While the US announcement of production cuts last week was premature, many n countries have since commented on their desire to participate, and we believe that a coordinated cut is now more likely than not. Assuming that a deal is reached, which is in no way certain, the key question will be whether its size and timing will improve global oil balances sufficiently to support prices above current levels. Our updated 2020 global oil balance suggests that a 10 mb/d headline cut (for an n effective 6.5 mb/d cut in production) would not be sufficient, still requiring an additional 4 mb/d of necessary price induced shut-ins. While this argues for a larger headline cut of close to 15 mb/d, we believe this would be much harder to achieve, since the incremental burden would likely need to fall on Saudi Arabia to be effective. Further, our price modeling suggests that Brent prices near $35/bbl already reflect such an outcome, with last week’s rally having brought crude prices to levels that likely slow the large-scale US production drop that are necessary to a deal in the first place. Net, while the prospect of a deal can support prices in coming days, we believe n this support will soon give way to lower prices with downside risk to our near-term WTI $20/bbl forecast. Ultimately, the size of the demand shock is simply too large for a coordinated supply cut, setting the stage for a severe rebalancing. While the path of the demand normalization will remain key to the subsequent price recovery, lasting supply cuts will matter too and could create upside risks to our $40/bbl October Brent forecast. Damien Courvalin +1(212)902-3307 | damien.courvalin@gs.com Goldman Sachs & Co. LLC Callum Bruce +1(212)902-3053 | callum.bruce@gs.com Goldman Sachs & Co. LLC Jeffrey Currie +44(20)7552-7410 | jeffrey.currie@gs.com Goldman Sachs International Oil Record cuts still not enough 8 April 2020 | 12:07PM EDT Investors should consider this report as only a single factor in making their investment decision. For Reg AC certification and other important disclosures, see the Disclosure Appendix, or go to www.gs.com/research/hedge.html. For the exclusive use of RWANI@HARTREEPARTNERS.COM b789af4ce4cb4dd6845eb024ed5bfb41 Record cuts still not enough While the US announcement of a production cut last week was premature, many n countries have since commented on their desire to participate, including OPEC+ members, Norway, Brazil and Canada. We believe a coordinated cut is now therefore more likely than not as the collapse in oil demand is creating either a strong incentive to act as revenues collapse or an actual need to shut-in as physical storage constraints get reached. Importantly, this applies to the low-cost producers that had embarked on a recapture n of lost market share just a month ago. Russia would likely need to cut production anyways in coming weeks given declining refining intake (piped or domestic), the inland nature of its production and its limited storage capacity. In the case of Saudi, the collapse in global oil demand will likely prevent it from placing its planned April supply of 12.3 mb/d, with reports of Indian refiners claiming force majeure and its own storage increasingly filled too. Admittedly a deal hasn’t been reached yet with formal broad discussions planed for n both Thursday, during an expanded OPEC+ meeting, as well as Friday with G20 energy ministers. The main points that could delay an agreement remain the size of the cut (and implicitly the reference level from which to cut production) and the participation of the US (with some producers making its participation a requirement but the US administration denying any plan to do so). We believe that these can delay but are unlikely to derail an eventual agreement. The ability for the US to participate in coordinated cut is limited by regulation (as we n discuss in this footnote1) although such hurdles are unlikely to prevent a deal. First, the US administration has expressed its willingness to use political/defense leverage or tariffs 2 on Saudi and Russia oil imports to obtain a cut. Second, and perhaps most importantly, the US can point to its contribution with upstream activity already dropping and the US Department of Energy forecasting on April 6 that production would be falling by 2 mb/d by year-end relative to its previous March forecast. 1 At the federal level, productions cuts could be achieved by export quotas (executive order provisioned for by the legislation that ended the export ban in late 2015) however this would effectively back up crude in the US, further depressing local prices initially. Production cuts on federal land could potentially be implemented but would impact some producers much more or lead to potentially elevated shut-in and restart costs in the case of offshore Gulf of Mexico production. State level restrictions would create issues as well, first as it is not clear whether all major producing states have the same authority as the Texas Railroad Commission in reducing output (Chapter 85 the state’s Natural Resources Code). Second, such state level action would likely violate the Commerce Clause that limits the ability for US states to enter into even implicit agreements with foreign countries. 2 Tariffs are admittedly of limited impact as they are ultimately borne by US consumers and would only affect 0.5 mb/d of imports vs. a current demand loss of 5 mb/d. 8 April 2020 2 Goldman Sachs Oil For the exclusive use of RWANI@HARTREEPARTNERS.COM b789af4ce4cb4dd6845eb024ed5bfb41 Assuming that a deal is reached - our base case now - the key question will be n whether its size and timing will improve global oil balances sufficiently to support prices above current levels. This is key, as a cut that would prove too little too late would lead to storage saturation and additional necessary production shut-in, with distressed producers driving physical crude prices and spot oil prices sharply lower. Our updated 2020 global oil balance suggests that a 10 mb/d headline cut would not be sufficient, still requiring necessary price induced shut-ins on top of such voluntary curtailments. To come to this conclusion, we update and refine our 2020 global oil balance (we n provide a full breakdown of this analysis in Section 2 of this report). First, we once again slightly lower our demand expectations as isolation policies get deployed in a still growing number of countries and with their lift date increasingly pushed back. We now expect a hit to April/May/June global oil demand of 22/16/9 mb/d (19/12/6 mb/d previously). Second, we introduce an estimate of peak monthly storage fill capacity, which starting from current stock levels is of 17 mb/d in April (with 20 mb/d likely achieve last week3 ), 13 mb/d in May and only 5 mb/d in June. From a base-case production perspective, we broadly maintain our last published production forecast from March 17 (which reflected steady declines in output but did not allocate the necessary shut-in), allowing us to estimate both the sizes of the voluntary and necessary shut-ins. 3 Kpler satellite data points to a 5 mb/d global onshore crude build, with historical demand collapses typically leading product stocks to build by 1.5x crude (so 7.5 mb/d) and floating storage estimated to have increased by 7 mb/d. Exhibit 1: Quickly rising oil inventories are leading more producers to consider cuts Crude and products on water (mb) Exhibit 2: The US administration can point to the sharp expected decline in US production forecasted by its Department of Energy as sign of participation to the cuts US oil production revision from monthly EIA Short-Term Energy Outlook 1200 1250 1300 1350 1400 1450 1500 1550 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2017 2018 2019 2020 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 Oct-19 Jan-20 Apr-20 Jul-20 Oct-20 Jan-21 Apr-21 Jul-21 Oct-21 Apr vs. Mar Apr vs. Feb -2 mb/d Source: Goldman Sachs Global Investment Research, Kpler Source: EIA, Goldman Sachs Global Investment Research 8 April 2020 3 Goldman Sachs Oil For the exclusive use of RWANI@HARTREEPARTNERS.COM b789af4ce4cb4dd6845eb024ed5bfb41 In terms of cut, we assume a 10 mb/d “headline” cut starting on May 1 allocated as n follows: 4.9 mb/d to core-OPEC from April levels - with all producers cutting instead from 1Q20 average levels - 1.0 mb/d from Russia, 1.0 mb/d from Canada and 1.5 mb/d from the US, and the remaining 1.5 mb/d from the rest of OPEC+ participants and a handful of new participants. In practice, such a cut could be presented as one of 10 mb/d although would represent an only 6.5 mb/d “effective” supply reduction relative to 1Q20 production levels. Our balance shows that such a cut would still leave 4.0 mb/d of supply in excess of n available storage capacity in both April and May having to be shut-in out of necessity. This points to a 10 mb/d “headline cut” still leading prices falling back below cash costs as local storage saturates, with WTI likely having to retrace below $20/bbl. For example, Platts Midland WTI prices at $24/bbl for May, just above cash-costs for most Permian producers and not a level where we would expect large shut-in forced by the upcoming inventory saturation. And while US production was down 0.6 mb/d in this week’s EIA data, pipeline scrape data from Genscape only points to declines of 0.3 mb/d, both well short of the several million barrels per day of required declines in the month of April.. Exhibit 3: We adjust our demand expectations down slightly to reflect further tightening in global lockdowns and a slower recovery in May kb/d demand impact from Covid19 Exhibit 4: Inventory builds are tracking in line with our balance and are quickly filling available storage Global inventory movements (mb) vs. implied by S&D balance. Effective capacity is storage capacity factoring in limits to the rates of stock builds. -25,000 -20,000 -15,000 -10,000 -5,000 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec OECD (kb/d) China (kb/d) Other EM (kb/d) Jet Total -100 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 Jan Feb Mar Apr May Jun Landed crude Landed products Crude on water Product on water Total GS Balance Effective Capacity Source: Goldman Sachs Global Investment Research, IEA, OAG Onshore petroleum product stocks are implied from crude to produce a declining crude/product stock ratio seen in previous demand shocks. Source: Goldman Sachs Global Investment Research, Kpler, IEA, EIA 8 April 2020 4 Goldman Sachs Oil For the exclusive use of RWANI@HARTREEPARTNERS.COM b789af4ce4cb4dd6845eb024ed5bfb41 While this would simply argue for a larger 15 mb/d “headline” cut (with “effective” n supply reduction of 10 mb/d), we believe it would be much more difficult to pro-actively allocate an incremental 5 mb/d of voluntary cuts across countries that do not have Saudi’s geological oilfield flexibility, with the Kingdom’s own ability to cut to 8 mb/d likely challenged. In fact, recent price levels would likely keep US crude prices at levels that disincentivize the large scale US production drop that backdrop the rationale for the cuts in the first place. As a result, an 10 mb/d “headline cut” that would still require additional large but short-lived price-induced cuts is likely the preferred outcome for core-OPEC, as it ensures the contribution of many producers (and higher OSPs even should Brent prices return to their lows). While the prospect of a deal can support prices near current levels in coming days, n we believe this support will soon give way to lower prices, due to both positioning and fundamentals. First and foremost, we don’t believe that a 10 mb/d “headline” production cut can shore up oil balances with lower prices still needed. Second, the sharp short-covering rally of the past week has already brought prices to the levels that we forecast in 3Q20 when the surplus would be ending, a still uncertain outcome. Third, downward price pressure in coming days will be exacerbated by the roll of the long WTI future positions underlying the recent increase in speculative buying4 , as visible on April 6 and in early 2009. Fourth, the potential inability to broker a cut sufficiently large to support seaborne crude prices suggests that Saudi producing at high levels instead to maximize cash flow would actually not be irrational, an outcome that appears mis-priced, with the WTI options market currently only assigning an 18% probably that prices reach $15/bbl by mid-May. Ultimately, the size of the demand shock is simply too large for a coordinated supply n cut, setting the stage for a violent rebalancing (see Top of Mind - Oil’s Seismic Shock page 9). Looking forward, our base-case remains unchanged, with a shift to a deficit 4 ETFs like the USO do not hold physical crude but instead crude oil futures that need to be rolled every month before their expiration. Such flows exert downward pressure on WTI prices and timespreads into the close as front-month contracts are sold to purchase second-nearby contracts. Exhibit 5: Even assuming a 10 mb/d headline cut (6.5 mb/d effective), forced shut-ins would still be required in April through June to balance the market Supply and inventory changes required to offset Covid-19 demand impact (kb/d). Note supply/stock lines are accumulative. Exhibit 6: Physical markets have however recently rallied, lessening the incentive to curtail output Physical crude assessments $/bbl -5000 -4000 -3000 -2000 -1000 0 1000 2000 3000 Persistent shutins Required shutins Core OPEC, Russia RoW (excl. shut-ins) 0 10 20 30 40 50 60 70 Jan-20 Feb-20 Mar-20 Apr-20 Bakken Midland LLS Mars WCS Hardisty WTS WTI Brent Source: Goldman Sachs Global Investment Research, IEA Source: Goldman Sachs Global Investment Research, Platts, ICE 8 April 2020 5 Goldman Sachs Oil For the exclusive use of RWANI@HARTREEPARTNERS.COM b789af4ce4cb4dd6845eb024ed5bfb41 by July and a gradual recovery in Brent prices (from lower levels than currently) to $40/bbl by October as inventories start to wind down. The path of the demand recovery will be of course essential to how this price recovery plays out. But after such a violent rebalancing, supply cuts will matter increasingly and could create upside price risks later this year. It will take time to restart the 10.5 mb/d of shut-in wells (6.5 mb/d voluntarily and 4 n mb/d out of necessity) which could also lead to permanently lost productive capacity. For example, if we assume (1) a recovery in OPEC/Russia production through 3Q-4Q20 back to 1Q20 levels, (2) that production declines expected outside of shut-ins (due to lower capex and decline rates) don’t reverse and (3) that 1 mb/d of the remaining production cuts take at least a year to return online, our oil balance would feature fully normalized inventory levels by early 2021 and point to Brent prices at $55/bbl. While such a rally could be delayed by a faster unwind of the cuts, it does illustrate that the violent rebalancing we expect in coming weeks can set the stage for a potential large price recovery later this year. Exhibit 7: In the short-term, both fundamentals and positioning (including USO roll) create downside price risks Dates of futures roll (shaded bars), WTI front month timespreads ($/bbl) Exhibit 8: A lagged unwind of the supply cuts would instead create risks that prices recover faster than our base case Change in global oil stocks from Dec-19 (mb). -10 -7.5 -5 -2.5 0 2.5 Oct-08 Nov-08 Dec-08 Jan-09 Feb-09 Mar-09 Apr-09 May-09 S&P GSCI roll dates WTI 1st vs. 2nd mo timespread -200 0 200 400 600 800 1000 1200 1400 Base case Assuming no 2H20 ramp Capacity Source: S&P, Goldman Sachs Global Investment Research Source: Goldman Sachs Global Investment Research, IEA 8 April 2020 6 Goldman Sachs Oil For the exclusive use of RWANI@HARTREEPARTNERS.COM b789af4ce4cb4dd6845eb024ed5bfb41 Framing the largest oil production cut in history We first assume a May 1 implementation of the cuts, consistent with media n reports. Despite the global oil market filling storage at a record pace, such a lagged implementation would be consistent with past cuts and the fact that April loadings are already committed. In fact, the geological difficulty of shutting production should leave producers who do not face immediate saturation to initially fall short of their commitment. For that reason, we assume that only core-OPEC cuts are fully implemented in May with all other countries only reaching their full commitment in June (except Canada where we assume forced shut-ins already in April). This start date will likely to be the easiest part of a deal to agree to, with the level off n which to base the cut (and the size of the cut as we discuss later) likely to prove much more contentious. Russia is proposing a cut from 1Q20 average levels likely reflecting its inability to increase production in April. Core-OPEC producers are instead advocating a cut from their April production surge level, allowing to optically boost their cut by 3.5 mb/d. This is ultimately semantics and our reference to a cut is relative to 1Q20 levels for all countries except core-OPEC, which will be off their April stated output. We initially consider a 10 mb/d headline cut, allocated as follows: 4.9 mb/d to n core-OPEC from April levels - with all producers cutting instead from 1Q20 average levels - a likely inevitable 1.0 mb/d from Russia, an unavoidable 1.0 mb/d from Canada and 1.5 mb/d from the US, and the remaining 1.5 mb/d from the rest of OPEC+ participants and a handful of new participants (Norway, Brazil). The cut for the latter group amounts to a 7% cut from 1Q20 levels. In practice, such a cut could be presented as one of 10 mb/d although would represent an only 6.5 mb/d effective supply reduction relative to 1Q20 production levels. 8 April 2020 7 Goldman Sachs Oil For the exclusive use of RWANI@HARTREEPARTNERS.COM b789af4ce4cb4dd6845eb024ed5bfb41 To determine the impact of such supply cuts, we first update our expected demand n impact from the coronavirus disruption. With isolation policies being deployed in still a growing number of countries and with their lift date increasingly pushed back, we are slightly downgrading once again our demand expectations, with an expected hit to April/May/June global oil demand of 22/16/9 mb/d (19/12/6 mb/d previously). We now expect that 2Q20 demand will be down 14 mb/d yoy and also lower our 3Q and 4Q demand to falling 3.3 and 1.1 mb/d yoy. Admittedly these are for once only modest downgrades as we wait for implied demand data to test whether our modeling off China’s path and country fixed effects proves valid. Exhibit 9: An announcement of a 10 mb/d cut would result in an effective 6.5 mb/d cut from 1Q20 levels GLOPEC agreement arithmetic (kb/d) Country Jan-20 Feb-20 Mar-20 1Q20 GSe Claimed^ Announced* Effective Saudi Arabia 9,739 9,683 10,300 9,907 11,200 12,100 -3,100 -907 UAE 3,027 3,040 3,300 3,122 3,600 4,030 -1,300 -392 Kuwait 2,710 2,658 2,800 2,723 3,050 3,150 -500 -73 Core-OPEC 15,476 15,381 16,400 15,752 17,850 19,280 -4,900 -1,372 Nigeria 1,760 1,789 1,780 1,776 1,800 2,300 -124 -124 Iraq 4,568 4,508 4,500 4,525 4,420 4,420 -317 -317 Other OPEC non-exempt 3,535 3,554 3,544 3,544 3,509 3,509 -248 -248 OPEC ex. Exempt 9,863 9,851 9,824 9,846 9,729 10,229 -689 -689 OPEC Exempt** 3,642 2,988 2,970 3,200 2,830 2,830 0 0 Russia 11,319 11,288 11,242 11,283 11,250 11,250 -1,000 -1,000 Mexico 1,761 1,746 1,752 1,753 1,753 1,753 -123 -123 Azerbaijan 769 759 766 765 765 765 -54 -54 Kazakhstan 1,972 1,998 1,945 1,972 1,900 1,900 -133 -133 Other non-OPEC+ 2,137 2,138 2,137 2,137 2,143 2,143 -150 -150 Non-OPEC+*** 17,958 17,929 17,842 17,910 17,811 17,811 -1,459 -1,459 US 12,740 12,710 12,720 12,723 12,390 12,390 -1,500 -1,500 Canada 4,731 4,775 4,661 4,722 3,800 3,800 -1,000 -1,000 Others^^ 7,600 7,483 7,446 7,510 7,230 7,230 -506 -506 New GLOPEC 25,071 24,968 24,827 24,955 23,420 23,420 -3,006 -3,006 Total GLOPEC 72,010 71,117 71,863 71,663 71,640 73,570 -10,055 -6,527 ^^ Norway, Brazil, Colombia, Argentina, Egypt, Indonesia, Trinidad Apr-20 Cut scenario Notes: OPEC is crude only as measured by OPEC Secondary Sources, Non-OPEC is IEA crude (inclusive of condensate, except for Canada which also includes Nonconventional oils). Core OPEC cuts from April levels, everyone else from 1Q20 average. * Assume 7% production cut for countries outside those highlighted to generate 10mb/d announcement. ** Exempt countries: Iran, Libya, Venezuela ***Sudan, S. Sudan , Brunei, Malaysia, Bahrain, Oman not shown.Trinidad not shown for New GLOPEC. ^ Equal to country stated production for those highlighted. Otherwise equal to GS April estimate. Source: Goldman Sachs Global Investment Research, IEA Exhibit 10: Nonetheless, the cuts would be historic in magnitude and take core-OPEC/Russia to 2011 production levels kb/d unless stated Dec-18 % Dec-19 % Apr-20 % Quota Last achieved Saudi 322 44% 167 47% 3100 53% 9000 May-11 UAE 96 13% 60 17% 1300 22% 2730 Oct-14 Kuwait 85 12% 55 16% 500 8% 2650 Oct-19 Russia 230 31% 70 20% 1000 17% 10250 Jul-11 Total 733 100% 352 100% 5900 100% 24630 May-11 Allocation of cuts (kb/d) Level Source: Goldman Sachs Global Investment Research, IEA 8 April 2020 8 Goldman Sachs Oil For the exclusive use of RWANI@HARTREEPARTNERS.COM b789af4ce4cb4dd6845eb024ed5bfb41 Big picture, a 10 mb/d cut would be dwarfed by the 19 mb/d average demand loss in n April-May. The second exogenous factor to gauge the potential efficacy of such a cut is therefore the ability of the global oil market to absorb the remaining excess in storage. To assess such storage capacity, we go through the historical peaks in crude and product storage for each DM and EM country and determine that, starting from current inventory fill levels, peak inventory inflows will likely be of 17 mb/d in April, falling to 13 mb/d in May and just 5 mb/d by June at which point global storage capacity of 1.2 bn barrels would be reached. Admittedly, such calculations are only global estimates and are bound to miss the local logistical saturation constraints that are starting to prove binding with for example Canadian Edmonton Syncrude currently trading at $6.67/bbl. Finally, to estimate the impact of the cuts on production, we start from close to our n last published production forecast from March 17 , which reflected steady declines in output but did not allocate the necessary localized shut-in given the magnitude of the fundamental uncertainty. For example, in the case of the US, we assumed that production would have declined by 1.1 mb/d between February and October, roughly 0.6 mb/d less than the EIA’s April STEO projection. Starting from this production path allows us to estimate both the relative sizes of the voluntary and necessary shut-ins. The only material update we make is to cut our Canada supply forecast to Exhibit 11: We adjust our demand expectations down slightly reflecting further tightening in global lockdowns Country level monthly breakdown of demand assumptions (mb/d) Country Demand Lag Intensity Recovery Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2020 USA 18.99 52 110% 70% 0.00 -0.28 -1.86 -6.29 -4.62 -1.79 -1.11 -0.76 -0.57 -0.49 -0.44 -0.42 -1.55 Japan 3.33 35 55% 70% 0.00 -0.09 -0.51 -0.50 -0.29 -0.19 -0.14 -0.11 -0.08 -0.06 -0.06 -0.06 -0.17 Korea 2.41 35 50% 70% 0.00 -0.05 -0.31 -0.33 -0.21 -0.13 -0.10 -0.07 -0.05 -0.04 -0.04 -0.03 -0.11 Canada 2.34 52 40% 70% 0.00 -0.01 -0.11 -0.28 -0.23 -0.13 -0.10 -0.07 -0.05 -0.04 -0.03 -0.03 -0.09 Germany 2.11 52 90% 70% 0.00 -0.03 -0.23 -0.57 -0.44 -0.27 -0.19 -0.14 -0.10 -0.08 -0.07 -0.05 -0.18 Mexico 1.84 52 35% 70% 0.00 -0.01 -0.08 -0.21 -0.16 -0.09 -0.07 -0.05 -0.03 -0.03 -0.02 -0.02 -0.06 France 1.53 48 140% 70% 0.00 -0.05 -0.32 -0.65 -0.48 -0.29 -0.22 -0.15 -0.11 -0.09 -0.06 -0.06 -0.21 UK 1.26 60 140% 70% 0.00 0.00 -0.08 -0.59 -0.49 -0.29 -0.19 -0.14 -0.10 -0.07 -0.06 -0.05 -0.17 Spain 1.15 48 150% 70% 0.00 -0.04 -0.27 -0.52 -0.39 -0.24 -0.17 -0.12 -0.09 -0.07 -0.06 -0.05 -0.17 Australia 1.03 60 40% 70% 0.00 0.00 -0.02 -0.13 -0.12 -0.06 -0.04 -0.03 -0.02 -0.02 -0.01 -0.01 -0.04 Italy 1.03 44 170% 70% 0.00 -0.05 -0.29 -0.46 -0.37 -0.24 -0.18 -0.12 -0.10 -0.08 -0.06 -0.05 -0.16 Other DM 6.14 52 72% 70% 0.00 -0.07 -0.55 -1.38 -1.07 -0.63 -0.45 -0.33 -0.24 -0.19 -0.15 -0.12 -0.43 China 12.59 0 100% 100% -0.47 -3.38 -1.88 -0.84 -0.42 -0.21 -0.11 -0.05 -0.03 -0.02 0.00 0.00 -0.62 India 4.93 60 75% 80% 0.00 0.00 -0.19 -1.22 -1.06 -0.51 -0.31 -0.21 -0.15 -0.13 -0.11 -0.09 -0.33 Russia 3.34 75 60% 80% 0.00 0.00 -0.05 -0.37 -0.60 -0.40 -0.23 -0.16 -0.11 -0.08 -0.06 -0.05 -0.18 Brazil 3.00 60 60% 80% 0.00 0.00 -0.09 -0.58 -0.50 -0.24 -0.16 -0.11 -0.08 -0.06 -0.05 -0.05 -0.16 Saudi 3.06 52 60% 80% 0.00 -0.02 -0.20 -0.58 -0.45 -0.23 -0.17 -0.12 -0.09 -0.06 -0.05 -0.04 -0.17 Iran 1.95 30 100% 80% 0.00 -0.10 -0.62 -0.55 -0.26 -0.17 -0.12 -0.09 -0.07 -0.05 -0.05 -0.04 -0.18 Indonesia 1.74 75 30% 80% 0.00 0.00 -0.01 -0.10 -0.16 -0.11 -0.06 -0.04 -0.03 -0.02 -0.02 -0.01 -0.05 Thailand 1.36 75 30% 80% 0.00 0.00 -0.01 -0.08 -0.12 -0.08 -0.04 -0.03 -0.02 -0.02 -0.01 -0.01 -0.04 Singapore 1.05 15 75% 80% -0.02 -0.14 -0.24 -0.16 -0.09 -0.06 -0.04 -0.03 -0.02 -0.02 -0.02 -0.02 -0.07 Other EM 15.84 60 41% 80% -0.02 -0.15 -0.59 -1.74 -1.68 -0.95 -0.58 -0.40 -0.29 -0.22 -0.18 -0.16 -0.58 DM 43.15 50 92% 70% 0.00 -0.67 -4.64 -11.90 -8.86 -4.37 -2.95 -2.07 -1.56 -1.26 -1.06 -0.95 -3.36 EM 48.87 44 66% 85% -0.50 -3.78 -3.88 -6.20 -5.33 -2.96 -1.82 -1.22 -0.89 -0.68 -0.55 -0.48 -2.36 Total 92.02 47 78% 78% -0.50 -4.46 -8.51 -18.10 -14.19 -7.33 -4.78 -3.29 -2.44 -1.93 -1.60 -1.43 -5.71 Jet 7.00 na na na -0.01 -0.59 -2.20 -3.50 -2.00 -1.50 -1.00 -0.80 -0.70 -0.60 -0.50 -0.45 -1.15 Grand Total 99.02 na na na -0.51 -5.05 -10.71 -21.60 -16.19 -8.83 -5.78 -4.09 -3.14 -2.53 -2.10 -1.88 -6.87 Published 25-Mar -0.51 -4.89 -10.40 -18.68 -11.84 -5.90 -3.61 -2.35 -1.67 -1.19 -0.86 -0.67 -5.21 Diff 0.00 -0.16 -0.31 -2.92 -4.35 -2.93 -2.17 -1.73 -1.48 -1.34 -1.25 -1.21 -1.65 Lag is days lagged impact vs China; Intensity is %peak impact vs China; Recovery is %decay rate from peak impact as proportion of China Source: Goldman Sachs Global Investment Research, IEA 8 April 2020 9 Goldman Sachs Oil For the exclusive use of RWANI@HARTREEPARTNERS.COM b789af4ce4cb4dd6845eb024ed5bfb41 reflect our expectations that a 1 mb/d of required shut-ins are currently being implemented. Having laid out our demand path, a maximum inventory fill rate and our March 17 n supply forecast (which reflected neither localized shut-in nor voluntary cuts), we can finally estimate the potential impact of voluntary production cuts. Our balance shows that the 10 mb/d cut that we assumed (worth an effective 6.5 mb/d) cut, would still leave 4.0 mb/d of supply in excess of available storage capacity in both April and May. This would point to prices having to trade back below cash costs as local storage saturates, with WTI likely having to retrace below $20/bbl. For example, Midland WTI prices at $25/bbl for May are not at levels where we would expect large shut-in forced by inventory saturation. Following last week’s rally this points to a sharp decline in prices in coming weeks, with likely extreme price volatility. Exhibit 12: We track the intensity of isolation measures globally and map this into our oil demand forecasts Y-axis = Covid-19 ‘intensity’ parameter (peak oil demand impact as % of what occurred in China), X-axis = Shutdown intensity index calculated from University of Oxford’s Blavatnik School of Government’s Government Response Tracker. Bubbles size correlates with size of country oil demand. Exhibit 13: Global capacity to absorb the surplus will shrink as local constraints are sequentially met` Global inventory absorption capacity (mb/d) by month 0% 20% 40% 60% 80% 100% 120% 140% 160% 180% 0 20 40 60 80 100 120 0 2 4 6 8 10 12 14 16 18 20 Apr May Jun Other EM S. Africa India Brazil China Other DM Germany France Canada Netherlands Korea Japan US Source: Goldman Sachs Global Investment Research, University of Oxford BSG Source: Goldman Sachs Global Investment Research, IEA, EIA, Kpler 8 April 2020 10 Goldman Sachs Oil For the exclusive use of RWANI@HARTREEPARTNERS.COM b789af4ce4cb4dd6845eb024ed5bfb41 Exhibit 14: We base-case a shallower US production decline than the EIA in order to investigate the level of global shut-ins required US crude production changes vs Mar-20 (kb/d) Exhibit 15: Even assuming a 10 mb/d headline cut, we estimate that more than 4 mb/d will be necessary in both April and May to prevent breaching storage capacity GS summarized oil balance; previous forecasts from March and revisions (kb/d) -2,000 -1,800 -1,600 -1,400 -1,200 -1,000 -800 -600 -400 -200 0 200 Mar-20 Apr-20 May-20 Jun-20 Jul-20 Aug-20 Sep-20 Oct-20 Nov-20 Dec-20 STEO GS 1Q20E 2Q20E 3Q20E 4Q20E Supply (17-Mar-20) 100,784 102,612 102,086 101,578 OPEC/Russia -488 -3667 -5042 -2857 Core OPEC + Russia 96 -2745 -3660 -1959 Iraq -78 -362 -553 -498 Libya -38 -403 -557 -167 Venezuela 7 100 80 83 Rest of OPEC -475 -257 -351 -316 Rest of World 148 -1954 -2032 -1158 US shale (black oil) -59 -692 -706 -305 Other US (GoM, AK., NGLs, Conv.) 38 -308 -290 -152 Canada 0 -615 -680 -589 Brazil 176 -233 -310 -208 North Sea 37 19 107 79 Kazakhstan 33 0 31 45 Elsewhere -76 -126 -184 -28 Net Impact -340 -5621 -7073 -4014 Supply (08-Apr-20,pre-shut-ins) 100,444 96,990 95,013 97,563 Shutins 0 -3572 -2000 -1000 Supply (08-Apr-20,post-shut-ins) 100,444 93,419 93,013 96,563 Demand (17-Mar-20) 96,920 96,930 101,135 102,486 China -429 -237 -294 -307 OECD -483 -7260 -1894 -1161 EM ex China -491 -4264 -1080 -619 Net impact -1403 -11760 -3268 -2087 Demand (08-Apr-20) 95,517 85,170 97,867 100,400 Imbalance old 3,863 5,682 951 -909 Net impact 1,063 6,139 -3,805 -1,928 Imbalance (post-shut-ins) 4,927 11,821 -2,854 -2,837 OECD government 0 530 330 0 China 847 316 -103 7 EM exc. China 1,088 1,840 -1,315 -1,135 Other EM/Floating 258 761 -520 355 EM total 2,193 2,916 -1,938 -773 OECD industry stocks 2,734 4,802 -3,246 -3,063 OECD Stock change (mb, vs end Dec-19) 251 688 389 107 Global Stock change (mb, vs end Dec-19) 450 1200 754 401 Source: Goldman Sachs Global Investment Research, EIA Source: Goldman Sachs Global Investment Research, IEA 8 April 2020 11 Goldman Sachs Oil For the exclusive use of RWANI@HARTREEPARTNERS.COM b789af4ce4cb4dd6845eb024ed5bfb41 GS Global Oil Supply-Demand Balance Exhibit 16: GS Oil Supply Forecasts (before production shut-ins) kb/d 1Q2019 2Q2019 3Q2019 4Q2019 Jan-20 Feb-20 Mar-20 Apr-20 May-20 Jun-20 Jul-20 Aug-20 Sep-20 Oct-20 Nov-20 Dec-20 2019 2020E yoy 20E Lower 48 crude 9,470 9,700 9,987 10,350 10,320 10,310 10,190 9,850 9,557 9,438 9,306 9,260 9,337 9,341 9,246 9,320 9,877 9,623 -254 Gulf of Mexico crude 1,847 1,927 1,817 1,943 1,910 1,890 2,020 2,030 2,020 1,990 1,980 1,920 1,900 1,810 1,980 2,030 1,883 1,957 73 Alaska crude 490 467 427 477 510 510 510 510 510 510 510 510 510 510 510 510 465 510 45 US crude 11,807 12,093 12,230 12,770 12,740 12,710 12,720 12,390 12,087 11,938 11,796 11,690 11,747 11,661 11,736 11,860 12,225 12,090 -135 US NGL 4663 4807 4800 4987 5150 5090 5130 4997 4875 4815 4757 4715 4738 4703 4733 4783 4814 4874 60 Lower 48 other 170 180 186.66667 166.66667 140 150 150 170 170 170 150 150 160 160 160 160 175.83333 157.5 -18 US ethanol 1,013 1,047 1,017 1,040 1,060 1,050 1,030 773 906 984 979 1,027 951 990 1,040 1,030 1,029 985 -44 Total US 17,653 18,127 18,233 18,963 19,090 19,000 19,030 18,329 18,038 17,907 17,681 17,582 17,595 17,514 17,668 17,833 18,244 18,106 -138 Canada 5,490 5,517 5,515 5,667 5,766 5,784 5,670 4,809 4,770 4,731 4,731 4,731 4,831 4,931 4,931 4,931 5,547 5,052 -496 Mexico 1,914 1,911 1,937 1,952 1,987 1,970 1,976 1,952 1,916 1,854 1,854 1,854 1,854 1,854 1,854 1,854 1,929 1,898 -30 Total North America 7,404 7,428 7,452 7,619 7,753 7,754 7,646 6,761 6,686 6,586 6,586 6,586 6,686 6,786 6,786 6,786 7,476 6,950 -526 Argentina 594 596 606 607 605 605 606 601 588 569 569 569 569 569 569 565 601 582 -18 Brazil 2,664 2,737 3,009 3,170 3,288 3,059 3,030 2,818 2,680 2,713 2,747 2,780 2,829 2,838 2,855 2,849 2,895 2,874 -21 Colombia 899 900 884 889 891 888 885 874 857 826 826 826 820 810 799 788 893 841 -52 Guyana 0 0 0 3 50 80 90 100 110 120 120 120 120 120 120 120 1 106 105 Other Latam 355 349 354 356 352 354 352 351 351 351 349 347 347 347 346 346 353 349 -4 Non-OPEC LatAm 4,512 4,582 4,853 5,025 5,186 4,986 4,963 4,745 4,586 4,580 4,611 4,643 4,686 4,684 4,689 4,668 4,743 4,752 9 Norway 1,774 1,568 1,651 1,950 1,963 2,066 2,068 2,010 1,989 1,902 1,902 1,902 1,952 2,002 1,993 1,976 1,736 1,977 241 UK 1,213 1,150 1,086 1,134 1,167 1,205 1,256 1,278 1,288 1,024 1,252 1,142 1,094 1,232 1,231 1,221 1,146 1,199 53 Other Europe 614 589 569 553 543 537 534 534 532 533 531 530 531 531 529 528 582 533 -49 Total Europe 3,601 3,308 3,306 3,638 3,673 3,808 3,858 3,821 3,809 3,460 3,685 3,574 3,578 3,765 3,753 3,725 3,463 3,709 246 Azerbaijan 800 743 762 756 770 760 767 766 739 712 712 712 732 737 737 737 765 740 -25 Kazakhstan 1,993 1,821 1,928 1,990 2,011 2,036 1,983 1,938 1,907 1,877 1,877 1,877 1,897 1,917 1,927 1,937 1,933 1,932 -1 Russia 11,668 11,501 11,573 11,590 11,653 11,623 11,577 11,585 11,102 10,618 10,618 10,618 10,918 11,068 11,268 11,393 11,583 11,170 -413 Other FSU 350 353 353 356 356 357 357 355 355 355 353 351 350 348 346 347 353 353 0 Total FSU 14,811 14,419 14,615 14,692 14,790 14,776 14,684 14,644 14,103 13,562 13,560 13,557 13,896 14,070 14,278 14,414 14,634 14,195 -440 China 3,878 3,907 3,877 3,864 3,859 3,861 3,866 3,859 3,822 3,919 3,849 3,803 3,796 3,808 3,801 3,798 3,881 3,837 -45 India 825 806 789 782 776 772 768 766 761 757 753 750 746 742 738 734 800 755 -45 Indonesia 791 747 762 753 753 752 752 752 752 752 752 752 752 752 752 752 763 752 -11 Malaysia 708 690 603 679 700 700 700 700 700 700 700 700 700 700 700 700 670 700 30 Australia 369 407 439 459 477 466 483 481 489 493 498 501 515 518 520 522 418 497 79 Vietnam 235 238 219 213 208 212 208 207 207 205 204 202 202 200 199 198 226 204 -22 Rest of Asia-Pacific 766 762 747 761 754 750 743 744 736 728 725 721 717 715 712 708 759 729 -30 Non-OPEC Asia 7,573 7,558 7,436 7,510 7,527 7,513 7,520 7,510 7,467 7,555 7,481 7,429 7,427 7,435 7,422 7,412 7,519 7,475 -44 Oman 978 978 979 979 982 983 982 986 951 914 914 914 914 914 914 914 979 940 -39 Qatar 1,979 1,956 1,956 1,955 1,975 1,974 1,974 1,973 1,973 1,973 1,972 1,972 1,971 1,971 1,971 1,970 1,962 1,972 11 Other Middle East 291 304 305 307 306 305 304 303 296 289 288 288 287 287 286 285 302 294 -8 Non-OPEC Middle East 3,248 3,238 3,240 3,242 3,263 3,262 3,260 3,262 3,220 3,176 3,174 3,174 3,172 3,172 3,171 3,169 3,242 3,206 -36 Egypt 642 637 630 620 615 615 611 609 592 573 573 573 573 573 573 573 632 588 -45 Ghana 198 198 200 201 196 194 192 189 187 186 183 181 180 178 176 175 199 185 -15 Other Africa 637 651 625 646 628 626 626 624 616 609 609 606 606 606 605 604 640 613 -27 Non-OPEC Africa 1,477 1,487 1,456 1,467 1,439 1,435 1,429 1,422 1,395 1,367 1,364 1,359 1,358 1,356 1,353 1,351 1,472 1,386 -86 Processing gains 2,349 2,349 2,349 2,349 2,365 2,357 2,383 2,320 2,226 2,207 2,210 2,166 2,230 2,273 2,296 2,265 2,349 2,275 -74 Biofuels exc. US ethanol 1,415 2,088 2,391 1,863 1,458 1,464 1,499 2,118 2,325 2,394 2,462 2,458 2,444 2,334 2,029 1,721 1,939 2,059 120 Total non-OPEC supply 64,044 64,583 65,330 66,369 66,544 66,355 66,272 64,931 63,853 62,793 62,815 62,528 63,072 63,388 63,445 63,344 65,081 64,112 -970 Non-OPEC ex. US Lower 48 & NGL 49,910 50,076 50,543 51,032 51,074 50,955 50,952 50,084 49,422 48,540 48,752 48,553 48,997 49,344 49,466 49,240 50,391 49,615 -776 Ecuador 527 530 547 513 530 535 535 522 515 498 498 495 488 481 475 468 529 503 -26 Venezuela 1,107 863 753 777 840 760 760 700 650 600 560 530 500 475 450 450 875 606 -269 Algeria 1,027 1,020 1,023 1,023 1,010 1,007 1,010 985 969 939 939 934 921 909 896 884 1,023 950 -73 Congo 340 347 337 307 300 308 310 300 291 283 283 283 283 279 275 271 333 289 -44 Gabon 213 220 207 210 170 191 205 200 191 182 182 182 182 182 182 179 213 186 -27 Angola 1,430 1,430 1,347 1,350 1,380 1,390 1,360 1,326 1,309 1,279 1,275 1,258 1,241 1,224 1,207 1,190 1,389 1,286 -103 Nigeria 1,690 1,723 1,813 1,697 1,680 1,709 1,700 1,735 1,713 1,652 1,652 1,646 1,674 1,702 1,704 1,707 1,731 1,690 -41 Equatorial Guinea 107 110 110 113 120 123 124 121 119 115 115 115 113 112 110 108 110 116 6 Libya 953 1,147 1,090 1,153 780 146 110 80 80 80 80 400 700 850 950 1,000 1,086 438 -648 Iran 2,740 2,407 2,187 2,113 2,060 2,049 2,067 2,050 2,050 2,050 2,050 2,050 2,050 2,050 2,050 2,050 2,362 2,052 -310 Iraq 4,697 4,730 4,787 4,633 4,540 4,480 4,472 4,417 4,314 4,209 4,209 4,209 4,309 4,359 4,384 4,434 4,712 4,361 -351 Kuwait 2,713 2,680 2,650 2,683 2,670 2,658 2,800 3,050 2,650 2,650 2,650 2,650 2,750 2,850 2,900 2,950 2,682 2,769 87 Saudi Arabia 10,060 9,760 9,487 9,910 9,720 9,683 10,300 11,200 9,000 9,000 9,000 9,000 9,500 9,800 10,000 10,200 9,804 9,700 -104 UAE 3,067 3,133 3,167 3,340 3,090 3,040 3,300 3,600 2,730 2,730 2,730 2,730 2,930 3,080 3,180 3,230 3,177 3,031 -146 Total OPEC Crude 30,670 30,100 29,503 29,823 28,890 28,079 29,053 30,285 26,583 26,267 26,223 26,481 27,641 28,352 28,763 29,121 30,024 27,978 -2,046 Total OPEC NGL 5,409 5,419 5,355 5,407 5,401 5,368 5,371 5,403 5,428 5,428 5,426 5,426 5,426 5,426 5,426 5,426 5,397 5,413 16 Total OPEC supply 36,079 35,519 34,858 35,230 34,291 33,447 34,424 35,688 32,011 31,695 31,649 31,907 33,067 33,778 34,189 34,547 35,422 33,391 -2,031 World supply before shut-ins 100,123 100,101 100,188 101,599 100,835 99,802 100,696 100,619 95,864 94,488 94,463 94,435 96,139 97,166 97,634 97,890 100,503 97,503 -3,000 Required shutins 0 0 0 0 0 0 0 -4015 -4096 -2604 -2000 -2000 -2000 -1000 -1000 -1000 0 -1,643 World supply after shut-ins 100,123 100,101 100,188 101,599 100,835 99,802 100,696 96,605 91,768 91,884 92,463 92,435 94,139 96,166 96,634 96,890 100,503 95,860 Non-OPEC Supply OPEC Supply Source: Goldman Sachs Global Investment Research, IEA, EIA Exhibit 17: GS Oil Demand Forecasts kb/d 1Q2019 2Q2019 3Q2019 4Q2019 Jan-20 Feb-20 Mar-20 Apr-20 May-20 Jun-20 Jul-20 Aug-20 Sep-20 Oct-20 Nov-20 Dec-20 2019E 2020E yoy 20E USA 20,649 20,652 20,988 21,116 20,851 20,427 18,459 13,745 15,779 18,727 19,782 20,183 20,393 20,516 20,590 20,626 20,851 19,173 -1,678 Canada 2,432 2,403 2,580 2,528 2,437 2,413 2,286 2,072 2,145 2,254 2,496 2,527 2,547 2,517 2,526 2,532 2,486 2,396 -90 Mexico 1,923 1,930 1,909 1,863 1,936 1,921 1,834 1,696 1,753 1,828 1,846 1,868 1,882 1,866 1,872 1,875 1,906 1,848 -58 North America 25,003 24,986 25,476 25,507 25,224 24,761 22,579 17,513 19,677 22,809 24,124 24,578 24,822 24,899 24,989 25,034 25,243 23,417 -1,826 Brazil 3,024 3,075 3,182 3,256 3,093 3,084 2,970 2,495 2,595 2,864 3,073 3,129 3,160 3,257 3,270 3,277 3,134 3,022 -112 Chile 371 371 371 357 359 357 347 321 330 345 349 353 356 351 352 353 368 348 -20 LatAm ex. Mexico, Brazil, Chile 3,338 3,360 3,382 3,234 3,281 3,254 3,114 2,903 2,946 3,098 3,220 3,262 3,287 3,196 3,206 3,212 3,328 3,165 -163 LatAm ex. Mexico 6,733 6,806 6,935 6,847 6,733 6,695 6,431 5,719 5,872 6,307 6,643 6,744 6,802 6,804 6,828 6,842 6,830 6,535 -295 OECD Europe 13,831 13,974 14,637 14,218 13,769 13,440 11,636 9,245 10,449 11,769 13,025 13,462 13,711 13,495 13,634 13,707 14,165 12,612 -1,553 Non-OECD Europe 750 777 783 788 768 764 741 689 710 752 772 781 786 794 796 798 775 763 -12 Total Europe 14,582 14,752 15,421 15,006 14,537 14,204 12,376 9,934 11,159 12,521 13,797 14,243 14,497 14,289 14,430 14,505 14,940 13,374 -1,565 Japan 4,086 3,407 3,420 3,828 4,016 3,908 3,442 2,697 2,951 3,068 3,166 3,210 3,241 3,705 3,714 3,716 3,685 3,403 -282 South Korea 2,627 2,479 2,582 2,647 2,659 2,594 2,310 2,144 2,293 2,375 2,470 2,499 2,520 2,580 2,587 2,591 2,584 2,469 -115 Australia & New Zealand 1,363 1,350 1,367 1,381 1,358 1,342 1,278 1