Road to Rebalancing: 3 Steps

Transcription

1 March 9, 217 1: PM GMT The Oil Manual Road to Rebalancing: 3 Steps A surprise US inventory build has triggered a sell-off in oil prices this week. Yet the case for further rebalancing in the rest of the year remains robust. We see global stock draws over the summer driving the forward curve into backwardation, and oil to finish the year higher than it started. Correction in oil markets, but case for medium term rebalancing stays robust recovery in three stages: After unusual stability in recent months, oil markets have corrected sharply lower this week. This was triggered by an unexpectedly large build in US inventories. However, the persistent rise in US drilling activity in recent weeks, weakness in gasoline demand, and record long positioning also played a role. In the short term, this may continue and oil could retrace its gains since the OPEC deal. Nevertheless, we think this will not prevent the further rebalancing of oil markets, which has been ongoing for some time already, during the remainder of this year. In this note, we set out our broad views over the medium term: First, inventory draws drive curve into backwardation - Brent to $62.5 by end- 217: Demand growth remains strong, and as long as prices do not fully recover, we expect this to remain the case. At the same time, OPEC discipline has been encouraging and there is little production growth in non-us/non-opec. US shale is recovering fast we forecast.7 mb/d growth 4Q16/4Q17 but not yet sufficiently to prevent meaningful inventory draws during 2Q-4Q. This should drive the forward curve into backwardation, with WTI reaching $6/bbl and Brent $62.5 by year-end. Second, OPEC discipline diminishes and shale accelerates - Brent still at $62.5 by end 218: At $6+, OPEC cohesion would probably be hard to maintain, and production would likely creep higher again. Also, shale production is set to accelerate further in 218 by ~1 mb/d 4Q17/4Q18. We expect those factors to fully offset demand growth and production declines elsewhere, leading to a broadly balanced market in 218. Although there will undoubtedly be volatility, we see oil prices finishing 218 at broadly the same level as they started. Finally, oil price converges to marginal cost, boosted by sulfur spec changes Brent to $7-75 by 219/2: By 22, approximately 1.5 mb/d of our demand estimate relies on projects that are yet to be sanctioned and that have breakeven oil prices around $7-75/bbl. In addition, changes to bunker fuel regulations will start to put a premium on low sulfur crudes, like Brent, and could drive up crude demand by as much as an additional ~1 mb/d, on top of base line demand growth. With this boost, we expect oil prices to reach $7-75/bbl by 219/2. MORGAN STANLEY & CO. INTERNATIONAL PLC+ Martijn Rats, CFA EQUITY ANALYST Martijn.Rats@morganstanley.com MORGAN STANLEY & CO. LLC Ashley L Petersen RESEARCH ASSOCIATE Ashley.Petersen@morganstanley.com Exhibit 1: We see stock draws during 2Q-4Q... Global supply/demand balance Q13 4Q13 3Q14 2Q15 1Q16 4Q16 3Q17e2Q18e Implied stock build (mb/d) Source: IEA, Morgan Stanley Research, e = Morgan Stanley Research estimates Exhibit 2:...and prices slowly recovering Oil price target ($/bbl) Jan-13 Jan-15 Jan-17 Jan-19 Source: Thomson Reuters Datastream, Morgan Stanley Research estimates Martijn Rats is head of the European Oil & Gas Research team and lead analyst on key stocks including Royal Dutch Shell, BP, Total and Eni. With this report he also assumes responsibility as the strategist for Morgan Stanley Research's oil price forecasts. Morgan Stanley does and seeks to do business with companies covered in Morgan Stanley Research. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of Morgan Stanley Research. Investors should consider Morgan Stanley Research as only a single factor in making their investment decision. For analyst certification and other important disclosures, refer to the Disclosure Section, located at the end of this report. += Analysts employed by non-u.s. affiliates are not registered with FINRA, may not be associated persons of the member and may not be subject to NASD/NYSE restrictions on communications with a subject company, public appearances and trading securities held by a research analyst account

2 Road to Rebalancing "The problem of oil is that there is always too much or too little" - Myron Watkins in 'Oil: Stabilization or Conservation' (1937) Summary Exhibit 3: Global supply/demand balance: inventory draws from 2Q. Global supply/demand balance Q13 4Q13 3Q14 2Q15 1Q16 4Q16 3Q17e2Q18e Implied stock build (mb/d) Source: IEA, Morgan Stanley Research Exhibit 4:...putting upward pressure on oil prices Oil price target ($/bbl) Jan-13 Jan-14 Jan-15 Jan-16 Jan-17 Jan-18 Jan-19 Jan-2 Source: Thomson Reuters Datastream, Morgan Stanley Research estimates Exhibit 5: The Dec 219 future appears well supported, having rallied only modestly since oil markets were 2 mb/d oversupplied December 219 Brent future ($/bbl) Dec-12 Dec-13 Dec-14 Dec-15 Dec-16 Source: Bloomberg 75. At double the low but still half the peak, the range of possible outcomes for oil prices seems particularly wide, with meaningful risks to both upside and downside. Still, the range of plausible outcomes is narrower, in our view, and more positively skewed. In this note, we set out our broad expectations for oil prices in the coming years this note is not focused on the short term. In summary, we expect inventories to draw at above-seasonal rates from sometime in 2Q onwards, persisting into 3Q and 4Q. On our estimates, global inventories will contract close to 28 mln barrels during 217, which will likely force the forward curve into backwardation. We see limited room for the long-end of the curve to fall, given its underpinning by marginal cost, and the fact that long-dated futures have risen only very modestly in the past year. This means the short-end will need to rally. We expect that this mechanism will be sufficient to drive WTI to ~$6/bbl, with Brent at ~$62.5 by year-end 217. After that, a number of balancing factors will likely kick-in; US shale will likely accelerate further, OPEC compliance should start to diminish, and demand growth could also slow. As a result, we see a broadly balanced market in 218, with prices likely remaining flat at $6/62.5 WTI/Brent throughout the year. Beyond that however, a period of renewed strength will likely commence. We estimate that ~1.5 mb/d of our 22 oil demand forecast will need to be supplied by projects that have not been sanctioned yet and that have break-even oil prices of ~$7-75/bbl. Over time, oil prices will need to incentivise these projects again. Also, in 22 new regulations governing the sulfur content of bunker fuel will kick-in. This has meaningful implications for crude oil: it will likely put a premium on 'sweet' (i.e. low-sulphur) crudes, like Brent. Also, to produce enough additional diesel and blending components, refiners may need to process as much as an extra ~1 mb/d of crude oil, according to a recent IHS study. Hence, this change should boost crude oil demand around that time, even relative to oil product demand, further tightening the market. Taking both factors into account, we see Brent reaching $7/bbl by end 219, converging to the long-run marginal cost of ~$75/bbl by end-22. In the remainder of this note, we set out our expectations for supply and demand that underpin the view described above. Specifically, we address these key questions: 1) Will demand growth slow? 2) How will US shale and other non-opec production respond? And 3) Will OPEC compliance stay strong? 2

3 Demand Concerns abound, but medium-term outlook remains robust Exhibit 6: The link between oil demand and population growth is yet to break... World population vs oil consumption Oil consumption (mb/d;rhs) Population (bn) Source: World Bank, BP Statistical Review Exhibit 7:...and population growth continues Global population (bn) e: 9 bn 223e: 8 bn 212: 7 bn 1999: 6 bn 1987: 5 bn 1974: 4 bn 196: 3 bn 184: 1 bn 1927: 2 bn Source: Worldbank Exhibit 8: Electric vehicle sales in Norway have soared to 29% of total car sales, the highest in the world. Yet, gasoline+diesel consumption has continued to rise to a record level. Norway: Transportation fuel demand vs electric vehicle registrations EV regs. (thou.) Gasoline+diesel cons. (kb/d; LHS) Note: Yellow line shows trailing 12-month average gasoline plus diesel consumption Source: IEA, Norwegian Road Federation (OFV) In the last few years, discussions about 'peak oil supply' have been displaced by fears over 'peak oil demand'. This is despite demand numbers that continue to come in strong: in response to lower prices, demand grew at above-trend rates in both 215 and 216. This shows little sign of waning yet: during Aug-Dec 216, oil demand increased 1.9 mb/d y/y, well above the 1-year trend rate of 1.1 mb/d per year. Whilst there are good reasons to assume demand for certain products (e.g. gasoline) will peak in certain regions (e.g. the US) at some point (e.g. over the next few years), we expect global demand to remain robust. We forecast growth of 1.5 mb/d in 217 and 1.4 mb/d in 218. There are a few reasons for this: The global economic outlook remains supportive: although the economic outlook is somewhat brittle, Morgan Stanley economists nevertheless see global growth gaining momentum in 217 on the back of faster US growth, stable DM growth and rebounding EM momentum. PMI's in Europe, the US and China all show improvements in the outlook for manufacturing, supporting oil demand, in particular for middle distillates. In particular, commodity exporters and petrochemicals will boost demand: China and India will continue to anchor oil demand growth in the next few years. However, this should be supported by two other sources: first, we see rising demand for natural gas liquids (NGLs) given the large expansion of petrochemical and ethane cracking capacity that is underway. Second, commodity exporters have seen their fortunes improve and will likely see their oil demand rise. We expect this to be the case for Russia, Saudi Arabia and Brazil. In particular, the latter should see a large reversal: after shrinking 11 kb/d in 216, we expect oil demand in Brazil to grow again by ~25 kb/d in 217. In the long term, oil demand has two powerful tailwinds population growth and rising GDP/capita: The world is adding 1 billion people every years, and over each of these periods global GDP per capita rises 35-4%. Exhibit 9 highlights how the latter has historically impacted oil demand: so far, there has been a typical upward-sloping relationship between GDP/capita and oil consumption. Around 85% of the world's population is only in the foothills of these curves. As GDP/capita grows in those emerging countries, this will likely put upward pressure on oil demand. To illustrate this effect, let's focus on China. In their recent BluePaper Why We are Bullish on China (13 Feb 217), our colleagues in Economics argued that the country's GDP/capita will likely increase 6% (or ~$3,9) in real terms over the period Over the last 1 years, China's oil consumption per capita has risen.34 barrels/person/year for every $1, increase in GDP per capital. This ratio has been remarkably stable. If this continues to hold, and our economists are right about their GDP forecast, China's per-capita oil consumption will grow by.34 x 3.9 = 1.33 barrels/person/year over the next 1 years. 3

4 Exhibit 9: As GDP/capita goes up, oil consumption tends to follow.85% of the world's population is only in the foothills of these curves Oil consumption per capita (barrels per person per year) Middle East Mature Asia U S & C anada 15 Australia & NZ 1 Russia W. Europe 5 Africa Other Americas India E. Europe China Developing Asia % of the world's population Source: World Bank, BP Statistical Review, Morgan Stanley Research Exhibit 1: If China ends up 'there' in 1 years, it would add 5 kb/d to global oil consumption per year Oil consumption per capita (bbl per person per year) China 'there' GDP per capita (thousand 21 US$) Source: BP Statistical Review, Worldbank, Morgan Stanley Research GDP per capita (thousand 21 US$) That would still leave China's oil consumption per capita at a much lower level than at many other regions when they reached similar GDP/capita. Still, even under this scenario, China's oil demand would grow by 1.86 bn barrels over the next 1 years, assuming a stable population of ~1.4 bn people. This equates to growth of 5 kb/d on average each year. In short, over the next decade China alone could account for nearly half the historical trend growth rate in demand of 1.1 mb/d. Another way to look at it is this: over the last few years, the global car fleet has typically expanded by ~4 million cars per year (net of scrapping), and this is unlikely to change soon. If each car drives 9, miles per year and achieves 4 mpg, this alone adds 6 kb/d to global oil demand each year. Growing fleets of trucks, ships and airplanes create additional demand on top of that. These numbers appear very optimistic, and the two examples above overlap, of course. Also, to be clear, our forecast for China's oil demand growth for 217 and 218 are much more modest at 3-34 kb/d per year. However, this does illustrate the support that oil demand gets from a growing population and rising average wealth levels. Energy efficiency and substitution by renewables can offset this to a substantial extent. However, they have so far been insufficient to prevent oil demand from growing. As we argued in From Molecules to Electrons: What Energy Transition Means for Oil & Gas Investors (5 Jan 217), we expect this to remain the case for the foreseeable future. 4

5 Non-OPEC Supply US supply is responding strongly... With oil prices rallying over the last 12 months, one source of supply that is set for a large comeback is US shale. Baker Hughes' oil directed rig count has already nearly doubled from the lows, increasing from 316 rigs in May 216 to 62 at the moment. In January alone, 44 rigs were put back to work in the four main shale basins the Permian, Bakken, Eagle Ford and the Niobrara and February looks to be on track for a similar increase. Exhibit 11: At current productivity, a oneoff addition of 1 rigs can add ~75 kb/d to production in two years Incremental production from a one-time increase in the rig count of 1 units (kb/d) Note: horizontal axis shows number of months after rig count increase Source: Morgan Stanley Research estimate This is already sufficient to cause a significant increase in production during the course of 217, especially because drilling productivity has dramatically improved: the time between spuds is shorter, IP rates have increased, and decline rates are being managed. At current productivity levels, an increase of 1 rigs relative to our base line expectation would add ~75 kb/d to production 24 months later, and still grow thereafter (this assumes five are added in the Permian, two in the Bakken and Eagle Ford each, and one in the Niobrara). In short, the rig count matters. Economics of drilling activity are attractive, but pace of growth will moderate. It is clear that drilling activity in shale will continue to increase. The economics are attractive at current oil prices, and the vast majority of US E&Ps are indicating plans to add rigs. However, we do assume a slow down from the recent breakneck pace. By end-218 we model ~6 horizontal rigs in the four main shale plays, up from 355 at the end of January - see Exhibit 12. The cost and productivity of those additional rigs is a key uncertainty however. Many of the improvements in recent years have come from pad drilling, which is already widely adopted by now. Achieving high levels of productivity requires 'walking' rigs with three mud pumps and a powerful top drive. Our industry sources tell us that there are ~1 such rigs still idle. At the current rate of rig count increases, those would be absorbed within a few months. Beyond that, rigs need to upgraded, but this is costly (~$6-7 million investment per rig) or productivity starts to suffer. Also, prices of frac sand are already rising fast, and with every rig requiring ~2 people (so ~5, people for our projected 25 rig increase), some wage inflation could kick in again as well. Those factors could drive up costs and slow the rate at which the rig count is increasing. Nevertheless, we estimate that shale oil output will increase by a relatively hefty ~25 kb/d in 217 to 5.2 mb/d, implying 'exit-to-exit' growth of ~67 kb/d between 4Q16 and 4Q17. This will likely accelerate further in 218, growing another ~1.1 mb/d to ~6.3 mb/d. It is possible that shale production will grow faster if more rigs return. However, that would start to impact the global supply/demand balance quite negatively, particularly in 218, which would lower oil prices and hence slow down drilling activity again. Also, at very high growth rates it becomes likely that some bottlenecks will eventually appear (e.g. take-away capacity, labour, proppant, etc), so we are weary of taking these estimates higher. 5

6 Exhibit 12: We expect the rig count to recover further Jan-12 May-13 Sep-14 Feb-16 Jun-17 Nov-18 Historic MS Forecast Source: Rystad Energy, Morgan Stanley Researcj Exhibit 13:...driving 4Q/4Q supply growth of ~67 kb/d in Jan-12 May-13 Sep-14 Feb-16 Jun-17 Nov-18 Historic MS forecast Source: Rystad Energy, Morgan Stanley Research Thousands Exhibit 14: The US has accounted for twothirds of global crude oil supply growth since 25 World crude oil supply excl. USA (mb/d) Jan-84 Jan-94 Jan-4 Jan-14 Source: IEA, Morgan Stanley Research Outside shale, we estimate that production of crude oil and condensate in the United States will decline modestly over the next few years. However, we see NGL production growing at a reasonable pace. These two effects are broadly of equal magnitude, which means our forecast for total oil liquids growth for the United States is roughly similar to our forecast for shale. By 218, our forecast for production growth in the US (+1.1 mb/d) becomes a high share of our oil demand forecast (+1.3 mb/d). However, it is worth pointing out that this has been the case for many years already. Exhibit 14shows conventional crude oil and condensate production outside the US over the last 3 years. Despite growing steadily in the initial 2 years of this chart, it has nearly flat-lined since 25. Only since early 215, when OPEC stepped-up its production, has non-us crude oil production grown again. Over the decade 25-15, the US met 76% of the growth in conventional crude oil supply in terms of global supply growth, the US has not been the icing; it has been the cake....but there is not much growth in the rest of non-opec The outlook for the rest of non-opec is very different, however. Given the absence of shale, the supply response to the brutal capex cuts of recent years is likely to follow a more traditional pattern. We forecast a small decline in production in 217 in non- US/non-OPEC, returning to a small amount of growth in 218. This relatively flat profile is the balance between a number of factors: Decline rates doing their work: In many basins around the world, decline rates of mature fields have been steadily increasing over time. Exhibit 15 and Exhibit 16 show this effect for the UK and Gulf of Mexico respectively: discoveries made in the 198s decline at a much slower pace after they reach peak-production than those discoveries in the 199s and 2s, and fields discovered in the last 1 years decline faster still. Although the oil price collapse of the last few years has overshadowed this effect, decline rates are still chipping away at supply from mature fields. Reduced infill drilling: To slow decline rates, oil companies drill additional 'infill' wells on existing fields. However, with less cash flow available, infill drilling has suffered too. Accurate data on this is only available in five countries Brazil, Mexico, US, Norway and UK but even this partial data shows a significant decline: across these countries, infill 6

7 drilling fell 4% in 216, and is set to decline another 18% in 217. Decline rates are notoriously difficult to estimate accurately, but this will likely have a negative effect. We estimate that production from currently producing fields in non-us/non-opec will fall 2-3% per year in 217/18, rising to 5-6% in 219/2. Lower contribution from new fields: Following the sharpest capex cuts in the recent history of the oil sector, the impact on new field startups is becoming visible. Data from Rystad Energy shows fields in non-us/non-opec that started in 214 reaching maximum output in 217 of 2. mb/d - see Exhibit 18. However, peak production from startups will fall steadily thereafter. Fields that start in 218 will have a maximum output of 1.6 mbd by 221. Exhibit 15: Recent UK fields decline a lot faster than older ones. UK North Sea: Production decline from peak Source: UK Oil & Gas Authority, Morgan Stanley Research Exhibit 16:.and this is visible in the Gulf of Mexico too, for example US Gulf of Mexico: Production decline from peak Source: BOEM, Morgan Stanley Research Exhibit 17: Where we have visibility, infill drilling has fallen 4% since 214 Infill wells drilled e Brazil Mexico Norway UK US Source: Rystad Energy, Morgan Stanley Research Exhibit 18: The contribution from new fields starting up is set to decline Production from new fields by startup year (mb/d) Source: Rystad Energy, Morgan Stanley Research Offset by some historical growth still coming through... Despite all this, production in non-us/non-opec is still benefitting from recent startups that are still ramping up. Examples of this are Brazil and Canada, which are set to grow strongly. Also, Russia and Kazakhstan are likely to return to growth once the current OPEC/non-OPEC agreement comes to an end....as well as simply 'doing a better job': Finally, many of the major operators have commented positively about the extra volumes they have been able to produce due to improved operating procedures. Shell, BP and Total have said that plant reliability has improved from mid-8% to mid-9%. This yielded an extra 1 kboe/d of oil equivalent for Shell, according to management, and an extra 8 kboe/d for BP and Total each. Assuming half of this is oil, that would equate to 13 kb/d from just these three companies, or 2-3% of their oil production. This can probably be extrapolated across a large part of the oil industry, and has compensated for the impact from decline rates and falling investment. Still, given what has been achieved already, we would be sceptical that similar improvements can continue to be made. 7

8 OPEC Supply Discipline has remained strong The late oil academic Robert Mabro once commented that OPEC should change its logo to a tea-bag "because it only works in hot water". With oil in the mid-$4s in November last year, this must have been the case, as the implementation of the recent OPEC/non- OPEC agreement has been very successful. Using Reuters data for February, production from the countries that are subject to quota fell to mb/d last month, down ~1.1 mb/d from the reference level from which countries promised to make cuts. This equals ~94% of the targeted cut of 1.2 mb/d. We believe OPEC is trying to achieve three objectives with these cuts: drive the forward curve into backwardation, reduce excess inventories, and cause oil prices to increase to ~$6/bbl. So far, the oil market is moving towards these targets but they have not been realised yet: oil prices are ~$5/bbl higher than just before the Vienna meeting in late November, but recently stuck at ~$5-55/bbl. Forward curves are mostly in backwardation from late 217/early 218 onwards but remain in contango at the front end, and OECD inventories have been drawing in recent months but are still ~3 mln bbl above the 5-year average. Given this combination of strong compliance but only partial success, we expect OPEC to extend its agreement into the second half. Exhibit 19: We assume that OPEC production will slowly return in 2H OPEC crude oil production (mb/d) Jan-13 Jan-14 Jan-15 Jan-16 Jan-17 Source: IEA, Morgan Stanley Research Nevertheless, we see OPEC production slowly rising in coming months see Exhibit 19- for a few reasons: First, we assume production growth in Libya and Nigeria. Both countries have suffered large unplanned outages, mostly due to social disruptions. However, it is likely that some of this will come back. We have assumed Libyan production increases slowly from ~7 kb/d recently to ~9 kb/d by 2H17. Also, we have pencilled in ~1,6 kb/d for Nigeria in 2H, up from its January level of ~1,44 kb/d. Second, we still anticipate normal seasonality in Saudi Arabia's production: Saudi Arabia usually produces ~3 kb/d more during June-August to meet extra demand for power generation. Despite the OPEC agreement, we still assume this normal seasonality in Saudi Arabia's production. Third, we expect that the oil market can absorb this: With this increase, we still see inventory draws of ~1 mb/d later this year. That will probably suffice to reduce the inventory overhang at an acceptable pace. Towards the end of 217, we expect OPEC discipline to start waning. Our current forecasts foresee OPEC production growing ~4 kb/d in 218 again, driven by the core Middle Eastern members Saudi Arabia, Iran, Iraq, the UAE and Kuwait. This is one of the key factors why we see a broadly balanced market next year. Medium-to-long term Over the medium term, a return to $7-75/bbl is still necessary... 8

9 Exhibit 2: By 22, approximately ~1.5 mb/d of supply will need to come from projects that have yet to be sanctioned with break-even prices of $7-75/bbl Oil supply (mb/d) Source: Rystad Energy, Morgan Stanley analysis Above $65 $55-65 $45-55 Below $45 Under dev. Producing A return of OPEC production, combined with growth from the most prolific shale plays, can keep a lid on oil prices for some time. However, ultimately some higher cost projects will be needed. We estimate that ~1.5 mb/d of our 22 demand forecast will need to come from projects that have not been awarded yet but that have break-evens around $7-75/bbl. Beyond 22, the reliance on this category increases rapidly. By 225, as much as 7-8 mb/d may need to come from projects with those higher break-evens. Using Rystad estimates, the vast majority of projects in the 'above-$65/bbl' category break-even at $7-75/bbl. Towards 22, prices will need to rise to a level that incentivizes these projects....and new bunker fuel regulations will likely provide an extra boost to demand, driving oil prices to those levels Late last year, the International Maritime Organisation (IMO) announced new regulations for bunker fuel (i.e. fuel for ships) to come into force from 22. Bunker fuel represents just 4-5% of total product demand. Still, this change should have a meaningful impact on oil prices, in particular those of benchmark crudes such as Brent. We suspect this will be sufficient to see prices reach the $7-75/bbl mentioned above. In short, the IMO's new regulation lowers the maximum amount of sulfur in bunker fuel from 3.5% to.5%. Shipping companies can deal with this in three ways. They can either: 1) install scrubbers, which remove part of the sulfur in the ship's emissions, 2) convert their ships to run on LNG, or 3) switch to low sulfur bunker fuel. The first two options face various practical obstacles and show little sign of being widely adopted, at least not in time for 22. This leaves switching to low sulfur bunker fuel as the likeliest option initially. However, this is set to create a 'scamble' for low sulfur fuel, which will impact oil markets - we suspect in the following ways: First,demand for crude oil will increase, possibly by ~ 1mb/d: To meet the additional demand for low sulfur fuel, refiners can expand their conversion capacity (e.g. build additional cokers, crackers). However, this requires huge investment and it is unclear whether the return on these investments will be sufficient over the long term. As an alternative, refiners can simply run more crude through their existing units. A recent study by IHS found that crude runs will need to rise by 1 mb/d to meet demand for low-sulfur fuels. This would tighten the market for crude oil, and hence benefit prices. Second, the price differential between 'sweet' and 'sour' crudes will widen. Demand for sweet crudes (i.e. crudes that naturally contain less sulfur) will rise relative to demand for sour crudes. Most benchmark crudes, such as Brent and WTI, are sweet crudes and will probably fetch a premium by then. In short, the IMO's specification changes disproportionately benefit the oil prices that most investors look at. Third, refining margins will improve: Starting in 219, refiners will need a price signal to produce enough low-sulfur fuel. As mentioned, this will lead to wider sweet/sour differentials, but also wider heavy/light spreads. These tend to be positive for refining economics. High conversion refineries with coking units, and those cracking sweet crudes, should see the greatest benefit. 9

10 Price conclusions & risk/reward discussion Recovery in 217, stable in 218, before a further rise to 22 Exhibit 21: Brent crude oil price forecasts ($/bbl) Bear case Base Bull case End End End End Source: Morgan Stanley Research Our base case forecast sees oil prices rise during 217 as inventory draws force the forward curve into backwardation. After that, a more challenging period ensues in 218 as declining OPEC discipline and rising shale output keep the market in broad balance. Towards 22 however, another period of strength emerges as oil prices converge towards long-term marginal cost, supported by the effects of the IMO specification changes. Still, there are meaningful risks to this trajectory. Below we discuss alternative scenarios that could lead to either higher or lower oil prices: The bear case - oil stays around current levels A number of factors come together in the bear case that drive oil prices down first, before recovering to current levels: Extra supply from lower outages: Approximately 1.3 mb/d of production from Libya and Nigeria are currently 'off the market' due to social unrest. Another.5 mb/d of production in the Neutral Zone (shared between Saudi Arabia and Kuwait) is currently shut in. We assume that part of this production comes back, but not all. Still, this could surprise on the upside. Tax competition amongst non-opec countries: One of the factors that has made US shale so competitive is the low tax rate that the US government levies relative to other oil producing countries. In the rest of non-opec, tax is usually the highest cost for operators by some margin. If other non-opec countries see capex flow away from their resources and towards shale, they might try to attract those investments back by offering lower tax rates. This has the potential to lower our estimate of long-run marginal cost substantially. 'Peak demand' fears cause an acceleration of supply: Irrespective of whether the peak in oil consumption is truly in sight or not, the expectation that it is might lead low-cost producers in the Middle East to increase production as much as possible. Their capacity to do so over the next few years is hard to gauge but could surprise on the upside. Shale costs continue to fall: Shale has continuously surprised on the upside, both in terms of its ability to produce as well as the cost of that production. If these upside surprises continue, our estimate of marginal cost may be too high. Stronger dollar: Morgan Stanley's economists are upbeat about the prospects of further dollar strength. A stronger dollar makes oil more expensive in other currencies, and can therefore lead to lower global oil demand. 1

11 The bull case - back to $95 Exhibit 22: Recent cuts to upstream investments have been large by historical standards Change in global upstream capex from two years prior (%) 1% 8% 6% 4% 2% % -2% -4% -6% Source: Rystad Energy, Morgan Stanley Research Oil markets have historically shown remarkable volatility, and the range of possible outcomes is often surprisingly wide. It may seem unlikely now, but when the following factors conspire, oil prices could get close to previous peak levels: Capex cuts have a dramatic impact on mature fields: The decline in upstream investments in 215 and 216 has been large by historical standards, comparable only with those seen in the aftermath of the 1986 oil price crash. The impact this will ultimately have is yet to be seen and could be more severe by 219/2 than we currently assume. If so, it would be hard to recover from that quickly. Excessive fears over 'peak demand' lead to under-investment: Most conventional oil fields need 1+ years to earn their full return. If the long run outlook for demand is uncertain, it could hold back investment in those long-term projects. If demand does not peak, investment may turn out to be insufficient. Demand surprises positively: Oil demand has accelerated during this recent downturn. With the global population still growing, and 85% of people living in emerging countries where GDP per capital is still rising, oil demand could surprise to the upside, including over the long term and especially with oil prices much lower than before. More regulatory changes are introduced: Regulations are usually designed to lower demand for dirty fuels, which often create the expectation that they are negative for oil demand overall. However, the opposite is often the case. Fuel regulations usually shift demand to cleaner fuels. To meet increased demand for the cleaner fuel, refiners often need to process more crude oil, as they can vary the different 'product yields' only within certain ranges. Following 'diesel-gate', the prospect for further regulatory changes are high. However, we think the chances are that this will increase rather than decrease demand for crude oil. Shale productivity reverses - high-grading becomes low-grading: Shale is a heterogeneous resource, and it is often difficult to extrapolate from one well to the next. Cost and production has surprised positively in the last few years but it is hard to separate the effect of high-grading (concentrating rigs in the best areas as drilling activity falls) from underlying 'like-for-like' improvements. Now that drilling activity is increasing again, it is possible that well productivity and cost metrics start to reverse again as inevitably some less prolific acreage will need to be drilled at some point and oil service costs increase again. Outages go from high to higher: Whilst the outages in Libya and Nigeria have lowered global supply, it is possible that unplanned disruptions increase further. There is no shortage of oil producers with the potential for unrest. Weaker dollar: The most recent recession in the US is now 8 years ago, which is long by historical standards. Whilst our economists' outlook is reasonably upbeat, a US recession remains a possibility. This could reverse the path of interest rate hikes and lead to a weaker dollar. Although this would likely lower oil demand in the US, it could raise demand elsewhere, depending on the severity of the recession. A weaker US dollar is usually positive for oil prices. 11

12 Balance Overview Exhibit 23: Global Balance Summary (mmb/d) YoY Change 216 1Q17 2Q17 3Q17 4Q Q18 2Q18 3Q18 4Q Demand OECD US Euro Non-OECD China India Non-OPEC Supply US Canada Russia OPEC NGLs/Condensates Call on OPEC Crude OPEC Crude Implied stock build/(draw) Source: EIA, IEA, Rystad, Morgan Stanley Exhibit 24: Global Balance (mmb/d) Q13 3Q13 1Q14 3Q14 1Q15 3Q15 1Q16 3Q16 1Q17e3Q17e1Q18e3Q18e Source: EIA, IEA, Rystad, Morgan Stanley Implied stock build/(draw) Demand Supply

13 Supply - Demand Overview Exhibit 25: OECD Demand Growth (annual, mmb/d) Exhibit 26: Non-OECD Demand Growth (annual, mmb/d) Mar-13 Sep-13 Mar-14 Sep-14 Mar-15 Sep-15 Mar-16 Sep-16. 1Q13 3Q13 1Q14 3Q14 1Q15 3Q15 1Q16 3Q16 Source: IEA Source: IEA Exhibit 27: Non-OPEC Supply Growth (annual, mmb/d) Mar-13 Sep-13 Mar-14 Sep-14 Mar-15 Sep-15 Mar-16 Sep-16 Source: IEA Exhibit 28: OPEC Supply Growth (annual, mmb/d) Mar-13 Sep-13 Mar-14 Sep-14 Mar-15 Sep-15 Mar-16 Sep-16 Source: IEA 13

14 Key Agency Revisions Exhibit 29: Oil Demand Growth (mmb/d) 1.7 Exhibit 3: Non-OPEC Supply Growth (mmb/d) Jul-16 Sep-16 Nov-16 Jan-17 Mar-17 May-17 Jul-17 Sep-17 Nov Jul-16 Sep-16 Nov-16 Jan-17 Mar-17 May-17 Jul-17 Sep-17 Nov-17 IEA EIA OPEC IEA EIA OPEC Source: EIA, IEA, OPEC Source: EIA, IEA, OPEC Exhibit 31: Call on OPEC 217 (mmb/d) Exhibit 32: Change in the Call on OPEC (mmb/d) Source: EIA, IEA, OPEC Jul-16 Sep-16 Nov-16 Jan-17 Mar-17 May-17 Jul-17 Sep-17 Nov-17 IEA EIA OPEC Jul-16 Sep-16 Nov-16 Jan-17 Mar-17 May-17 Jul-17 Sep-17 Nov-17 Source: EIA, IEA, OPEC IEA EIA OPEC 14

15 Prices & Differentials Exhibit 33: Crude Prices($/bbl) Exhibit 35: WTI vs Brent ($/bbl) Exhibit 37: Dubai vs Brent ($/bbl) OPEC Basket Brent WTI Jan-15 Jul-15 Jan-16 Jul-16 Jan-17 Source: Bloomberg Source: Bloomberg Source: Bloomberg Exhibit 34: Nigeria vs Brent ($/bbl) Exhibit 36: Angola vs Brent ($/bbl) Exhibit 38: Heavy - Light Diff ($/bbl) Qua Iboe Bonny Light Cabinda Kissanje LLS - Maya Urals-Brent (RHS) Source: Platts, Bloomberg Source: Platts, Bloomberg Source: Platts, Bloomberg 15

16 Forward Curves and Time Spreads Exhibit 39: Brent Forward Curve ($/bbl) Exhibit 41: WTI Forward Curve ($/bbl) Exhibit 43: Dubai Forward Curve ($/bbl) Feb-8 Mar-1 Mar-8 Feb-8 Mar-1 Mar-8 Feb-8 Mar-1 Mar-8 Source: Bloomberg Source: Bloomberg Source: Bloomberg Exhibit 4: Brent Time Spreads ($/bbl) 1 Exhibit 42: WTI Time Spreads ($/bbl) 15 Exhibit 44: Dubai Time Spreads ($/bbl) Brent 1-12 Brent 1-6 WTI 1-12 WTI 1-6 Dubai 1-12 Dubai 1-6 Source: Bloomberg Source: Bloomberg Source: Bloomberg 16

17 Oil vs Other Things Exhibit 45: Brent Crude vs Inverse TWD 14 Exhibit 46: Brent Crude vs Inflation vs Base Metals Dec-11 Jun-12 Jan-13 Jul-13 Feb-14 Aug-14 Mar-15 Oct-15 Apr-16 Nov-16 May-17 Brent ($/b) TWD (RHS) Source: Bloomberg 2 1. Dec-11 Jun-12 Jan-13 Jul-13 Feb-14 Aug-14 Mar-15 Oct-15 Apr-16 Nov-16 May-17 Brent ($/b) Base Metals Spot Price Index Inflation Expectations (RHS) Source: Bloomberg 17

18 Refining Margins Exhibit 47: Rotterdam - Brent Cracking ($/bbl) Source: Thomson Reuters Exhibit 48: Rotterdam - Brent Hydroskimming ($/bbl) Source: Thomson Reuters Exhibit 49: USGC - WTI Cracking ($/bbl) Source: Thomson Reuters Exhibit 5: USGC - Brent Cracking ($/bbl) Source: Thomson Reuters 18

19 Exhibit 51: Med - Urals Cracking ($/bbl) 8 Exhibit 52: Med - Urals Hydroskimming ($/bbl) Source: Thomson Reuters Source: Thomson Reuters Exhibit 53: Singapore - Dubai Cracking ($/bbl) Source: Thomson Reuters Exhibit 54: Singapore - Dubai Hydroskimming ($/bbl) Source: Thomson Reuters 19

20 Product Cracks Exhibit 55: Refined product cracks by region Source: Platts, Bloomberg 2

21 Product Cracks Exhibit 56: Gasoline US vs WTI ($/bbl) Source: Bloomberg Exhibit 57: Diesel US vs WTI ($/bbl) Source: Bloomberg Exhibit 58: Heating Oil US vs WTI ($/bbl) Source: Bloomberg Exhibit 59: Gasoline NWE vs Brent ($/bbl) Source: Platts, Bloomberg Exhibit 6: Diesel NWE vs Brent ($/bbl) Source: Platts, Bloomberg Exhibit 61: Fuel Oil NWE vs Brent ($/bbl) Source: Platts, Bloomberg Exhibit 62: Gasoline Singapore vs Brent ($/bbl) Source: Platts, Bloomberg Exhibit 63: Gasoil Singapore vs Brent ($/bbl) Source: Platts, Bloomberg Exhibit 64: Fuel Oil Singapore vs Brent ($/bbl) Source: Platts, Bloomberg 21

22 Inventories - OECD Exhibit 65: OECD Total Liquid Stocks (mmb) Exhibit 67: OECD Total Crude Stocks (mmb) Exhibit 69: OECD Total Product Stocks (mmb) 3,2 3,1 3, 2,9 2,8 2,7 2,6 2,5 2,4 Jan Mar May Jul Sep Nov Range Source: IEA Exhibit 66: OECD Americas Total Liquid Stocks (mmb) 1,25 1,2 1,15 1,1 1,5 1, Jan Mar May Jul Sep Nov Range Source: IEA Exhibit 68: OECD Europe Total Liquid Stocks (mmb) 1,7 1,6 1,5 1,4 1,3 1,2 1,1 Jan Mar May Jul Sep Nov Range Source: IEA Exhibit 7: OECD Asia Oceania Total Liquid Stocks (mmb) 1,8 1,1 47 1,7 1,5 45 1,6 1, 43 1, , , ,2 Jan Mar May Jul Sep Nov Range Source: IEA 8 Jan Mar May Jul Sep Nov Range Source: IEA 35 Jan Mar May Jul Sep Nov Range Source: IEA 22

23 Inventories - United States Exhibit 71: US Crude - Total (mmb) Range Y Avg Exhibit 72: US Crude - Cushing (mmb) Range Y Avg Exhibit 73: US Gasoline (mmb) Range Y Avg Exhibit 74: US Distillate (mmb) Range Y Avg

24 Inventories - ARA Exhibit 75: ARA Gasoline (mmb) Range Y Avg Source: PJK, Bloomberg Exhibit 76: ARA Gasoil (mmb) Range Y Avg Source: PJK, Bloomberg Exhibit 77: ARA Jet-Kerosene (mmb) Range Y Avg Source: PJK, Bloomberg Exhibit 78: ARA Fuel Oil (mmb) Range Y Avg Source: PJK, Bloomberg 24

25 Inventories - Singapore Exhibit 79: Singapore Light Distillates (mmb) Range Y Avg Source: International Enterprise, Bloomberg Exhibit 8: Singapore Middle Distillates (mmb) Range Y Avg Source: International Enterprise, Bloomberg Exhibit 81: Singapore Residues (mmb) Range Y Avg Source: International Enterprise, Bloomberg 25

26 United States - Crude Oil Exhibit 82: Crude Oil Stocks (mmb) Range Y Avg Exhibit 83: Crude Oil Production (mmb/d) Jan-15 May-15 Aug-15 Nov-15 Feb-16 Jun-16 Sep-16 Dec-16 Mar-17 Weekly Monthly Exhibit 84: Net Crude Imports (mmb/d) Range Y Avg Exhibit 85: Crude Runs (mmb/d) Range Y Avg

27 United States - Gasoline Exhibit 86: Gasoline Stocks (mmb) Range Y Avg Exhibit 87: Long-term Gasoline Stocks (mmb) Jan-7 Jan-9 Jan-11 Jan-13 Jan-15 Jan-17 Exhibit 88: Gasoline Product Supplied (mmb/d) Range Y Avg Exhibit 89: Gasoline Product Supplied (% YoY) 15% 1% 5% % -5% -1% -15% Jan-7 Jan-9 Jan-11 Jan-13 Jan-15 Jan-17 YoY % Change 4 per. Mov. Avg. (YoY % Change) 27

28 Exhibit 9: Gasoline Production (mmb/d) Range Y Avg Exhibit 91: Gasoline Imports (mmb/d) Range Y Avg Exhibit 92: Gasoline Stocks Days of Supply (4 wk avg) Range Y Avg Exhibit 93: Gasoline Yield (%) 7% 68% 66% 64% 62% 6% 58% 56% 54% 52% 5% Range Y Avg

29 United States - Distillate Exhibit 94: Distillate Stocks (mmb) Range Y Avg Exhibit 95: Long-term Distillate Stocks (mmb) Jan-7 Jan-9 Jan-11 Jan-13 Jan-15 Jan-17 Exhibit 96: Distillate Product Supplied (mmb/d) Range Y Avg Exhibit 97: Distillate Product Supplied (% YoY) 4% 3% 2% 1% % -1% -2% -3% -4% Jan-7 Jan-9 Jan-11 Jan-13 Jan-15 Jan-17 YoY % Change 4 per. Mov. Avg. (YoY % Change) 29

30 Exhibit 98: Distillate Production (mmb/d) Range Y Avg Exhibit 99: Distillate Exports (mmb/d) Range Y Avg Exhibit 1: Distillate Stocks Days of Supply (4 wk avg) Range Y Avg Exhibit 11: Distillate Yield (%) 34% 32% 3% 28% 26% 24% 22% 2% Range Y Avg

31 Positioning Exhibit 12: Managed Money Crude Positioning (' lots) Jan-14 May-14 Sep-14 Jan-15 May-15 Sep-15 Jan-16 May-16 Sep-16 Jan-17 Longs Shorts Net Source: ICE, CFTC, Bloomberg Exhibit 13: Producer/Merchant Crude Positioning (' lots) Jan-14 May-14 Sep-14 Jan-15 May-15 Sep-15 Jan-16 May-16 Sep-16 Jan-17 Longs Shorts Net Source: ICE, NYMEX, CFTC, Bloomberg Exhibit 14: Net Long Positioning - Gasoline (' lots) Managed Money Producer Source: NYMEX, CFTC, Bloomberg Exhibit 15: Net Long Positioning - Distillate (' lots) Managed Money Producer Source: ICE, CFTC, Bloomberg 31

32 Supply/demand balance Exhibit 16: Global demand and non-opec supply (mb/d) 1Q15 2Q15 3Q15 4Q Q16 2Q16 3Q16 4Q Q17 2Q17 3Q17 4Q Q18 2Q18 3Q18 4Q Demand OECD US Europe Japan Canada Mexico Other Non-OECD China India Russia Brazil Middle East Other Total demand Supply Non-OPEC Crude + Condensate United States Shale production Alaska Other Lower Canada Mexico UK Norway Russia China Kazakhstan Azerbaijan Brazil Colombia Indonesia Other Source: IEA, Morgan Stanley Research estimates 32