Actionable insights — A "Bridge to Nowhere" Fuel

1. Screen for the gap between a commodity's physical standing and its financial standing

1. Pick the commodity and establish its physical share of the job it does — a hard, checkable, single number from official data (share of generation, share of transport fuel, share of a feedstock). Not a narrative about the future; the fraction of the work it is doing today.
2. Anchor that number against the substitute everyone assumes is replacing it, in the same unit. A ratio ("nearly double all renewables combined") does more work than either level alone, because it pre-empts the "but it's being displaced" reply.
3. Separately measure how investors are positioned — a bullish-sentiment or positioning series, stated as a percentile of its own history with the obvious outlier carved out ("lowest this decade, outside of Covid"). Carving out the outlier is what stops the reader dismissing the extreme as a known crisis.
4. Declare a candidate only where those two readings point in opposite directions: indispensable in the physical world, un-owned in the financial one. Symmetry matters — an indispensable commodity that everyone already loves is not a candidate, and a hated commodity doing no real work is a value trap.
5. Name the mispricing in one sentence you could be held to. The label is the falsifiable claim: the market is pricing this as X; here is why it is not X.
6. Only then go looking for the equities. The screen identifies the commodity gap; the security selection is a separate step (insights 3-6) and must not be smuggled in here.

• The bullishness series turning up off the low — the signal is the extreme, and a screen like this expires quietly as positioning normalises without any price move. Watch too for the physical share slipping: if gas's share of generation starts falling materially, the whole left-hand side of the gap erodes.

2. Deflate the price before you call a commodity cheap

1. Never assess a commodity price in nominal terms across more than a couple of years. Convert to real (inflation-adjusted) terms first — the cost of drilling, steel, labour and capital has moved, so a flat nominal price is a falling real one.
2. Express the result as a percentile of its own history over a stated window, rather than as a level. A percentile is comparable across commodities and immune to arguments about what "normal" is.
3. Choose the window to include at least one full cycle and say where it starts, so the reader can check whether the start date flatters the conclusion.
4. Read the real percentile as a statement about the producer's economics: a bottom-quintile real price means the marginal barrel or Mcf is being sold near or below its true replacement cost, which is a supply-side argument, not a sentiment one.
5. Use it as corroboration of the positioning extreme, not as a substitute. Positioning tells you what people believe; the real price tells you what the industry is being paid. Two independent readings agreeing is the point.

• The real price climbing out of the bottom quintile while sentiment stays depressed — that is the sweet spot, and the sequence in which these mean-revert (price first, sentiment later) is what leaves the equities behind and creates the entry.

3. Distinguish a bottleneck discount from a value discount — they have different cures

1. When an asset's realised price is far below the benchmark price, ask why before treating it as cheapness. The usual answers are: quality (it is worth less), demand (nobody wants it), or logistics (it cannot get to anyone who does).
2. Test for logistics with the tell-tale signature: prices that go negative, and physical product being destroyed rather than sold. Both are irrational for any product that can reach a market, so their presence localises the problem to transport.
3. Ask whether the excess supply is involuntary. Associated gas comes up with the oil whether the operator wants it or not, so producers cannot fix the glut by cutting output — which is exactly why the discount persists past the point a normal market would clear it.
4. Note the crucial asymmetry: a value discount needs the market to change its mind, while a bottleneck discount needs a pipe to be finished. The second has a construction schedule; the first has no date at all.
5. Quantify the trapped volume as a share of the national market, so the fix is sized. A basin curiosity and 12½% of national output are different investments.

• Regional basis differentials (Waha vs Henry Hub) narrowing as lines come into service — the single cleanest, publicly-observable confirmation that the thesis is paying. Also watch flaring-volume disclosures, which fall as the discount closes.

4. Prefer catalysts with a construction schedule over catalysts that need a re-rating

1. For a contrarian idea, list the possible catalysts and sort them by whether anything physical has to happen. Sentiment shifts, multiple expansion and "the market realising" have no timetable; a pipeline, a terminal or a plant does.
2. Verify the projects are under construction, not proposed, permitted or announced. The distinction is the whole value of this catalyst class — capital already sunk by a third party, on a schedule that does not depend on your view.
3. Confirm the counterparty is spending its own money for its own reasons. Midstream builds because the tariff economics work; the producer gets the benefit without funding it, so the catalyst survives a producer capex cut.
4. Check the demand at the far end will exist when the pipe opens — otherwise you have solved the logistics problem into a glut. State the demand claim explicitly.
5. Use the completion window to set the holding period, and hold yourself to it. This is the discipline the method buys you: a dated thesis can be marked wrong, which an "eventually the market will see it" thesis never can.

• In-service dates slipping, and the second-order risk that all the lines land at once — a step-change in delivered supply can depress the destination price even as it lifts the source price, so track where the marginal molecule ends up, not just that it moves.

5. Choose the expression whose payoff does not require your price forecast to be right

1. Having identified a mispriced commodity, list the ways to own it and ask of each: what has to happen for this to work? A pure producer needs the benchmark price to rise. That is a forecast.
2. Look instead for a beneficiary whose gain comes from a structural change in what it receives — closing a discount, removing a cost, monetising something currently wasted — at an unchanged benchmark price.
3. Prefer gains that need no incremental capital: volumes already being produced, now sold at a real price rather than flared or given away, drop through with no new drilling and no dilution.
4. Frame the result as an earnings event, because that is what gets modelled and re-rated by others, rather than as a commodity view, which does not.
5. Then treat the commodity upside as free optionality stacked on top — depressed sentiment and a bottom-quintile real price pay you extra if they mean-revert, and cost you nothing if they don't.
6. Sanity-check the downside symmetrically: if the pipe opens and the price falls, does the basis relief still cover it? That is the question that decides whether this is genuinely forecast-free.

• Quarterly realised-price disclosures and the gap between realisations and the benchmark. If realisations converge on the benchmark, the kicker is arriving regardless of what the benchmark itself does — the exact thing to verify before adding.

6. Check that the named beneficiaries actually sit where the catalyst lands

1. After a thesis names several companies, map each one to the specific geography or mechanism of the catalyst. A shared commodity is not a shared catalyst.
2. Split the names into those that receive the datable, mechanical benefit and those that only receive the general one (sentiment, demand, the commodity price). Size and rate the two groups differently.
3. Ask whether the fix is neutral or adverse for the second group. Unlocking stranded supply helps the basin that was trapped and adds competing volume for everyone else — an honest thesis states that tension rather than letting the shared headline paper over it.
4. Read what the author actually rated. A name given a mechanism, a chart and a "likely beneficiary" line is an argued call; names appearing in a list of who has suffered are context, however familiar.
5. Where the second group is already held, treat the piece as a reiteration of standing conviction, not as a fresh entry — and don't let the freshness of the argument re-rate an old position.

• Whether the next Buy-list update rates any of the three, and at what level — that is where a "likely beneficiary" either becomes a position or stays commentary. Also watch Appalachian basis once the Permian lines fill, as the read-across for the names that are not the beneficiary.

7. Keep your objection to a boom separate from your read on what the boom must buy

1. When you are bearish on a spending cycle, state precisely which leg you object to: the financing, the accounting, the valuations, or the underlying physical demand. These fail independently and on different timetables.
2. Isolate the claims on the cycle that are physically settled and paid in cash — the electricity, the fuel, the concrete, the copper. Those get paid whether or not the accounting holds up, and typically before the equity does.
3. Use the boom's own capex figures as the demand estimate for those inputs, even while disputing the returns on that capex. Consistency requires only that you not use the same number as evidence for both claims.
4. Position accordingly: short or avoid the leg you dispute, own the input the cycle cannot avoid buying. This is what makes a bearish macro view actionable rather than merely defensive.
5. State the hedge honestly — if the spending is cut, the input demand falls too. The claim is seniority in the capital stack of the boom, not immunity from it.

• A genuine capex cut — the one event that collapses both legs at once, and (per Jul-26) the one Hay himself flags as potentially sending a hyperscaler's stock "straight up." That would be bullish the disputed leg and bearish this one, so it is the specific scenario in which this trade and the house's market view stop agreeing.