Description

Summary

There have been 6 write-ups of Valero in VIC since 2002, including a short recommendation in 2012! The most recent one in September 2020 has aged really well, compounding at 44% since, although the exit recommendation in February 2021 locked in a 72% total return for Bruno677.

Over time, you could have owned Valero simply because it is a best in class, low-cost operator in a mission-critical industry for the world economy (cars, trucks, machines and ships do not eat crude oil but only refined product), with strategically located refineries in the US and UK, and do well anyway: despite the cyclicality of the oil and refining industries, Valero has compounded at 15.5% since its IPO in 1997, and 16% over the last decade. There have been a few times when it took several years for the stock to get back to previous highs: 1998-2001, 2007-2014, 2018-2022; but even during these, investors received a (mostly) growing dividend, which has itself compounded at 15% since IPO.

The stock is up 16% YTD but it’s had an equivalent pull-back since April’s all-time high.

I’m conscious of the myriad of short-term drivers of refiners performance, but looking beyond the likely normalization of refining margins following a couple of extraordinary years, I believe Valero currently offers attractive double digit annualized returns over the medium term for investors. 

EVs and oil demand

The world consumes 103 million barrels of oil per day. There are more than 400 types of crude oil produced globally, making it useful only when refined into value added products like gasoline, diesel, jet fuel, residual (bunker), naptha, ethane, etc. Two thirds of global refined product demand are for transportation, and of this, 40% is gasoline – mostly burned in internal combustion engines (ICE) – and another 40% is diesel, used in trucks, machinery, buses and trains.

The world wants to decarbonize transportation by going electric.

Shell thinks there will be up to 275 million EVs by 2030, 7x more than in 2023, on a fleet of ca. 1.6 billion cars, with declines in oil demand accelerating after 2030 at a pace of 4% per year in their worst case.

Source: 2024 (Shell) ESG Strategy update

Bernstein thinks there can be 260 million EVs by 2030 and ca. 500k electric trucks (on a fleet of ~70 million), but expects demand for oil products to decline at 2% annually to 2050.

It seems that, as EVs penetrate the global car fleet, oil demand should decline between 30% and 50% in a worst-case scenario by 2050 from today's 103MMbdp.

There is evidence supporting the view that gasoline demand does fall as EVs replace ICE vehicles. However, the impact on fuel products as a whole is not as straightforward.

Norway offers real-life data for both arguments. In 2023 Rystad (an energy consultancy) published a report showing EV penetration going from 20% of vehicle sales in 2015 to 90% in 2023, with TOTAL fuel product demand declining ca. 12%.

Source: Is Norway's Love For EVs Enough To Put A Dent In Fuel Demand? | OilPrice.com

So even though fuel product demand for passenger cars fell by 24% between 2015 and 2022, it continued growing in other vehicles: buses, trucks, vans.

Source: Is Norway's Love For EVs Enough To Put A Dent In Fuel Demand? | OilPrice.com

An important consideration: 90% of new vehicle registrations are electric, but 2/3rds of the vehicle fleet in Norway still use diesel or gasoline and another 12% are hybrids (Wikipedia). Even a country like Norway, with its multi-year advantage in electrifying its car parc, is still years (decades?) away from eliminating fossil-fuel products in transportation.

China offers another example. EV share of passenger new car sales were close to 35% towards the end of 2023:

Yet crude demand has been growing through the period depicted in the chart above.

Source: Bloomberg

Thinking through Norway and China’s examples, the conservative assumption is to expect oil demand to decline in the next few decades, but there is a possibility that it falls at a slower pace than predicted.

This creates a difficult environment for the refining industry to navigate.

What if refining capacity declines faster than oil demand?

Refining is capital intensive, competitive at a local and global scale, operates on thin profit margins and the assets may last longer than they may be profitable (in many cases). Similar to what is happening to offshore-oil rig owners: knowing that oil demand will fall in the next couple of decades, will an investment in new refining capacity earn a return on billions of dollars of capital that compensates for its cost?

In 2016, Shell had 22 refineries. In 2020 it had 14 left and as of 2023, has plans to reduce the portfolio to 6 sites by 2025. BP had 16 refineries in 2015, 11 in 2018 and is currently converting a few to renewable diesel, as well as shutting capacity down in others. TotalEnergies had interests in 25 refineries (including 12 operated) in 2015, and is down to 7 as of 2023.

Most oil majors are on a similar path. Valero itself started life in 1997 with one refinery and grew thanks to buying 13 that were disposed of by oil majors (BP used to own and operate 41 refineries in the 1980s!). Covid accelerated refinery closures around the world. Between 2020 and 2023, nearly 4.5 MMbpd of refining capacity shut down, mostly in OECD countries.

The stringent regulatory environment targeted at fossil fuels has also accelerated European closures. This suggests industry operators are not expecting a return that justifies spending as much capital as needed by a refinery and are instead building refineries that process food waste or convert agricultural products into sustainable fuels.

There is a scenario – typical of any capital cycle – where the capacity to supply refining products to the world declines faster than the demand for such products.

Valero expects global oil demand growth to outpace new refining capacity additions from 2025, creating a sweet spot for structurally higher refining margins.

Replacement cost and Terminal Value

In a world short of refining capacity, the value of assets-in-place should be higher than their carrying value. Operators like Valero may enjoy structurally higher margins thanks to the complexity, geographical position and low-cost nature of their operations.

The US refining system is the most complex in the world, at 11.5 NCI. Valero’s average NCI is 11.8.

Source: 2024-04 (Tupras) Investor Presentation

The US refining system as a whole enjoys structurally lower operating costs than other markets, in part thanks to abundant natural gas resources.

Source: Valero 2022-TCFD-Report-FINAL-digital-pages.pdf (q4cdn.com)

This advantage was worth $2.5 per barrel of throughput as of 3Q23 compared to European refiners. This was a record high due to Europe’s recent energy crisis, but the advantage has been at least $1/bbl for a long time. Valero also has lower OPEX per barrel of throughput than domestic peers:

Even in a world of diminishing oil demand, Valero would likely be your last-refiner standing.

A question worth asking then is what the replacement cost of a profitable refinery is?

There is wide range of potential values, all educated estimates given there has not been a brand new refinery built in Western Europe or the US in the last 50-plus years. For reference:

• The Palgrave Handbook of Energy Economics (published in 2022 puts the cost of a complex refinery (equipped with catalytic cracking, visbreaking and gasoline units) built in Europe and with nameplate capacity of 160,000 barrels per day of throughput at around $6 billion, which would be likely higher when considering emission regulations for the refinery itself and the product quality. Valero has 14 refineries with total throughput capacity of 3.15 million barrels per day and complexity of 11.8 (high). Adjusting for size, the average Valero refinery would cost ~$8.4bn to replicate, implying $118bn of replacement value for the business.
• Southern Energy Partners – a private company – has plans to build a 250,000 barrels per day refinery in Oklahoma with a view to open in 2027 at an estimated cost of $5.6bn. This is a long shot (Barron’s, May 2023). Going back in time to 2005, Glenn McGinnis from Arizona Clean Fuels had already spent six years and several million dollars trying to build a 150,000 refinery near Phoenix at an estimated cost of $2.5 bn (New York Times, May 2005). It did not happen. The same attempt today will cost 35% more than the equivalent 2005 effort. For Valero, this implies $5bn per refinery, or $70bn replacement value for the business. Factor in energy transition, environmental restrictions, local permits and public antagonism, and any attempt in the late 2020s or 2030s to build a new refinery anywhere in Europe or the US is almost guaranteed to carry a higher price tag.
• The CEO of Tupras, the leading Turkish refiner majority owned by the Koc conglomerate, mentioned in early 2023 that the cost to replace its 560,000 barrels per day system in Turkey would cost at least $19.5bn, or $35,000 per barrel of throughput capacity. Tupras owns one of the most complex refineries in Eastern (and Western) Europe. The equivalent for Valero would be $110bn of replacement value (Hosking Partners).
• Other examples include Mexico’s Dos Bocas refinery, set to start operations in 2024. With 340,000 barrels of daily throughput capacity, it will come at an actual cost of $20bn (vs. original estimates of $8bn!) and about 3 years later than initially planned and 7 years after planning began.

Calculating Valero’s terminal value based on a multiple of EBITDA (~5.2x on average in the 2006-2019 period), which sell-side prefers, suggests Valero’s valuation is discounting margins of around $12.5/bbl, which seem high compared to its historical $9/bbl average since 1996. 

However, adjusting the terminal value to reflect the likely replacement cost of Valero’s assets, the implied refining margin in Valero’s current $151/share is close to its historical average at $9.5/bbl.

Source: Valero

Consensus expects refining margins of $13.65/bbl in 2024 and $12.75/bbl in 2025. Under a base case of $14, $13, $11, $11 for 2024-2027 fading to $10/bbl in 2028, Valero trades at ~35% discount to its intrinsic value and can deliver decent double digit annualized returns if bought at current prices. Dividends (regular and specials) and buybacks – which combined have been 7.4% on average over the last couple of decades – support expected returns:

Source: Valero, Bloomberg

I would buy Valero at current prices.

Risks?

Apart from the obvious valuation risk embedded in my choice of terminal value, there are a few:

• The opening of the Transcontinental Canadian pipeline, which offers a route for Canadian WCS exports to Asia and may balance the flow of WCS barrels into the US mid-continent and Gulf Coast. For years, US refiners have bought WCS at a significant discount to Brent and WTI benchmarks due to volumes clogging pipeline capacity, forcing producers to ship crude via rail. If the spread narrows permanently AND refiners cannot replace heavy Canadian crudes with other alternatives such as Mexican Maya, Venezuelan or middle eastern heavies, margins could suffer. Valero thinks they can replace WCS volumes with Alaska North Slope into their west coast refineries if they must, and their Gulf Coast system could take crude from anywhere too. Perhaps California becomes an export hub for US refining products into Asia in the near future?
• Valero is a consensus buy. I’m not much of a contrarian here but merely expect Valero’s historical outperformance to continue.
• The world is past peak oil demand? The IEA’s latest oil report (Slowing demand growth and surging supply put global oil markets on course for major surplus this decade - News - IEA) forecasts an 8 million-barrels-per-day surplus in 2030! I’d take this forecasts with a (large) pinch of salt however. Use IEAs own coal use estimates for example: in 2018, the IEA expected coal use to peak in 2023 at 7.7GTpa (Gigatonnes per annum). However, the forecast increased to 7.9GTpa by 2023 in their 2019 WEO and to 7.95GTpa in their 2021 WEO. In 2023 global coal use hit 8.5GT, an all-time high. I’d expect the IEA to underestimate the stickiness of oil just as they have underestimated coal’s. See Deutsche Rohstoff’s on oil demand below:

Source: 2023-10 (Deutsche Rohstoff) Capital Markets Day presentation

• Environmental regulations in most developed markets are forcing penetration of renewable fuels into the mix, displacing hydrocarbon derived fuels in the near future:

Source: Valero Investor presentation

The race to build renewable refining capacity has resulted in lower renewable refining margins. Valero’s DGD refinery has been selling renewable diesel at a 90% average premium over Ultra-Low Sulphur Diesel since 2018. This premium peaked in 2020 at 206% and came down to 9% in 2023. As more capacity comes to market, the premium will disappear and may turn into a discount versus traditional diesel, potentially displacing demand for the latter product in markets where it is readily available.

• Net Capacity additions could be higher than expected globally. As with Dos Bocas in Mexico, the record is not great for projects being on time, on budget and ramping up as planned, which mitigates the risk.
• Finally, demand surprising to the upside may keep refining margins higher for longer, and Valero printing more than decent free cash flows for a few more years.

Catalyst

- Stronger oil demand growth

- Accelerated refinery closures/delayed new capacity additions (or cancellation of projects)