of a country’s debt repayment d^culties or as equivalent to a country’s default.
Table 2 below contains a summary of the studies that tried to explain the
occurrence of a debt rescheduling. The common idea in these works is that a
limited number of ...nancial, macroeconomic, or socio-political indicators can be
identi...ed as the main determinants of debt repayment behaviour. Saini and Bates
(1984) provide a survey of the development of the quantitative approaches to
“country risk analysis”, where the existence of a probability of debt rescheduling
is one of the possible “manifestation” of such lending risk. They presented the
emergence of probit and logit models as the most used estimation techniques.
The choice of which variables are best to use to predi ct debt rescheduling
has been discussed in this literature at length, so that di «erent approaches have
been developed to predi ct the probability of LDC’s debt rescheduling. These are:
a “balance sheet approach”, a “macro approach” and a “structural approach”.
According to the “balance sheet approach” .nancial variables are considered
more relevant to explain the probability of a debt rescheduling. Lloyd-Ellis et
al. (1989) include three sets of variables in a logit model used to predict the
probability of a debt rescheduling. These are the traditional “ratio variables” (as
the debt service to export ratio, the foreign exchange reserves to import ratio, the
rate of growth of per capita GDP and the rate of growth of imports) and the so
called “balance sheet variables” (as the ratio between short, or medium, or long-
term debt over total borrowing from the banks, the proportion of each country’s
debt relative to total bank lending, total bank borrowing relative to bank deposit,
the ratio between the unallocated or undisbursed credit over total banks’ lending).
Number and value of current reschedulings are included in order to “capture”
a general attitude towards rescheduling. With both a sample of 27 countries,
during 1977-’81 and of 59 countries, from 1977(II) to 1985(I I), they discover that
bal ance sheets vari ables are more signi.cant than ratio variables (the only one
that is signi.cative, the rate of growth of exports, has the “wrong”, positive,
10