Your post - off and on - has occupied my thoughts for the last day. I've held Delek, Valero, and Phillips 66 previously even 1) I tend to invest more upsteam and 2) I think hydrocarbons are causing catastrophic global warming and the governments in the world will decide to leave a chunk of what's accessible in the ground and convert energy to renewable electric sources.

Investment 1 has worked out well, but 2 has been not worked out well (far more disappointing for my grandchildren's sake than any financial hit).

----

I don't love creating suspense. I would sell partially, or start to hedge, in your shoes.

First, I don't think my point about the downstream's terminal value is controversial. Otherwise they could invest in operations, like you said. They had $20B-ish after Speedway and could have done so, but chose to buy back shares. One can talk about the regulatory and environmental hurdles, but AI-focused datacenters are clearcutting all over the planet. If MPC or its peers saw an opportunity they would add capacity. They don't believe in their own business prospects.

Further evidence is the insider transactions, they're net sellers 10:1 from what I see. The pattern is similar with VLO and PSX. Short ratios are higher than I'd expect for otherwise cheap-ish companies given the overpriced alternatives. I didn't have time to research it in depth, but Gemini indicates some bearish option activity in MPC, VLO, and PSX, too.

If there is a recession they have large fixed costs and no control over the crack spread. I think betting on a recession is looking smarter lately.

They have gone through a lot of their buyback authorization already (70%).

That said, they just aren't expensive and the lack of investment in capacity while the US population increases probably means good things for them. It's been a fascinating question to think about, and I think my answer says more about myself than Marathon.