Instruments and Hedging

Electricity has more traded instruments than its physical properties should allow, and each of them exists because some risk in the preceding chapters had to go somewhere.

Blocks, because hours are not fungible

A barrel is a barrel whenever it is delivered. A megawatt-hour at 6pm in January and a megawatt-hour at 3am in April are different products with different prices, so power forwards trade in shaped blocks rather than as a single commodity.

The three standard American time block products are on-peak, conventionally sixteen hours a day (usually 7am to 11pm) across five weekdays excluding the NERC holidays, and written 5x16, and around the clock, written 7x24, with an off-peak block making up the difference. A trader hedging a daytime industrial load buys on-peak. A baseload plant sells around the clock. The distinction sounds administrative and is the beginning of every real hedging problem, because almost no physical position matches a standard block exactly.

The congestion hedge

Chapter 10 left a generator with a problem. It settles at its own node, the market trades at a hub, and the difference between them moves for reasons outside its control.

Financial transmission rights (FTRs), called congestion revenue rights in some markets, are the instrument built for exactly that. A right between two points pays its holder the congestion component of the price difference between them. A generator stuck behind a constraint buys the right from its node to the hub, and when congestion widens the discount on its energy, the right pays the difference.

The elegant part is where the money comes from. Chapter 10 described the congestion rent the grid operator collects when a constraint becomes actively limiting, taking more from load than it pays to generation. That rent funds the rights. The physical market generates precisely the surplus needed to settle the financial claims written against it, which is why the two were designed together and why a nodal market without transmission rights would be close to unhedgeable.

Rights are allocated through periodic auctions, and their revenue adequacy depends on the operator making available roughly the same network it sold rights against. When outages or derates shrink the network, collected rent falls short of the claims, and the shortfall is prorated back to holders. A hedge that pays less than it should in exactly the conditions it was bought for is a known weakness of the design.

Betting on the two-settlement gap

Chapter 9 described a market that clears twice, once a day ahead against a forecast and once in real time against reality. Virtual bids let a participant with no generation and no load trade the difference between them.

A virtual supply offer sells energy in the day-ahead market with no plant behind it, and buys the position back at the real-time price. A virtual demand bid does the reverse. Neither party ever produces or consumes anything.

The reason operators permit this is convergence. If the day-ahead price at a location sits persistently above real time, virtual sellers arrive until it does not, and the two prices are pulled together by people whose only interest is the gap. A recurring divergence between the forecast market and the physical one is an arbitrage, and letting speculators close it produces a better day-ahead schedule than the operator would get by insisting only physical parties may participate.

The long-dated problem

Everything above hedges months or a few years. Financing a power plant requires certainty over fifteen or twenty, and no liquid forward curve extends that far.

The power purchase agreement fills the gap. In its physical form a buyer contracts to take output at an agreed price. In its virtual or financial form, which is what most corporate renewable buyers sign, no electricity changes hands: the generator sells into the market as normal, and the two parties settle the difference between the market price and the strike. That structure is a contract for difference, which is what Chapter 19 described Europe and China both adopting as public policy, arrived at independently by private counterparties for the same reason.

Two risks make power purchase agreements harder than they appear: basis (location) and shape (volume).

Basis risk is the first. The contract references a hub and the project settles at a node, so the developer bears the difference. Chapter 10 gave this its due, and it has damaged more renewable projects than construction overruns.

Shape risk is the second, and it is what makes hedging a wind farm harder than hedging a nat gas plant. A generator selling a fixed block has to deliver that block. A wind farm produces what the weather provides, so in hours when it generates less than it sold, it must buy the shortfall at the prevailing price, and the hours it underproduces correlate with the hours everyone else does too, which are the hours prices are highest. Selling a flat hedge against a variable output creates an exposure that is small most of the time and severe precisely when it matters, which is why proxy generation structures and volume-following contracts exist.

The instrument for load that never arrives

Weather derivatives close the loop back to Chapter 7. They settle against accumulated heating or cooling degree days at a named weather station, and they exist because the largest uncertainty facing a retailer is how much its customers will want rather than what the power costs.

A retailer that has hedged its expected volume perfectly is still exposed to a mild winter, in which it holds contracts for energy its customers did not buy. Price hedges do nothing about that, since a mild winter tends to lower the price as well, leaving the retailer long energy it cannot sell in a falling market. A degree day contract pays out in exactly that case, which makes it a hedge on quantity in a market where quantity is decided by weather.

Which is the pattern across this whole chapter. Every instrument here exists because one specific property of electricity, established in Part One, created an exposure that no ordinary commodity contract could carry.