4.3 Sequential investment

There will often be an intermediate route between investing immediately and delaying all investment in order to learn more about the payoff. For example, it might be possible to undertake limited investment in the meantime, while learning more about the project payoff before undertaking the rest of the investment. Indeed, the outcome of the partial investment might reveal information about the desirability of the full project. For example, the early stages of a gradual rollout of ultra-fast broadband will reveal information about the level of demand for the services it provides. If this demand is high, then the full network can be built, with users experiencing slightly delayed delivery. In contrast, if the demand revealed by the early stages is low, then the rollout of the remainder of the network can be abandoned or delayed until demand rises, allowing the government to avoid (or delay) uneconomic investment. However, investing in stages has a cost that may outweigh these benefits. Specifically, building a large network initially will allow it to fully exploit economies of scale in network construction. The decision to build the network in stages or all at once must trade off the benefits of quick, low-cost construction against the cost of investing without first gathering potentially valuable information.

As in the case considered above, for decision making to result in an optimal allocation of resources, the cost of projects needs to incorporate the full opportunity cost. In this case, the full opportunity cost includes the incremental capital expenditure plus the value of the learning option destroyed by investment.

The value of a learning option destroyed by investment depends on the amount of information that can be learned if the investment is delayed (more precisely, the amount by which uncertainty can be reduced by further investigations). Real options analysis provides a rigorous approach to estimating the value of learning options. Various models of learning options appear in the academic literature and have been used to identify the sources of option value and the implications for resource allocation.[9] These can be used to evaluate staged investment—allowing learning to occur before the full resources are committed to a project—against alternative policies.

Phased investment has another potential benefit, which is subtly different from the one just considered. As in the case just discussed, there is an option value in delaying (or slowing down) investment. In this case, however, the source of the option value is the flexibility to respond to different market conditions. For example, by conducting staged investment a decision-maker can suspend construction, if the price of inputs rises, if recession dampens demand, or if new technology supporting a different approach to investment has emerged, and accelerate it, if input costs fall, if demand rises, or if a new technology does not emerge. The difference is more than one of semantics. In the situation considered previously, the decision-maker could use the opportunity to delay to learn more about a project: if nothing was done while waiting, then nothing would change while waiting. Here, there can be value in the option to simply wait and do nothing and wait for market conditions to improve.

The payoff to typical infrastructure investment is sensitive to variations in market conditions, whether it be fluctuations in demand for ultra-fast broadband or longer-term variation in electricity carried over the national grid. The cost of building such infrastructure will typically exhibit economies of scale. In such circumstances, investment involves a trade-off between exploiting the economies of scale by investing in a small number of large expansions and retaining flexibility by investing in small increments as and when they are needed so as to avoid unplanned investment in excess capacity.

Greater volatility means that investment should involve larger, relatively infrequent, steps.[10] By raising the value of delaying investment, greater volatility motivates the decision-maker to set a more demanding investment test. This in turn changes the trade-off between scale and flexibility, because it reduces the costs of “overinvesting” in additional capacity—that is, even if demand growth is lower than expected after expansion, the fact that the investment threshold is so high at the time of investment means that the decision-maker may still be able to recover the capital expenditure incurred. The optimal policy is to take greater advantage of economies of scale and raise the scale of each expansion of capacity.

Volatility can vary between industries, so that the appropriate scale and timing of the ultra-fast broad band roll-out, for example, can differ from the rate at which the national electricity grid should be upgraded, which can differ from the rate of investment in the roading network. Furthermore, as we will see below, the nature of contractual arrangements between the funding authority and the party undertaking the investment influences the allocation of demand risk, which affects volatility. Thus, the precise form of a contract can affect the scale and timing of investment that occurs as a result of the contractual arrangement.[11]