Author: Brynne Kelly (w/Lee Taylor technical levels)
The impairment to demand caused by covid-19 upended the petroleum complex and caused producers, refiners and end-users across the globe to scramble. Prior to recent events, oil markets had been adjusting to a backdrop of slower demand growth in developed economies to focus on growing demand in developing economies. Countries like the US that had once relied primarily on petroleum imports were now focused on building out export capacity to provide outlets for production growth in excess of marginal demand.
Investment dollars were drawn towards opportunity, specifically the premium being paid by Asia and other developing markets for oil and oil products. In the US, companies poured money into export infrastructure located along the Gulf Coast, based on widening benchmark futures spreads.
As the dust settles on a less robust demand-growth picture and capacity overhangs, we are entering an era of optimization of the lowest cost and most efficient resources. We have seen oil markets lack general momentum over the last 2 months as they try to cling to old patterns and signals (i.e. inventory, curve structure, etc.). This implies a bias towards short-term optimization vs long-term optimization.
Short-term optimization is a reaction to disruptions caused by things like hurricanes or drone strikes. Long-term optimization is a recognition that we are entering a period where the least efficient/lowest margin players are being squeezed out. Long-term optimization asks questions like:
How does refinery capacity match up to products demand?
Will downstream product demand be fulfilled by local refineries or by product imports?
During long-term optimization, spread margins (i.e. refining cracks, global oil spreads) are lower for longer until they effect the desired change. Symptomatic of this period are false flags of spot market tightening used as an all-clear signal. However, it's important to remember that any spot market tightening is against a backdrop of restrained capacity (bloated inventory, production cuts, reduced refinery operations...). There will be temporary mis-matches in supply/demand as each of these are dealt with.
While oil markets hang their hat on OPEC's ability to smooth out supply/demand shocks, other markets take optimization much more literal. To see what real organized optimization looks like, we turn to electricity markets.
Optimization in Electricity Markets
Electricity markets are constructed around regional independent system operators that match supply, demand and logistical/transportation constraints.
"Today's power industry is far more than a collection of power plants and transmission lines. Maintaining an effective grid requires management of three different but related sets of flows – the flow of energy across the grid; the exchange of information about power flows and the equipment it moves across; and the flow of money between producers, marketers, transmission owners, buyers and others. ISO/RTOs play an essential role in managing and enhancing all three of these flows." (Wikipedia).
In a power pool, all generating companies offer price-quantity pairs for the supply of
electricity. This forms an aggregated supply curve. The offered prices can be based on
predetermined variable costs (such pools are referred to as Cost Based Pools) or the
generators can be free to offer any price they like (such pools are referred to as Price Based Pools). On the demand side, the market operator may forecast demand and dispatch generating units against this. This is called a one-sided pool. In more sophisticated pools (two-sided pools), the market operator may dispatch on the basis of a demand curve created from price-quantity bids made by the buyers on the market, such as distribution companies and eligible consumers (Figure 1).
The objective of an optimization function is maximization of economic surplus subject to generation and transmission capacity constraints. Detailed representation of a network makes it possible to take losses, parallel flows, voltage stability and reliability criteria into account. As a result, first-order conditions associated with the maximization of economic benefits under constraints yield prices, which represent the change in the total cost (as defined by market participants’ bids and offers) of meeting system energy requirements caused by a change in load or generation at each node.
This makes OPEC supply cuts look like child's play! OPEC deals with one side of the equation (supply) and leaves the rest up to wholesale market prices. notwithstanding the constraints of a global oil market versus a regional electricity market, if we took lessons from the electricity market described above, we would see how inefficient the oil markets are by comparison in optimizing the entire complex in real time. This doesn't mean that electricity markets are perfectly optimized - they have their own problems due to being quasi-regulated. However, it's worth thinking through what a 'system optimization' looks like. Oil markets use OPEC to 'balance' the system, but that only focuses on supply. The rest of the complex now has to balance itself and spreads are sending the signal. The signal being: Production isn't the only excess capacity out there.
Spread Markets Send Optimization Signals
Absent a formal optimizing body in the petroleum complex, producers, refiners and consumers optimize their individual bottom lines. Initially, balance sheet strength and access to capital can be confused with 'most efficient provider'. In practice however, these two factors delay optimization. In low price environments, the focus is heavily weighted towards outright prices in relation to commodity producers.
We are seeing this play out on steroids since the pandemic wreaked serious havoc on market prices in early 2020. Oil is a highly competitive industry at every point across the supply chain. The swift reduction of capacity utilization across the board will be more complicated to put back together as the various actors attempt to bring capacity back online. Absent a major event, markets will only allow the most efficient, cost-effective resources back into the game.
This will be led by the demand-side of the equation which rests squarely on refiners. At the moment, oil is making it's way towards the highest priced markets which is generally governed by refining margins. This is where benchmark oil spreads come in to play. To highlight this, we look at BP's posted 'Refining Marker Margins'.
* The Refining Marker Margin (RMM) is a generic indicator. Actual margins realised by bp may vary significantly due to a variety of factors, including specific refinery configurations, crude slate and operating practices. For example being based on a single regional marker crude, the RMM does not include the impact of the differential between Canadian Heavy crude and WTI.
BP provides further detail regarding the the benchmark crude used in each region's RMM calculation and the regional capacity weightings:
Given this information, we see where optimization comes in: landing a different crude grade from the same location or a similar crude grade from a different location at a price that is cheaper than the benchmark used above.
For example, a refiner in NWE could benefit from importing WTI at a landed cost less than current Brent prices. This is where the WTI/Brent futures spread comes in to play...
The continued narrowing of this spread over the last month is a reflection of shipping rates, but it's also a nod to the fierce competition amongst crude suppliers. The market is not going to allow an obvious price advantage to WTI vs Brent (meaning that the spread plus shipping costs is moving towards equilibrium). The next molecule brought back on-line will have to compete in this environment. In this way, oil spreads are going to force a thoughtful return of production.
Beyond NWE, we also see that the spread between Dubai and Brent futures remains compressed (black line below) and is not providing a means to clear the inventory overhang.
So, refining margins are compressed, benchmark oil prices are compressed and, as we have covered in previous reports, inventories are at or above 5-year highs. This should ultimately result in only the most cost-effective barrels being brought back on-line first rather than reckless investment in supply growth.
Underneath all of this, we also have refined products themselves fighting for the right relative value. About the only spread that we have seen 'rally' over the last month is the US spread between gasoline and distillate. The 12 month RB/HO spread curve has rallied almost $4/bbl in favor of gasoline since the start of September.
This is what optimization looks like: allocating refining capacity based on products demand. If the petroleum complex has truly entered a longer-term optimization phase, a lower supply/demand intersection needs to work it's way across the complex. The next step from here would be lower refining margins for longer. This will move the focus from technology gains in production to technology/efficiency gains in the refining model. Collectively, the most efficient refineries with the highest margins would supply product demand growth. Capacity additions would only arise if they were technically more competitive.
EIA Inventory Statistics
Weekly Changes
The EIA reported a total petroleum inventory DRAW of 9.80_million barrels for the week ending September 18, 2020. Commercial Crude oil inventories posted a weekly DRAW of 1.70 million barrels.
Year-to-date, total inventory additions stand at a BUILD of 86.10 million barrels (vs 95.90 last week).
Commercial Inventory levels of Crude Oil (ex-SPR) and Refined Products are slightly above the 5-year average in oil, slightly within the 5-year average for gasoline and still above the 5-year average for distillate.
Lee Taylor - Technical Levels
To Return Next Week!