What’s driving oil prices, and where are they heading?
Our forecasts suggest short-term upside and longer-term downside to the oil price. This reflects an objective assessment of a wide array of demand, supply and risk indicators.
Forecasting oil prices? Not so easy!
The relationship between global economic activity and oil demand is, in fact, weak and does not fully explain oil price movements.
It’s generally assumed that global economic activity drives oil demand and, hence, prices. In fact, looking back over the past 35 years, such a relationship is weak and does not fully explain oil-price movements. One reason is that the energy intensity of the global economy fell by one-third between 1990 and 2015 (see Chart 1).
Chart 1: Energy intensity of various economies over time. Expressed in MegaJoules (MJ)/ GDP in purchasing power parity (PPP) US dollars 2011. Source JP Morgan May 2019.
Why is this relationship between economic growth and oil prices changing? One reason is technology, especially shale-oil production. This has allowed non-OPEC producers, particularly in the U.S., to increase their market share. As U.S. production has climbed, US import demand has fallen, and US exports have risen from zero to over three million barrels per day (about 3% of global supply). U.S. supply is determined not by OPEC meetings, but by each corporate producer, based on internal profit dynamics. Moreover, U.S. shale-oil producers can add or remove supply much more quickly than most global sources.
A second reason is the fictionalization of commodity markets. In recent years, financial-market participants have become important players in this sector. They have been able to drive the price of oil above and below fundamentally justified ranges for periods of time. For example, commodity trading advisors typically rely on momentum trading. Other financial market players are driven by "roll yield" – the profit or loss incurred by closing out a futures position as it nears expiration, and opening a new futures position with an expiry date further away. For other money managers, oil’s unique qualities make it an attractive diversifier, helping offset geopolitical risks elsewhere in a portfolio.
Another contributor to the inefficiencies in oil-market pricing is the poor quality and timing of data. A UBS report in 2014 showed that the average miss on the International Energy Agency’s (IEA) oil market balance was one million barrels per day. This magnitude of error explains 90% of the variability in the market and renders the IEA’s current estimates virtually useless. Their accuracy is only helpful after revisions are completed two years later!
Building robust forecasts
Given the difficulties in making sense of oil market data, how can we build useful oil market forecasts? In our own analysis, we overcome the data issues by: seeking objective data from multiple sources; seeking contrarian advice; and using a range of market data to negate those data signals for which confidence is low.
Using these principles, we create a proprietary energy scorecard. This includes analysis of more than 50 commodity research sources from a mixture of industry bodies and investment bank specialists. Such breadth of research means we can identify factors that have historically shown a close relationship with global oil prices, and which can therefore be useful forward indicators. We group indicators into six major categories: valuations, demand, supply, technical analysis signals, externalities and risk. We assign a score to each, indicating a positive or negative price outlook, as follows.
Valuation indicators include consumer price inflation, the world output gap and the US dollar trade-weighted index. Additionally, we take account of the "new project cost-curve" – the price which, if exceeded, would incentivize a big increase in production. Currently, valuations are above historical ranges but not in previously established danger zones. Oil prices are below, but approaching levels that would trigger new production.
Demand indicators include guidance from our ASI Research Institute specialists. We also use economic surprise indices from the U.S., China and EMEA, which often represent data not already assimilated into the oil price. These and other demand indicators imply solid oil demand growth of 1.4% in 2019. However, demand is typically lower in the first half of the year than the second half, so we are heading into a seasonally strong period.
Supply indicators are positive for higher oil prices. Supply is tightening because of falling production in Venezuela and sanctions against Iran. Added to this are pipeline capacity constraints in the U.S., OPEC production cuts and low inventories in Europe arising from the Russia-Europe pipeline outage.
Price signals are also positive. As an example of one measure we follow, the Saudi Aramco premium, or discount to the official selling price, is currently at its highest level since 2014.
Externalities include supply disruptions, geopolitical tensions, the effects of sanctions and cartel activity. Currently, these also point to upward pressure on oil prices over the summer.
Risk - lastly, there is a range of political risks that could materially affect supply or demand. Most of these are scoring negatively at present. For instance, if the U.S.-China trade dispute escalates, our expectations for oil demand and pricing could prove to be too high.
Our current oil market view
On a three-month time horizon, we remain positive on the oil sector outlook. Fundamentals justify a price range of $60-70 per barrel for Brent crude. We are entering a period of strong seasonal demand, while elevated geopolitical tensions and limited spare capacity will constrain supply. Our findings suggest the oil market will be undersupplied until October, when it will shift into oversupply going into 2020. On that basis, we expect oil prices to be in the lower part of their trading range in the coming year.
Trading in commodities entails a substantial risk of loss. Projections are offered as opinion and are not reflective of potential performance.
Projections are not guaranteed and actual events or results may differ materially.