We have often pointed out that ASX futures prices don’t always appear to behave rationally. In this post we look at the spread (difference) between the ASX futures prices at Benmore and Otahuhu.
ASX futures contracts are referenced to Benmore (BEN) in the Waitaki Valley in the centre of the South Is, and to Otahuhu (OTA) in Auckland. Most of the trading on the ASX is in quarterly baseload contracts, from this quarter out to the fourth quarter of 2025. The spread in prices between the BEN and OTA contracts in each quarter should, in theory, reflect expectations of the difference in the quarterly average spot price at these two nodes. There are different types of spreads in futures markets, so we’ll call the OTA-BEN spread a ‘nodal spread’.
Near the end of the current quarter, futures prices always converge to be very close to the actual average spot prices at BEN and OTA, and the nodal spread converges to the actual price. But when we look at the ASX, we see that the nodal spreads are never less than $20/MWh beyond the current quarter. The chart below shows the nodal spreads at the close of trading on 10th March, and they range from $22/MWh to $33/MWh in round figures.
So, if futures prices reflect expectations of future spot prices, then spot spreads between OTA and BEN must be in this range, right?
Wrong! When we look back over most of the past decade to 2014, the first year the HVDC link was upgraded, removing restrictions on HVDC flows and leading to smaller spreads. What we see is that the average spreads are consistently well below $20/MWh.
The chart above shows the nodal spreads in spot prices by month from 2014 to the start of the fourth quarter of 2021. Now we have to exclude the first quarter of 2020 from the averages shown, because the HVDC link was under repair and only one pole (half) was available, which means that spreads between OTA and BEN were higher than one would normally expect.
But with this exclusion, what we see is that the average nodal spread is considerably less than $20/MWh. Up until near the end of 2018, they averaged $6.9/MWh, but since gas supply problems became a major feature of the market, they’ve averaged $16.4/MWh (excluding Q1 of 2020).
What does this tell us about futures prices?
Maybe futures traders expect the nodal spreads to be higher, and are constantly astonished when they aren’t? Possible, but it doesn’t seem unlikely: futures traders can’t be that naive.
Maybe futures traders consistently expect spot prices at BEN to be lower than they actually are, or spot prices at OTA to be higher than they actually are? But that’s just the same as expecting the spread to be greater than it is.
Maybe BEN is traded at a discount to expected spot prices. Or is OTA being traded at a premium? Or could it be both? This seems a more likely explanation, especially as the research we’ve done in the past shows that some quarters tend to trade at a premium or discount to expected spot prices.
For example, if futures buyers are eager to reduce their exposure to spot prices during periods when the spot prices might spike higher – think winter when demand is highest – then they might be prepared to pay a premium to futures sellers to achieve this. Or if sellers have higher costs and risks during these periods, maybe they need to sell at a premium, which buyers are prepared to accept. Higher costs and risks could be a factor when we have dry periods, in which northward flows on the HVDC link reduce, and we get more southward flows.
The problem with this explanation is: If the premium is the same at BEN and OTA, then the nodal spreads in the futures prices should not be as wide as they are. This suggests that any premiums or discounts, relative to expected spot prices, are different at BEN and OTA. Perhaps the risk is perceived as being higher at OTA and less at BEN, resulting in OTA trading at a premium to BEN. Or maybe the BEN risk is perceived to be less than the OTA risk.
Premia in futures prices is a fascinating topic, and upcoming futures posts will look at this in a wider context. But in the meantime, what do the futures nodal spreads indicate for buyers and sellers? Suppose you wanted to hedge the nodal spreads between OTA and BEN for some reason, for example, if you generate (sell power) in the North Is and retail (buy power) in the South Is: you could do this by selling futures contracts at OTA and buying futures contracts at BEN, both in the same quarter. If you do this more than a quarter in advance, it is probably safe to conclude that you will pay a premium for this hedging strategy, on average, of at least a few dollars per MWh. That might not sound like much, but relative to the actual nodal spreads, it is a significant percentage premium to pay.
There are other ways to hedge the nodal spreads, so you could look at buying Financial Transmission Rights between BEN and OTA, but that’s a topic for a future post.