Skip to content

Financial Implications of South African Divestment

Literature › Works

Blake R. Grossman; William F. Sharpe (1986). Financial Implications of South African Divestment. Financial Analysts Journal, Vol. 42, No. 4. CFA Institute.
Type journal article
Scope Comparative and methods
Method econometric
Period studied 1960–1983
Regions United States (NYSE), South Africa (as investment context)
Themes sanctions divestment, macro fiscal finance debt, methods data sources
Citation key grossman1986

Argument

The article asks only about the minimum financial cost to a US institutional investor of divesting stocks of companies doing business in South Africa. It argues that a limited Sullivan non-compliance divestment barely changes the investable universe or performance, and that complete divestment into a value-weighted portfolio of all South Africa-free NYSE stocks is cheap. Transaction costs are re-estimated from actual passive trades, and returns are decomposed with a cross-sectional factor model into a 'South Africa factor' and a small-stock effect. The authors conclude that the divested portfolio would historically have matched or beaten the NYSE on a risk-adjusted basis, and that this is likely to continue largely because of its small-stock bias. The effectiveness of divestment in changing apartheid is explicitly set aside.

Evidence and method

CRSP monthly returns and valuations for all NYSE stocks, 1960-1983; the 1984 IRRC list of firms with South African operations; Sullivan compliance ratings; World Trade Academy lists (1959-1981) for an ex ante check; Loeb's broker survey of trading costs; 36,463 actual Wells Fargo Investment Advisors passive index-fund transactions (Nov 1984-May 1985); monthly cross-section regressions on 15 factors.

Findings

  • Sullivan non-compliance firms are a small share of NYSE value, while the full list of South Africa-linked firms is a large share. (p. 4)

    At the end of 1983, such firms had a combined market value of $570 billion-37.1 per cent of the total value of stocks listed on the NYSE. Sullivan non-compliance companies accounted for only 7.1 per cent

  • The South Africa-free universe is smaller in average firm size, which is its key difference from the NYSE. (p. 5)

    The largest U.S. firms tend to have operations in South Africa; the average SAF firm is thus nearly 30 per cent smaller than the average NYSE firm

  • Using actual passive-trading costs, the initial cost of divesting a $1 billion portfolio is much lower than earlier estimates. (p. 9)

    Estimated divestment costs for a $1 billion portfolio were 0.41 per cent of the overall portfolio value, considerably lower than the 0.95 per cent obtained using the Loeb assumptions.

  • Wagner's earlier estimate rested on an equal-weighting assumption that inflated costs. (p. 7)

    Wagner estimated divestment transaction costs to be 6.0 per cent of the amount divested, or 2.28 per cent of overall portfolio value.

  • The equal-risk divested portfolio outperformed the NYSE historically. (p. 10)

    This "equal risk" SAF strategy outperformed the NYSE by 0.187 per cent annually over the period.

  • The South Africa factor was positive overall but unstable, significant only before 1976, and negative in the early 1980s. (p. 11)

    From 1960 through 1975, SAR stocks tended to outperform SAF stocks. From 1976 through 1979, the two groups performed approximately equally well. And from 1980 through 1983, SAR stocks underperformed SAF stocks.

  • An ex ante check using World Trade Academy lists finds no major ex post selection bias. (p. 12)

    Differences between the ex ante and ex post curves are small enough to allow us to reject the hypothesis of major ex post selection bias in the previous results.

  • The small-stock effect outweighed the cost of excluding South Africa-related stocks, giving a counterfactual premium of 0.513 per cent a year. (p. 12)

    Had the South Africa factor equaled zero, the SAF strategy would have beaten the NYSE by an additional 0.326 per cent, or a total of 0.513 per cent

  • Equilibrium reasoning bounds the expected return premium on South Africa-related stocks under maximum divestment. (p. 14)

    the maximum likely divestment-induced change in the differential between Es and Ef would be 0.310 minus 0.107, or 0.417 per cent per year.

  • Even under maximum divestment the authors predict the SAF portfolio will outperform and recoup costs quickly. (p. 14)

    one might reasonably predict that a SAF portfolio will outperform the NYSE by at least 0.358 per cent (= 0.513% - 0.155%) per year, even under conditions of maximum possible divestment.