Submitted on March 27, 2017

Our Price Path is a forecast of electricity spot prices at over 200 nodes on the national grid.  The latest version was released last week and covers the period from April 2018 to March 2033.

By Greg Sise, 22 March 2017

The forecast consists of a distribution of forecast prices over the full forecast horizon, and the “headline” price path is the weighted average of the 860 scenarios included in the forecast data.  To take a much simplified example, suppose we had three scenarios for the average price at Otahuhu for 2018 consisting of $100 with probability 15%;  $85 with probability 30%;  $70 with probability 35%;  $45 with probability 20%.  Then the weighted average headline price path forecast for 2018 would be $100 x 0.15 + $85 x 0.3 + $70 x 0.35 + $45 x 0.2 = $59.15/MWh.

Price Path

The forecast is based on fundamental models of the factors that drive, and will drive spot prices into the future, and the scenarios include the impact of variability in inflows into the nation’s main hydro lakes based on 86 years of data going back to 1931.

While the forecast is for the electricity spot market, the drivers include inputs from the natural gas market (gas price path), the carbon market (carbon price path), the global aluminium market, the global oil market, the global methanol market, the foreign exchange markets, global steel markets, global coal markets, and key parameters including domestic inflation, and how New Zealand’s GDP will grow.  We also observe and predict how the technology markets relevant to electricity will evolve over time, for example the cost of wind generation, the cost of geothermal generation, the cost of solar PV, and the costs of battery technology for use in solar systems and in electric vehicles.

There is obviously no certainty about the future, which is why the price path includes a wide range of possibilities, as contained in the 860 scenarios.  It would be much easier for us just to model the scenarios without ascribing any probabilities to them, thus requiring our forecast subscribers to come to their own conclusions about how to weight the scenarios to produce a price path.

But this is not what we do.

Our subscribers want to know what we think the future holds, at least in a probabilistic sense, which places a duty of care on us that we take very seriously (often in the middle of a sleepless night!), and which disciplines our entire forecasting process.  A key part of this is monitoring our forecast accuracy and adjusting and improving the forecasts over time.

So we always present our views on the likelihood of our forecast scenarios, and include the weighted average headline price path.  But subscribers are always free to disagree and to create an alternate price path from the forecast scenarios that we provide, using their own probability assessments.  For example, a subscriber may view the weighting we put on our “Low Carbon” scenarios as being too low, so they could increase the weighting on this to create their own Price Path based on the modelled scenarios.

However, a more common approach amongst subscribers is to take the headline Price Path “as read”, but then they test their energy strategy using percentiles from the Price Path data.  For example, they might be looking to build new generation, and look at the 25th percentile of the Price Path to ask the question:  if prices turn out lower than the Price Path forecast, is our project still viable?  Because the future is uncertain, it is wise to consider the range of scenarios that we provide in each Price Path, not to just focus on the headline numbers.