Portfolio theory and cost-effectiveness analysis: A further discussion

Portfolio theory has been suggested as a means to improve the risk–return characteristics of investments in health-care programs through diversification when costs and effects are uncertain. This approach is based on the assumption that the investment proportions are not subject to uncertainty and that the budget can be invested in toto in health-care programs. In the present paper we develop an algorithm that accounts for the fact that investment proportions in health-care programs may be uncertain (due to the uncertainty associated with costs) and limited (due to the size of the programs). The initial budget allocation across programs may therefore be revised at the end of the investment period to cover the extra costs of some programs with the leftover budget of other programs in the portfolio. Once the total budget is equivalent to or exceeds the expected costs of the programs in the portfolio, the initial budget allocation policy does not impact the risk–return characteristics of the combined portfolio, i.e., there is no benefit from diversification anymore. The applicability of portfolio methods to improve the risk–return characteristics of investments in health care is limited to situations where the available budget is much smaller than the expected costs of the programs to be funded.