Description

BP Midstream is a recently IPO’ed pipeline MLP. We think that management’s IPO guidance of 15% distribution CAGR over 4 years is highly achievable and offers upside. Beyond the LQA distribution yield of 6.2%, the management case and the upside case imply a highly attractive dividend discount rate of 10% to 13% at the current stock price when assuming that distributions will be flat after the initial growth phase. Given this favorable valuation, we expect a flat or increasing multiple over the next two years resulting in annual returns at or in excess of 21% (6% current yield plus the 15% distribution growth).

Overview

- BPMP is a sponsored MLP that will initially own North American onshore and offshore pipelines

- BP owns the general partner and 54% of the limited partnership units

- Initial pipeline volume will consist of 79% crude oil, 19% refined products and 1% natural gas

- Over time, BP plans to sell (“drop down”) additional pipeline, terminal, refining logistics and fuel distribution assets to BPMP

- Similar to other sponsored MLPs, BP also owns incentive distribution rights (“IDRs”) on BPMP. BPMP unitholders will receive 100% of distributions below $1.21/unit. After that, BPMP will have to pay increasing incentives to BP. If distributions exceed $1.58, BPMP will have to pay BP 50% of every incremental cent of distribution it generates

Initial assets

Assets by DCF (“Distributable cash flow”) contribution

- The BP2 pipeline transports crude oil to BP’s Whiting refinery under FERC-regulated tariffs. BP has recently finished a multi-billion modernization project at the Whiting refinery and expects to further increase capacity from 325kbdp to 350kbpd by 2020. Capacity at BP2 was also expanded from 240kbpd to 475kbpd leaving ample room for growth from the current utilization of 61%. Even if this throughput growth does not materialize, BPMP will get 2-3% of growth at BP2 through its minimum volume commitments that will grow from 303 kbpd in 2018 to 310kbpd in 2019 and 320kbpd in 2020. The initial dividend is based off 303kbpd at BP2, the lowest possible level of EBITDA BP2 can generate.

- The Mars pipeline (28.5% owned) and the Mardi Gras pipeline system (20% owned by BPMP and 80% owned by BP) will benefit from increasing Gulf of Mexico volumes. New offshore projects that have come online recently or are expected to come online before 2020 and that are connected to the Mars and the Mardi Gras pipeline system amount to around 700 kbpd of production capacity. This compares to current GOM production of 1,600 kbpd and throughput on the Mars pipeline of 470 kbpd.

- Management has said that future dropdown assets will have similar organic growth characteristics to these initial assets

Investment thesis

1) BPMP owns stable assets and is very cheap at a LQA distribution yield of 6.2% and an implied lifetime dividend discount rate of 10% to 13%

2) At today’s valuation, BPMP is highly attractive under management’s conservative IPO case and our upside case

                - In the base case, BPMP will grow distribution at 15% for 5 years

                - Our upside case results from a larger dropdown portfolio ($1bn EBITDA) and results in 12.5% annual growth for 8 years or 21% growth for 5 years

3) Recent fears on FERC’s elimination of tax allowance in MLP pipelines’ cost-based rates are overblown and have made a cheap sector even cheaper. BPMP’s entire tax allowance represents 0.5% of FCF.

4) BPMP is also attractive on a relative basis compared to its closest peers (PSXP, VLP, SHLX). It trades inline on current distribution yield and EV/EBITDA yet has structurally higher long-term growth due to:

- Smaller size that leads to bigger impact from future accretive dropdowns from BP

- Current distribution being at MQD levels (“minimum quarterly distribution”) whereas peers have to pay 50% of distribution growth to their general partner

- No leverage today with a $600m credit facility from BP in place at a cost of debt at 3% (Libor + 85bps)

- Superior organic growth of 5% to 6% due to minimum volume commitments and ramping customer volumes

- A dropdown portfolio that we estimate exceeds $400m in EBITDA, which is the level BPMP requires to grow distributions at a CAGR of 15% for 5 years

5) In MLP-Land, new is always better. New MLPs and newly acquired assets are almost always the most attractive as older assets often face contract cliffs or become stranded. Hence BPMP with its strong organic growth looks more attractive than MLPs that acquired or built a lot of assets from 2008 to 2012.

6) While sponsored MLPs with IDRs have fallen out of favor and investors are now concerned about “conflicts of interests” with the sponsor, we think that in the first two years IDRs and BP’s 54% ownership of the LP units actually provide great alignment with LPs as the sponsor tries to grow distributions beyond the 50% tier

7) Sponsored MLP peer companies MPLX, PSXP, VLP and SHLX have done extremely well fundamentally yet underperformed due to excessive valuations pre 2016. If BPMP does half as well as these peers, returns are likely to be favorable given its attractive valuation today

Performance of comparable sponsored MLPs

- Distributions for sponsored peers have grown every single quarter since IPO

- VLP and SHLX still have substantial dropdown backlogs, so they will likely continue to grow beyond 3x their initial distribution

- MPLX has been the worst performer of the group with 2.3x growth of the initial distribution due to the value-destructive acquisition of MarkWest in 2015 (more details below)

- However even this “worst case scenario” of 2.3x growth is still above the base case for BPMP (15% for 4 years or 1.75x the initial distribution)

- Peers’ underperformance has been caused by excessive valuations in 2014 at sub 3% yields

Peer valuations today

Pushback #1 – Size of dropdown portfolio

- BPMP has provided very little detail on the size of its dropdown portfolio

- We think that BPMP doesn’t need more than $400m of droppable EBITDA in order to CAGR distributions at 15% for 5 years  

- This uncertainty provides mostly upside as we estimate the dropdown portfolio to range anywhere from $500m in EBITDA to $1.5bn

- Management has said that except for one chemical plant, the entire US business of BP would qualify as MLP-eligible income. The largest and most relevant drop down assets are the remaining 80% of the Mardi Gras pipeline system, other onshore pipelines, logistics assets at BP’s 3 US refineries, terminals, the US fuel distribution business, BP’s 48% interest in the Trans Alaska pipeline and BP’S GOM production hubs.

- One of the largest assets is the US fuel distribution business. Marathon Petroleum dropped a very similar business into MPLX recently that is generating $600m of EBITDA. On its website, BP says that the business delivered 11bn gallons of fuel in 2016. At 5 cents/gallon (similar to competitors), this suggests over $500m of EBITDA from this asset alone. The Citi analyst estimates $835m for this asset

Pushback #2 - Cost of capital

- One concern investors might have is that cost of equity and cost of debt for these vehicles have increased

- While this is true, we don’t think it is significant and does not merit a re-rating from a 3.0% to a 6.0% current yield

- Furthermore any headwind from higher cost of capital will likely be negated by BP lowering its dropdown multiple

- As can be seen below, in the worst case scenario of an increase of 400bps in cost of debt and 500bps in cost of equity, the distribution CAGR would fall from 18% to 12%. However assuming BP changes the dropdown multiple, that would already take the CAGR back up to 15%

- Dropdown multiples are not static as the sponsor needs to ensure dropdowns are accretive in order to grow the IDRs and to enhance the value of its majority LP ownership

Pushback #3 - IDR structures

- While IDR structures represent conflicts of interests, we think that at current valuations the concerns around IDRs are very misplaced and conflicts were actually more significant when MLP cost of equity was much lower

- One of the best examples of IDR conflicts was the MPLX acquisition of MarkWest in 2015. At the time, MPLX was trading at a sub 3% yield and the sponsor MPC used that currency to acquire the much larger MarkWest. While the deal was accretive in year 1, it was very likely dilutive to LPs in years 3 to 5 (as the dropdown pipeline became much smaller percentagewise to the now larger MLP). The deal was massively accretive to MPCs IDRs as the LP unit count of MPLX more than tripled and as IDRs per LP units increased. MPC management would have been crazy not to do that deal

-  At the time, MPLX had already grown distributions by 70% from IPO levels. At current distribution levels, BP has no incentive to do a similar deal as their IDR payments would stay at 0. In order to do a similar “worst case scenario deal”, BP would have to first grow distributions by at least 50%

Key risks

- As explained above, we do not see the IDR structure as a risk at the current valuation. However we do think that parent company control represents some risk to sponsored MLPs. Sponsor control results in two risks: 1) MLP abandonment if the parent company gets acquired and 2) if the MLP trades poorly for inexplicable reasons, the sponsor can just buy back the float at an insignificant premium

- Minimum volume commitments expiring in 2021. We think this is a minor risk as BP2 will be much a smaller percentage of total assets by the time the MVC expires. Furthermore, the BP2 and BP1 pipelines (a future dropdown candidate) are the only pathways to the Whiting refinery, BP’s most valuable refinery. Therefore combined volumes on both pipelines are less likely to decline.

Catalyst

- Announcement of second quarterly distribution indicating implied growth rate

- First dropdown later this year

- Management disclosing more information on dropdown portfolio