fiscal finance debt, methods data sources, comparative theory, business and state, sanctions divestment
Citation key jefferis2000

Argument

The article asks whether stock markets in three southern African economies track economic fundamentals, which bears on whether developing-country stock markets can allocate capital efficiently. It estimates cointegrating relationships and error correction models of real stock indices on domestic and foreign real GDP, real interest rates and real exchange rates, using quarterly data for 1985-95 (Botswana 1989-96). The JSE shows the clearest long-run link to fundamentals (positive for GDP and the real exchange rate, negative for the long-term interest rate). The Zimbabwe results are partly perverse, which the authors attribute to heavy controls and structural adjustment. The Botswana index is stationary, so it is not weak-form efficient over the period, and only short-run relationships are estimated. The common finding is the role of domestic growth. Foreign influences run through trade variables rather than capital flows, which suggests little capital-market integration. The authors link this to market size, openness and liquidity.

Evidence and method

Quarterly national stock indices plus IFS series (CPI, exchange rates, T-bill and bond rates, GDP) from the Bank of Botswana and Zimbabwe's CSO. Market descriptive data come from IFC Emerging Markets Factbooks and Morgan Stanley (1993). GDP for Botswana and Zimbabwe is linearly interpolated to quarterly frequency (Diz 1970). Tests used are ADF and Phillips-Perron unit roots, ADF residual and Johansen cointegration tests, and error correction models.

Findings

  • The JSE is the oldest African exchange, founded with the gold discovery and historically dependent on the mining sector. (p. 8)

    The Johannesburg Stock Exchange is the oldest stock market in Africa, having been established in 1887 shortly after the discovery of gold on the Witwatersrand.

  • Conglomerate ownership structures, which originated in the mining houses, leave most JSE shares unavailable for trading and keep liquidity low. (p. 8)

    Because of this, most of the shares in companies quoted on the JSE are effectively unavailable for sale.

  • After 1994 the JSE became more exposed to international shocks, for example the 1998 Russian/Brazilian crisis. (p. 9)

    the overall share index fell by 30 per cent during August 1998 alone.

  • The JSE's long-run cointegrating relationship has the expected signs for the real exchange rate, GDP and the long-term real interest rate. (p. 16)

    which shows that the real stock market index is positively related to the real exchange rate and real GDP, and negatively related to the long-term interest rate.

  • The South African error correction model shows fast adjustment to the long-run relationship. (p. 17)

    its relatively large magnitude (70 per cent) shows that there is rapid adjustment to the long-term equilibrium relationship each quarter.

  • The Zimbabwe short-run model fits well statistically but adds little economic insight. (p. 22)

    Although it has a high R2 (74 per cent), it does not appear to provide much additional insight into the determinants of real stock returns in Zimbabwe.

  • The Botswana real index is stationary, so cointegration methods cannot be used, and the market is judged not weak-form efficient. (p. 24)

    Unit root tests show that the real stock market index for Botswana is I(0).

  • South African growth benefits Botswana's market, so the two economies are complements. (p. 24)

    It indicates that South African economic growth has a positive effect on Botswana (and probably the region in general)

  • The JSE shows the closest link to fundamentals of the three markets. (p. 25)

    The Johannesburg Stock Exchange demonstrates the closest relationship with those economic variables that were expected to have an impact.

  • No long-run link between the JSE and US GDP or real interest rates; international influence runs through the real exchange rate. (p. 25)

    there is no long-term relationship with either US GDP or real interest rates.

  • South African interest rates do not affect Botswana's market, indicating limited capital mobility between the two countries despite close trade ties. (p. 27)

    indicates that capital mobility between the two countries is limited, despite the close economic relationship in other respects.