New 2020 sulphur regulations for global shipping

Transcription

1 Macro research IMO2020 Report Wednesday 14 March 2018 New 2020 sulphur regulations for global shipping Gasoil to HSFO price spread to rise above $450/bl IMO s 2020 sulphur deadline a done deal It is now impossible for the IMO to push the 2020 global 0.5% sulphur limit to a later date without breaking its own rules. From January 2020 it will be illegal to run a ship using fuel containing more than 0.5% sulphur without operating an exhaust cleaning gas system. Fewer than 2000 ships will have a scrubber in 2020 Currently, some 18,000 bulk carriers, crude oil carriers and container ships account for 76% of the world s DWT capacity. Together, they likely consume close to 4 m bl/d HSFO. By 2020, we expect less than 2000 ships to possess a scrubber. We therefore expect that present HSFO demand will fall sharply from 4 m bl/d to as little as m bl/d. Demand will instead switch to higher quality oil products such as ultra-low sulphur fuel oil (ULSFO 0.5%) or Gasoil. Gasoil to HSFO spread likely to widen to over $450/ton As demand for HSFO almost evaporates in 2020 and instead shifts to ULSFO 0.5% or Gasoil, we forecast global refinery upgrading capacity utilization will be stretched to its maximum. We therefore expect the Gasoil to HSFO 2020 price spread to widen to more than $450/bl. Bjarne Schieldrop Chief analyst commodities (47) Bjarne.schieldrop@seb.no Historical Gasoil 0.1% to HSFO 3.5% product price spread - USD/ton Source: SEB, Bloomberg

2 Macro research : IMO2020 Report Wednesday 14 March Contents New 2020 sulphur regulations for global shipping... 1 Gasoil to HSFO price spread to rise above $450/bl... 1 Executive summary... 3 The IMO 2020 Sulphur deadline is a done deal Likely less than 2000 ships with scrubbers in HSFO to Gasoil spread to rise above $450/ton... 3 Paper scrubber, competitive hedge or plain bet... 3 Foreword... 4 Meeting with clients in Europe, Singapore and Hong Kong... 4 Background... 5 The graveyard of dirty oil... 5 IMO ending high sulphur emissions from global freight... 5 Scrubbers? No thank you! We want fuel compliance!... 6 Competitive concerns - What do others do?... 6 Risk-reward deemed far from good enough... 6 IMO will they or won t they?... 7 The IMO s sulphur regulation is very different from the BWT regulation. 7 Focus shifts from IF to HOW... 7 IMO wants change, not havoc and disruption... 7 Who should pay? Shipowners, refineries or charterers?... 8 Who should pay? Shipowners, refineries or charterers?... 8 Catch Investing in a scrubber... 9 The need for switching fuel m bl/d of out of today s 4 m bl/d of HSFO to disappear to What can refineries do to 2020? Surplus of bitumen, asphalt and sulphur Historical prices & spreads for HSFO 3.5%, Gasoil 0.1% and Brent. 13 Historical prices and spreads for ULSFO 0.5% and Gasoil 0.1% The HSFO to Gasoil Coker link Compliant fuel to be Gasoil 0.1%, or ULSFO 0.5%? Recent price developments Down the scrubber route unless they are forbidden Summing up The road ahead... 28

3 Macro research : IMO2020 Report Wednesday 14 March The IMO 2020 deadline is a done deal. It may be softened peripherally by transitional measures but strong price signals are needed and wanted by the IMO to facilitate change. Global shipowners do not want to install scrubbers. They wish to be fuel compliant. Demand for HSFO is likely to drop from 4 m bl/d to as little as m bl/d in 2020 We expect global refinery upgrading capacity utilization to be pushed to its limit in 2020 and that the Gasoil to HSFO price spread will in response widen to more than $450/ton Executive summary The IMO 2020 Sulphur deadline is a done deal. Legally it is now impossible for the IMO to change its decision before January Nor, since its decision in October 2016, have there been any requests from IMO member states for any alteration. Strong price signals in the form of a wider ULSFO 0.5% to HSFO or Gasoil product price spread are needed to drive the desired shift to lower sulphur emissions. The IMO will likely favour a wider spread and support such an outcome. They will however also probably ease the impact of worst possible scenarios, by softening transitional measures to avoid unnecessary transitional havoc and disruption. Likely less than 2000 ships with scrubbers in We expect that fewer than 2000 ships will have a scrubber in Consequently, demand for HSFO will decrease dramatically in 2020, likely to less than 0.8 m bl/d and probably as low as m bl/d, representing barely 10% of today s marine HSFO consumption of 4 m bl/d. Demand will instead switch to compliant fuel with less than 0.5% sulphur content, i.e. either ULSFO 0.5% or Gasoil 0.5%. This will push the global refining system s upgrading capacity to the limit as m bl/d of HSFO suddenly need to be upgraded to ULSFO/Gasoil. Ripple effects of this development will likely be felt across the whole oil product sector and further impact pricing of different crude slates. There is a real risk of a repetition of events in 2008 when strong Gasoil demand and limited and fully utilized refinery upgrading capacity led to an upward spiralling price dynamic between higher and higher Gasoil prices and higher and higher light sweet crude oil prices. HSFO was then also priced at a deep discount to crude HSFO to Gasoil spread to rise above $450/ton The 2020 HSFO to Gasoil price spread traded at only $220/ton on a forward basis in the early autumn of Since then it has blown out to as much as $350/bl at the start of 2018 while currently trading at $320/ton. The spread is typically and historically proportional to the Brent crude oil price. It averaged $308/ton between 2011 and 2014 when Brent crude averaged $108/bl, and $447/ton in 2008 when Brent crude averaged $98/bl. Compared to the current forward 2020 Brent crude oil price of $57/bl, the 2020 HSFO to Gasoil price spread is today trading expensively certainly from an historical perspective, but is still more then $100/ton below an historical spread extreme of $420/ton observed at a crude oil price of around $57/bl. In 2008 the spot spread blew out to an extreme of $700/bl for a short period. We expect worldwide refinery upgrading capacity utilization to be pushed to its limit in 2020 causing the Gasoil to HSFO price spread to widen to over $450/ton. Paper scrubber, competitive hedge or plain bet Our main view is that the HSFO to Gasoil price spread is likely to blow out to more than $450/ton. Buying the forward 2020 Gasoil to HSFO product spread can be utilized as a virtual scrubber for shipping fleets without scrubbers. This will provide such shipowners with an extra source of income proportional to the incremental income advantage (due to a higher spread) for scrubber ships and as such offer a competitive hedge for nonscrubber ships compared to scrubber ships. A long Gasoil vs. short HSFO position for 2020 delivery may also be taken as a plain bet for a wider spread under the new regulatory regime.

4 Macro research : IMO2020 Report Wednesday 14 March Foreword This report reflects our involvement in the IMO 2020 sulphur emission issue following meetings with more than 100 shipping clients and refineries over the past 12 months. It summarises what we have learned over the past year and our views and reflections on the situation. It is not an in-depth technical study trying to pin-point global refinery capacities vs. product needs in A little more than a year ago the IMO decided that from January 2020 it will not be allowed to run ships globally with HSFO fuel unless an exhaust scrubber has been installed. Otherwise, a ship will have to use a fuel with maximum 0.5% sulphur content. Today s HSFO typically has a 3.5% standard product specification. The actual volume weighted sulphur content in the 4 m bl/d HSFO consumed for marine propulsion today does however contain closer to 2.5% sulphur on average. Two detailed studies have been made concerning the projected balance between marine fuel demand in 2020 under the new rules vs. global refinery product capacities. The IMO s Delft study concluded that there would be sufficient compliant fuel available in 2020 to satisfy the new regulations. As a result, they decided in October 2016 that the new regulations would take effect in January The Ensys Energy study however came to a completely opposite conclusion. It determined that the world s refinery system would be put under extreme stress to such an extent that it would send shock waves through all product markets as well as crude oil prices and grades. Meeting with clients in Europe, Singapore and Hong Kong To understand more clearly how the IMO s decision will impact fuel markets and the shipping industry in 2020, we decided to meet with market participants to hear what they thought and were planning to do. Over the past year, we have travelled throughout both northern and southern Europe, to Singapore and Hong Kong, to meet more than 100 shipping companies, refineries and technical experts, to learn how shipowners, charterers and the refining industry are preparing for the IMO 2020 deadline. This report summarises what we have learned over the past year and our present reflections and conclusions. We are grateful to all those clients who have taken time to meet with us, who have helped us understand the issues at stake. Our increasing expertise enabled us to contribute more to these discussions as we progressed.

5 Macro research : IMO2020 Report Wednesday 14 March Background The graveyard of dirty oil The global shipping fleet has been running on heavy fuel oil with high sulphur content since the early 1970 s. Today this fuel is typically labelled HSFO or HSFO 3.5% (3.5% sulphur). It is the bottom of the oil barrel, the residue of simple refining. It is the dirtiest and heaviest product produced by refineries worldwide. Basically, it is a waste product with little use outside the shipping market except (and less and less in recent years) in power production. For this reason, the world s shipping fleet has been dubbed the graveyard of dirty oil. Generally, it is the old oil refineries in Europe, Africa and Latin America who produce HSFO. These are simple refineries with mostly straight through processing. Therefore, crude oil used is basically just split into its different hydrocarbon components through simple distillation and/or vacuum distillation. There is very little modification or aftertreatment of the molecules in the crude oil used. Conversely, the world s most modern refineries today are highly complex, using many after treatment stages that modify molecules at high temperatures and pressure. Consequently, they produce very little heavy residuum [ residue ]. Instead, long, heavy molecules are cracked apart and converted to much more valuable middle distillate products. For the world s old refineries, the residuum [ residue?] (the heaviest products from simple refining) is basically a waste product which is sold at a discount to Brent crude and consumed by the global shipping fleet. This symbiosis, which has lasted since the early 1970 s, is now set to change or end. Unless the global shipping fleet installs scrubbers, they will no longer be able to handle waste products from old refineries after January 2020, at least not in its current form. New refineries are in general built with after treatment modules so that they hardly produce any heavy residue. It is mostly all cracked away and instead converted to higher value middle distillate products. Thus as today s old refineries IMO ending high sulphur emissions from global freight In October 2016, the IMO decided that from January 1, 2020 ships will be forbidden to use fuel with more than 0.5% sulphur worldwide unless they are equipped with a scrubber that cleans out the sulphur in the exhaust. However, the IMO did not announce any other fuel guidelines, so generally fuel can be very dirty and of low quality provided it contains no more than 0.5% sulphur. The decision was long overdue and had been in the pipeline for since The end of HSFO (without scrubber) in global shipping had been decided a long time ago. As of October 2016, the only thing left to decide was whether the deadline would be January 2020 or It was all down to the fuel availability study, the Delft study which the IMO had ordered and had received in the autumn of This study concluded that to the best knowledge of various experts there would be sufficient fuel in required qualities to implement the new regulations from January So, armed with the Delft report, the IMO finally decided that the new regulations would be enforced from January The Delft study has however been strongly contested by many who argue that the transition will not be easy at all. A comparable in-depth report prepared by Ensys Energy concludes in total opposition to the Delft study that an abrupt and complete switchover from one day to the next will place the global refining industry under extreme duress with consequences rippling throughout both oil products and crude markets. Further that it will result in a large expansion in the Gasoil to HSFO product price spread as well as in the price spread between high quality crude slates versus lower quality crude slates. The Ensys study is not the only study or report arguing that this will be problematic. Indeed, most reports we have studied agree with its authors to a greater or lesser extent. We also concur much more with the Ensys study than its Delft counterpart. As such, we expect a large and significant impact on global fuel and crude markets in 2020.

6 Macro research : IMO2020 Report Wednesday 14 March Scrubbers? No thank you! We want fuel compliance! One thing which became very clear to us as we travelled around the world was that hardly any shipowners we met really wanted or planned to install scrubbers on their ships. The list of why nots was almost endless. The almost unanimous response was: We want to be fuel compliant. We don t want to install scrubbers on our ships. Shipowners do not want scrubbers. They want to be fuel compliant. Their greatest concern is sunk cost capex on scrubbers which limits their ability to pass on such expenses to charterers. Following several tough trading years, many shipowners have little cash to spend on scrubbers. No one needs to act so long as no one acts. A level playing field remains intact as most shipowners do not install scrubbers. Charterers will bear most of the fuel costs. Shipowners look at each other. What matters is what other shipowners do. Whether fuel costs are high or low is less important. The idea of installing scrubbers on board every ship in the world very clearly did not come from the shipowners. They for sure don t want scrubbers. They want to be fuel compliant. The key reason for this view is that generally it is the charterer who indirectly pays for the fuel as a pass through cost. It is much easier to pass on the specific fuel cost to charterers than to demand compensation for capex spending on scrubbers. For shipowners a scrubber means capital expenditure, less free space on a ship, more maintenance, greater crew competence, higher fuel consumption and uncertain sludge disposal costs. It also means new legal obligations for exhaust pipe sulphur emissions. If shipowners run fuel compliant vessels without a scrubber, the fuel supplier is legally obliged to ensure proper fuel quality and sulphur content. If a ship on the other hand runs with a scrubber it is instead the shipowner who is legally responsible for the sulphur content in the exhaust gases and thus that the scrubber works properly the ship-owner may be fined. In other words, almost unanimously shipowners do not wish to use scrubbers. Competitive concerns - What do others do? Also clear from our many meetings was that most shipowners were not very concerned about higher fuel costs during the 2020 transition. What they really bothered about was what other shipowners were doing. Fuel cost is mostly passed on to charterers, so long as everyone else is doing the same thing within the shipowner community they will all be in the same boat - so to speak. Internal competition within the shipping market will be on a level playing field with everyone just as well or badly off. What shipowners were and still are concerned about is that a significant share of the global fleet is embarking on installing scrubbers, placing others at a considerable competitive disadvantage. Risk-reward deemed far from good enough Generally, shipowners we met thought the risk-reward for installing a scrubber was well short of justifying capex on them. Such devices are still regarded as technically immature, still at the test stage. Shipowners expect scrubber costs will continue to decrease over time. They were not confident about the firmness of IMO s 2020 deadline. They feared that they would not be able to pass the capex over to the charterers, that the Gasoil to HSFO price spread would not be wide enough to make the scrubber installation profitable, that it could become very expensive to dispose of sludge from the scrubber and that new environmental regulations governing CO2, PM and NOx in a few years would mean that a scrubber installation today would be useless in a few years. All in all that it was better to wait and see has been and probably still is the general view within shipping.

7 Macro research : IMO2020 Report Wednesday 14 March IMO s 2020 sulphur regulation deadline appears immoveable unless someone can prove that the market will run into insufficient fuel availability in IMO will they or won t they? A year ago, there was huge uncertainty whether the 2020 deadline would be pushed forward or not. The shipping industry s experience with the ballast water treatment system regulations made it easy to assume that there would be endless delays, and that sulphur regulation deadlines would also be repeatedly postponed. At that stage most companies we met felt uncertain about the 2020 deadline. Clearly, this view changed in the summer of Going forward, the discussion switched from if to how. A key factor for the IMO concerning the 2020 deadline decision involved the fuel availability study. How the 2020 deadline would impact costs and fuel prices per se was not at the heart of what was decided. Instead, the most important question was whether physically there would be sufficient fuel available to enable the global shipping fleet to operate from If it was to run solely on expensive Gasoil then that was probably fine in the eyes of the IMO provided there was physically enough Gasoil for the fleet to run on. So, if it was shown there would not be physically enough compliant fuels in 2020 on which to run the global shipping fleet then the IMO might possibly have shifted the deadline to This is however very difficult to prove and as such the 2020 deadline is today fixed. Legally, it would take a minimum of 22 months to change the 2020 deadline. It is therefore too late to alter the deadline now consistent with IMO rules. So far, the IMO has not yet received proposals from any of its members to change the deadline. By IMO s own laws there is no longer any way that the January 2020 deadline can be pushed back in time. To change that date will require an amendment to the MARPOL Annex VI treaty. First, this would involve a proposal for such an amendment. Next, the proposal would have to circulate for six months before adoption. Afterwards, it could not come into force before at the earliest 16 months after adoption, i.e. the amendment could not take effect until 22 months had passed. No such proposal by the IMO s members has been received since the 2020 deadline decision was announced in October The IMO s sulphur regulation is very different from the BWT regulation A key difference between the IMO s sulphur regulation and the ballast water treatment system regulations is that the sulphur regulations do not require any capex installations onboard a ship. The IMO does not state that a scrubber is needed but that it can be used if desired; otherwise compliant fuel should be utilized. The explicit need to install capex on ships to comply with the ballast water treatment regulations was one of several key reasons for extensions. Unless lack of fuel availability can be proved such excuses are hard to use in the case of sulphur regulations. As far as we can see, the 2020 sulphur regulation deadline is near rock solid absent proof of lack of fuel availability. Focus shifts from IF to HOW As the market began to accept fully the 2020 sulphur deadline, its focus shifted from IF to HOW. How is it going to play out? What will happen to fuel prices? What can refineries do and at what price? Will compliant ULSFO 0.5% be of acceptable quality and quantity and will it be available worldwide? Will it be possible to buy HSFO globally for those few ships that are running with scrubbers? Will there be a lot of cheating? How will the regulations be monitored and how will they be enforced? Will there be waiver options unless there are no available fuels to be bought etc. Many such questions remain unanswered today. IMO wants change, not havoc and disruption Effecting change 1) IMO s 2020 deadline decision 2) Forward 2020 pricing reaction 3) Spot price reaction in ) Capex spending Shipping/Refineries Strong product price signals in terms of a wide Gasoil to HSFO product spread needed in order for capex spending to be undertaken It is important to remember that the IMO s decision is a regulatory decision on emissions. The IMO wants change. It wants it so strongly that it may be willing to accept some disruption to get there. It has of course no desire for disruption if it can avoid it. As such there is always the risk that the IMO looks kindly at softening measures which help to avoid disruptions so long as they do not prevent changes. This could come in the form of different exceptions and waivers for 2020 if fuel is not available. The process of driving change is sequenced as follows: 1) IMO s 2020 decision, 2) Forward based price reaction 3) Fuel spot price reaction once we get to 2020 and lastly 4) Market participant reaction in terms of capex spending on either scrubbers or refinery upgrades. Neither shipping nor refineries are willing to make necessary investments before the market sends the sufficient price signals. Today, these are beginning to emerge on a forward basis.

8 Macro research : IMO2020 Report Wednesday 14 March If the market solution is a refinery and fuel compliant solution, then charterers will be handed the bill for higher costs for the new regulations directly or indirectly through increased fuel prices passed on to them. If scrubbers are the market solution then shipowners are unlikely to be able to recover all of their scrubber capex. Over the longer term however the charterers will of course have to pay sufficiently high freight rates to incentivise the building of new ships equipped with a scrubber. Shipowners and refineries both fear that the other part will invest so to render their own possible investment unprofitable. Who should pay? Shipowners, refineries or charterers? Who should pay? Shipowners, refineries or charterers? It has been clear for several years that eventually the day would come when HSFO could no longer be used without restriction in global seaborne transportation. Transport by sea will certainly become more expensive under the new regulations. The big question is - who should pay for it? Shipowners, refineries or charterers? Clearly, in the long term, shipowners will always have to pay higher environmental regulatory costs. If the market solution for the global shipping market under the new regulations is fuel compliance, then charterers will almost directly and immediately pay for the new regulations through higher fuel costs which normally are directly passed on to them. The higher fuel costs for ULSFO or Gasoil will in time become a payment to the refineries for key investments they need to convert HSFO into compliant fuels. If the future solution mainly involves the use of scrubbers, then the world s shipowners will need to pay for the scrubbers and both their maintenance and operating expenditure. In a retrofitting of today s shipping fleet, it is unlikely shipowners would be able to recover all capex on scrubbers. Over time however it is clear that charterers will implicitly have to accept a freight rate high enough to pay for the building of new ships including a scrubber. Otherwise, new ships will not be built, at least not sensible, economically viable vessels. There are however significant skews in the two solutions. It takes many years for a refinery to upgrade their physical equipment by adding cokers or hydro-crackers. Generally, it only takes a couple of weeks to retrofit a ship with a scrubber. Therefore, if the market sends positive investment signals to both refineries and shipowners, the latter have the capacity to react very quickly. Therefore, we think both action and reaction may be taken by shipowners rather than by refineries. Catch 22 When the IMO opened up to allowing both fuel compliance and scrubbers it created a stale-mate situation. The refinery sector is unlikely to undertake multi-billion-dollar investment decisions if it turns out that shipowners will end up installing scrubbers across the fleet and therefore consume all HSFO anyhow. Then, there would be no need for refineries to invest heavily to upgrade units. Amongst a myriad of issues, shipowners are reluctant to install scrubbers because there is a chance that the refinery sector may be able to produce a compliant ULSFO 0.5% product at only a small mark-up to the HSFO price making it unprofitable to install a scrubber. Consequently, both shipowners and refineries are afraid that the other party will render their investments unprofitable and are therefore reluctant to undertake necessary investments for the upcoming transition.

9 Macro research : IMO2020 Report Wednesday 14 March Investing in a scrubber The economics of a scrubber investment are straightforward: we assume an investment of USD 3.5m for the scrubber and 200 days of sea time per year. Payback in years for scrubber investment at various price spreads Probable long term average with ref Brent $50-80/bl Synthetic ULSFO to HSFO Historical average (ref Brent $50-80/bl) Synthetic ULSFO to HSFO 2020 Synthetic ULSFO to HSFO 2020 Gasoil to HSFO spread SEB 2020 forecast 2020 Gasoil to HSFO spread Type - Tonnes Fuel spreads Handysize 25 VLCC 50 VLCC 75 Container 100 Container 200 Container The 2008 intra year high Gasoil to HSFO spread Source: SEB Our assumptions are probably unrealistically simple as the scrubber investment cost probably scales with the size of the ship with fuel consumption ranging from tonnes per day. There are in addition a lot of other uncertain costs associated with a scrubber installation like: higher fuel consumption, sludge disposal costs, more maintenance, loss of space, Installing scrubbers on the world s shipping fleet. Many ships are not suited for a scrubber. The sum of crude tankers, Bulkers and Container ships is close to 18,000 and accounts for 76% of the world s DWT. So installing scrubbers on these 18,000 ships would cover much of today s consumption of HSFO. Possible feasible scrubber installations could however amount to close to 40,000 ships World Shipping Fleet Category Number of ships mdwt Average DWT per ship Estimated need for scrubber Number of ships suited for scrubber Crude tankers 2, % 2,017 Product tankers 8, % 5,882 Chemical tankers 3, % 1,843 Other tankers % 405 Bulkers 11, % 7,779 Combos % 12 LPG carriers 1, % 1,016 LNG carriers % 0 Containerships 5, % 4,131 Multi-purpose 3, % 955 General cargo 15, % 4,520 Ro-Ro 1, % 1,662 Car carriers % 782 Reefers 1, % 1,458

10 Million barrels per Day Macro research : IMO2020 Report Wednesday 14 March Category Number of ships mdwt Average DWT per ship Estimated need for scrubber Number of ships suited for scrubber Offshore AHTS 4, % 1,404 World cargo fleet 59, Others 34, % 5,187 World Fleet 94,171 1,925 20,438 41% 39,054 Source: Clarksons The need for switching fuel Global demand for HSFO in 2012 was estimated at 228 m ton/year. Since then the global shipping fleet has grown significantly. However, it has also moved into slow steaming mode meaning ships consume 20% less fuel than when they were fast steaming. Data on current global shipping fleet consumption of HSFO is hard to come by. This makes it very difficult to project total demand for marine fuels in Consequently, there is a very wide spread in projections for total marine fuel consumption and composition in Our best guestimate is that current marine consumption of HSFO largely the same as in 2012 since the increased fleet size is countered by its slow steaming. Demand for HSFO in 2020 will collapse to as low as 0.3 m bl/d vs. today s estimated demand of 4.0 m bl/d. Average forecasts project 2020 HSFO demand of 0.8 m bl/d. Demand will switch to ULSFO 0.5% and Gasoil. We believe few ships will run with scrubbers on January 1, The base case in the IMO s Delft Energy study was that by 2020 there would be 3,800 ships with scrubbers, and that these would consume 36 m tons of HSFO per year or 0.62 m bl/d. We expect there will be less than half that number, i.e. below 2,000 ships out of a global shipping fleet of 94,000 ships. As such it looks more like demand for HSFO could fall to as low as 0.3 m bl/d in If we consider the various forecasts by different companies such as BP and Shell as well as projections made by agencies and independent research, we can deduce the following: On average they expect HSFO demand of 0.8 m bl/d, Gasoil 0.5% demand of 1.4 m bl/d, and non-scrubber compliant ULSFO 0.5% demand of 2.8 m bl/d. There is however a very wide projection for total marine fuel consumption in On average the different forecasts projects a total marine fuel demand of 5 m bl/d. Fuel demand projections for 2020 (m bl/d) HFO LSFO/Blends Gas Oil IMO 2012 CE Delft BP Shell PIRA IEA 2020 RobinMeech Average Source: SEB, Delft, Ensys, BP, PIRA, IEA, Shell, Robin Meech 3.2 m bl/d of out of today s 4 m bl/d of HSFO to disappear to 2020 Today we estimate global marine consumption of HSFO at 228 m ton/year and marine gasoil consumption of 64 m ton/year. This is equal to the IMO s Delft study assumption for marine fuel demand in 2012 of 4 m bl/d of HSFO production and consumption and 1.3 m bl/d of MGO consumption. The average market forecast we have seen expects 3.2 m bl/d of today s 4.0 m bl/d HSFO demand to disappear overnight to 2020.

11 Million barrels per Day Macro research : IMO2020 Report Wednesday 14 March That is a massive hit to demand to a marine oil product which only constitutes 4% of global oil products. Very little demand for HSFO in 2020 This means that in an already very small oil product market there will hardly be any demand left for the product in If it is possible to convert the projected 3.2 m bl/d of surplus HSFO in 2020 then of course it would not be all that problematic. That is however the thing, it is not all that easy to convert 3.2 m bl/d of HSFO to a non-scrubber compliant ULSFO 0.5% fuel. In order to better be able to compare the different 2020 forecasts we have normalized them all to a total 2020 marine fuel consumption of 5.5 m bl/d. Fuel demand projections 2020 normalized to 5.5 m bl/d Delft 2012 Shell 2020 BP PIRA 2020 Average 2020 IEA 2020 HFO 3.5% ULSFO 0.5% (Blends) Gas Oil 0.5% or lower RobinMeech 2020 Delft Base 2020 Source: Source: SEB, IEA, BP, Delft, PIRA, Robin Meech In this normalized picture we have that the most extreme is Shell s projection which projects demand of only 0.3 m bl/d of HSFO and only 1.7 m bl/d of ULSFO in I.e. it projects that two out of four m bl/d of today s HSFO demand will shift all over to Gasoil in In terms of demand for HSFO in 2020 Shell s projection is consistent with the assumption that less than 2,000 ships will have scrubbers by 2020 which equally implies HSFO demand of only 0.3 m bl/d. SEB s projection that less than 2000 ships will have a scrubber is actually consistent with Shell s very low projection of only 0.3 m bl/d HSFO demand in Demand shifting from HSFO 3.5% to ULSFO 0.5% What stands out very clearly is a general projection that there will be a large shift to nonscrubber compliant ULSFO 0.5% fuel. This fuel is still on a test-bed stage and very few participants have been able to test it. The general view of the coming ULSFO 0.5% is that it will be a hybrid, blended fuel and not a straight through refined product. As such it cannot be easily mixed between different suppliers of the fuel and it is also assumed to be quite unstable so that it can be difficult to use for ships running tramp trading. If there really will be demand for 3 m bl/d of ULSFO or if refineries really will be able to supply 3 m bl/d of ULSFO we do not know. Both sides of the equation could be restrained. What is clear is that there will not be a lot of HSFO demand in 2020 and that demand instead will shift to non-scrubber compliant ULSFO 0.5% or Gasoil 0.5% marine fuel.

12 Macro research : IMO2020 Report Wednesday 14 March There is no time for the world s refineries to install additional upgrading units by 2020 in response to the IMO s 2020 decision, at least not above what they have already planned to install. The complexity of the world s refinery industry makes it very difficult to estimate and predict what it is really able to do in What can refineries do to 2020? One of the major unknowns in the IMO maze is the question of what refineries really can do by the 2020 crossover deadline. What can they do with their current refinery equipment, and additional equipment to be installed between now and 2020? The refinery industry has been very tight-lipped about what it can or cannot do. Generally, it has said that to install a new major installation such as a coker or deep conversion unit needed to convert HSFO to middle distillates, it will take five years to progress from planning to completion. So, what will be installed by 2020 will have very little to do with the IMO s 2020 deadline decision in October New refinery upgrading capacity from now to 2020 will basically be due to plans and investments taken ahead of the IMO s 2020 decision in October One problem with the refining industry is that the huge refinery machines located worldwide are not especially homogenous. It is possible to create general refineries that can do everything, but this is not economical. It is much better to tailor each refinery to what it is specifically required to do, i.e. the crude oils it is required to convert and products to make. There are more than 300 different qualities of crude oil worldwide and a myriad of different products and needs. Consequently, there is also a very wide range of different refineries. This is also why there is little general literature on the global refinery system because it is so difficult to generalize. Even refineries themselves find it hard to maintain a strong, clairvoyant view on their business outlook. What will refinery margins look like one year from now? Based on communications with several refineries over the past 10 years, most, as far as we can see, do not hold especially strong views on forward cracks and product spreads. The IMO has only imposed a requirement on sulphur content to be less than 0.5%. The many other quality requirement aspects of the fuel are unchanged. This gives refiners a lot of freedom to make dirty ULSFO 0.5%. If IMO requirements had also been imposed on other quality aspects, they would have had to upgrade HSFO to Gasoil instead. Sulphur is tied much deeper in the molecular structure of HSFO hydrocarbons. This reason, as well as its possessing a range of impurities, makes it difficult and expensive to remove sulphur from HSFO. If you cut the barrel differently during the vacuum distillation stage to obtain slightly less sulphur and heavy molecules and impurities, then that cut can be run through existing desulphuring units. This helps to create ULSFO 0.5%. A clear message from one complex refinery we met with was that one good thing about bout the IMO regulation was that it was solely a regulation on sulphur content and not on the many other quality aspects of the marine fuel. In other words, compliant marine fuel in 2020 can be quite dirty and heavy in many respects. What matters is that it does not contain more than 0.5% sulphur. If there had also been a regulation on the many other quality aspects of the fuel itself and not just sulphur, then it surely would have presented a serious challenge for the global refining industry. Then they would have had no choice other than to convert 4 m bl/d of HSFO into cleaner middle distillates using deep conversion units like cokers. However, the IMO has only set a new standard for sulphur content for shipping fuel and not many other quality aspects of the fuel. According to refiners we have met with, this confers considerable freedom on global refineries on how they can modify the HSFO to fuels that can be used under the new regulations. It is not so difficult to get down to 0.5% sulphur provided there are no guidelines and requirements governing the many other aspects of fuel quality measures was the response we got from one complex refinery. They had however no estimate of the aggregated global refinery capacity to produce ULSFO 0.5%. A key problem with HSFO is that it is very difficult and expensive to de-sulphur the fuel directly. Running the fuel through existing de-sulphuring units quickly clogs up the catalysts and halt the process, partly due to the high level of impurities in the dirty fuel. It is however also attributable to sulphur molecules being tied much deeper into fuel molecules in heavy fuel oil than they are in Gasoil making them harder to remove. Therefore, existing de-sulphur units are unsuitable to de-sulphur HSFO. While it is possible to build de-sulphur units to treat HSFO, they are however as expensive as deep conversion units. So, it is much better to build deep conversion units in order to convert low-quality fuel oil to high quality middle distillates than to convert low-quality HSFO 3.5% to ULSFO 0.5%, which is still a low-quality fuel. However, according to the complex refineries we spoke with, there are other routes to create ULSFO 0.5%. This involves different kinds of after-treatments and blending. One issue highlighted was that if you cut the barrel differently in the vacuum distillation stage so that the cut contains somewhat less of very heavy elements and somewhat more of lighter elements, the cut may then be run through a normal desulphurisation unit to lower the sulphur content. What emerges is still a fairly heavy, dirty fuel but with much less sulphur.

13 Macro research : IMO2020 Report Wednesday 14 March Subsequently, a significant amount of blending of different fuels takes place, to produce an acceptable fuel lately dubbed ULSFO 0.5%. So far, there has not been a unified fuel quality description of this new fuel, adding to market unease. ULSFO 0.5% is predicted to be fairly unstable and best suited for steady freight activity between two ports Maersk has stated that they plan to run on fuel compliance, most likely by operating on hybrid ULSFO 0.5% products. Another very unclear issue is the stability of new ULSFO 0.5% qualities. Since typically they will be blended fuels rather than straight through processing products, they will likely be more unstable. Arguably, these ULSFO 0.5% fuels will be better suited for fixed routes link container liners. They can burn off the fuel continuously. They rarely risk sitting in one harbour for long with the possibility of fall-outs in blended fuels. Freight activity on fixed routes like this will also be able to source a stable quality of ULSFO 0.5% from a few suppliers, as they typically only trade between two locations. Container company Maersk has clearly stated they plan to target fuel compliance with ULSFO 0.5% and not with scrubbers. Their argument is very clear: Scrubbers occupy space, add complexity requiring skilled on-board operatives, generate waste that must be processed or disposed off at a cost, and give rise to higher fuel consumption. Bulk carriers that run tramp trade could have issues with the new ULSFO 0.5% fuels. They will trade from location to location and are subject to the risk they may receive very different qualities of the ULSFO 0.5% they need. We understand these cannot be mixed. Therefore, fuel tanks will need to be cleaned before they can be re-filled with ULSFO 0.5% from a different supplier. Bulk carriers also typically risk being land-locked for periods with the risk of a fall-out in unstable ULSFO 0.5% fuel. How unstable the ULSFO 0.5% will turn out to be remains to be seen. It is hard to gauge at this stage with little of these fuels on the market yet. There is very low visibility on how much ULSFO 0.5% the world s refineries will in fact be able to produce in Sulphur removal from HSFO by the world s refineries is likely to create the need to dispose of 3.6 m ton of sulphur A 2020 surplus of HSFO/residue also means a surplus in the bitumen/asphalt market. The European bitumen market is already in surplus. We expect more downside pressure. Shipowners holding back on scrubber investments as they don t know what the fuel spreads will be in the future It is very unclear what volume of ULSFO 0.5% the world s refineries can actually produce. Those we have spoken with have said they have no clear views on global capacity and magnitude, emphasising once again the highly complex and variable nature of the global refining business. Surplus of bitumen, asphalt and sulphur The new maritime sulphur limit means that sulphur needs to be removed from HSFO either before it is burned or after. Assume that sulphur must be removed from 3.2 m bl/d of HSFO before it is burned, this is equal to a reduction of sulphur in 182 m ton of HSFO from average 2.5% sulphur to 0.5%. This is equivalent to sulphur production of 3.6 million ton of sulphur annually. To our understanding the global sulphur market is already saturated. It may therefore be problematic for the world s refineries to get rid of the sulphur in The refineries may thus get a sulphur disposal problem in Bitumen which is used in asphalt is basically a variant of the old refineries residue production. i.e. it is part of the HSFO complex. Therefore, a 2020 market overflowing with HSFO also means a surplus of bitumen. The European bitumen market is already in surplus with prices heading downwards, partly due to the increasing tendency to build roads using concrete. The IMO 2020 event is likely going to lead to yet higher bitumen surplus and thus even lower bitumen prices. Historical prices & spreads for HSFO 3.5%, Gasoil 0.1% and Brent Much of the current market discussion is centred on the HSFO to ULSFO/Gasoil spread. The value of a scrubber investment will be much lower if the spread is low rather than high. This is one of many reasons why many ship-owners have held back in terms of investing in scrubbers because they are uncertain whether the spread will be wide or narrow. They can see what it is today and what it has been in the past, but what will it really be in the future? And, if the spread turns out to be wide in 2020 then how long will it stay that way? One year, two years or longer? So what is the correct spread to look at and to study? The HSFO to Gasoil spread or the HSFO to ULSFO spread? In the previous section it is clear that the general market expectation is that there will be a large shift from HSFO to ULSFO with ULSFO accounting for 3 m bl/d of supply and demand in I.e. the expectation is that there will be a refinery fuel compliant solution and that this solution will primarily be ULSFO. As such the

14 Gasoil 0.1% to HSFO 3.5% spread in USD/ton Gasoil 0.1% and HSFO 3.5% in USD/ton Macro research : IMO2020 Report Wednesday 14 March most important thing would be to investigate the HSFO to ULSFO spread. However, historically and also currently there is no such thing as a ULSFO product, price or contract. Market looking at Gasoil to HSFO spread as there are no historical index for ULSFO 0.5% prices The lack of an historical ULSFO price is probably why most focus has centred on the HSFO to Gasoil price spread instead. As an alternative we have created a synthetic ULSFO 0.5% index constituted by 44% LSFO 1.0% + 56% Gasoil 0.1% which we will get back to in the next section. Gasoil typically has a larger multiplier relationship to Brent crude oil than HSFO 3.5% HSFO 3.5% and Gasoil 0.1% prices versus Dated Brent crude oil HSFO 3.5% Gasoil 0.1% Linear (HSFO 3.5%) Linear (Gasoil 0.1%) y = 8.3x y = 5.6x Brent crude oil in USD/bl Source: SEB, Bloomberg This of course naturally means that the historical relationship between the Dated Brent crude oil price and the Gasoil 0.1% to HSFO 3.5% spread has a multiplier of 2.7 times the Brent crude oil price: Gasoil 0.1% to HSFO 3.5% spread in USD/ton vs Dated Brent crude oil in USD/bl 800 Gasoil 0.1% to HSFO 3.5% spread in USD/ton Linear (Gasoil 0.1% to HSFO 3.5% spread in USD/ton) y = 2.7x Dated Brent crude oil in USD/bl Source: SEB, Bloomberg More clearly sorted as averages for each 10-dollar price slot of Dated Brent crude oil prices we have the following:

15 USD/ton USD/ton Macro research : IMO2020 Report Wednesday 14 March Gasoil 0.1% to HSFO 3.5% spread in USD/ton vs Dated Brent crude oil in USD/bl Gasoil 0.1% to HSFO 3.5% spread in USD/ton Dated Brent crude in USD/bl Source: SEB, Bloomberg The Gasoil to HSFO 3.5% has historically averaged $256/ton when Brent has been in the range of $50-80/bl What we see is that historically when the Brent crude oil price has traded in the range of $50/bl to $80/bl then the Gasoil 0.1% to HSFO 3.5% has averaged $256/bl. However, what we see in the daily dots is that the price product price spread has traded all the way up to $450/ton on individual days even when the Dated Brent crude oil price has traded in the range of $60/bl. If we instead sort average prices and spreads by year back to 2001 we get the following: Yearly averages for Gasoil 0.1%, HSFO 3.5% and the spread in USD/ton 1, Gasoil - HSFO spread in USD/ton HSFO 3.5% in USD/ton Gasoil 0.1% In USD/ton Source: SEB, Bloomberg The Gasoil to HSFO spread has rarely been above $300/ton for a full year except in 2008 when it averaged $447/ton What we see is that the spread has rarely been much higher than about $310/ton for a full year. This happened from 2011 to 2014 when Brent crude oil average $ /bl and the spread was close to that as well also in 2006 and The real stand-out year was 2008 for which the spread averaged $447/ton. That was a year with strong middle distillate demand and a very tight gasoil market. Gasoil prices and Brent crude oil prices then spiraled higher and higher until Brent crude oil reached $148/bl. In an historical perspective the current forward pricing of the 2020 Gasoil to HSFO spread at $314/ton is basically above all historical normal years. It is just trading lower than the extreme year If we also take into consideration that forward crude oil

16 USD/ton Macro research : IMO2020 Report Wednesday 14 March prices are only trading in the range of $56-57/bl for 2020 and 2021 it looks on the face of it expensive. Yearly average Gasoil mark-up to Brent and HSFO discount to Brent in USD/ton Gasoil mark-up over Dted Brent crude oil in USD/ton HSFO disscount to Dted Brent crude oil in USD/ton Source: SEB, Bloomberg Firstly however we think that 2020 is really going to be an extreme year in terms of the product spread in question. As such the relevant comparison in our view is the year We are not necessarily forecasting a Brent crude oil price spike towards $150/bl even though that is very possible. Refinery upgrading capacities will be stretched to their limits. Product price spreads will widen out The market is already pricing in an unusually tight Gasoil market in Historically such situations have led to a spike in Brent crude oil prices What we do feel confident about is that the global refinery upgrading capacity will be stretched to its limit. To all our understanding of the product price spread dynamics this means a significant widening in the product price spread. Historically such events have also implied a rippling effect into the pricing of the different crude oil slates. If the market is in need for more middle distillates and refinery upgrading and and conversion units have been maxed out then the next solution is to use more light sweet crude oil. Thus the spiraling price between Gasoil prices and Brent crude oil prices in such historical events. What really stands out in the last graph is the mark-up in Gasoil prices over Brent crude oil prices on a forward basis from 2020 to This mark-up is trading above all the historical yearly averages except for This forward mark-up for Gasoil over Brent is signaling a real market concern that the Gasoil market will be tight for this period. And this is exactly the recipe for a spike in the light sweet crude oil prices as well. Thus implicitly the forward market prices are implying a fairly deep concern for a significantly tight Gasoil market in 2020 to This is also visible when we look at Gasoil and HSFO prices in percentage terms versus Dated Brent crude oil prices:

17 Percent disscount and mark-up versus Dated Brent crude -47% -44% -42% -45% -37% -38% -38% -30% -32% -35% -37% -36% -40% -38% -37% -26% -26% -27% -25% -28% -29% -27% -30% 6% 13% 14% 11% 10% 11% 10% 11% 21% 18% 16% 23% 20% 18% 18% 15% 23% 21% 24% 26% 27% 27% 28% Macro research : IMO2020 Report Wednesday 14 March Gasoil 0.1% and HSFO 3.5% mark-up and discount to Dated Brent crude oil in percent 40% 30% 20% 10% 0% -10% -20% -30% -40% -50% -60% HSFO 3.5% disscount in % to Dtd Brent crude Gasoil 0.1% mark-up in % to Dtd Brent crude Source: SEB, Bloomberg In this perspective we see that the forward market price for Gasoil for 2020 to 2023 in percent versus Dated Brent crude oil prices is trading at levels above all historical years. There is very little middle distillate content in new ultra-light US shale oil. A tight Gasoil market could be the result There are no historical prices for ULSFO 0.5% A synthetic ULSFO 0.5% given by 44% LSFO 1.0% + 56% Gasoil 0.1% In addition to the IMO 2020 event there is also another reason to be concerned about the supply of middle distillates in 2020 onwards. On the face of it the current oil market situation here and now looks fine. Global oil demand is growing strongly on the one hand while US shale oil production is growing comparably strongly on the other hand. This looks balanced and good. This may however not be the case. The reason is that crude slate produced by US shale oil contains hardly any middle distillates. It is ultra-light and almost only contains light end products like gasoline and naphtha. Thus we may get a situation towards 2020 and the following years where the high level overall balance in the global hydrocarbon liquids market is balanced while there is actually could emerge a growing deficit in middle distillate products. Historical prices and spreads for ULSFO 0.5% and Gasoil 0.1% Though not proven at all in terms of feasibility the average forecasts are projecting that the lion share of today s 4 m bl/d HSFO demand shifts to ULSFO 0.5%. No such product exists today and no such historical benchmark exists either. As such it is a little bit difficult to review and analyse. The general expectation is though that the ULSFO fuel will trade quite tightly towards Gasoil 0.5% prices in 2020 as it will be priced relative to the alternative compliant fuel. From historical LSFO 1.0% prices and Gasoil 0.1% prices we have created a synthetic ULSFO 0.5% index price given as a mix of 44% LSFO 1.0% and 56% Gasoil 0.1%. The historical price picture for this ULSFO index may give some indication of where the future ULSFO 0.5% price will trade in the longer term while it may be far away from the prices we ll see in the first years from 2020 onwards.

18 USD/ton Product spreads in USD/ton Macro research : IMO2020 Report Wednesday 14 March Average Gasoil to HSFO and synthetic ULSFO 0.5% to HSFO spreads in USD/ton Gasoil to HSFO 3.5% spread in USD/ton ULSFO 0.5% to HSFO 3.5% spread in USD/ton Average Brent crude oil in USD/bl Source: SEB, Bloomberg The synthetic ULSFO 0.5% has averaged $150/ton above HSFO in a Brent crude reference of $50-80/bl A synthetic ULSFO is now priced at more than $200/ton above HSFO for 2020 delivery and beyond What we see is that when Brent crude oil prices have ranged from $50/bl to $80/bl the synthetic ULSFO 0.5% mark-up over HSFO 3.5% prices has averaged $150/ton. In comparison the comparable spread between Gasoil 0.1% and HSFO 3.5% averaged $256/ton for the same Brent crude oil prices. So in a historical perspective the synthetic ULSFO 0.5% index has been $150/ton more expensive than HSFO 3.5% while Gasoil 0.1% has been $256/ton more expensive when Brent crude oil prices have averaged $50/bl to $80/bl. Assume that the product market and refinery system is not stressed in 2020 onwards and that product price spreads trade according to normal historical patterns and that a freight market segment is set by non-scrubber ships running on ULSFO. Then the few scrubber-ships running on HSFO would only fetch a freight premium equal to $150/ton times the number of tonnes per day of fuel consumption. On a forward basis the current market pricing is for Gasoil to average $289/ton above HSFO from 2020 to 2023 while the our synthetic forward ULSFO 0.5% is priced $216/ton above the HSFO prices. It looks like this: Yearly averages for synthetic ULSFO 0.5%, HSFO 3.5% and the spread in USD/ton Synthetic ULSFO 0.5% mark-up over HSFO 3.5% in USD/ton HSFO 3.5% in USD/ton Synthetic ULSFO 0.5% USD/ton Source: SEB, Bloomberg

19 Macro research : IMO2020 Report Wednesday 14 March The HSFO to Gasoil Coker link There is a very strong and fundamental link between HSFO or refinery residue on the one hand and Gasoil (middle distillates) on the other hand. The link goes through refinery upgrading units like cokers and crackers. These are often labelled deep conversion units as they go deep into the molecular structure of the long and heavy residue molecules and break them apart at high temperature and pressure into smaller and lighter molecules. The opex of a typical Coker conversion unit has been given $29/ton while the opex + capex cost is Cokers are very expensive and costs approximately one billion USD for a unit able to convert 40,000 bl/d of residue to middle distillates. Thus to upgrade 4 m bl/d of HSFO one would need to spend $100 bn in capex. To our understanding such a unit typically has a Short Run Marginal Cost (SRMC) or opex of $29/ton residue converted and $86/ton all in cost (opex + capex) or Long Term Marginal Cost (LTMC). If that was all there was too it then the typical spread between HSFO and Gasoil should typically have been around $86/ton. The historical average price spread between the two products from 2001 to the start of 2018 was however $237/ton. The cost of converting HSFO to Gasoil in a coker unit is proportional to the HSFO price since 30% of the volume is lost to Pet Coke in the process. The key reason is that when HSFO is converted to middle distillates in a coker then 30% of the feedstock is lost in the conversion process and instead falls out as Pet Coke which typically only fetches a price of $60/ton. For each ton of HSFO converted to middle distillates there is thus an added cost on top of the all-in opex + capex cost for the coker operation stemming from conversion losses. For each ton of HSFO converted there is thus an additional cost due to the conversion loss equal to: Eq1: Feedstock Coker conversion loss: 0.3*(HSFO price Pet Coke price) However, since one ton of HSFO is only converted to 0.7 ton of Gasoil or middle distillates one needs to convert 1/0.7 = 1.43 tons of HSFO in order to get one ton of Gasoil. Thus one needs to calculate the conversion cost of converting 1.43 tons of HSFO: Total conversion cost in order to convert 1.43 ton HSFO to 1 ton Gasoil with Pet Coke =$60/ton: Eq2: 1.43 * [0.3 * (HSFO price Pet Coke price) + 86] = 0.43*HSFO + 97 Putting in the historical average HSFO price of $340/ton and a Pet Coke price of $60/ton we get that Eq2 is equal to $243/ton. That is very close to the historical average HSFO to Gasoil spread of $237/ton over the given historical period. If we look at the Gasoil price as a function of the HSFO price we get that Gasoil = HSFO plus conversion cost in Eq2, or: Gasoil = HSFO * HSFO + 97 = 1.43* HSFO + 97 If we define the Short Run Marginal Cost (SRMC) of conversion as conversion losses + opex and Long Run Marginal Cost (LRMC) of conversion as conversion losses + opex + capex we get the following equations where the gasoil price is written as a function of the HSFO price: SRMC: Eq3: Gasoil(HSFO) = (1/0.7)* [HSFO *60] = 1.43*HSFO + 15 LRMC: Eq4: Gasoil(HSFO) = (1/0.7)* [HSFO *60] = 1.43*HSFO + 97 The Gasoil to HSFO is proportional to the HSFO price. Higher HSFO price means a wider spread What is clear both when one looks upon this from the mathematical point of view as well as the historical price spread point of view is that the Gasoil HSFO price spread is proportional to the price level of the HSFO. The higher the HSFO price is the higher is the conversion loss and the higher is the price spread between Gasoil and HSFO. When we graph the theoretical SRMC and LTMC as given by Eq3 and Eq4 on the basis of the historical HSFO prices and also graph the actual HSFO and Gasoil prices we get:

20 Gasoil 0.1% in USD/ton Macro research : IMO2020 Report Wednesday 14 March Gasoil 0.1% vs HSFO prices and SRMC and LRMC Coker conversion costs - USD/ton Historical Gasoil 0.1% prices in USD/ton Theoretical SRMC Cocker cost Theoretical LT Cocker cost Fueloil 3.5% in USD/ton Source: SEB, 37th IAEE International conference 2014 What is apparent is that the historical price points not at all are glued neither to the SRMC nor to the LTMC, though very few price points are below the SRMC line which of course makes a lot of sense. The price spread between HSFO and Gasoil becomes exponential when conversion needs exceed installed conversion capacity. It did so in The driving force for where the price spread settles is given by the need for conversion of HSFO from day to day since When the need is very low the utilization rate of the global coker conversion units is very low. This again means that market is cleared on the SRMC which is the lower blue line in the graph. If the conversion need is high so that all coker conversion units in the global refinery system are maxed out then the marginal price setting goes above the LTMC line of conversion which is the green line in the graph. There are other marginal conversion units than cokers which are less economic and less optimal for the treating and conversion of HSFO. These have much higher treatment costs. In extreme cases when the global conversion capacity is totally maxed out but the market is still asking for more the price spread can go exponential. This is typically what happened in 2008 when the Gasoil to HSFO price spread widened to $700/ton. In the next graph we show the historical relationship between HSFO price on the x-axis versus the Gasoil 0.1% to HSFO price spread on the y-axis. We have also drawn the theoretical SRMC and LRMC conversion cost lines as given in Eq3 and Eq4 using $29/ton and $86/ton respectively as the constants within the equation. The dots which lay above the light green line are again typically situations where upgrading needs have exceeded coker capacity making it necessary to move to more expensive and uneconomical conversion units, driving up the conversion cost and therefore the Gasoil to HSFO price spread.

21 Gasoil 0.1% to FO 3.5% spread in USD/ton Macro research : IMO2020 Report Wednesday 14 March The Gasoil to HSFO price spread is already trading high vs. theoretical long term coker conversion costs. This is a sign that the utilization rate of the world s refinery system is expected to be significantly stressed in 2020 and HSFO prices vs Gasoil to HSFO spreads, SRMC, LRMC Coker conversion costs - USD/ton Historical Gasoil to HSFO spread in USD/ton Short term conversion cost Long term conversion cost Forward spreads Fueloil 3.5% in USD/ton Source: SEB, 37th IAEE International conference 2014 However, the Gasoil to HSFO price spread for 2020 and 2021 is still priced within a reasonable historical range. It is nowhere nearly priced alongside the severely stressed event of 2008 which saw the spread move up to $ /ton. The price of HSFO may need to fall towards the comparable price of coal to burn surplus HSFO in power stations instead of coal in 2020 and With thermal coal today trading at $85/ton and with 40% more energy content in HSFO per ton, this implies a potential decrease in HSFO towards $120/ton. At the time of writing the 2020 HSFO price in the ARA region is $220/ton. In the above graph we have also pinpointed the market prices for forward Gasoil to Fuel oil price spreads for the coming years to What we see is that only the balance of the year for 2018 is trading below the LTMC of conversion with respect to the market price of HSFO. The other years from 2019 to 2021 are trading above. This shows that the market has started to price in a wider spread as well as a lower HSFO price. What we have experienced historically is that the highest price spreads have occurred when the HSFO price has been high. This is of course obvious from the equations above which shows that the spread is proportional to the HSFO price. The most extreme price points are from 2008 when the Brent crude oil price traded up to $148/bl. It was very strong demand for Gasoil and middle distillates and the Brent crude oil price spiralled upwards hand in hand with higher Gasoil prices. As the global refinery upgrading capacity was maxed out the Gasoil to HSFO price spread expanded to $700/ton at the most. What is important to remember from the 2008 incident is that it was still allowed to use HSFO, it was just not so much need for it. Also, the real need in the market was on the Gasoil side which again led to the need for maximum upgrading of HSFO to Gasoil. What will be different in 2020 and 2021 is that it will no longer be allowed to use HSFO in marine freight without a scrubber. Thus rather than moving from the lower left hand side corner in the graph to the upper right hand side we think that the 2020 price points should head to the upper left hand side corner of the graph. The IMO 2020 event should give us a very low HSFO price of about $120/ton, and a fairly high Gasoil price and a high spread. It will basically be impossible to store a significant running surplus of HSFO over some time in Therefore, if demand is insufficient and there are not enough conversion capacities in the refining industry action will need to be taken. In our opinion, the last such option is to burn the surplus HSFO in on-shore power plants instead of coal. There are probably sufficient such power plants available equipped with on-shore sulphur scrubbers where the HSFO can be burned. There has been discussion whether refineries would in fact have to pay to get rid of their surplus HSFO or residue. We do not think it will be necessary for refineries to pay to dispose of their surplus residue or HSFO. However, to burn the surplus HSFO in a power station, the HSFO price will have to fall further so it becomes competitive vs. today s much cheaper coal prices. HSFO typically contains some 40% more energy per ton than normal thermal power station coal. With a coal price currently at $85/ton it means that the HSFO price will need to fall to around $120/ton to be competitive with coal and therefore be burned in a coalfired power station. At the time of writing, the price of HSFO delivered 2020 is about $220/ton. If the power station option needs to be utilized in order to handle a running surplus of HSFO in 2020, it

22 Macro research : IMO2020 Report Wednesday 14 March means that the 2020 HSFO price has potentially $100/ton further downside before it finds solid support at the coal cost level. If it really turns out that the market will need to burn HSFO in power stations in order to get rid of the surplus, then it also means that the total liquids supply in the global oil market is shrinking by a comparable amount. It signifies an overall tighter global oil market with higher oil prices. This again pushes up the Gasoil to HSFO price spread based on the relationship described earlier. If cokers are used to convert a 3.2 m bl/d HSFO surplus in 2020 to Gasoil then 1.2 m bl/d of this will be lost to Pet Coke in the process. This results in a comparable tightening of the overall oil market. Product price spreads should revert to historical norms as markets eventually normalizes. Probably not before 2025 We assume that HSFO consumption is 4 m bl/d. If demand for HSFO is only 0.8 m bl/d in 2020, 3.2 m bl/d will need to be converted to either ULSFO or Gasoil. If cokers were to upgrade these 3.2 m bl/d of HSFO to middle distillates or more specifically Gasoil, 1.2 m bl/d of this would be lost to Pet Coke during the process. This means a comparable tightening in the overall supply of liquids in the oil market, increased crude oil prices, and again a higher price spread between Gasoil and HSFO. Compliant fuel to be Gasoil 0.1%, or ULSFO 0.5%? One thing is that product price spreads are likely to blow out in 2020 and for some years following that as the global refinery system upgrading capacity is stretched to its limit. Over time however the product price spreads are likely going to revert to the spreads we have witnessed historically for different levels of Brent crude oil prices. The following is a table of historical average product prices for different price levels of Brent crude oil. We assume that these historical price spreads reflects refinery upgrading economics over time. Thus as the global refining system adjusts over time one should comparable spreads to emerge again. Historical prices for Brent crude and products and spreads in USD/bl and USD/ton Source: SEB, Bloomberg Assume that a normal oil price is to be in the range of $50-80/bl in the years to come. Then ships using Gasoil 0.1% as compliant fuel should expect to pay a mark-up in fuel price over HSFO in the ball-park of $250/ton. Ships which instead run on a blended mix between LSFO 1.0% and Gasoil 0.1% (44% & 56%) should expect to pay a fuel premium of $150/ton over HSFO. Complex refineries are however stating that they can do more than just blend LSFO and Gasoil. They claim that they can do different kinds of upgrades of HSFO as well as partial desulphurization by cutting the barrel differently in the vacuum distillation process. Some have also claimed that they have come up with desulphurization methods for HSFO directly (Rigby Refining LLC) though we have no details of this process in terms of quality, cost or possible volume of magnitude. In the longer term beyond 2025 the ULSFO 0.5% grade should probably trade only $50-100/ton above HSFO However, it seems clear that the complex refineries can do more than just plain blending of LSFO and Gasoil. As a consequence the price of ULSFO 0.5% products should end up being cheaper than our synthetic ULSFO 0.5% index extrapolated from LSFO and Gasoil composition. This means that the spread between the future ULSFO 0.5% product and HSFO should be less than $150/ton in the Brent crude oil price slot of $50-80/bl range. A wild guess would be that it ends up somewhere between a $50-100/ton mark-up above the HSFO with Brent ref $50-80/bl.