ABSTRACT
The Arrow and Lind theorem has been fundamental for suggesting a risk-neutral approach for managing sovereign risk. As already tentatively indicated by Arrow and Lind (1970), yet not picked up by practice and theory, it does however not apply for a number of countries subject to high natural disaster risk and with lesser means at disposal for spreading or pooling the risks. Especially the risk spreading capacity of governments and the resulting individual cost being negligible is debatable. In reality, external aid or loans have often to be sought post-disaster to somewhat continue with business as usual. Crucial about risk preference is that if governments are perceived to be risk neutral and thus risk and potential losses are only included as expected value, there is no incentive for a decision maker to evaluate risk financing mechanisms as the price of risk financing mechanisms is usually higher than the expected loss.
