For two years in succession oil production growth has not had to meet demand growth in its entirety.
From the first quarter of 2017 to the first quarter of 2018 the market was able to pull roughly 590 000 barrels from drawing down on OECD inventories in order to meet a large portion of the 1.7 million barrels a day (Mbblsd) demand increase over the period. In 2018, the mantle was then handed to Saudi Arabia to pull in supply from spare capacity – oil that should really only be tapped during times of extreme tightness. It is not a source of supply that is sustainable. This reduction in spare capacity was a large contributor to oil moving above $80bbl but was ultimately a significant portion of the volume that drove the oversupply in the market.
However, the extent of the inventory overbuild this time is not a patch on what the market experienced in 2016 – the 2019 first quarter peak in inventories is likely to be below 1% (c.20 million barrels), versus an almost 8% (220 million barrels) breach in 2016. We believe that we will be back below the five-year average in the second quarter of 2019. That is not to say, however, that the market cannot continue to pull from inventories and spare capacity in the second half of 2019, but the drawdown is unlikely to continue coming from the Saudis, since their inventory cover is already -30% below their 2011-2016 average (almost 15% below their previous low in 2011) and they have just clawed back much-needed spare capacity via the OPEC cuts. Calling the bottom (or top) is always difficult, but in 2016 prices started turning nine months before the top in inventories.
As we progressively move through 2019 it will be increasingly harder to ignore crude quality when trying to determine where the overall oil price is headed – particularly as we head closer to the implementation of IMO2020, the law requiring ships to use less sulphur in their fuel, thereby forcing c.97% of the vessels to switch from sulphur heavy residual fuel to diesel. Over 80% of supply growth over the last two years has emanated from light oil (United States shale oil) and a barrel of light oil will produce twice as much gasoline but almost half as much distillate (including diesel) as an equivalent barrel of Brent oil. Consequently, refiners are finding it increasingly difficult to produce the required diesel to satiate the market. We have, even before the effects of IMO 2020 are felt, seen a progressively tight distillate market and a looser gasoline market – diesel storage in the US hit 10-year lows in November 2018 while gasoline storage (as at December 2018) was approaching 10-year highs. This has sent the prices of diesel v gasoline to their widest spread in four years.
This divergence in prices is likely to continue widening as vessels begin to switch to diesel (by the third and fourth quarters) and refineries begin to prepare by building distillate inventories (the first half of 2019). The fuel that they previously used (residual fuel oil) will likely price down to a point where it has to find another market (likely to be the power market, backing out coal/gas). Consequently, we could see a big demand surprise towards the end of 2019, going into 2020, of crude oil. One Texas refiner we recently visited envisages a two million barrel demand spike. If we see just half of this amount of additional demand growth, then the market could be significantly tighter than forecast.
We see the potential for a much more constructive oil price dynamic in 2019, helping stocks higher and revenue higher for production companies. However, we continue to envisage a market that is reluctant to spend on significant supply growth (something that, in itself, is a positive macro driver for the oil price) so service stocks may continue to be challenged.
Differentials are likely to extend higher, particularly in the second half of the year when IMO 2020 increases the demand for distillates and further exacerbated as we see the completion of Permian pipelines in the fourth quarter of 2019 unlocking supply of lighter oil. The risk of Iranian sponsored disruptions could also be a wild card, particularly if waivers are dropped in the second half of the year.