LDC Debt
Secondary markets

Transactions on the secondary market in LDC debt are estimated to have grown from almost zero in 1985 to more than $70 billion in 1989. Prices on this market throw important light on the pattern of default risks, and proposed solutions to the LDC debt crisis must take such transactions into account.

In Discussion Paper No. 459, Research Fellow Daniel Cohen and CEPR Director Richard Portes present the preliminary conclusions of an econometric analysis of the secondary market prices of long-term debt for seven countries whose debt is traded extensively and three countries for which reliable information on the price of short-term debt and the relationship between the two is available.

Cohen and Portes find that more than 85% of the price volatility of long-term debt is driven by the same variables. The elasticity of the average price of such debt with respect to the interest rate (Libor) is unity, and the individual elasticities for the seven countries range from 0.23 to 1.67 with a mean of 0.74. Pooled time-series data show an elasticity of unity, which suggests that investors perceive debt as the present value of an indefinite sequence of future payments. The other `common factors' are not obviously correlated with world macroeconomic factors and can be interpreted as the `systemic risk' of LDC debt, i.e. a set of factors common to the indebted countries only.

The unit elasticity of the price of debt with respect to Libor suggests that this `systemic risk' does not represent a risk of `once-and-for-all' default, but rather a measure of changes in the incentive to service debt over time. A decomposition of the long-term price into the short-term price and the relative price of long-term versus short-term debt supports this view, since the `systemic risk' and Libor both explain the relative price much better than the short-term price.

Further, Cohen and Portes find little or no correlation between the prices of long-term debt and debtors' exports. It seems implausible that fluctuations in oil prices have no effect on Mexican or Venezuelan debt prices. By decomposing long-term debt into its component parts Cohen and Portes find that for Brazil and Mexico (but not for Argentina) the export prices significantly influenced the relative price of long-term versus short-term debt.

Cohen and Portes conclude that the short-term debt price is influenced neither by `systemic risk' nor by idiosyncratic economic factor and that it must therefore be driven by an idiosyncratic non-economic factor, such as local political risk. If long-term debt service is conditional upon the service of short-term debt, long-term debt payments are scaled on the country's resources, but the decision to service the debt is contingent upon the service of short-term debt (driven by an idiosyncratic political risk) and upon a `systemic risk' that characterizes the developing countries as a group.

The Price of LDC Debt
Daniel Cohen and Richard Portes


Discussion Paper No. 459, September 1990 (IM)