# When crude is cheap but diesel and gasoline are scarce, the profit sits with the companies that do the refining, not the ones pumping oil — refiners' margins on each barrel widen sharply. The Saudi pipeline outage tightens finished-fuel supply further, an

_AI-generated trading idea · BULLISH · CRAK, MPC, VLO_

> Canonical page: https://commonquant.ai/research/for-you/when-crude-is-cheap-but-diesel-and-gasoline-are-scarce-the-p--2eb4078f-a97f-4be8-beef-ed9a61c5b657

When crude is cheap but diesel and gasoline are scarce, the profit sits with the companies that do the refining, not the ones pumping oil — refiners' margins on each barrel widen sharply. The Saudi pipeline outage tightens finished-fuel supply further, and industry experts say no one can quickly add refining capacity. This makes refiners a different and arguably purer way to play the energy shock than oil majors or tankers, because they benefit from the bottleneck itself. Record diesel prices with no near-term supply relief gives this trade a durable momentum driver rather than a one-day spike.

## Idea

When crude is cheap but diesel and gasoline are scarce, the profit sits with the companies that do the refining, not the ones pumping oil — refiners' margins on each barrel widen sharply. The Saudi pipeline outage tightens finished-fuel supply further, and industry experts say no one can quickly add refining capacity. This makes refiners a different and arguably purer way to play the energy shock than oil majors or tankers, because they benefit from the bottleneck itself. Record diesel prices with no near-term supply relief gives this trade a durable momentum driver rather than a one-day spike.

## Advanced Analysis

### Verdict: The Refining Thesis Is Real, But the Trade Isn't — Wait for the Pullback

The thesis is being confirmed in reported numbers, not just headlines: Valero's gross margin jumped to 12.3% in the quarter ended June 30, 2026 from 6.3% the prior quarter, net margin more than doubled to 8.4%, and Marathon's free cash flow rebounded to $9.1B in its most recent quarter from $208M. Per the Bloomberg and Yahoo Finance pieces from September 13, 2026, the Saudi pipeline outage has diesel at record highs with no quick capacity relief — the supply-shock driver is live. But the entry conditions are nowhere close: all three names trade above their 20-day averages with stochastics in the 80s–90s when the rule needs a close below the average with a stochastic at or below 35, and the setup never fired across 1,235 evaluated daily bars over 60 months. Insiders at MPC (about $4.0M net selling) and VLO (about $5.7M net selling, reporting period ended June 30, 2026) add a dated caution while revenue at both firms fell year over year. This is a wait-and-watch, not a buy — do not front-run an oversold-pullback setup from overbought readings.

#### Conviction score breakdown

Composite score computed by the server from the applicable evidence-tier dimensions.

| Measure | Value |
| --- | ---: |
| Thesis support | 72/100 |
| Trade readiness | 30/100 |
| Risk quality | 55/100 |
| Trigger proximity | 10/100 |
| Fundamentals trend | 65/100 |
| Score | 46/100 |
| Composite Score | 46/100 |
| Evidence Tier | rules\_not\_triggered |

#### Decision scenarios

Bull, base, and bear cases synthesized from the cited evidence tier. Likelihoods are rounded evidence-weighted judgments, not statistically calibrated forecasts.

| Measure | Value |
| --- | ---: |
| Evidence Tier | rules\_not\_triggered |

### Trade now: the pullback hasn't arrived — everything is overbought, so there is nothing to buy yet

\*\*Nothing is live today. This is a wait, and the wait is a long one.\*\*

The idea argues that cheap crude plus scarce diesel and gasoline funnels profit to refiners (CRAK, MPC, VLO), and industry experts per the thesis say no one can quickly add refining capacity. That may be right — but the strategy's entry is a pullback-and-reclaim setup, and right now the market is at the opposite extreme. CRAK closed at $65.65, well above its 20-day EMA of $61.79, with a 14-day stochastic of 91.0 and RSI of 78.8. MPC sits at $395.93 versus a $372.90 EMA, stochastic 81.9. VLO closed at $390.42 — its range high — versus a $360.36 EMA, stochastic 90.4. The entry requires price below the 20-day average, a stochastic at or below 35, and then a close back above that average. Not one of those conditions is close: the stochastic leg alone is roughly 47 to 56 points away from triggering on every name.

\*\*What wait means concretely.\*\* CRAK needs to fall roughly $3.86 to reach its EMA at $61.79 (and the stochastic needs to fall from 91 to 35); MPC needs about $23 of downside to $372.90; VLO about $30 to $360.36. Until price is under the 20-day average \*and\* the stochastic prints at or below 35, no entry is evaluated. Do not front-run the setup — buying stochastic readings in the 80s and 90s is the entry's mirror image and carries the worst of the risk profile.

\*\*Risk framing.\*\* The strategy sizes each position to risk 2.3% with a hard stop at a 2.3% loss and a take profit at 4.5%, capped at 25% per position — roughly a two-to-one reward-to-risk once an entry fires. On valuation grounds the thesis has supporting color: VLO's Q2 gross margin rose to about 12.3% from 6.3% the prior quarter, with net margin up to about 8.4%, though quarterly figures come from reported company data without full consecutive-period coverage. One caution per the ownership filings: MPC insiders show net open-market selling of about $4.0M and VLO insiders about $5.7M net selling in the latest reports — useful bearish context while we wait for the entry, not a reason to skip it.

\*\*Honest note on the setup's tuning.\*\* The rules were evaluated on real daily bars but did not open an entry across the evaluated history — this is a watch-list setup, not an active signal. A bounded optimization was requested, but no robust setup was established (the sensitivity evaluation exceeded its time budget), so the entry conditions stand as written and the wait is simply for the market to meet them.

#### CRAK price and trigger map

Uses the idea timeframe and keeps price levels on the price axis.

| Measure | Value |
| --- | ---: |
| Ticker | CRAK |
| Timeframe | 1d |

#### MPC price and trigger map

Uses the idea timeframe and keeps price levels on the price axis.

| Measure | Value |
| --- | ---: |
| Ticker | MPC |
| Timeframe | 1d |

#### VLO price and trigger map

Uses the idea timeframe and keeps price levels on the price axis.

| Measure | Value |
| --- | ---: |
| Ticker | VLO |
| Timeframe | 1d |

### Refining economics are doing exactly what the thesis predicts — and Q2 2026 proves it

The core thesis — that the profit pool migrates to refiners when crude is cheap but diesel and gasoline are scarce — is being confirmed by the latest reported quarter, not just by the Bloomberg and Yahoo Finance pieces on record diesel and the Saudi pipeline outage. Valero's gross margin nearly doubled from 6.3% in the quarter ended March 31, 2026 to 12.3% in the quarter ended June 30, 2026. Net margin more than doubled from 3.9% to 8.4%, and return on equity jumped from 5.3% to 14.9% — a +9.6 percentage point swing in a single quarter. That is the margin-widening the thesis calls for, and it happened inside the last two reported quarters, not years ago. The cash generation backs it up. Marathon Petroleum posted $4.0B in net income for the fiscal year ended December 31, 2025, with a return on equity of 23.4% — placing it in roughly the 87th percentile of 129 Energy-sector peers for that metric. Its free cash flow for the most recent quarter (ended June 30, 2026) came in at $9.1B, a sharp rebound from $208M in the prior quarter. For a thesis premised on bottleneck economics driving outsized cash returns to the refining layer, that trajectory is the single most decision-relevant datapoint in the file. The idea's direction is deliberately not a naive buy-at-any-price momentum play. The compiled entry waits for a pullback-and-reclaim: price below the 20-day average with the stochastic oscillator oversold, then a close back above the 20-day average with a retracement-level low. On the supplied daily bars, that combination has not fired across 1235 evaluated bars over the last 60 months (nor across the 493-bar 24-month or 246-bar 12-month windows). That is a patience requirement, not a defect — and it means the reader is looking at a watch-list setup rather than an already-consumed signal. The trade structure caps risk at roughly 2.3% per position against a take-profit near 4.6%, so the reward-to-risk asymmetry is baked into the setup rather than assumed. The refining-capacity constraint the thesis leans on is corroborated by the cited news: a Bloomberg piece from September 13, 2026 titled "Refining Bottleneck Sends Diesel to Record High" and a Yahoo Finance piece the same day framing the Saudi pipeline outage as a worsening threat to gas prices. Industry experts quoted there say no one can quickly…

#### MPC Free cash flow

Free cash flow trend from CommonQuant fundamentals/XBRL data; +486.2% from first to latest point.

| Measure | Value |
| --- | ---: |
| 2009-12-31 | $-436000000 |
| 2010-06-30 | $-339000000 |
| 2010-09-30 | $-631000000 |
| 2010-09-30 | $-292000000 |
| 2010-12-31 | $1000000000 |
| 2011-03-31 | $672000000 |
| 2011-06-30 | $435000000 |
| 2011-06-30 | $-237000000 |
| 2011-09-30 | $1827000000 |
| 2011-09-30 | $1392000000 |
| 2011-09-30 | $1684000000 |
| Latest Value | $1684000000 |
| Change Pct | $486.23853211009174 |
| Ticker | MPC |
| Timeframe | reported periods |

#### VLO Gross margin

Gross margin trend from CommonQuant fundamentals/XBRL data; +410.1% from first to latest point.

| Measure | Value |
| --- | ---: |
| 2016-12-31 | \0.011115663701608531% |
| 2017-03-31 | \0.033621164798824175% |
| 2017-06-30 | \0.023826829600690495% |
| 2017-06-30 | \0.04713759324166442% |
| 2017-09-30 | \0.0238799786944428% |
| 2017-09-30 | \0.06850012732365673% |
| 2017-12-31 | \0.011811023622047246% |
| 2017-12-31 | \-0.08657926644437709% |
| 2018-03-31 | \0.04016793373425621% |
| 2018-03-31 | \0.05669639602420472% |
| Latest Value | \0.05669639602420472% |
| Change Pct | \410.05857631335374% |
| Ticker | VLO |
| Timeframe | reported periods |

#### MPC sector percentile check

Ranks MPC against 95 companies in its sector using CommonQuant fundamentals.

| Measure | Value |
| --- | ---: |
| Free cash flow | \96.84210526315788th percentile |
| Return on equity | \86.82170542635659th percentile |
| Gross margin | \13.432835820895525th percentile |
| Revenue growth (YoY) | \40.909090909090914th percentile |
| Ticker | MPC |
| Sector | Energy |
| Peer Count | 95 |

### Scores

- **Conviction score breakdown:** 46
- **Thesis support:** 72
- **Trade readiness:** 30
- **Risk quality:** 55
- **Trigger proximity:** 10
- **Fundamentals trend:** 65

### Watch items

- **CRAK — Stochastic (14) on CRAK**
- **CRAK — CRAK close vs 20-day EMA**
- **MPC — Stochastic (14) on MPC**
- **MPC — MPC close vs 20-day EMA**
- **VLO — Stochastic (14) on VLO**
- **VLO — VLO close vs 20-day EMA**
- **MPC — MPC insider net open-market activity**
- **VLO — VLO insider net open-market activity**
- **VLO — VLO gross margin (Q2, %)**
- **CRAK — CRAK annual dividend per share**

## Key details

- Symbols: CRAK, MPC, VLO
- Timeframes: 1d
- Tags: \#canonical-demand, \#cluster-version:1, \#direction:bullish, \#entity-kind:instrument, \#entity:CRAK, \#entity:MPC, \#entity:VLO, \#horizon:unspecified, \#intent:research, \#symbol:CRAK, \#symbol:MPC, \#symbol:VLO

## Community

- Upvotes: 1
- Views: 0
- Copies: 0
- Cosigns: 0

## News sources

- [Refining Bottleneck Sends Diesel to Record High](https://www.bloomberg.com/news/videos/2026-09-13/refining-bottleneck-sends-diesel-to-record-high-video) — Bloomberg
- [From Bad to Catastrophic: Saudi Pipeline Outage Threatens to Make Gas Prices Even More Brutal](https://finance.yahoo.com/energy/articles/bad-catastrophic-saudi-pipeline-outage-153110199.html) — Yahoo Finance

## Related

- [CRAK trade ideas](https://commonquant.ai/stocks/crak)
- [MPC trade ideas](https://commonquant.ai/stocks/mpc)
- [VLO trade ideas](https://commonquant.ai/stocks/vlo)
- [Latest market news](https://commonquant.ai/news)

## About CommonQuant Research

Ideas and Advanced Analysis are generated with fresh, LLM-selected news headlines and grounded in SEC XBRL fundamentals and peer percentiles. Strategies built from these ideas are automatically backtested (12-month and 5-year, walk-forward tuned) and stress-tested before they can go live, then monitored against the news hourly while they run.
