Generation Investment and Security of Supply

So far, this paper has focused on competition and market power. As noted in the introduction, these considerations are subtly related to the issue of generation investment and security of supply. [15]

Security of supply is somewhat difficult to define when looked at from an economic point of view. What does it really mean? After all, people don’t often find reason to worry (at least in industrialised countries) about security of supply in other areas of the market economy – even areas that, like electricity, are “crucial” to daily life, such as food. For example, there’s little discussion or concern about security of supply of apples. We take it for granted that, in the apple market, supply and demand interact, prices fluctuate and there are (almost) always apples available to those willing to pay the market price. That is, the market for apples always “clears”: the price fluctuates so that supply equals demand.

Figure 3: A Wholesale Electricity Market with Elastic Demand

In an ideal wholesale market, price spikes indicate scarcity and offer incentive for new investment.

Indeed, this is what should happen in an electricity wholesale market if demand were sufficiently elastic. Consider Figure 3, which depicts an ideal wholesale market with fairly elastic demand (not typically seen in real world electricity markets). The demand curve is drawn with a downward slope to indicate the responsiveness of consumers to hourly price changes. The demand curve D_low represents low (or normal) demand and the demand curve labelled D_high represents high (or “peak”) demand (for example, a cold winter evening when a large fraction of consumers turn on electric heaters). In this ideal market, prices rise during periods of high demand and consumers cut back consumption to the point where supply equals demand – that is, the market price “clears” the market. The high market price acts as a signal to investors, sending information about the optimal amount of long-term investment. New firms will enter the market if they believe that the market price will cover investment costs.

In practice, investment decisions are very sensitive to the prices received in rare high-demand periods.

Some analysts (California Public Utilities Commission 2005, Cramton and Stoft 2006 and Joskow 2006) have recently argued that – in many wholesale markets around the world – there are not adequate incentives to underpin enough long term investment. In some wholesale markets, it appears that prices are not spiking to high enough levels to support long term investment. In particular, “peaking” capacity runs only a few hours per year and so capital costs must be recovered in those few hours. That is, price must rise above the marginal cost of running these units and they must earn significant scarcity rents in a small number of hours. As a result, investment decisions are very sensitive to price conditions in rare high-demand hours. Cramton and Stoft call these inadequate scarcity rents the “missing money” problem.

Figure 4: A Wholesale Electricity Market with Inelastic Demand

Why might wholesale prices not rise high enough to provide adequate scarcity rents for optimal investment? One reason is that many countries have implemented price caps – partly in response to concerns about market power. These may be straightforward administrative measures (i.e., no bids are accepted above a pre-announced price level) or may involve certain interventions (such as operation of a designated government-owned generation plant at a pre-announced level).

However, Cramton and Stoft (2006) and Joskow (2006) – looking primarily at the US experience – argue that another set of issues would lead to a “missing money” problem and inadequate generation investment, even in the absence of price caps. The combination of inelastic demand and inelastic supply (when capacity is near its limit) produces situations when the market is not able to clear. A stylised representation of this type of situation is presented in Figure 4. Here, the demand curve is shown as vertical (or perfectly inelastic).[16] What happens when demand jumps – eg, only for a number of hours in a single evening – from D_low to D_high? In the case depicted in Figure 4, most consumers aren’t aware of what is happening to prices on an hourly basis – so they do not respond to the rising prices on this particular evening. In order to keep demand from outstripping supply and destabilising the transmission network, the system operator (SO) typically intervenes, administratively setting prices and rationing electricity. In extreme situations (as shown in the “high” case in Figure 4) this intervention takes the form of managed ”rolling” blackouts. However, in practice, the SO’s intervention usually begins when some pre-determined “reserve margin” line has been crossed. Joskow (2006) offers some examples of how the SO’s actions can artificially depress prices in these cases, but for the purposes here it is enough to keep the “blackout” situation depicted in Figure 4 in mind.

Some economists contend that crucial peak-period investment signals are often distorted: the inelasticity of supply and demand forces the system operator to intervene in the wholesale market.

Cramton and Stoft put it this way: the market is simply unable to “choose” the efficient level of long-term capacity in the decentralised way that other markets do. Again think of the apple example. In the absence of any distorting government intervention, the market price (determined by supply and demand) sends signals to investors about the “right” amount of long-run investment. However, inelastic demand effectively requires the SO to set prices during occasional “crisis” periods when demand is high (or supply is temporarily low). Cramton and Stoft argue that it is thus the SO (or the SO’s procedural rules) that effectively sets a target for long-run investment. They lament the fact that this is not recognised by policymakers and the SO, who thus make the implicit decisions about investment “with eyes closed”.[17] This is not to say that the SO should be disparaged for intervening. After all, it is the SO’s primary responsibility to maintain the grid’s physical stability. Instead, the economic distortions may be better described as an unintended consequence of the SO’s technical responsibilities.

It might be tempting to conclude that – if prices are not spiking high enough to support investment – policymakers should err on the side of ignoring market power. This is not justified. Instead, policy should promote wholesale market institutions that encourage competition and “good” price fluctuation (ie, driven by legitimate scarcity rents, not market manipulation).

Policy should focus on improving incentives for investment.

There are different approaches to improving the performance of wholesale markets in stimulating optimal long term investment. First, policymakers could focus on ‘fixing imperfections’ – by removing price caps (and instead mitigating the scope for market power through other measures described above), improving demand responsiveness, reducing regulatory uncertainty and carefully evaluating SO procedures. Improving demand responsiveness is particularly important, although there may not be much that can be done beyond waiting for technological advances that allow consumers to affordably monitor prices.[18] Second, Cramton and Stoft and Joskow recommend that policymakers explicitly recognise the need for a long-term investment (or “resource adequacy”) program. Such a program might involve establishment of a parallel “capacity market” that would offer payments to generators to replace “missing” scarcity rents and improve incentives for long-term investment. Of course, like other areas of electricity policy, the devil is in the details. Implementing a resource adequacy program that avoids creating new distortions is a challenge. Finally, policymakers should be mindful of producing a stable, credible regulatory regime that reassures investors that any justified scarcity rents won’t be removed by regulatory changes in the future. Generation investments last for a large number of years – so investors have good reason to be sensitive to the possibility of future regulatory changes.