The Brady Plan has triggered a number of proposals to reduce the face value of LDC debt, but there is a gap perhaps very large between the debt relief nominally granted to a debtor and the relief given up by the creditors in terms of the net transfers the debtor is expected to pay. In order to assess the real reduction in the debt burden when the nominal debt is cut, it is necessary to know not only the market value of the debt but also its marginal price, which is usually much lower than the observed market price.

In Discussion Paper No. 460, Research Fellow Daniel Cohen develops a valuation formula for determining the relationship between real and nominal debt reduction. Cohen uses this model to show that a `buy-back' of debt will probably only be advantageous to the country concerned if its marginal price is less than half the observed market price; that the value (to creditors) of guaranteeing payment flows against external shocks to the debtor's economy may be no more than 25%; and that creditors are willing to trade off a 1% increment in the rate of growth of payments against a 15% reduction in their level.

Cohen notes that the ratio of net transfers to outstanding long- term debt to private creditors paid by the `highly indebted countries' (HICs) was approximately 5% in 1987. In theory, neglecting the effects of risk, this should be equal to the difference between the interest rate on the debt and the growth rate of transfers, which should itself be equal to the growth rate of the economy as a whole. On these assumptions, 5% seems plausible, but the average discount observed in December 1987 was about 50%. Cohen investigates the how risk may account for this discrepancy and relates it to various debt relief proposals. In particular, he concludes that the deal negotiated for Mexico in July 1989 was advantageous for Mexico but disadvantageous for the multilateral agencies that helped to finance it.

In Discussion Paper No. 461, Cohen considers the relationship between debt, investment and growth in the LDCs. The average annual growth rate of the LDCs fell from an average of 7% in the period 1974-80 to 3.5% in the 1980s, and the decline for the HICs was even more spectacular: from 7% to 1.1%. Direct evidence on Latin American countries also suggests that the attempt to service this debt created tremendous domestic pressure which eventually led to reductions in both investment and the growth rate.

Most empirical studies that support this `negative' influence of debt on growth hinge on a comparison of the rates of growth and investment prevailing in the 1970s with those prevailing after the 1982 debt crisis. Cohen maintains, however, that such comparisons overlook the significant change of regime between the two periods: whereas in the years from the first oil crisis to the early 1980s the LDCs enjoyed relatively free access to world financial markets and faced an extremely low indeed perhaps negative real interest rate, the abrupt rise in world interest rates closed off most countries' access to these financial markets after 1982.

Cohen argues on theoretical grounds that the initial (direct) effect of this regime switch, which may be characterized as a return to financial autarky, is unambiguously adverse to growth, regardless of the countries' level of indebtedness. The impact of the second effect that the HICs were not only prevented from borrowing new resources but also had to repay their debt precipitously is ambiguous, however, since it depends upon the efficiency of the rescheduling process, and in particular on whether the service of the debt crowded out or crowded in investment in comparison with the level for a `financially autarkic' economy.

In the empirical section, Cohen calculates the bench-mark that would prevail under full financial autarky and compares it with the actual rate of LDC investment in the 1980s. He finds that the stock of debt accumulated in the early 1980s was not a good predictor of investment, which refutes the conclusion of a naive `debt-overhang' argument that the nominal stock of debt in itself deters growth. Cohen also shows that investment was crowded out by the service of the debt (relative to the `financially autarkic' level) and that this negative correlation between investment and debt service is almost identical to that between investment and inflows of foreign capital in the 1960s. In both periods, the impact of the foreign financial sector on investment was low: one percentage point of GDP brought in or taken out was associated with an opposite change in investment equal to one-third of the foreign financial flow, so the bulk of the adjustment was borne by saving rather than investment in each case. A level of debt service equal to 3% of GNP (almost an upper bound to the adjustment experienced in the 1980s) should therefore reduce investment by 1% of GNP below the `financially autarkic' level, so that writing off the debt would foster little investment unless accompanied by a large inflow of foreign capital.


A Valuation Formula for LDC Debt
Slow Growth and Large LDC Debt in The Eighties: An Empirical Analysis
Daniel Cohen

Discussion Papers Nos. 460-1, September 1990 and January 1991 (IM)