The Future of the National Electricity Market

As I was preparing a presentation on the National Electricity Market (NEM) for the Economic Society of Australia (QLD), the federal government announced the National Energy Guarantee (NEG) proposal.

While policy details are missing, I have come to a view that the NEG, while it may counter some of the impact of the Renewable Energy Targets (RET) on electricity markets, cannot decrease wholesale electricity prices and can only increase (at least in the short run) retail prices, and there is no guarantee (pardon the pun) that it will deliver efficient outcomes.

Below is a summary of my (preliminary) analysis. I focused exclusively on the NEM and have not discussed the implications of the NEG for other parts of the electricity system, such as networks.

Why the NEM is not working as designed

The NEM was designed, and subsequently implemented in 1998, so that competition, rather than administrate fiat, would became the mechanism to price wholesale energy in a system characterised by coal as baseload, gas as shoulder/intermediate and hydro (and other more expensive fuels) as peak. The NEM is an energy only, gross pool with mandatory participation. That is, the NEM accounts for all the trade in wholesale electricity with generators (above a minimum size) as sellers and large consumers (including retailers who buy electricity on behalf of small consumers) as buyers.

There is no demand-side bidding in the NEM. Generators submit bids specifying how much they are willing to supply at various prices, these bids are added up, and the market price is determined so that the added bids equal the system demand. Market prices are determined for every half-hour for each of the five regions of the NEM. The Australian Energy Market Operator (AEMO) settles the financial transactions for all of the electricity traded on the NEM on the basis of these spot prices. Generators and retailers often insure themselves from volatility in the spot price by entering into hedge contracts.

The NEM worked well for a number of years. This can be checked, for example, by comparing the wholesale prices with estimates of the long run marginal cost (LRMC) – conceptually, the LRMC is the (optimal) levelised cost of meeting an increase in demand over an extended period of time. In a competitive market with large fixed costs, economic theory suggests that the market price will converge to the LRMC, ensuring that generators who invest in new capacity will recover their costs (including a return on capital). The comparison of prices to the LRMC supports the notion that the NEM was well-designed and worked well for a period.  

Today, however, there is no reason to expect prices in the NEM to converge to the LRMC. The reason, in a nutshell, is the Renewable Energy Target, which requires large electricity buyers to purchase Renewable Energy Certificates (RECs). This means that renewable generators recover their total costs, including the cost of capital, not only through the wholesale energy market, when they generate electricity and are dispatched, but also from the revenue from selling RECs. As this revenue is not directly linked to the prices determined in the NEM, it will not induce truthful bidding in the wholesale market in the same way as hedging contracts do.

The existence of hedging contracts, which also provide revenue to generators above and beyond what they earn in the wholesale market, changes their bidding behaviour in the right direction (i.e., closer to true costs), and consequently the market price. In particular, it is well known in the economics literature that producers who hold a forward contract for some part of their possible production offer a lower price in the market (or if the market has a Cournot market structure, offer a higher quantity). In game theory jargon, a firm’s optimal reply is more aggressive (closer to its costs) when it holds a forward contract, and this leads to a lower equilibrium market price.

To summarise, while generators in general recover their costs through both the wholesale market and hedging contracts, renewable generators have an additional source of revenue that it is not directly linked to the wholesale market price. Thus, the NEM price differs from the LRMC of the system because some generators have an avenue to recover costs that is not linked to the wholesale market price. Moreover, the extraordinary growth in renewables represents prima facie evidence that the revenue from the RECs has been sufficient to promote investment despite the distortions in the NEM.

The issue is bad market design and policy, not the existence of renewables

The participation of renewable energy in the market by itself is not to be blamed for wholesale market prices not reflecting the LRMC of supply. Of course, efficient investment in renewable energy is essential to achieving the goals of lower emissions and affordable electricity.

Instead, the source of today’s electricity market woes likely goes back to the 1997 Howard Government climate change strategy, and its ‘Mandatory Targets for the Uptake of Renewable Energy in Power Supplies’. This was legislated in 2000 as the Mandatory Renewable Energy Target, and subsequently amended by Labor with distortions increasing over time. 

Perhaps the adoption of RECs was justifiable as a first, temporary step towards developing an energy policy to address the multiple goals of affordability, emissions reduction and security of supply. However, successive governments have failed in developing such a policy, and relied on the continued use of the RECs, with market distortions accumulating over time, leading to where we are today.

To see why more appropriate policies would likely have led to difference results, consider for example what might have happened if the Howard government, or the subsequent Labor and Coalition governments, had adopted instead an Emissions Trading Scheme (ETS). In this case, the market would have determined the appropriate carbon price to meet our emissions target. This carbon price, in turn, would have impacted on generators’ marginal costs, and their bids in the NEM and wholesale market price, making coal relatively more expensive to gas, and gas relatively more expensive to renewables.

An Emissions Trading Scheme, accompanied by a phasing out of the RET, would have likely led to gas becoming the transition fuel to an electricity system where renewables would necessarily play a larger role over time, but where investment, both in generation and in transmission, would have been driven by market dynamics linked to fundamentals like costs, including emission costs, rather than by the wedge between renewables and non-renewable revenues created by the RECs.

The NEG: Two wrongs don’t (necessarily) make a right

The NEG policy announced by the government is based on an eight-page advice provided by the Energy Security Board (ESB). (http://www.coagenergycouncil.gov.au/sites/prod.energycouncil/files/publications/documents/Energy%20Security%20Board%20ADVICE....pdf) . The gist of the ESB proposal is the imposition of an obligation on electricity retailers to hold forward contracts with generators to achieve particular reliability and emissions targets. The advice suggests that retailers will have the flexibility to create a portfolio that achieves these target in a way that minimises their costs, although no details are provided on how this may be achieved.

Under the proposal, the Australian Energy Market Commission (AEMC) would determine the system wide reliability standard, and the AEMO would then determine how that standard can be specified in terms of operating requirements for each of the regions of the NEM. In turn, electricity retailers would secure forward contracts with generators for dispatchable resources covering a predetermined percentage of their forecast peak load. Also under the proposal, the emissions target will be set by the government, and retailers will be required to meet an emissions guarantee either through contracts with existing generators or with generators to develop new capacity.

This high level description of the proposed mechanism should make it clear that it is essentially a RET-like mechanism, applied to both emissions and reliability. It is an attempt to counter the financial incentives provided by the RECs to renewable generators by providing financial incentives to dispatchable energy. It follows that the proposed mechanism has the same drawback as the RET in that it is not directly linked to market prices. 

Thus, as this additional revenue does not seem linked to the market price, and instead seems to be linked to dispatchable electricity, NEM prices will not reflect this additional revenue. As a result, NEM prices will continue to be distorted away from the LRMC. Therefore, the NEM cannot be relied upon to ensure that efficient investment in generation (and transmission) takes place, as prices lower than the LRMC imply that investments will not be recovered.

Whether the proposed mechanism will ensure additional investment in non-renewables will depend upon the magnitudes of the additional payments. These additional payments may lead to the postponement of the retirement of existing plants, or if substantial enough to investment in new capacity, whether it is efficient or not to do so. These payments will be driven by the choices made by regulators in defining the reliability guarantee, not the market.

There are also important implementation challenges. For example, it is not clear how the proposed forward markets will be established. The letter of advice refers to bilateral contracts, and seems to rely on the analogy with existing bilateral hedging contracts to support a view that new bilateral contracts will emerge as soon as the guarantees are defined.

There are a number of potential difficulties with this view. First, the current hedging market evolved over time, as retailers better understood how best to ensure against price volatility, including for example through vertical integration with generators. In contrast, to be useful to avoid a deepening of the electricity sector current woes, the implementation of the reliability guarantee will be subject to a much shorter timeframe

Second, the existence of asymmetric information and transaction costs suggests that bilateral trading is not the most efficient way to organise the sale of the RET-like reliability ‘certificates’. This means that a more structured market will need to be designed to aid price discovery and to promote efficient trading.

Designing such a market may offer an opportunity to better integrate the proposed capacity payments with the NEM. However, this is not a trivial exercise – think about how long it took to develop and implement the NEM. Based on experience, it would likely take more than two years to design an effective market to achieve efficient contract and NEM prices, so that market dynamics can lead to efficient investment decisions. Given the likely long leading times for establishing a new market for reliability and emissions payments, transition arrangements will be needed, but have not been contemplated in the letter of advice of the ESB. 

Finally, as the new payments will not impact on NEM prices, and they constitute an additional cost to retailers, it is difficult to understand how the proposed mechanism can lead to a reduction in retail prices.

Conclusion

It is far from clear that the mechanism proposed by the ESB will work in achieving the multiple goals of security of supply, emissions reduction and affordability. I am also skeptical, for the reasons given above, that bilateral contracts will emerge unaided to ensure that retailers can efficiently provide the emissions and reliability guarantees that underpin the mechanism.

Even if such bilateral contracts do emerge, the nature of the mechanism does not guarantee that capacity payments for reliability will be sufficient to ensure new investment and greater reliability in the long run. A more promising avenue, which will take time and effort to accomplish, would involve designing a market that can integrate new capacity markets into the NEM.