Negative $1,000 per megawatt-hour at noon. Positive $15,000 per megawatt-hour by six in the evening. Same grid, same day, same summer. That is not a malfunction. That is the National Electricity Market doing exactly what it was designed to do — and that design deserves more scrutiny than it usually gets.
I’ve been watching NEM spot prices for long enough to stop being surprised by the swings themselves. What still catches me is how consistently people misread them. The standard explanation — renewables are intermittent, therefore prices are volatile — is true in the way that saying ‘pitches take spin’ explains a Warne wrong’un. Technically accurate. Misses almost everything.
The market structure that invites extremes #
Start with the basics. The NEM dispatches electricity in five-minute intervals. Generators submit offers — how much power they’ll provide and at what price — and AEMO’s dispatch engine stacks those offers from cheapest to most expensive until demand is met. The last offer accepted sets the clearing price for everyone in that interval. This is a uniform-price auction, and uniform-price auctions have a well-documented property: they reward strategic bidding.
Generators know this. A plant that can move fast — an open-cycle gas turbine, a big battery — can hold capacity back until late in the dispatch stack and collect the elevated clearing price. This is not illegal. It is rational. The AEMC and the AER watch for conduct that crosses into market manipulation under the National Electricity Rules, but the line between smart bidding and gaming is genuinely contested, and the regulator has said so in published determination documents.
The market price cap currently sits at $16,600/MWh. The market floor is negative $1,000/MWh. That’s a $17,600 range. No other commodity market most Australians interact with has anything like that spread in a single trading day.
Rooftop solar and the midday collapse #
Here’s where the transition adds a new layer. Australia now has more rooftop solar per capita than any comparable grid in the world — well past 20 gigawatts of installed capacity as of earlier this year, according to the Clean Energy Regulator’s published generation data. On a clear winter day, that rooftop fleet pushes system demand down to levels the grid simply wasn’t built to absorb at low cost.
Thermal generators face a choice: run at minimum stable load and bleed money at negative prices, or shut down and face the cost and time penalty of restarting. Many curtail. Some rooftop solar is now involuntarily curtailed by network operators managing voltage and stability constraints. The result is a midday price trough that keeps getting deeper. Australia’s rooftop solar success is real — but it has a balance sheet entry on the other side of the ledger that the marketing tends to skip.
Then the sun drops. Demand rises — people get home, turn on air-conditioning, cook dinner. The rooftop fleet falls off a cliff. Whatever thermal capacity sat on the sidelines during the cheap hours now has significant market power. Prices spike. This daily pattern, the so-called duck curve, is no longer a forecast. It’s the operational reality on most NEM days.
The role of gas and the scarcity premium #
Gas peakers are the swing suppliers. They exist to fill exactly the gap the duck curve creates. The honest read is that without them, the evening ramp would be significantly worse. But gas peakers also have every commercial incentive to maximise the price at which they clear. When there are only a handful of fast-response plants available to cover a three-gigawatt demand swing in forty minutes, those plants have pricing power. Full stop.
Fuel cost matters too. East-coast gas prices have stayed elevated since the supply disruptions that followed Russia’s invasion of Ukraine in early 2022, which tightened the global LNG market and exposed just how exposed the domestic market was to export-parity pricing. Gas generators pass that cost through. When their input cost rises, their offer prices rise, and the dispatch stack shifts upward. The connection between a spot LNG cargo and your electricity bill is real, even if it’s indirect.
Follow the money and you’ll find that the vertically integrated gentailers — the AGL and Origin-scale businesses — are partly hedged against this by owning generation on both sides of the fuel type. But smaller retailers with no generation portfolio are almost entirely exposed to spot. When spot spikes, their margins compress or invert. The Default Market Offer provides a retail price anchor, but it doesn’t insulate retailers from wholesale risk in real time.
Weather, interconnectors and the cascade effect #
Victoria and South Australia share an interconnector. New South Wales and Queensland share one. When a heatwave runs across multiple states simultaneously — as they do, because weather fronts don’t respect regulatory boundaries — peak demand rises everywhere at once, interconnector flows hit their transfer limits, and each region is thrown back on its own resources. Constrained regions with thin supply margins see prices move fast.
Add a transmission outage or a large generator trip and the scarcity signal goes from loud to deafening. AEMO publishes its Electricity Statement of Opportunities each year laying out exactly where reserve margins are tightest. The honest read of the last couple of editions is that South Australia and Victoria are carrying less slack than is comfortable, particularly as coal exits accelerate. AGL’s Loy Yang A closure timeline is part of that picture — AGL’s coal exit strategy has real implications for how the evening ramp gets covered in Victoria over the next few years.
The cascade effect is underappreciated. A price spike in one region pulls flows from a neighbouring region, tightening supply there, which can lift prices in the second region too. The NEM’s interconnected design means contagion is baked in. That’s not a flaw — it’s the point of an interconnected market — but it does mean localised scarcity can become regional scarcity quickly.
Batteries are changing the shape but not taming it #
The big battery buildout is real and accelerating. Projects like Akaysha Energy’s Waratah Super Battery in New South Wales — part of BlackRock’s expanding NEM portfolio — are designed specifically to charge during the negative-price trough and discharge into the evening ramp. In theory, that arbitrage compresses both extremes: it lifts the floor by absorbing excess generation and caps the peak by adding supply at the critical moment.
In practice, batteries are also commercial assets. Their owners maximise revenue. When a battery operator can see that supply is critically short and a price spike is imminent, holding dispatch for a few more intervals is commercially rational. The same market structure that rewards gas peakers rewards batteries. The technology is different; the incentive is the same.
I’ll be direct: anyone who tells you that enough batteries will smooth the NEM into a placid, low-volatility market is probably selling something. Volatility is, in part, how batteries earn the returns that justify their capital cost. Compress the volatility completely and you compress the revenue case. The pumped hydro versus batteries debate often loses sight of this — both technologies need price signals to function as commercial investments, and price signals in the NEM are, by design, violent.
What the quarter-price mechanism actually does #
From October 2021, AEMO moved to five-minute settlement — that is, the dispatch price and the settlement price became the same interval. Before that change, the settlement price was the average of six five-minute dispatch prices, which blunted some volatility at the settlement level even when dispatch prices spiked. Removing that averaging was the right call for market efficiency, and the AEMC argued it would improve incentives for fast-response storage. It also meant the full force of every dispatch price swing now hits financial settlements directly.
This is not a complaint about the reform — it was correct policy. But it does mean that anyone looking at settlement data post-2021 and comparing it to pre-2021 volatility needs to account for the rule change. Apples and oranges, as the saying goes. The AEMC’s five-minute settlement documentation explains the mechanics in full if you want to go deep.
The gap between wholesale and retail that confuses everyone #
Here is the thing that gets lost in almost every public conversation about electricity prices. Most households don’t pay the spot price. They pay a flat or time-of-use retail tariff set months in advance. Retailers hedge their wholesale exposure — through financial contracts, through owning generation — so that retail prices move slowly even when spot prices swing wildly. The volatility is real, but it’s largely absorbed in the commercial layer between generators and consumers.
The people who feel it directly are large industrial customers on market-linked contracts, small retailers without adequate hedging, and the generators themselves. The case for pricing reform often focuses on retail tariff structures, which is fair, but the wholesale volatility question sits underneath all of it.
The NEM was built on the principle that volatile spot prices would send efficient investment signals. Build the right capacity and prices moderate. Don’t build it and prices scream until someone does. Whether that signal is still being heard clearly — given the policy overlays, the capacity schemes, the state government interventions — is, in my view, the more interesting question heading into the back half of this decade.
The swings aren’t going away. The question is who absorbs them, and whether the current answer to that question is the right one.
— Marcus Wren, Editor
Photo by gio shravan on Unsplash