← Analysis page  ·  Rory Johnston hub  ·  Research hub

Actionable insights — the most unbalanced oil market ever

How a barrel counter checks the numbers: measure a reroute by its incremental exports, size a price spike from the gap demand has to close, test "demand destruction" against end-use activity, keep a deficit separate from a smaller deficit, follow a policy shock to its second-order effects, and read speculative positioning as a risk skew.
2026-SEP-18 · Contrarian Codex interview (Patreon) · Rory Johnston · ▶ Watch (Patreon) · full analysis · transcript
How to read this page: each insight is a method Johnston actually used in this interview, with the boxed line showing how it played out. No securities were named, so these are commodity-market methods. The Patreon video has no timestamps, so nothing is deep-linked.

1. Measure a reroute by its incremental exports, not its nameplate capacity

The repeatable method
  1. When a bypass route is said to "offset" a lost chokepoint, start with its nameplate capacity, then subtract what it already carried before the disruption.
  2. Confirm the result with the change in net exports at the bypass terminal (pre-disruption versus now), not with pipeline capacity.
  3. Check whether some of the "recovered" chokepoint flow is the same producer's volume switching back between routes. That is reshuffling, not new supply.
Here: East-West/Petroline is 7 mb/d nameplate, but Red Sea net exports only went from ~2 to ~5.5 mb/d, a ~3.5 mb/d swing. After the Houthi ban, part of Hormuz's rise to ~12 mb/d was Saudi volume moving back from the Red Sea to the Gulf.
Watch for

2. Check a flow number's window and what is driving it

The repeatable method
  1. Before accepting a headline flow figure, ask which window it covers (7-day average? a 48-hour burst?).
  2. Separate one-off releases (captive cargoes leaving at once) from what fresh loadings can sustain. Leave the one-off out of your comparisons.
  3. Judge an official claim against what was flowing on the date it was made. Hitting the number later does not make it true at the time.
Here: the post-MOU >15 mb/d spike was 150m+ bbl of trapped cargo leaving at once; today's 11.98 mb/d 7-day high is backed by 10+ mb/d of fresh loadings (~60% of pre-war). Chris Wright's earlier "10 mb/d" was an exaggeration when he said it.
Watch for

3. Estimate a price ceiling from how much demand has to disappear

The repeatable method
  1. Take the gross supply loss and subtract every known offset (reroutes, strategic-reserve releases). What remains is the demand that price has to destroy.
  2. Compare that gap with the deficits behind earlier price peaks and the prices those peaks reached.
  3. If the gap is several times larger than any earlier one, the price needed to destroy that demand is well above past peaks. Then decide whether crude or refined products will do the rationing, because product prices are what actually cut demand.
Here: 8–10 mb/d left after all offsets (COVID-peak scale) versus the 2–3 mb/d deficits behind past $130–150 peaks gave his $200 call. The missing variable was China's discretionary import cut.
Watch for

4. Test "demand destruction" against end-use activity, product by product

The repeatable method
  1. When a large importer's apparent demand collapses, first strip out what is simply the mirror image of lost exports from its suppliers.
  2. Compare it with peers on the same supply shock. If they recover and it keeps falling, the move is a choice, not a shortage.
  3. Build a balance for each product and set each one against its end-use activity data (jet vs flights; diesel vs road freight, ~70% of diesel use).
  4. When activity rises while apparent demand falls sharply, the gap is an unseen inventory draw. That draw is finite, so it predicts a return to buying.
Here: China cut imports 5.4 mb/d while the rest of Asia recovered; road freight was up >3.5% YoY but apparent diesel demand was down 20%. His estimate: ~100m bbl commercial + ~100m strategic crude drawn, with diesel similar. China is now back as a buyer.
Watch for

5. A smaller deficit is still a deficit — follow inventories, not the change

The repeatable method
  1. Separate the sign of the balance from its day-to-day change. While it is negative, stocks keep falling.
  2. Check physical benchmarks against futures (Dated Brent vs prompt futures) for how tight the physical market really is.
  3. Expect tightening to continue until the balance actually turns to surplus, then expect a sharp repricing toward the surplus level.
Here: prompt Brent $107.50 vs Dated Brent $131.77, near April's highs; "the market will continue tightening until we actually get into an actual surplus," then $50 or even $40 by mid-to-late 2027.
Watch for

6. Follow an export ban through to its second-order effects

The repeatable method
  1. Size the exports the ban removes against the size of that product's market, not the whole oil market (diesel is ~1/4–1/5 of crude, so the impact is 4–5x larger in proportion).
  2. Trace what happens at home: the barrels that can no longer leave build up in storage, local prices and cracks collapse, refiners cut runs, and output of other products falls.
  3. Name who loses indirectly: suppliers whose main buyers are the refiners that will cut runs.
  4. Check the political timing: a measure meant to help before an election must land early enough to show up in pump prices.
Here: a US diesel export ban would remove ~1–1.5 mb/d from global diesel, send US Gulf Coast cracks negative while global cracks are above $100, possibly force Gulf Coast gasoline imports, and crush Western Canadian crude values. It would need to happen within ~3 weeks to matter before the midterms.
Watch for

7. Read speculative positioning as a skew in risk, not a forecast

The repeatable method
  1. Each week, pull CFTC Commitments of Traders across all the major contracts for a commodity (for crude: ICE and CME Brent, WTI, Dubai), not just one.
  2. Scale net speculative position as a % of open interest so readings compare across time.
  3. High longs with low shorts means little buying power left and a lot to liquidate: expect sharper drops than rallies, whatever the fundamentals say.
  4. Expect the drop to need a trigger (any fundamental headline). The flush resets positioning, and then the fundamental trend can resume.
  5. Near political pain levels ($110-ish Brent), expect jawboning to add to downside volatility.
Here: crude net spec ~4.5% of OI, near April's highs; diesel ~9%, the highest since 2024; gasoline also stretched. He expects a $10–15 flush (e.g. on a partial East-West restart), then a more sustainable rise through year-end.
Watch for

Methods distilled from the members-only Contrarian Codex Patreon interview with Rory Johnston (Commodity Context) for personal study. Not investment advice.