Actionable insights — The Diesel Squeeze

1. Read the derived product, not the headline commodity

1. For any commodity with a headline price everyone quotes, identify the downstream product the economy actually consumes — the thing that gets bought after processing (crude → diesel; gas → power; ore → refined metal; hogs → pork).
2. Track the spread between the two — the processor's margin. The spread, not the raw input, is where a processing constraint shows up, and processing capacity is the slowest thing in any commodity chain to add.
3. Establish the spread's long-term normal range first, so a move can be measured in multiples rather than described as "high."
4. Escalate only on a settle outside that range, not a print. A spike is a dislocation; a spread that stays outside its range is a structural signal.
5. Confirm with the trend, not the level — a rising long moving average (200-day) says the constraint is still tightening.

• The crack spread's weekly settle versus its $15–$30 band; the 200-day slope; and any other processing spread breaking range (refining margins, smelter treatment charges, power spark spreads, crush spreads).

2. Diagnose which link is constrained — a bottleneck is not a shortage

1. When the downstream price runs away from the upstream one, do not call it a shortage of the raw material. Ask which single link in the chain — extraction, processing, transport, storage — is the binding constraint.
2. Test with inventories at the constrained link: if stocks of the finished product are falling while the processors are running flat out, the constraint is in processing, not supply of the input.
3. Test with spare capacity: locate where the world's idle processing capacity actually sits, then ask what policy (quota, export ban, sanction) keeps it off the market. Idle capacity that cannot legally reach the market is not spare capacity.
4. Reject fixes aimed at the wrong link. New input supply does nothing for a processing bottleneck — and check the lead time before crediting it at all.

• Weekly EIA distillate stocks versus the 5-year seasonal band; refinery utilisation running high with falling product stocks (the signature of a bottleneck); changes to China's refined-product export quotas; further strikes on Russian/Ukrainian or Middle East refining assets.

3. Score the constraint's durability with a driver checklist before sizing anything

1. Enumerate the drivers that could produce the tightness. The piece's four: (1) geopolitical disruption — capacity destroyed or exports banned; (2) very low inventories; (3) seasonal demand peaks (e.g. harvest); (4) inability to ramp capacity quickly.
2. Mark each as transient or structural. Seasonality reverses on a calendar; destroyed capacity and multi-year build times do not.
3. Count how many are live simultaneously. One transient driver is a trade; several structural drivers overlapping is a regime.
4. Weight the capacity-ramp driver heaviest — it is the one that sets how long the other three can persist. "You simply cannot magic up new refining capacity overnight."
5. Add adjacent-system stress before concluding. A constraint in one energy vector raises the value of the others, so check whether the neighbouring markets are tight too.

• EU storage percentage entering the withdrawal season; TTF forward curve; announcements of new refining capacity (and their commissioning dates, not their press dates); data-centre interconnection queues adding backup-generation demand.

4. Stress-test your own headline number — separate the regional artefact from the global signal

1. Before building a thesis on a record print, seek out the strongest deflationary explanation for it and state it in full, from a credible source who disagrees.
2. Ask specifically whether the extreme is regional — a local refining or pipeline dislocation — rather than a system-wide constraint. Physical markets are chronically local.
3. If the caveat holds, demote the single print and promote the trend: rebuild the argument on the moving average and the settle, both of which survive a one-day regional squeeze.
4. Keep the caveat in the write-up. A thesis that has already absorbed its best counter-argument is the one that survives contact with a mean-reverting week.

• Whether the spread narrows once the named regional dislocation clears (a pipeline restart, a refinery back from turnaround) — that is the test of whether the caveat or the trend was right.

5. Trace the squeeze into the cost line of what you already own

1. For each holding, ask what share of its operating cost the squeezed input represents. Put a number on it — a percentage of the cost metric the sector actually reports.
2. Check whether that number is already showing up in company guidance. Guidance revisions are the confirmation that the macro spread has reached the micro P&L.
3. Then ask the second-order question: does the output price rise faster than the input cost? Cost inflation is only a thesis-breaker if it outruns the revenue line.
4. Look for the supply-side kicker: higher costs make marginal new projects harder to sanction and stretch lead times — which reinforces the very scarcity driving the output price. In a supercycle, the cost squeeze is partly self-cancelling.
5. Use the answer to screen, not to exit: prefer producers whose realised price gains outrun the cost pass-through.

• AISC guidance revisions in quarterly results; the gap between realised metal price growth and AISC growth (the margin, not the cost, is the signal); project sanction decisions deferred on cost grounds — bullish for the commodity, bearish for that specific developer.

6. Verify the official inflation print against physical prices you can observe yourself

1. Do not take CPI/PPI/PCE as the measurement of what is happening. Assemble a small basket of directly observable physical prices in the affected chain and track them independently.
2. Prefer prices with a public, high-frequency, un-hedonic series: retail fuel per gallon, cereal and soft-commodity futures, freight rates.
3. Look at the dispersion, not just the average — a state-by-state or region-by-region spread reveals whether the pressure is broad or pocketed.
4. Note the decoupling explicitly: a wide processing spread can keep consumer prices rising even while the headline input falls. That is exactly the case an aggregate index will understate.

• Retail diesel and gasoline versus crude (the decoupling test); soft-commodity futures against weather/El Niño forecasts; sovereign import decisions (India's sugar buying) as confirmation that a shortage is real rather than speculative.

7. Classify the inflation before predicting the policy response

1. Decide whether the inflation is cost-push (higher input costs) or demand-pull (excess demand). The two look identical in the index and imply opposite policy efficacy.
2. If cost-push, note that the standard tool cannot reach the cause: rates can suppress demand, but no policy rate refines a barrel of diesel or rebuilds an inventory.
3. Enumerate the policymaker's full option set and eliminate on political feasibility — raise taxes, cut spending, default, or inflate. What survives elimination is the forecast.
4. Convert that into a near-term call and a direction of travel: what they will do at the next meetings, and what the constraint forces over the cycle.
5. Read the bond market as the confirming instrument: persistent cost-push keeps nominal yields higher for longer, which itself worsens the fiscal arithmetic and narrows the option set further.

• Which component drives the CPI beat — energy/food/transport (cost-push) versus services/wages (demand-pull); the long end of the curve steepening on inflation-plus-fiscal rather than growth; official language shifting from "transitory supply factors" toward tolerance of an overshoot.

8. Don't let a nominal yield veto a real-asset position — check the real rate and the liquidity path

1. When "high bond yields are bad for gold" is offered as a reason not to own it, restate it precisely: the claim only holds for real yields, not nominal ones.
2. Ask which of the two the inflation regime is actually pushing. Cost-push inflation raises nominal yields and the price level — so the real yield can fall while the nominal one rises.
3. Add the liquidity path: an economy run hot raises the probability of QE, balance-sheet expansion and other support, all of which accelerate debasement independent of where the policy rate sits.
4. Then check the demand base: buyers seeking protection from an eroding currency do not price off the yield curve at all.
5. Conclude only on the conjunction — real yields falling or liquidity expanding is sufficient; you do not need both.

• The 10-year real (TIPS) yield versus the nominal; any resumption of balance-sheet growth or bill-buying; central-bank and retail physical demand series holding through nominal-yield spikes. Cross-check against the same conclusion reached monetarily on 2026-AUG-30.

9. Fix the balance sheet before the security selection

1. Treat a supply-driven inflation regime as a liquidity problem before an allocation problem — the cost shock reaches your household budget on the same schedule it reaches the P&L.
2. Clear bad debt first: floating-rate liabilities compound the damage in a higher-for-longer nominal regime.
3. Hold a defined liquidity reserve — at least six months of cash or liquid assets — sized in months of spending, not as a percentage of the portfolio, so it scales with the very costs that are rising.
4. Hold the insurance leg as physical gold and silver, distinguished from any producer exposure; it is there for debasement, not for return.
5. Only then take the producer risk, screened by insight 5 — well-capitalised names in genuinely supply-constrained areas.
6. Extend the same contingency thinking beyond the portfolio (a "Plan B" — second residency or passport) where the risk being hedged is jurisdictional rather than financial.

• Your own reserve measured in months of current (inflated) spending rather than last year's; floating-rate exposure repricing; the gap between the physical insurance leg and the producer leg drifting as the producers outperform.