“Are there any risks in using CFDs to hedge wholesale electricity prices?”
In recent weeks we looked at the challenges of price volatility for buyers and sellers trading on the wholesale electricity market. We also saw how a hedge, in the form of a Contract for Difference (CFD), can help reduce price volatility. You’ll recall that a CFD works to fix total price because the counterparties agree to exchange sums of money based on the difference between a fixed contract price and a floating wholesale price.
Standard CFDs are written with five fixed elements: start and end dates, a volume (typically in MW), a fixed price and a location to reference the floating price. Here’s an example:
Start date:1 July 2022
End date: 30 September 2022
Volume: 2.5 MW
Location: HAY2201 (the node on the electricity transmission grid just North of Wellington at one end of the cable linking the North and South Islands)
So, in principle, with this CFD a generator selling 2.5 MW into the wholesale market can expect to receive total revenue for Q2 2022 of $828,000 ($150/MWh). Likewise, an electricity retailer or consumer buying 2.5 MW at spot price in the wholesale market will expect to pay a total of $828,000 for offtake.
So, with a CFD I am guaranteed a fixed price for all the electricity I sell into or buy from the wholesale market!
Well, perhaps not!
It turns out the CFD only works perfectly (as described in principle above) under limited circumstances. For a generator this is when the volume and location in the CFD match exactly the physical quantity and point of injection of power production. For the consumer or retailer, the physical offtake must also match the exact quantity and location written into the CFD. Whenever the traded location and volume cannot match the CFD, the hedging effect falls away from being 100% effective: which leads me to introduce the very important concept of ‘residual risk’.
It is easy to see that production from a solar array or windfarm will not be able to produce a steady 2.5 MW of electricity day and night, rain or shine. A large industrial consumer may need to adjust production levels at short notice in response to customer orders, and for a retailer with geographically dispersed customers the wholesale prices paid for offtake across a number of locations may not move in perfect step with the price at the CFD reference location.
The effect of wholesale prices at different locations not moving in parallel is illustrated in the chart to the right which shows wholesale prices by half hour for a randomly selected winter day (Monday 20 July 2020). You’ll notice that up until 7 a.m. prices in Dunedin (green line) were highest while the lowest prices were in Auckland (yellow line). As the morning demand peak approaches, Auckland and Dunedin change places; a pattern maintained until after the evening peak. During the middle of the day prices drop back with Dunedin and Wellington tracking close to each other but Auckland remains noticeably higher until after the evening peak.
Now let’s see how our example CFD referenced to Wellington would have performed on this day. The chart to the right shows the total unit price (including purchases or sales at spot prices, plus the cashflow from the CFD) by half hour for a steady 2.5 MW of electricity sold into or purchased from the wholesale market in Dunedin, or Auckland, or Wellington.
As you might expect the CFD works very well for a wholesale trader in Wellington (grey line), providing a consistent $150 MW/h throughout the whole day.
The outcome is not so stable for Auckland and Dunedin.
During the early part of the day prices across the country are lower than the $150/MWh contract price so the buyer pays the seller the difference. (remember that amount to be paid under this CFD is based on the difference between the contract price and the spot price in Wellington). So, during the early hours, when the wholesale price is lower in Wellington than in Dunedin, the difference between contract and spot price is greatest for the spot trader in Dunedin, causing the total Dunedin half hourly unit price to rise above $150/MWh. The situation reverses when the spot price in Dunedin falls below that in Wellington.
Also, after 7 a.m. Auckland quickly reaches the highest spot price. This is most evident during peak demand periods, when the total unit price in Auckland tops out at $165.04 for the most expensive half hour.
In this example the average price across the whole day was $150/MWh for Wellington, $149.88/MWh for Dunedin and $155.56/MWh in Auckland. Due to the location of generation in New Zealand, there is generally a wholesale price gradient sloping downwards from north to south, however this example still illustrates the residual risk of using a CFD to hedge wholesale trading in locations remote from the CFD’s refence location.
So, what! I can safely ignore the locational risk simply by hedging at the same place as I trade electricity, Right?
Consider an independent retailer with a residential customer base restricted to the Wellington region. The profile of a daily average 2.5 MW of residential demand in the Wellington region on a Monday in mid-July might look something like the next chart.
The area in orange indicates half hours when the retailer’s customer demand is greater than the 2.5 MW volume in the hedge.
Let’s see how the CFD would perform for our retailer (we’ll just change the contract price to $100/MWh for the sake of simplicity in this example).
The bars in the chart to the right represent the total half hourly price in $/MWh of electricity the retailer takes from the grid to supply residential customers under the CFD. The segments in orange highlight half hours where the hedging effect of the CFD is less than prefect and the total price is greater than the $100/MWh contract price.
What is happening here is, despite average demand for the day matching the volume in the CFD (2.5 MW), actual demand changes by half hour throughout the day. When periods of highest demand align with the periods of highest wholesale price any half hourly demand greater than 2.5 MW is left ‘exposed’ to the wholesale price. The dotted line represents the total (spot + CFD) cost per half hour. The increases when unhedged demand coincides with high spot prices clearly reduce the effectiveness of the CFD.
In this example the total price averaged over the day is $103.55/MWh; an increase of 3.5% over the $100/MWh that a perfect hedge would have provided.
With a $3.55/MWh (3.55%) increase over the contracted price, does this mean a CFD is a poor tool for managing price risk?
Well, the total price for the day without any hedging would be $142.39/MWh – So, I’ll leave you to answer that question.
1. For nearly all practical purposes a standard CFD will be effective but not perfect at eliminating the effects of wholesale price volatility.
2. For a standard CFD the residual risks arise wherever the physical volumes traded in the wholesale market vary in quantity and location from the corresponding criteria in the CFD.
There are several ways the residual risks associated with the trusty CFD can be mitigated and we’ll return to some of the most common in a later blog. First though, consider that there are times when a CFD performing just as expected, nonetheless delivers a sub-optimal outcome. This most often results when the view of the future at the time the contract was written materially differs from what actually takes place. Which brings us to the central role played by forecasts in hedging. A topic for the next blog.