The Treasury

Global Navigation

Personal tools

6.3 Public sector risk

Section 4 highlighted the point that private sector measures of risk may not automatically carry over to the public sector. In particular, the CAPM (and other market-based models of risk-premiums) assume that the decision-maker is willing to freely substitute between different assets in order to balance the risk and return of the overall portfolio in accordance with the decision-maker's appetite for risk. For example, this definition of risk means that governments are comfortable offsetting the risk of say a health intervention 'failing' with an improvement to communications infrastructure 'succeeding', and vice versa, as long as the expected net effect on the risk and return of the overall portfolio is favourable.

However, some people may argue that governments select public sector investments on the basis of their ability to deliver specific economic and social goals. Under this view, the relevant concept of portfolio could be defined more narrowly as the set of possible investments that contribute to the delivery of a particular economic or social goal. If this view is considered reasonable, governments might be willing to make investments so as to offset risks within any given portfolio, whilst being much less willing to substitute freely between portfolios.

Furthermore, there are reasons to expect that the public sector is better placed to manage risk than the private sector. For example, public sector risk is spread much more thinly than most private sector risk in the sense that, for any given project considered in isolation, the success or failure of that project does not affect the government's ability to deliver its other economic and social commitments in the same way as it would for most businesses in the private sector. The government also has an implicit taxpayer guarantee, given its constitutional power to tax, and can issue highly liquid low-risk debt. These options are not usually available to private companies.

These points raise the question of whether taking a market-based SOC as the basis for recommending public sector discount rates is appropriate. Specifically, this approach assumes that market-based models for pricing risk are also applicable to how the government prices risk. Again, this amounts to imposing a potentially significant assumption, at least for some projects, which ought to be scrutinised.

The SRTP approach does not explicitly account for risk through the discount rate. Rather, it usually relies on separate risk analyses such as scenario analysis, sensitivity analysis and/or Monte Carlo simulations to take account of project-specific risk relating to particular cost items and benefits.[24] Whether this is the best way to account for the overall risk that is faced by the government is a question for debate.

Notes

  • [24]That is, specify probabilities for uncertain outcomes and cash flows, then statistically sample from these distributions to obtain probability distributions and confidence intervals over the range of all possible outcomes.
Page top