Electricity is the only major traded commodity that cannot be stored at grid scale, so a power position is not a quantity of energy but a stack of commitments to specific hours and locations. This article explains why the hour — not the megawatt-hour — is the unit of exposure, and why a flat power book is rarely flat.
Electricity is the only major traded commodity that cannot be stored at grid scale. Oil can wait in a tank, gas in a cavern, grain in a silo; a megawatt-hour cannot wait anywhere. The consequence is that a power position is not a quantity of a thing. It is a set of promises to deliver or take energy in particular hours at particular points on the network, and a megawatt-hour in one hour is not the same asset as a megawatt-hour in the hour before or after it. A power book that nets those hours together reports a number that is arithmetically tidy and commercially empty.
This article sets out how a power position is actually shaped: why the unit is the hour and the location rather than the megawatt-hour, why the price of an hour is set by the last unit needed rather than by the cost of what you bought, why a negative price is the market clearing rather than a malfunction, and why a book that looks flat on volume is usually long one hour and short another. It is written for the people who have to read and defend a power book: traders running a position across delivery hours, the risk and middle office who report it, and anyone who has watched a flat-looking position produce results that should have been impossible.
If the distinction is new to you, start with Power has no warehouse, so a megawatt-hour is a claim on an hour. If you already run a power book and want the part that goes wrong quietly, start with Netting a book across hours discards the position while keeping the volume.
Power has no warehouse, so a megawatt-hour is a claim on an hour
Every other commodity can wait; electricity cannot. The object of power trading is not a stock of energy but a promise to deliver or take energy at a particular hour, and the position exists only for as long as that hour has not yet passed.
A barrel of oil is a thing you can hold and hold again. A megawatt-hour is a rate sustained for a period — power flowing over one hour — and once the hour is gone the product is gone. There is no warehouse where an unsold megawatt-hour sits waiting for a better bid, and no inventory you can carry forward to next week. Because the product cannot be stored, supply and demand must be brought into balance continuously, and every trade is therefore a claim on a moment rather than on a pile of goods.
This is the fact from which everything else in power trading follows, and it is worth stating plainly because most habits from other commodities do not survive it. In a storable market, a position is something you own and can be patient with. In power, a position is something you owe, and owe at a time. The time is not a delivery detail attached to an otherwise ordinary quantity; it is the product itself.
A megawatt-hour in one hour is not a megawatt-hour in another
Because energy cannot be moved between hours, two megawatt-hours in different hours are two different products, not two units of the same product.
In oil, a barrel you are long in June can be compared with a barrel you are long in July on the same curve, and the difference between them is a spread you can reason about and often trade. In power, the hour you are long is not power at a slightly different date; it is a different market with a different balance, and the overnight hour and the evening peak can clear at prices that have almost nothing to do with each other.
What ties the hours together is not a commodity but a machine: the same generators, lines and demand shape that make one hour cheap are what make the next hour dear. A position that is long the early-morning hour and short the evening hour is not approximately flat; it is two live positions wearing one volume. The habit of adding megawatt-hours together and calling the result a position is where most power books first go wrong.
You nominate at a node for an hour; there is nothing to load or ship
Physical delivery of power is a schedule entry, not a shipment. Nothing is loaded, shipped or stored; you nominate a quantity at a connection point for a specific hour, and performance is measured against that nomination.
A physical oil trade ends in a vessel or a pipeline parcel. A physical power trade ends in a nomination: an instruction to the system operator that, in a given hour, this quantity will be injected at this point or withdrawn at that one. The cargo is an entry in a schedule, and the schedule is the whole of the physical delivery.
Two consequences follow. First, there is no quality dispute of the kind that occupies physical oil settlements — power at a node is power at a node, and the question is whether the flow matched the schedule, not what the flow was made of. Second, and more important for the position, the trade's exposure is defined by where and when the schedule says, which is often not where and when the trade was negotiated in aggregate. A trader who buys a flat block of power for a month has not bought a pile of energy; they have bought a position in every hour of that month, and the book only shows the hours to whoever has gone and laid them out.
The same hour is a different product on either side of a constraint
A power price is locational. When a transmission line reaches its limit, the same hour clears at different prices on each side of the constraint, so a megawatt-hour is pinned to a node as well as to an hour.
Oil can be moved from where it is to where it is worth more, so a single location's price stays tethered to the rest of the world by the cost of moving it. Electricity moves only along lines, and lines have limits. When a line is full, the cheap power on one side cannot reach the expensive demand on the other, and the market separates: the two sides clear at different prices for the same hour.
The consequence for a position is the same as the hour's. Two megawatt-hours in the same hour but on opposite sides of a congested line are two different products, and netting them is as wrong as netting across hours. A power position is defined by three things — the hour, the node, and the direction — and any report that drops one of the three to make the arithmetic add up has discarded the position.
Every hour clears at the cost of the next unit, not the cost of yours
The clearing price of an hour is set by the most expensive unit that has to run to meet demand, and it is the same price for everyone who clears — whatever their own cost was.
In most commodity markets the price is found by bargaining over a physical thing. In power the price is found by stacking. To meet the demand of an hour, the system operator runs generators from the cheapest to the dearest until enough are running, and the offer of the last unit brought on — the marginal unit — becomes the price every generator and every buyer in that hour clears at. A plant with very low running costs does not sell its power at its own cost; it sells at the marginal unit's price. A buyer does not pay the average cost of what was running; it pays the marginal price.
This is why the market price of power can be so far from the cost of producing it, in either direction. The unit of the trade — a megawatt-hour — is priced by a mechanism that has nothing to do with what your particular megawatt-hour cost to make. Your cost is a fact about you. The clearing price is a fact about the last unit the system needed.
There is one more way the stack can clear, and it is the subject of the next section: when supply still exceeds demand even at a price of zero, the hour clears below zero.
A negative price is the market working, not a malfunction
When supply can be reduced only at a cost and demand is not there to take it, the market clears below zero: producers pay to keep injecting. A negative price is a clearing price like any other.
A negative price looks like an error to anyone raised on storable commodities, where a good cannot cost less than nothing because you could always just hold it. That option is exactly what power lacks. A generator that cannot cheaply ramp down, or a renewable plant whose fuel is already there, will pay to keep running rather than bear the cost of stopping, and in an hour with more supply than demand the only price that balances the two is a negative one.
Two things follow for anyone running a power book. First, a negative price is not a data problem to be corrected away; it is the market's answer to a real condition, and a book that refuses to believe it will mis-mark exactly the hours that matter. Second, the same mechanism that produces a negative price in one hour can produce an extreme positive price in the next, because both come from the same source: the product cannot wait, so every hour must clear at whatever number balances it, however far that number is from the last one.
Netting a book across hours discards the position while keeping the volume
A power book that is long one hour and short another can report zero volume and a live position. Netting megawatt-hours across hours adds up the quantities and cancels the exposure, because the hours are different products.
Take a book that is long a quantity of power in the overnight hours, when demand is at its lowest, and short the same quantity in the evening peak, when the last unit is setting the price. Sum the two and the volume is zero. Read the book by the hour and it is long a market that can clear near zero and short a market that can clear at the scarcity price. The netted line reports the first; the position is the second.
This is the power-specific version of a general trap, and it is worth naming precisely because the arithmetic is so persuasive. The unit of exposure in power is not the megawatt-hour; it is the combination of hour, node and direction, and any aggregation that discards those fields discards the position while keeping the volume.
The same trap runs through any commodity whose unit is not fungible. For physical crude it is grades and delivery windows rather than hours, and the failure looks identical: the netted number is right and the position is gone. Crude Oil Trading: Why the Barrel You Hedge Is Not the Barrel You Own
The useful test is the same one as in that case: ask of any net figure what single trade would close it. If the answer is a single trade, the netting was probably legitimate. If the answer is "several, in different hours and at different nodes, at prices we would have to go and discover", then the report has aggregated across a risk rather than netting within one.
The trades sit in the system; the hourly shape usually sits beside it
A power trade is captured as a contract; the hourly and nodal shape it produces is usually rebuilt in a spreadsheet beside the system, because the shape is a schedule of commitments rather than a property of the trade.
The contract records the headline terms: the volume, the period, the node or zone, the price. What it does not naturally hold is the shape — the trade spread out hour by hour, or node by node, with the position each hour leaves behind. A flat block for a month is a position in every hour of the month, and the profile of that position is a working artifact that has to be assembled from schedules, price feeds and nominations and kept current as each hour is traded away or passed.
That arrangement is not careless. It reflects that the facts which define the shape — the nomination, the constraint on a line, the clearing price of the last unit — arrive from outside the desk: from operations, from the system operator, from the exchange and from the price feeds. A spreadsheet is where someone can bring them together quickly. The consequence is the familiar one: the shape is right when the person maintaining it has received every fact that changed it, and there is no signal when they have not. A book that is a day behind its nominations looks exactly like a book that is flat, and only the second one is.
The same quantity, three different hours
| The hour | What sets its price | What a long megawatt-hour there is exposed to | When the exposure ends |
|---|---|---|---|
| An overnight hour | Demand at its lowest, supply that cannot cheaply turn off | A price that can clear near zero or below | When the hour passes |
| The evening peak hour | Demand at its highest, the last units that must run | A price set by scarcity, far above the overnight hour | When the hour passes |
| The same hour across a congested line | The clearing on each side of the constraint | Two different prices for the same hour | When the hour passes |
The third column reads the same in every row. The fourth column is three different exposures. The column that decides what to hedge is the fourth, and it is decided entirely by the hour and the node, not by the volume.
What a power position line must state before it can be hedged or settled
| Field | Why it is load-bearing |
|---|---|
| The delivery hour — or shorter interval — of each side | A megawatt-hour in one hour does not offset a megawatt-hour in another |
| The delivery node or zone of each side | The same hour clears at different prices across a constraint |
| Whether the trade is a flat block, a peak or off-peak block, or an hourly schedule | Determines what the position looks like hour by hour |
| The shape of the position across the delivery period | The exposure is the shape, not the total |
| Which market session each side clears or settles in | Day-ahead, intraday and real-time are different prices |
| The imbalance exposure left after scheduling | Being short in real time pays a price the day-ahead did not foresee |
A line carrying all six can be read by someone who was not in the room when the trade was agreed. A line without them can only be interpreted by whoever maintains the schedule.
Questions people ask about this
Why is power trading different from trading oil or gas?
Because the product cannot be stored. Oil and gas can wait in a tank or a cavern while you decide when to sell; electricity must be consumed or absorbed at the instant it is produced. Everything that makes power trading unusual — the hourly granularity, the extreme swings, the negative prices, the locational pricing — follows from that one absence. You are not trading a thing you can hold; you are trading a claim on an hour at a node.
Can we store power, and does it matter for our position?
Only at the margin. Batteries and pumped storage exist, but they time-shift energy rather than hold a stock the way a tank holds oil. For the purposes of a trading book, the practical answer is that a megawatt-hour cannot be carried forward, which means every position is a time position. It matters because a storable market lets you wait out a price; power makes you settle at whatever the hour clears.
What does it mean that our book is flat on volume?
Very little, by itself. Flat volume means the quantities you bought and sold add to zero, not that you are not exposed. A book long the overnight hour and short the evening peak is flat on volume and fully long a low-price market and short a high-price one. The test is to lay the position out hour by hour and node by node and see whether it is still flat. If it is not, the flat volume was hiding the exposure rather than describing it.
Why do we see a negative price in our book?
Because that hour had more supply than demand, and some of that supply could not cheaply be turned off. In a storable commodity the price cannot go below zero because you could always hold the good; power cannot be held, so the only price that balances the hour is one where the producer pays. A negative price is not a data error; it is the clearing price of that hour, and it should be marked, not smoothed away.
Why is the same hour worth different amounts at different places?
Because electricity moves only along lines, and lines have limits. When a line is full, cheap power on one side cannot reach the demand on the other, and the two sides clear separately. The same hour at two different nodes can be two different prices, and the difference is the cost of the constraint. A position report that averages them has replaced a real locational position with a single number that did not trade anywhere.
What should a power position report show?
The shape, not just the total. Each line should carry the delivery hour or interval, the node or zone, the block or schedule type, the market session it settles in, and the imbalance exposure left after scheduling. The point of the report is to show the book the way the market actually sees it — as a stack of positions in specific hours and places — rather than as a single quantity that happened to net to a convenient number.
Can we run a power book in spreadsheets?
Many companies do, and the spreadsheet is rarely the thing that fails. What fails is that the facts which shape the book — nominations, constraints, clearing prices, the profile of each block — arrive from different places and keep moving as each hour passes, so the spreadsheet holds a snapshot of what somebody transcribed rather than a view of what is true now. The scale at which this stops working is not a volume threshold; it is the point at which no single person can still hold every open hour of the book in their head.
Where this lands in a trading system
Nothing above is an argument for buying software. It is an argument about which facts a power position is made of and where each of them first exists. Parts of it map onto system capabilities and parts of it do not.
The first conclusion is that the physical trades and the hedges against them have to be held together before a position can be laid out hour by hour at all. That is the job of a CTRM/ETRM system providing complete trading lifecycle management for commodity traders, integrating physical trade, financial hedging, risk control, and settlement in one platform; Time Dynamics' Fusion is one, and the parts of it that bear on this article are that complete physical trade lifecycle management and execution and advanced trading management for futures and derivatives markets sit in the same place, with comprehensive risk assessment and monitoring capabilities applied across both.
The second conclusion is that the facts which shape a power position — the nomination, the constraint on a line, the clearing price of each hour — mostly arrive from outside the trading desk. That is a different problem, and it is the one X-Ray addresses: a non-invasive data processing and analysis platform, whose collection toolkit XDK performs non-invasive automated data collection from databases, Excel files, and web interfaces, and which is built to collect data without disrupting existing systems or workflows. The last property is the relevant one here, because the spreadsheet holding your hourly shape is working, and a plan that starts by taking it away from the person who maintains it will not survive the first volatile week.
The third conclusion is that the same book has to be read several ways — by volume, by hour, and by node — and that the hourly view is the one most often rebuilt by hand. X-Sheet does one-click generation of personalized reports with Excel-like interface.
And here is where it stops. None of this decides which hours to be long or short, and none of it tells you what the evening peak will clear at. It will not predict where a line will congest, it will not choose your nomination for you, and it does not know that a constraint has moved until someone tells it. What a system can settle is narrower and duller than the problem in this article: where each fact about the shape is captured, how soon after it becomes true, and whether the person who needs today's hourly position can see it without asking the person who maintains the schedule. That is a real part of the gap. It is not the whole of it, and it is worth being clear about which part you are buying.
A note on the evidence in this article
This article contains no statistics, and that is deliberate. The observations behind it — that power cannot be stored at grid scale, that the hour and the node are the units of exposure, that a negative price is a genuine clearing price — come from how electricity markets work, which we are not in a position to publish here as a citable figure. A number nobody could check would make the argument look stronger and would make it worth less.
So the argument runs on mechanism, and each claim is written to be tested against your own book rather than taken on our authority. Three tests are proposed above and they are the intended use: lay the open book out hour by hour and node by node and compare it with the volume-only view, ask of any net figure what single trade would close it, and check whether the hourly shape you are trading against is current with your latest nominations. If your own book contradicts what is written here, your book is the better evidence.
A flat power book is not a flat position. The hours are the position.