Timera take on the 1st UK capacity auction

Posted on January 12, 2015January 15, 2020 by David Duncan

The first UK capacity auction was concluded just in time for Christmas, with 49.3 GW of capacity procured at a clearing price of 19.40 £/kW. The auction results gave little in the way of Christmas cheer for most UK generators, with the clearing price close to half that of market consensus expectations. However the first UK auction marks the start of an important transition of Europe’s larger power markets towards market based mechanisms to remunerate flexible capacity.

First auction headlines

The conditions that set up the downward price pressure in the auction stemmed from a relatively low government capacity target that saw an ‘oversupply’ of existing capacity. Existing capacity volume (54.9 GW) exceeded the procured volume in the auction (49.3 GW) by 5.6 GW. 2.8 GW of new capacity was successful in obtaining capacity agreements despite the low auction clearing price. But this meant 8.4 GW of older existing coal and CCGT plants failed to secure a capacity agreement, leaving plant owners in a precarious position.

The auction outcome is being heralded as a success by the government (‘capacity procured cheaply for the consumer’). But it is unclear whether the first auction has done anything to improve UK security of supply over the critical period of market tightness from 2015-18. In fact the outcome may have exposed one of the key weaknesses in the capacity market design, where capacity is procured on the basis of uncertain forecasts of conditions four years in advance. This leaves little ability for the capacity market to respond to market tightness over the next 3 years.

National Grid has published plenty of data on the 1^st auction outcome. Rather than duplicating any of this analysis, our intention in this article is to focus on the key lessons learned from the auction and the implications for evolution of the UK power market.

How did the auction results match our expectations?

Prior to the 1^st auction we published a ‘First Auction in Focus’ report. The following are excerpts from the summary section of this report which give a quick overview of our analysis ahead of the auction:

Marginal plant: 3 key plant types are likely to drive the 1^st auction outcome (older coal, older CCGT, low capex peakers).
Pricing: Our analysis indicates a 1^st auction capacity price around 30 £/kW if participants bid rationally to recover costs. But the 1^st auction outcome will come down to EM (&CM) expectations (diverse range likely across players).
Downside risk: The low 1^st auction target and ‘Fear of Missing Out’ dynamics may lead to a lower clearing price than expected. These factors could easily combine to reduce the 1st auction clearing price by 5-10 £/kW.
Older plant: 6-8 GW of older CCGT/coal to be unsuccessful in 1^st auction → most of this uneconomic without capacity returns.
Energy market impact: It’s likely a significant volume of older capacity is closed/mothballed as a result, supporting wholesale energy market generation margin recovery.

The auction outcome was broadly consistent with these expectations. However the downward price pressure dynamics were somewhat stronger than we had expected, resulting in a lower clearing price (19.40 £/kW) and higher level of unsuccessful older plant (8.4GW).

We also expected a higher volume of coal capacity to be successful in securing 3 year refurbishment agreements with a view to covering the costs of IED capex. The fact that a number of existing coal plants missed out on refurbishment agreements raises a query as to the economics of IED cost recovery if these plants are going to remain open in the 2020’s.

There was around 1 GW of new build peaking & unproven DSR capacity which was in line with our expectations. But we were surprised by two factors relating to new build CCGT:

The Carrington CCGT project did not secure a capacity agreement despite already being under construction. Sunk cost dynamics were perhaps trumped by expectations of higher capacity returns in the future.
One of the other new CCGT projects, Carlton Power’s Trafford plant (next door to the Carrington plant), was successful in the auction. This suggests that there may be some unique benefits to this project (e.g. low capex, synergies with the Carrington project construction) as well as one of the parties to the project taking a very optimistic view of the evolution of the wholesale energy market.

The success of the Trafford CCGT project raises an important question going forward. Was this an anomaly, or can we expect significant volumes of new build CCGTs at lower capacity prices (e.g. < 40 £/kW). We suspect the former. We also note that it is one thing to secure a capacity agreement, but this does not guarantee an ability to raise the capital, secure offtake contracts and construct and commission the plant.

Lessons from the auction results

While the specific bids of each plant have not been released, Grid has published a representation of the supply stack shown in Chart 1.

Chart 1: 1^st auction supply stack

Source: National Grid

Some useful conclusions can be drawn from this supply stack, particularly by focusing in on the margin (where supply and demand intersect):

As expected, around 30 GW of existing capacity did not require capacity support and was bid into the auction at zero price.
The supply curve looks to be quite steep above the marginal clearing price. For example if another 4 GW of capacity had been procured, the clearing price would have risen towards 35 £/kW.
There looks to be a reasonable level of price support (5-6 GW of capacity) below the clearing price around the 15-20 £/kW level. This is likely to represent owner bidding to reflect the direct fixed cost recovery requirements of thermal plant.
The ‘blockiness’ of the supply curve above 35 £/kW suggests the more expensive tail of the stack was dominated by a number of larger thermal units (both existing and new build).
The finer granularity definition of capacity volumes in the 20-35 £/kW bid range suggests that there may have been a reasonable supply of smaller scale peaking assets sitting above the clearing price, but below many of the larger thermal unit bids.

From a capacity perspective the auction result headlines are that:

8.4 GW of existing plants failed to secure agreements: comprised of
1. 3.9 GW of older CCGTs (Barry, Brigg, Killingholme A&B, Peterborough, Corby, Deeside & Peterhead)
2. 4.5 GW of older coal plants (Rugeley, Eggborough, Ferrybridge and 1 unit of Fiddlers Ferry & West Burton)
2.6 GW of new plants secured agreements:
1. The 1.8 GW Trafford CCGT project
2. 0.9 GW of smaller scale peaking plants (e.g. diesel gen sets, reciprocating engines)
3. 0.2 GW of unproven DSR capacity

The subsequent actions of plants that failed to secure a capacity agreement are likely to be as important as the auction outcome itself.

Implications for UK power market evolution

Without any capacity price support, the economics of many of the older coal and CCGT plants that failed to secure an agreement do not look healthy. This is particularly true of the 3.9 GW of older CCGTs which have been suffering cash losses for several years now. Owners have been holding on for capacity payments. So it is likely that a weak auction clearing price will crystallise plant economics. Expect plant closures and mothballing over the next two years.

Somewhat ironically, if a substantial volume of plant that was unsuccessful in the auction closes (e.g. 4-5 GW), the auction may have actually worked to undermine security of supply over the next three years rather than improve it. In this scenario it is likely Grid will take action to secure additional reserve capacity. This brings the new Supplemental Balancing Reserve (SBR) contracts into focus. While Grid can use SBR as an emergency capacity back stop, SBR contracted assets are removed from competing in the wholesale energy market supply stack. This means plant closures are likely to drive higher rents and generation margins in the energy market despite SBR contracting (as we set out here).

In a sense SBR may become a temporary capacity market over the 2015-18 period, albeit one that lacks transparency and a clear set of rules. As we have said many times previously, once the government starts overlaying complicated market interventions, the unintended distortions that result tend to generate the requirement for further intervention.

Welcome back

Posted on January 5, 2015January 15, 2020 by David Duncan

Happy New Year and welcome back. The festive season is often a quieter time for energy markets. This has certainly not been the case over the last month. Brent crude oil has continued its precipitous decline, falling another 15 $/bbl since early December to under 53 $/bbl, almost halving in value over the last three months. The shockwaves of this move are still feeding through into the global LNG and European gas markets. But there can be no doubt now that we are moving into a more structural period of hydrocarbon oversupply. This will have far reaching implications for LNG spot pricing & supply growth, oil-indexed gas contract prices, hub price dynamics and gas vs coal plant competitive balance. The commercial and market implications of this shift will be a key theme for this blog in 2015.

The other major event in European power markets over the last few weeks was the inaugural UK capacity market auction. There are some interesting lessons to be learned from the auction results, given a policy shift across Europe towards market based mechanisms to remunerate capacity. We will publish our first feature article on Monday 12^th January which will explore the auction outcome and implications for wholesale energy market evolution.

In the meantime we wish our readers all the best for a prosperous 2015.

2014 themes and the way forward into 2015

Posted on December 8, 2014January 15, 2020 by David Duncan

With Christmas approaching this is our last article of 2014. We will be back with more in early January. To finish the year, it has become a bit of a tradition for us to look back at the key market and commercial themes of the past year. Then to draw on these themes to look forward into the next year.

2013 looking forward

In Dec 2013, we published our end of year article with Brent crude around 110 $/bbl and Asian LNG spot prices above 18 $/mmbtu. At the time there was a strong consensus view that the gas and oil markets were tight and projected to remain so well through this decade. We finished 2013 with a list of potential energy market ‘shocks’ to consider coming into 2014:

A period of surplus in the spot LNG market causing Asian spot prices to fall to a level where flexible supply flows back into European hubs.
Another period of significant disconnect between European gas hub prices and oil-indexed contract prices (similar to 2009-10).
A major slowdown in Chinese economic and industrial growth, e.g. reducing gas import demand and inhibiting a policy shift from coal to gas-fired generation.
A significant fall in gas prices relative to coal, shifting the competitive balance back towards gas-fired generation.
A prolonged period where oil prices fall back below 80 $/bbl.
A pronounced policy shift away from support for low carbon generation capacity.

This was not a list of forecasts but a set of scenarios that we thought warranted prudent consideration. Most of these shocks have either transpired or sound a lot more plausible now than they did at the end of last year. In fact weakening oil and LNG prices have become a reality in 2014 with a profound impact on energy portfolio exposures. The knock-on effects are being felt across European gas and power markets and will likely determine the way forward into 2015.

2014 looking back

LNG market:

Both the short term and the longer term balance in the LNG market have shifted towards oversupply in 2014.

The LNG spot market is an increasingly good barometer of shorter term global gas market balance. A 50% decline in spot prices since January provides a good indication of the profound shift that has taken place across this year. The big move lower came in Q2 with prices falling towards 10 $/mmbtu. Cargoes flowed back into Europe as the Asian diversion arbitrage dried up and as LNG was sold into European hubs as a market of last resort. Any theories that this was a seasonal phenomenon have been dashed by renewed spot price weakness in Q4.

Perhaps more importantly the longer term balance also appears to be shifting towards oversupply. The LNG market has now weathered the ‘drought’ of new supply across the 2011-14 period. From 2015 new liquefaction projects, particularly in Australia and the US, start to ramp up. Not all of these projects are fully contracted suggesting more pressure to come on spot prices.

Towards the back end of the decade Russian pipeline gas poses an increasing threat. Russia signed framework agreements for a massive 68bcm of gas exports to eastern and western China in two separate deals this year. That is likely to put a substantial dent in China’s LNG import growth appetite from later this decade. It also puts Russia in the box seat to undercut LNG producers in making incremental sales to China going forward.

These factors are shifting the LNG market balance back towards buyers. A Japanese buyer (Chubu Electric) recently signed a 20 cargo deal with a partially spot indexed structure. Look out for an evolution in LNG contract structures and pricing terms if market oversupply continues.

European gas markets:

Events at European gas hubs in 2014 have been strongly connected to the LNG spot market. In a year where demand has remained relatively weak and pipeline supplies robust, an increase in LNG imports has weighed on prices.

The fall in hub prices over the summer of 2014 was on a similar scale to the slump in spot LNG prices. European hubs acted as key price support for the global gas market across the summer. But hub prices fell well below long term oil-indexed contract prices. This oil vs hub price divergence has eased into the winter, but it foreshadows problems that may lie ahead for suppliers who remember the pain of 2009 & 2010.

Chart 1 provides an illustration of historical TTF price evolution through the post-Fukushima phase of market tightness (2011-13) and the transition towards oversupply in 2014. It also indicates the current state of play in the forward market, with TTF pricing up towards oil-indexed contract levels in the winter periods.

Chart 1: TTF historical and forward price benchmarks

Source: Reuters

The events of 2014 have also seen the start of a recovery in seasonal price spreads and volatility. Interestingly this has occurred against the backdrop of falling hub prices, illustrating that oversupply does not necessarily reduce gas flexibility value. In the past wider spreads and higher volatility have tended to coincide with periods of higher hub prices. The pickup in spreads and volatility has also seen a renewed interest in gas flexibility products such as storage capacity.

As the year draws to a close, the European market remains focused on the threat of Russian supply cuts. But with robust gas storage levels across Europe and pronounced weakness in spot LNG import prices, this threat looks much more manageable than it appeared to be earlier in the year.

European power markets:

One of our themes in 2014 has been on the implementation of capacity markets across Europe. It is becoming clear that capacity markets will be the mechanism of choice to support thermal capacity in a world of low carbon support.

The UK is the first large European power market to lead the way with an auction this month. France is next and Germany looks to be following later in the decade. Market designs are likely to differ significantly and these differences may have profound impacts on generation margins and wholesale power price dynamics. The interaction between capacity and energy market pricing is a big story for power station owners and investors going forward.

We have also focused this year on the evolution of thermal power station value in Europe. The world of guaranteed baseload running has gone. As renewable output increases, gas and coal plants are increasingly playing a mid-merit and peaking role. These assets are essentially strips of options on the clean spark or dark spread. Asset ownership will likely evolve over time towards players who are able to understand and monetise the risk associated with this optionality.

Power asset transactions have started to pick up in 2014. Change in ownership patterns are suggesting a shift from utilities towards independent generators and investment funds. This is partly driven by balance sheet constraints and partly by a continuing negative outlook on generation margin returns. There have been some big portfolios on the block e.g. Vattenfall’s German assets and E.ON Spanish assets (recently purchased by Macquarie). E.ON’s announcement last week that it will spin off a separate listed entity containing its generation assets and trading business is another indication of utility appetite to shed generation asset exposures. We expect more to follow in 2015.

Looking forward into 2015

We ended last year with a list of potential shocks that could threaten market consensus. A number of these shocks have been delivered in 2014 (driven by falling oil, LNG and gas hub prices). We suspect that the market transitions taking place in 2014 will continue to shape events through 2015.

OPECs battle for market share against a backdrop of weak demand suggests there may be a sustained period of lower crude prices. Crude price recovery may require significant volumes of marginal supply to be driven out of the market as a result of lower prices (e.g. US shale oil).

For the LNG and European gas markets, lower crude prices are already a ticking fuse given lagged long term contract oil-indexation. This may be compounded by a spot market oversupply of gas, particularly coming into next spring. There is currently a lot of gas stored in anticipation of spot market tightness over the winter. If that fails to transpire, stored gas and increasing LNG flows may send spot prices spiraling lower in Q1.

As new LNG production ramps up in 2015 there looks to be a key risk that the trends of 2014 continue, potentially leading to a gas glut similar to that of the 2009-10 period. Important for European suppliers will be to what extent gas hub prices disconnect from oil-indexed contract prices again.

For European power markets, the factors above may lead to some rebalancing in gas vs coal plant competitiveness. Across the 2012-14 period coal prices have slumped to levels below the long run marginal cost of production. This has moved coal plants up the merit order to have a clear variable cost advantage over CCGTs. But if gas prices decline relative to coal prices in 2015, this may support an increase in gas plant load factors and margins, particularly in markets with a higher proportion of gas-fired capacity (e.g. the UK).

All of the above is hypothesis based on the impact of a continuation in the trends set in place in 2014. But the future is no more certain now than it was at the end of last year. We finished 2013 by emphasising the importance of stress testing asset & portfolio returns to understand the potential impact across a plausible range of market outcomes. That logic applies just as much now as it did then.

Happy Christmas

2014 has seen our readership approximately double, with our base of regular readers rising towards 10,000. We have again been widely published in industry journals and the press as well as speaking at a several conferences. We appreciate your support and look forward to bringing you more in 2015. In the meantime we wish you all the best for a relaxing break over the Christmas period.

Market interconnectivity and the next 6 months

Posted on December 1, 2014January 15, 2020 by David Duncan

The term ‘big price move’ is used too liberally in relation to energy markets which have a relatively high level of ambient volatility. But last week’s decline in the Brent crude curve was a big move by any standards. This decline in Brent will have profound repercussions for an already weakening LNG market. And these will feed through into European gas and power markets in 2015.

OPEC’s Saudi led Middle Eastern producer block has thrown down the gauntlet to non-cartel producers. By maintaining production targets and implicitly accepting associated price declines they have initiated a battle for market share with both the US and Russia. As a result the front month Brent contract crashed through 70 $/bbl on Friday. And more importantly, the front 3 years of the Brent curve has re-rated towards $80/bbl.

The move lower in Brent will act to drag down long term LNG & European gas contract prices with several months lag. This increases the chances of a substantial decline in European gas hub prices in 2015. It also opens up the potential for a return to supply glut conditions (similar to 2009-10).

In this environment of ‘tectonic’ movements in energy prices, market interconnectivity plays an increasingly important role. European gas hubs may play an important role in stemming the declines in spot LNG prices. In turn European power markets may act to provide key support for European gas hub prices.

An update on the LNG spot price fall

Last week we looked at some of the drivers of this renewed slump in LNG spot prices. The lead up to the last few Christmas periods has become associated with robust demand and rising prices. But this year Asian buyers have ample supply, storages are full and portfolio players are long LNG. As a result spot prices have crashed back to pre-Fukushima levels.

Chart 1 is an update of Reuter’s Asian LNG spot vs UK NBP price chart that we showed last week. The differential between these two prices is falling towards zero. That means Europe is becoming a much more attractive place to ship LNG, despite it being winter. This is going to be a key chart to watch over the winter as an indicator of:

The volume of LNG import flow into Europe (both spot cargoes and contracted European supply which cannot be economically diverted).
The price of incremental LNG import volumes if there are supply issues in Europe over the winter (e.g. Russian interruptions or major infrastructure outages).

But it also may have important implications for European power markets.

Chart 1: Asian spot LNG vs UK NBP prices

Source: Reuters

The fall in the differential between spot LNG and European hub prices shown in Chart 1 suggests European LNG import volumes may pick up substantially across this winter. In the absence of:

a recovery in LNG spot prices and/or
a major European shock this winter (e.g. Russian interruption, infrastructure outage, prolonged cold spell)

An increased flow of LNG imports will put downwards pressure on hub European hub prices. As Q1 develops, storage withdrawals are likely to add to that price pressure. This could precipitate a Q1 slump in European gas hub prices.

Power market support for gas prices

As Asian spot prices declined into summer 2014 reducing the premium over NBP, LNG imports into Europe increased, acting to drive down hub prices. A pickup in power sector demand (given weaker hub prices) provided some important support for NBP gas prices across this period.

Gas hub prices fell to a level where CCGTs started to displace coal-fired generation capacity in the merit order. The UK is Europe’s canary in the coal mine here, given the dominance of gas-fired capacity in the supply stack. But this effect may become wider spread across European power markets in 2015 if gas hub prices are really under pressure from oversupply.

If a similar or more significant hub price decline plays out in 2015, hub prices may again need to fall to a level where CCGTs are displacing coal in European merit orders. So the variable cost competitiveness of newer CCGTs vs older coal stations may become a very important dynamic to watch in 2015. The UK is a key market to watch but this is relevant for gas fired generation in Continental markets as well (e.g. Netherlands, France, Spain).

Gas price declines may be good news for beleaguered CCGT owners, through acting to increase plant load factors and generation margins, particularly in the UK where the carbon price floor increases gas plant competitiveness. We will come back and have a look at this dynamic in early 2015 when we have a clearer view of the winter supply/demand balance. In the meantime it may be worth keeping an eye on the LNG spot vs NBP price differential as a useful indicator of things to come.

A shift in LNG market balance

Posted on November 24, 2014January 15, 2020 by David Duncan

Since the Fukushima disaster, LNG has been a seller’s market. But the LNG market balance has undergone a sharp transformation in 2014. The summer slump in LNG spot prices sent shock waves through the global gas market. All eyes have been on the approaching winter as a barometer of the LNG supply/demand balance. But after a Q3 recovery, spot prices have slumped again over the last few weeks, falling below 10 $/mmbtu. This is suggesting a structural rather than a seasonal oversupply of gas is looming.

Renewed LNG spot slump into winter

Chart 1 courtesy of the team at Reuters, shows Asian spot LNG vs UK NBP prices. There was a sharp Q3 Asian LNG spot price recovery from summer lows of around 10 $/mmbtu. But the recovery into winter has proved short lived. Over the last six weeks spot prices have plunged from above 14 $/mmbtu to break through the summer lows to levels below 10 $/mmbtu towards the end of last week.

Chart 1: Evolution of Asian spot LNG prices

Source: Reuters

The Asian market looks to be well supplied into the coming winter. Temperatures have been mild to date and there has been a notable absence of the strong incremental hedging volumes that have been common in Q4 over recent years. In fact there has been very little interest from large Japanese and Korean buyers as prices have slumped, given they are already well supplied via long term contract volumes.

In addition the LNG spot market faces an overhang of ‘floating storage’ volumes. Portfolio players bought spot cargoes during the summer price slump with a view to selling into higher winter prices. This has proved to be a painful strategy and the overhang is contributing to downward price pressure.

LNG market players are also aware of the impact that the recent spot crude price slump will soon have in pulling down long term LNG contract prices. The majority of Asian LNG contract volumes are indexed to the JCC Japanese crude marker, typically with a several month time lag. This will mean a strong downward pressure on long term contract prices into the spring of 2015. And it will likely cap any recovery of LNG spot prices even if demand picks up over winter, as buyers have the ability to call on contract flex in preference to entering the spot market.

The impact on Europe

The Asian spot price slump reduces the risk around a tight winter in Europe. Concerns have been that a prolonged cold spell and Russian interruptions could cause very high and volatile hub prices across winter.

The threat of large scale Russian interruptions to European supply this winter is unlikely anyway in our view. But if this did occur, or if there were more major infrastructure issues (a higher probability in our view), the UK is particularly vulnerable to a sharp increase in marginal import prices over a cold winter. The price spikes of Mar 2013 are an illustration of what happens when the UK NBP hub needs to price up to attract spot LNG cargoes.

But weaker Asian spot LNG prices reduces this threat in two ways. Firstly, weak Asian prices will cut-off cargo diversion arbitrage plays from European supply contracts, meaning a higher flow of LNG into the NBP and TTF hubs, as well as the potential for higher volumes of Qatari LNG diverted to the UK. Secondly, if Europe (particularly the UK) does need to price up to attract additional LNG imports it is likely to be at much lower price levels than the last three winters.

We wrote previously that we would be watching this winter closely as a barometer of transition in the global supply/demand balance. The renewed slump in spot LNG prices suggests the balance in the global gas market is shifting back towards the buyers.

Spot vs forward price dynamics: UK gas case study

Posted on November 17, 2014January 15, 2020 by David Duncan

Commodity markets are plagued by confusion as to the relationship between spot and forward prices. There are good reasons. The relationship is not determined by a clean mathematical formula, as it is for example in interest rate markets. In fact it is hard to cleanly define a theoretical relationship between the physical delivery of a commodity and the trading of forward contracts in advance of delivery. This is particularly the case for gas and power markets given challenges with storing these commodities. Instead it is more useful to make observations about the spot vs forward relationship as it is observed in practice.

UK NBP gas spot vs curve animation

The energy industry is typically more focused on analysis of spot prices than forward prices. This is understandable in as much as spot prices drive physical portfolio dispatch and optimisation decisions. But forward price curves are much more relevant from a value monetisation and risk management perspective. It is these prices against which the majority of portfolio exposures are hedged.

One of the problems with analysing forward curves is that a single static curve on its own can be quite a bland snapshot of market conditions. A simple animation provides a more insightful view. Chart 1 shows an animation of the evolution of the relationship between spot and forward curve prices in the UK NBP gas market since Jan 2011. We have illustrated spot prices in the chart with the month-ahead rather than day-ahead contract to remove some of the noise.

Chart 1: UK gas spot price and forward curve evolution

Source: Timera Energy (based on ICE NBP gas futures EoD Settlements). Note, the animation may not work in all browsers (particularly older ones).

As well as being somewhat mesmerising (take care after a long Christmas lunch), the animation illustrates some important curve dynamics. Take two examples:

Spot wags the curve: A characteristic that is particularly common in gas and power markets is the influence of spot price movements in driving ‘parallel’ shifts along the forward curve. Look how prices evolved across 2012 as an example. There are practical physical forces that act to connect the two, e.g. time spread arbitrage via physical storage. But these are not formulaic and interact with a number of other drivers.
Spot price shocks: This year’s summer gas hub price shock provides a good case study of a more extreme disconnect between spot and forward prices. As spot prices began to fall in Q1 2014 the forward curve was dragged down. But as the spot continued to slump into Q2 the forward curve held up (albeit at a lower level to the start of the year). In other words a pronounced curve contango opened up representing a substantial shorter term physical oversupply of gas into NW European hubs (driven by surplus LNG, robust pipeline flows and storage injection limitations). However beyond the current year the back end of the forward curve retained a linkage to oil-indexed contract prices.

These simple case studies illustrate perhaps the two most commonly observed features of spot vs curve behaviour in gas and power markets. But they are by no means the basis for trying to develop an academic theory that comprehensively captures this relationship.

Curve behaviour and market maturity

Prompt vs forward curve behaviour is closely linked to market maturity and the level of commercially optimised intertemporal flexibility (e.g. storage, swing, production flex). Some factors that act as a useful barometer for the maturity of a forward curve include:

Forward horizon – how far ahead of delivery can I trade contracts
Liquidity & transaction costs – what is my access to contract liquidity and my cost of moving in and out of positions (measured by narrowness of bid/offer spreads and market depth)
Contract types & granularity – what types of products are available for me to trade e.g. to manage price shape & non-linear exposures, particularly important for markets with inherent shape and/or flexibility of underlying exposures
Dynamics of forward curve movements – to what extent do different portions of the forward curve move independently from each other
Availability of derivatives – that can be used to manage price risk along the curve with-out having to manage the complexity of physical delivery

As an illustration of different stages of forward curve maturity it is useful to step away from the NBP gas market example and compare:

The Brent crude curve – liquid several years ahead of delivery, tight bid/offer spreads, a range of time spread and options contracts traded and pronounced independent movement along different sections of the curve driven by arbitrage constraints
The UK power curve – limited liquidity out to 3 seasons ahead of delivery, relatively high bid/offer spreads, a limited range of standard contracts and a strong parallel shift relationship between spot and curve movements

As well as the UK power curve being less mature than Brent, these dynamics also reflect a greater difficulty in access to physical arbitrage opportunities for power, given limited storage ability.

Some practical observations on curve behaviour

One myth worth dispelling is that forward prices represent a market prediction of future spot prices. Forward contract prices represent the value today at which paper contracts change hands for delivery of gas over a defined period in the future. The market for forward contracts has its own supply and demand dynamics, driven primarily by the hedging of forward portfolio exposures, but also by other factors such as speculative trading flows.

Spot price events clearly influence the trading of forward contracts as can be seen in the parallel shift case study. This is both via:

Physical arbitrage between spot and forward prices (e.g. “the more cheap spot gas I have to buy and inject into storage now, the more I can sell forward against a future higher price”)
Precipitating rebalancing of supply and demand for forward delivery (e.g. “the price at which I can buy gas today impacts my pricing of gas for delivery in the future”).

Physical arbitrage is the strongest practical linkage between spot and forward prices and an important driver of curve backwardation and contango. Portfolio flexibility options such as swing contract take, storage inventories and production flexibility are all increasingly being managed against forward curve price shape. In more technical jargon, intertemporal flexibility is being optimised against forward time spreads. It is the constraints around optimisation of this flexibility that most closely determines the evolution of forward curve shape. While this is a relationship that can be analysed empirically it is not one that lends itself to a theoretical formula.

Tighter UK market, higher spreads & volatility

Posted on November 10, 2014January 15, 2020 by David Duncan

National Grid has now published its view on the UK power market supply & demand balance going into the current winter. The situation looks very tight. Grid’s numbers show a de-rated system capacity margin of just 4.1%. In volume terms that is only 2.3 GW. That is less than the capacity of one of the UK’s larger power stations (e.g. Drax, Longanett).

This does not mean a significant threat of a lights out scenario. There are several units currently on outage that are expected to return early in the winter. Grid also has some additional contracted reserve capacity behind this 2.3 GW margin. It can use this to respond in an emergency situation (e.g. to prevent brownouts or blackouts). But because the majority of this reserve capacity is effectively removed from the wholesale energy market, it won’t act to dampen rises in power prices, generation margins and price volatility if the UK experiences a cold winter.

Winter 2014/15 balance

Grid has published two important pieces of information in the last two weeks which add clarity on how tight the current winter looks:

Grid’s 2014/15 Winter Outlook shows detailed projections of UK market supply/demand balance.
Grid has also released details of the volume of Supplemental Balancing Reserve (SBR) capacity recently contracted for the coming winter, showing SBR contracts totalling 1.1 GW of de-rated capacity (covering 2 older CCGTs, an oil plant and some demand side response).

The winter supply & demand balance and 4.1% capacity margin is illustrated in Chart 1.

Chart 1: UK winter 2014/15 supply vs demand balance

Source: National Grid

Forecasting system capacity margins is not a precise science. Uncertain factors such as weather, outages and interconnector flows mean that Grid needs to make a range of assumptions in order to forecast system margins. For example, plant capacities are de-rated to reflect average planned & unplanned outages. Wind levels and temperatures are projected based on historical conditions.

If these uncertain factors combine to produce a more benign outcome as they did last winter (which was both wet and windy), then winter may pass by uneventfully, even with a tight system capacity margin. However if these uncertain factors land the other way, then it may be a very tight winter indeed as we have set out previously.

While at a headline level it may appear that a 4% capacity margin means the market will rely on the operation of a single large station, things are likely to be different in practice. In any given hour, uncertain factors may combine to provide a higher effective level of capacity than Grid assumes, e.g. higher than average plant outputs and higher import levels through the UK’s interconnectors with France, the Netherlands and Ireland. But factors may also conspire to reduce capacity, e.g. multiple forced outages across gas/coal/nuclear plants or conditions of little or no wind (Grid assumes quite an optimistic average 20% output from wind).

What is clear on an historical basis is that tighter system capacity margins coincide with periods of higher prices, higher spreads and higher volatility.

Impact on prices and spreads

The inverse relationship between system capacity margins and prices/spreads/volatility is more than a theoretical linkage. There is a practical mechanism that drives it. A tight system capacity margin acts to increase prices because there are less plants competing to provide the marginal MW required to balance the market. This increases the scarcity rents (the difference between market price and the highest short run marginal cost of the plant at the margin) that accrue to generators.

Historically these scarcity rents have acted as a price signal to encourage new build. But the mechanism for providing a price signal for capacity in the UK market is changing. The signal to retain an adequate system capacity level going forward will primarily come from:

Pre 2018: Grid’s tendering of SBR capacity, which acts to support older existing gas/coal/oil plants that would otherwise close.
Post 2018: The Capacity Market, which is likely to favour development of low capex but high variable cost peaking plant.

Neither of these mechanisms are likely to encourage significant volumes of efficient new CCGT plant to increase energy market competition at the margin. SBR contracted plants are removed from the energy market and used only as emergency backup. In other words SBR plants are shifted away from the margin to the very top of the merit order.

Rather than supporting efficient new CCGT build, the first capacity auction is likely to crystallise the weaker economics of a number of older existing coal and CCGT plants. Capacity already existing or under construction (55.8 GW) exceeds the max demand level (50.1 GW at 0 £/kW capacity price) by 5.7 GW. That means at least 5.7 GW of existing plant are likely to be unsuccessful in the first auction (more likely 7+ GW). This volume may be higher if low capex new peaking plant proves to be competitive (although this new capacity will also sit at the back of the merit order). It is reasonable to assume that a significant portion of existing plants will close without capacity price support. That will also act to reduce competition at the margin in the energy market.

Chart 2 shows a simple stylised illustration of the likely impact of this reduced competition on energy market pricing.

Chart 2: Stylised price duration curve view of reduced competition at the margin

Source: Timera Energy

The chart shows a simplified view of price duration curve, the shape of which is distorted in order to illustrate a point. The effects we are talking about are harder to visualise using real market data. The grey line represents short run marginal cost for the system across a given year. It increases to the left to reflect higher marginal cost plants setting system prices in times of peak net system demand. It is also in those periods that scarcity rents are highest given reduced competition at the margin.

The shift from the black line (normal PDC) to the red line (tight PDC) illustrates the impact of reduced competition at the margin. Reducing the number of plants competing at the margin (via SBR and plant closures) acts to increase mid-merit and peak rents. In practice this is achieved by higher and more volatile power prices.

Changes in the UK market design have implemented mechanisms (SBR and the Capacity Market) that allow the government and system operator to more closely control the level of backup capacity. SBR in particular may become a key tool to ensure the lights remain on over the next three years. However these mechanisms are unlikely to stop a pronounced rise in spark spreads, dark spreads and volatility over periods of market tightness.

Full commercial analysis of the 1^st auction: For a comprehensive analysis of the 1st auction and its commercial implications you can purchase Timera Energy’s First Auction in Focus briefing report.

The report provides a more detailed analysis of competition between plant types and the marginal pricing outcome in the first auction. Analysis covers the impact of energy market expectations, going forward costs, price maker/taker status, refurb option dynamics and new build competitiveness. The report also explores the key interaction between capacity and wholesale energy market pricing dynamics and the relationship between 1st and subsequent auction outcomes (e.g. T-1 and 2nd T-4). It concludes with a set of key commercial considerations on market dynamics going into the first auction.

For a report prospectus and more details please contact david.stokes@timera-dev.positive-dedicated.net.

A quick check on gas hub liquidity

Posted on November 3, 2014January 15, 2020 by David Duncan

Trading liquidity is the oxygen that is supporting the evolution of Europe’s gas hubs. Liquidity growth has been self reinforcing. With higher transaction volumes, price transparency increases and transaction costs decrease. This increases the attractiveness and reliability of hubs as a means to manage gas portfolio hedging and optimisation.

Some strong trends have emerged as hub liquidity has grown over the last decade. European gas trading has evolved around the UK NBP and Dutch TTF virtual trading points. While prompt liquidity has emerged at a number of other locations (e.g. Zeebrugge, NCG, Gaspool, PEG, PSV), forward liquidity remains focused on NBP and TTF. This is a function of the strength of price convergence across European hubs (illustrated in Chart 1).

Chart 1: Month-Ahead price evolution at major European gas hubs

Source: Timera Energy (LEBA prices)

Prices between the different Continental hubs can diverge over the shorter term (e.g. within-month) as a result of locational supply and demand factors (e.g. weather, LNG flow). But structural divergences in prices beyond the prompt horizon are rare (with the notable exception of the French PEG Sud and Spanish AOC hubs). There has been a particularly strong correlation between TTF and the important German NCG hub.

This has reinforced the focus on TTF as the hub of choice for forward trading. TTF liquidity has also been helped by relatively low trading costs (e.g. narrow bid/offer spreads and exchange trading fees) and a greater range of tradable products that help the portfolio exposure management and asset monetisation. Forward portfolio exposures are mainly hedged against TTF with locational basis risk managed via prompt trading at other hubs.

The dominance of TTF as the most liquid Continental hub is illustrated clearly in Chart 2 which shows the evolution of traded volumes over recent history. While volumes at other Continental hubs have expanded, they are small relative to TTF given the focus of trading on prompt portfolio optimisation.

Chart 2: Liquidity evolution at major European hubs

Source: Quarterly Report on European Gas Markets (Vol 7) – European Commission

Since the onset of the financial crisis in 2008, European hub liquidity has almost doubled. Liquidity evolution was given a big boost by the 2009-10 global gas glut period of oversupply. This was the result of two main factors:

As hub prices fell below oil-indexed contract levels, suppliers were strongly incentivised to use cheaper hub gas in their portfolios wherever possible.
Surplus LNG supply flowed into Europe as a market that offered a relatively robust price signal and forward liquidity to support cargo sales.

These factors are again at work as the global gas market balance has shifted back towards oversupply in 2014.

A pronounced increase in traded volumes can be seen in Q1 and Q2 2014 in Chart 2. This period coincides with an increase in the flow of spot LNG cargoes to Europe as Asian spot prices have declined. It is also a period over which European hub prices have fallen sharply below oil-indexed contract strike prices. If the gas market is transitioning towards a period of oversupply, these factors are likely to support further growth in hub liquidity.

First UK capacity auction in focus

Posted on October 27, 2014January 15, 2020 by David Duncan

Note: Several hours after this article was published DECC announced that they had revised the 1st auction capacity target down. The article has now been updated to reflect this new information.

It is now 7 weeks and counting until the first UK Capacity Market auction. The introduction of a Capacity Market represents the largest structural change to UK power market design since the NETA market replaced the pool in 2001. Large sums of money are involved. Capacity payments to generators are expected to be in the order of £1.5-2.5 billion in the first auction alone. And the outcome of the auction will almost certainly result in the development of new power plants and the closure of existing assets. It will also reshape pricing dynamics in the wholesale energy market. So what can we expect?

One of the challenges in analysing the Capacity Market (CM) has been the level of uncertainty behind a number of key market drivers. It doesn’t matter how good your telescope is if it’s misty. However the situation improved substantially about a week ago when the final list of pre-qualified plants for the 1^st auction was published. This provided a clear read on participating plants & players, de-rated capacity volumes and plant status (e.g. existing, refurb or new build), as well as a view on which plants have chosen to opt out of the market.

There is still considerable uncertainty around several CM drivers, the most important of which is individual player expectations of the evolution of future plant energy margins. But the information available now is about as good as it is going to get before the auction. And there is plenty to work with in order to draw some powerful commercial conclusions.

Pre-qualification – what have we learned?

Publication of pre-qualification data has laid bare details of the 80 or so market participants and their assets which will compete to provide the government’s target level of 48.6GW of de-rated capacity for delivery in 2018/19. A summary of prequalified capacity supply versus demand curve ranges is shown in Chart 1.

Chart 1: Final pre-qualified capacity for 1st auction (MW)

Source: Timera Energy

Some of the more important facts on pre-qualification are as follows:

Just over 67GW of de-rated capacity prequalified, leading to the obvious initial conclusion that upwards of 15GW of capacity is likely to miss out in the 1^st auction.
Scotland’s largest power station, the Longannet coal plant (~2GW), has opted out of the CM leaving the option open for Scottish Power to close the plant prior to 2019. Plant economics are being eroded by increasingly unfavourable transmission charges given its northerly location.
A number of other existing coal plants (Cottam, West Burton, Eggborough and Ratcliffe) have prequalified on the basis of 3 year refurbishment options, to cover plant efficiency upgrades and ensure IED emissions compliance.
EDF Energy has chosen to submit all its existing nuclear assets for refurbishment, although it is not exactly clear how this refurbishment will enhance the UK’s capacity position. This has been somewhat of a surprise and has the potential to become quite politically sensitive (given existing concerns around support for EDF via the Hinkley Point CfD and the Carbon Price Floor).
A wide range of smaller scale peaker plants (e.g. diesel gen sets, reciprocating engines, small scale gas turbines) have prequalified on 15 year terms. Although this provides an interesting dynamic, the overall volume is still small relative to conventional thermal power assets. Only slightly over 3GW of new build peakers are participating. There is also slightly less than 1GW of Demand Side Response (DSR), presumably mostly in the form of new back up peaking generators.
7.8GW of new CCGT has also pre-qualified on 15 year terms across 9 participants (SSE, ESB, Centrica, Scottish Power, Intergen and several other specialist developers). It however remains to be seen how much of this will be competitive enough on a cost basis to displace other capacity.

It is also important to note that the government revised down its 1^st auction capacity target level to 48.6GW (initially set at 50.8GW) on 27th October, to account for opt out decisions (e.g. Longannet).

Analysing the auction outcome

Analysis of CM dynamics has been a key area of focus for Timera Energy across 2014. As a result we have developed a strong capability to model pricing interaction between the UK capacity and energy markets. To keep this article focused we do not provide a detailed description of our modelling methodology. However there are two principles behind our analytical approach that is important to set out upfront.

The first principle is that what looks like a complex mess of different plants, players, costs and expectations can be substantially simplified by focusing on the plants and cost structures that are likely to drive marginal pricing. In other words by working out which subset of plants are likely to drive the intersection between supply and demand.

The second important principle behind our modelling is having a healthy respect for uncertainty. There are a couple of important and interrelated CM drivers that will influence the auction outcome, yet remain uncertain up until the auction day:

Energy margin expectations: The margin required to support existing and new power plants can be broken into three categories: (i) capacity (ii) energy and (iii) reserve margin. Plant bidding levels (determining capacity margin) in the auction are directly influenced by player expectations on the evolution of energy and reserve margin over the period to 2018/19 and beyond. With poor forward market liquidity past 2016, there is there is scant quantified market price signal on forward energy margin, and expectations on this for the future may differ significantly. Uncertainty around these expectations is a reality that has to be recognised and confronted when analysing CM auction dynamics.
Going forward losses: Some existing plants (e.g. older coal and CCGT assets) will suffer losses from weak generation margins in advance of delivering capacity in 2018/19. Players can recover for these losses by applying for price maker status (shortly prior to the auction). This allows them to bid above the price taker threshold for existing assets (25 £/kW), in order to attempt to recover for losses incurred by remaining open. Loss levels in turn come back to energy margin expectations. We can do some sensible analysis of the impact of losses on capacity bidding, but ultimately a significant degree of expectation based uncertainty remains.

In our view, trying to forecast these factors at an individual plant/player level tends to give credence to a spurious level of detail in CM analysis. We prefer to recognise the uncertainty involved around these factors and try and analyse how it may impact the auction outcome.

1^st auction supply stack

In order to analyse CM dynamics it is necessary to construct a representation of the supply stack. The supply stack is driven by the costs of providing capacity in 2018/19, with the CM being designed such that players have quite strong incentives to bid plants at true cost.

The task of developing the supply curve is helped by the fact that good cost benchmarks are available by plant type. The important benchmarks are the fixed costs of existing plants, the upgrade capex required for refurbishment plants and the capex costs for new build. The more difficult task is netting off energy margin expectations to derive capacity bids (as described above). As a result, energy margin expectations become a key focus variable for running scenarios.

Chart 2 shows a stylised scenario view of the 1st auction capacity supply stack with the demand curve overlaid. We have constructed this chart to provide a simplified representation of some of the CM pricing dynamics (described below). As such it is illustrative and should not be interpreted as a forecast of the auction outcome.

Chart 2: Stylised scenario view of 1^st auction supply stack

Source: Timera Energy

For the purposes of this chart, we have set up our CM model in ‘aggregated’ plant mode showing the supply curve grouped into tranches of capacity types based on cost of provision. Several of the key capacity categories around the margin are labelled. In practice we undertake detailed analysis at a more granular level, but it is easier to visualise the CM and draw high level conclusions at an aggregated level.

Some background in order to understand what the chart is showing:

The demand curve (black line) reflects the new 48.6GW auction target level. It is downward sloping from a min volume of 47.1GW (at the 75 £/kW price cap) to a max volume of 50.1GW (at zero capacity price).
The inflection point of the demand curve is at the government’s 49 £/kW estimate of net new entry cost (net CONE).
The chart also shows the price taker threshold at 25 £/kW which caps the level at which existing plants can bid in the absence of going through the process of applying for price maker status (to recover going forward losses).
With our supply stack model set in ‘aggregated’ mode, plants are grouped into 25 basic categories for which bids are defined based on cost and energy market expectations. These categories are shown for the illustrative scenario by different coloured sections of the supply stack.
Some capacity types are made up of multiple tranches, e.g. existing CCGTs are split into 3 tranches (T 1-3) as we have explained previously here, with the oldest assets (T3) split into two subcategories to differentiate between the impact of going forward losses on capacity bids.

Marginal pricing setting

As we described above, analysing the CM becomes easier if you focus in on the range of the supply stack where the marginal price is likely to be set. A good case can be made for a lower price bound around the fixed costs of older CCGT plants (~20 £/kW). There is 6-8GW of CCGT built in the early/mid 1990’s that is currently making close to zero margins in the wholesale energy market. This capacity (and several GW of older coal capacity) will be very likely to close at a capacity price below that level. There also looks to be a reasonable supply of new build gas and peaking plants above 50 £/kW that is likely to act as an upper bound in the first auction. It is the range in between these two bounds where the 1^st auction is likely to be fought out.

The supply stack representation in Chart 2 illustrates three key categories of plants that are likely to be competing to drive marginal pricing within this range:

Older coal plants: 7GW of less efficient existing coal plant are bidding for 3 year refurb contracts to comply with IED requirements. In addition there are several more GW of other older existing coal stations that will likely apply for price maker status to recover going forward losses (with a view to closing by 2023 given IED constraints).
Older CCGTs: Many of the 6-8 GW of older, less efficient CCGT that are currently making very little energy margin are also likely to apply for price maker status to recover going forward costs. The interaction between the bidding of these CCGTs and existing coal plant is likely to be a key factor driving the auction outcome.
New small scale peakers: Low capex small scale peaking plants under advantageous 15 year capacity agreements (supporting attractive leverage structures) are likely to be the most competitive form of new build. Given the number of different players and technologies, some of these plants should feature around the margin. But relatively low volumes mean that they are less important than the two categories above.

Our conclusions from previous articles we have published on the CM were that (i) refurb and going forward costs will be key drivers of the 1^st auction outcome and (ii) low capex small scale peakers will be competitive given leverage opportunities under 15 year agreements. Our analysis since release of the prequalification data only reinforces these initial conclusions.

A noteable exclusion from the list of key marginal price setting plants is the 7.8 GW of new CCGT capacity that is participating in the 1^st auction. We assume ESB’s Carrington CCGT will be built regardless given it is already under construction. But the developers of other new CCGT projects face a key challenge in putting a price on energy margin from which to imply a CM bid. The source of this energy margin pricing falls into two main categories. Projects will either need to sign a third party market tolling agreement (e.g. IPP developers) or internalise the energy margin risk in their portfolio (e.g. utility developers). Both categories are likely to face the issue of heavily discounted energy margin as a result of the pain being caused by current weakness in sparkspreads. As a result it will take a very cost competitive new CCGT project to compete with existing assets and smaller peakers.

The first vs subsequent auctions

After focusing in on marginal price drivers, it is useful to take a step back and consider the 1^st auction from a more strategic perspective. There are strong market design incentives to bid plant in a cost reflective manner. But first auction bidding behaviour will also be influenced by expectations of future auction outcomes.

If a plant bidding for a 3 year refurb or 15 year new build contract is unsuccessful, it has the option to try again next year. But the 1^st auction looks attractive given greater competition next year (e.g. from interconnectors) and expectations that the government may introduce less favourable rules around longer term capacity agreements (e.g. via a price duration curve mechanism). This is likely to incentivise a more aggressive bidding stance in the first auction, particularly for new build plant chasing favourable 15 year terms.

For a number of existing gas and coal plants this year’s auction is likely to pose an existential challenge. Plant losses mean that assets need to bid as price makers to recover going forward costs. Yet owners run the risk of missing out in the auction if bids are high, necessitating plant closure. Ultimately first auction bidding strategy is linked to the complex assessment of plant abandonment economics. This may skew plant owners to take a more aggressive stance on assumptions such as energy margin recovery expectations (e.g. 2015-18 as the market tightens) to support lower bid levels. This effect and the new build one described above may result in downward pressure on the 1^st auction capacity price.

Market pricing dynamics

Ultimately it is existing power plants that are going to be the main drivers of the first auction outcome. This is the logical result of an auction target level that does not require the delivery of incremental capacity volumes. The government’s target level relies heavily on DECC/Grid’s assumptions of peak demand reduction between now and 2018. If these turn out to be optimistic then it will mean substantial ‘top up’ buying of capacity or DSR in the year ahead auction. This is likely to favour developers of small scale peaking assets (with low capex and short lead times). It may also support contracting of supplementary reserve in the meantime.

There are 10-12GW of existing older coal and gas plants that currently have weak economics given low spark and dark spreads. Some of these plants will miss out in the 1^st auction and subsequently close. This will have an important knock-on impact for pricing and volatility in a tightening energy market over the next three years.

This is particularly the case if low capex but high variable cost small scale peakers are successful in displacing significant volumes of existing gas/coal plant. That would act to support power price and volatility levels (and hence plant energy margins). Small scale peakers may be cheap on a capex basis but they are very expensive on a variable cost basis and hence will have a limited impact in dampening power prices relative to larger conventional plants.

It is this interaction between capacity and wholesale energy pricing dynamics that is going to be critical going forward. The capacity market will determine the level and type of system capacity. This will then drive price shape, scarcity rents and volatility in the energy market. As a result, the first capacity auction marks the start of a transformational change in the UK power market and the dynamics of generation investment returns.

Full commercial analysis of the 1^st auction: For a more comprehensive analysis of the 1st auction and its commercial implications you can purchase Timera Energy’s First Auction in Focus briefing report.

The report provides a more detailed analysis of competition between plant types and the marginal pricing outcome in the first auction. Analysis covers the impact of energy market expectations, going forward losses, price maker/taker status, refurb option dynamics and new build competitiveness. The report also explores the key interaction between capacity and wholesale energy market pricing dynamics. It concludes with a set of key commercial considerations on market dynamics going into the first auction.

For a report prospectus and more details please contact david.stokes@timera-dev.positive-dedicated.net.

German recession, power prices & generation margins

Posted on October 20, 2014January 15, 2020 by David Duncan

After being the poster child of post crisis European recovery, Germany is suddenly facing the combined threat of deflation and recession. German GDP contracted in Q2 and industrial production has slumped across the summer. German exports have also sharply declined in August despite a weakening Euro. This is not a positive backdrop for German power demand. But there are factors ahead that may prevent a weakening German economy from translating into further declines in power prices and generation margins.

German power prices: a conspiracy of events

German wholesale power prices have fallen more than 60% since they peaked in Q3 2008. The three main factors behind this decline are:

New build: In the last 5 years approximately 10GW of wind, 30GW of solar and 10GW of new thermal (hard coal, lignite and gas) capacity has been commissioned. This can be compared to a little over 10GW of retirements (the majority of which are nuclear closures). There is also another 6.5GW of capacity (mostly hard coal) still under construction which will be commissioned over the next 2 years. The pace of new build has led to a substantial capacity overhang which has helped to drive higher variable cost thermal plant out of merit. In addition, the increase in solar and wind output has acted to flatten intra-day price shape pulling down peak prices.

Coal price slump: Gas-fired generators have suffered the double edged sword of a surge in low cost new capacity and an erosion of competitiveness as a result of falling coal prices. With gas-fired plant removed from the margin, wholesale power prices are now predominantly set by coal fired plant. As a result the post financial crisis decline in coal and carbon prices has fed through into a steady decline in wholesale German power prices as shown in Chart 1.

Weak demand: After a snap back recovery from the aftermath of the post-Lehman shock in 2008-09, power demand in Germany has also steadily declined. This is due to a combination of increasing embedded generation, energy efficiency measures and weaker post crisis industrial demand. Chart 1 shows the recent downturn in the IFO German business confidence index which is pointing to further declines in industrial demand.

Chart 1: 5 years of German power price declines (source Reuters)

German generation margins looking forward

After 5 years of events that have consistently reinforced the downward trend in German power prices, it is easy to extrapolate a pessimistic view into the future. This is particularly true if you subscribe to the theory that Germany and its neighbours are falling back into recession. But there are several factors that are working against a further substantial decline in power prices and generation margins.

Lignite on the margin: While hard coal plants dominate German marginal price setting, the factors described above are increasing the number of hours where lignite comes on to the margin (e.g. in periods of high renewable output and low demand). The variable cost of coal plant is driven predominantly by the cost of imported coal (a global traded commodity). However the variable cost of lignite plants is determined by local extraction costs at strip mines. It is harder to define the true variable cost of lignite plant than it is for coal plant, but it is in the order of 25-30 €/MWh. This is not far below the current level of year-ahead power prices (see Chart 1) and it represents an important support level going forward.

Coal prices: For the moment, API2 coal prices remain the most important factor driving German power prices. We wrote last week about how global coal prices had fallen below the long run marginal cost of new production and how export supply was being curtailed as a result. In the medium term these factors are likely to stem the fall in coal prices (although further price declines are certainly a possibility in the near term). But exchange rate movements are diluting the effect of falling global coal prices for German importers. The Euro has fallen almost 10% over the last six months, substantially offsetting the API 2 coal price decline across this period. If the US dollar continues to rise against the euro, this will act to support German power prices. A stabilisation in forward German gas and coal fired generation margins this year is evident in Chart 2.

Chart 2: German forward dark and spread margins

Plant closures: Under recently instated rules, plant owners planning to retire assets must submit an application to be approved by the German Federal Network Agency (BNetzA). Around 7GW of capacity has already applied for permanent closure, with BNetzA anticipating a total of more than 11GW of retirements by 2018. The volume of mothballing and retirements may increase significantly over the next few years given the inability of generators to cover costs at current dark and spark spread levels (shown in Chart 2). The large utility players in Germany continue to rationalise their generation portfolios as they face balance sheet constraints. Germany’s third largest generator Vattenfall may well sell its power portfolio and exit the German market all together.

Capacity mechanism support: Unlike the UK, Germany is in no rush to implement a new capacity market. Ample interconnection and the current capacity overhang are supporting security of supply, even if there are some increasingly acute localised transmission stress issues. The Ministry of Economic Affairs and Energy has commissioned three studies that investigate German power market design. The Ministry is expected to release a green paper setting out its thoughts on market design next month. This will then go to public consultation with a white paper expected in Sep 2015. There is pressure on Germany to implement a capacity mechanism as part of a broader European solution, despite the UK and France already having acted unilaterally to support capacity returns. While the nature of the solution is as yet unclear, it is a reasonable assumption that a capacity mechanism will be in place later this decade in order to support thermal generator fixed costs.

Power asset investment?

Both utilities and funds are treating the German power market with extreme caution. This is understandable when you look at the current market landscape. The important question is to what extent this is reflected in market expectations, investment decisions and asset prices. Sentiment on thermal generation assets in Germany is now very negative. Coal plant margins have been eroded with stiff competition from both renewable and thermal generation new build. Gas plants have been pushed out of merit, now representing ‘out of the money’ options that incur a substantial cost of carry (in the form of plant fixed costs).

But we have described a set of factors above that may act to support energy margins. In addition generators may benefit going forward from the implementation of a capacity mechanism and increasing reserve margin (as transmission stress rises). The key challenge investors face in valuing thermal generation assets is understanding how these factors drive asset margin risk/return distributions.

The German power market is also a key driver of pricing and generation margins in neighbouring markets such as France, Netherlands and Belgium. So the logic in this article extends across North West Europe, although the capacity situation varies by country. It appears to us that there may be some interesting generation investment opportunities in Continental Europe over the next two to three years. The challenge will be to find the right assets in the right locations at the right price.