# Washington Buys Into Venezuelan Oil: A Very Big Deal That May Not Matter Much

> Notes from a Columbia Energy Exchange conversation on shortcuts versus institutions, and how to read a headline when the investment cycle outlasts the political one. Educational only, not investment advice; no tickers or price targets.

Published: 2026-09-09
Locale: en
Tags: energy, geopolitics, national oil companies, reading the news, Venezuela


![Dusk over an aging Lake Maracaibo oil field, corroded pumpjacks standing among shallow-water pilings in the foreground, the water receding toward the lights of a new drilling rig far away](/covers/columbiaenergy-2026-09-08-luisa-palacios-and-bob-mcnally-on-why-the-us-venez-cover.png)

> Things have their root and their branches; affairs have their beginning and their end. To know what comes first and what comes last brings one near the Way.
>
> —— *The Great Learning*, in the *Book of Rites* (pre-Qin China; translation mine)

## What the episode is about

On the 8 September 2026 episode of Columbia Energy Exchange, host Jason Bordoff sat down with two friends of the house: Luisa Palacios, former chairwoman of Citgo and a scholar of national oil companies, and Bob McNally, founder of Rapidan Energy, author of *Crude Volatility*, and a veteran of the White House NEC and NSC.

The subject was the late-August announcement: a deal involving some 65 billion barrels of Venezuelan reserves and a 35% U.S. government equity stake in a company called North American Blue Energy Partners (NABEP). In the same week, Chevron pledged $7 billion to double its Venezuelan output, Italy's ENI advanced a giant heavy-oil field agreement, and GE Vernova signed on to repair the electricity grid. All told, Venezuela signed away development rights over more than 150 billion barrels of resources in seven days.

The two guests spent an hour neither cheering nor dismissing. They took the deal apart and asked what the terms are, what law it rests on, and who is carrying which risk. What I took away was a demonstration of how to reason with structure when the facts are still incomplete.

## The main points

**One: the two announcements do not match, and the mismatch sits on the core terms.** The White House fact sheet says a hundred-year oil lease. The Venezuelan government describes a 25-year contract covering 17 fields. Luisa's point is that neither "lease" nor "concession" exists in Venezuelan law — concessions ended with the 1975 nationalization. What the law permits is a joint venture or production-sharing arrangement with PDVSA or a state entity, capped at 25 years and renewable. So the "hundred years" is not a negotiating win; it is a word with no counterpart in the other side's legal system.

![Two horizontal bars of very different lengths: the top one is labeled one hundred years, the bottom one is a quarter as long and labeled twenty-five years, and below them a crossed-out dashed empty box stands for a contract form that does not exist in the other side's law.](/figures/venezuela-lease-term-mismatch-en.svg)

**Two: nobody can say who holds the equity on the U.S. side.** Bob, who disclaims being either a lawyer or a financial engineer, notes that the American counterparty is the Pentagon's Office of Strategic Capital, whose statute says it does not take equity stakes, and whose representatives have said so publicly. The workaround circulating is a penny warrant — not equity today, an instrument convertible to equity later. His question stands unanswered: who holds it, under what authority? Congress will not ratify this, and the documents carry a provisional expiry at 12:01 p.m. on 20 January 2029.

**Three: the actual barrels are somewhere else.** This was the part I most wanted to write down. Luisa tallied the same week's other deals: Chevron, $7 billion over five years; ENI, a 35-billion-barrel heavy-oil field with production expected by year-end; Colombia's GeoPark, roughly 16 billion barrels. Together those exceed the reserves signed to NABEP — and they are publicly listed operators with balance sheets, track records and disclosed accounts. NABEP was, two weeks ago, a small firm known to a handful of people inside Venezuela's oil sector. Half of its 17 fields sit in the Orinoco belt with no infrastructure whatsoever.

![Two bars compared by height: the left one is a single block of reserves awarded to an unlisted small company, while the right one is stacked in layers from two listed companies and extends upward with a dashed line past the left bar.](/figures/reserves-vs-operators-en.svg)

**Four: ask whose risk is being de-risked.** The official story is that American involvement makes Venezuelan investment bearable. Luisa's read: the arrangement de-risks NABEP, not Venezuela. She worries it may stall the work that matters — amending the hydrocarbons law, writing the regulations, standing up a regulator — after which investment arrives on its own. Her line, repeated three times across the hour: there are no shortcuts.

![Two paired before-and-after risk bars: the company bar shrinks sharply while the country bar barely moves, and three unchecked boxes below list the institutional work that has not yet been done.](/figures/derisk-for-whom-en.svg)

**Five: a 35% stake buys liabilities, not just assets.** A government agency taking a minority position in another country's extractive industry inherits that company's governance, operational and environmental record. NABEP operates around Lake Maracaibo, a body of water already wrecked by decades of spills. Luisa acknowledged the follow-up White House notes — U.S. veto over the board, an American majority on it, U.S. law governing — and answered that these are the floor, not the ceiling. When you are the most important operator in a country, you publish your investment and production plans. And if NABEP is a vehicle rather than an operator, then the outsourcing arrangements need naming: to whom, chosen how, under what process.

**Six: the pendulum bites from both sides.** Bob's prediction: this brand of state capitalism does not survive the current president inside the Republican party, while it gains favour among Democrats — who will apply it to different companies, different sectors, different reasons. He mentioned a joke tweet about AOC and Greta Thunberg discovering that the U.S. government can direct what oil company boards do. Jokes aside, once a precedent exists, the tool stays on the shelf for whoever comes next. What he has told clients for years: the largest political risk to oil companies is not climate policy, it is a cash-hungry government looking at their balance sheets.

**Seven: stretch the timeline and this ranks low.** Bob's word was "nothing burger" — for this year and the next few, the deal barely moves oil balances. Luisa agrees she sees no meaningful production relief in one to two years. Venezuela currently produces around 1.2 million barrels a day, up from about 800,000. What actually draws capital is elsewhere: the IEA has been walking back the peak-demand-by-2030 narrative that many people invested against for five years, and as it recedes, an underinvestment gap in the early 2030s comes into view. Bob's framing: we are going to need those two to three million barrels a day in 2033.

## Going further

### "Such a big headline — why shouldn't I just follow it?"

Four announcements in one week, comparable in headline size, an order of magnitude apart in executability.

The way to take them apart is not to judge good or bad, but to ask three questions: **who signed, who operates, who pays.** For Chevron, the answer to all three is Chevron, funded off its own balance sheet. For NABEP, the signatory is a company with no published financials, the operating capability is unknown, and the money is a projected $100 billion with no named source.

![A three-row, two-column comparison table where all three cells in the left column are filled with the same party and all three in the right column are question marks, making the gap between filled and empty obvious.](/figures/three-positions-check-en.svg)

The mistake I have made is reading an announced number as a committed number. For a figure to become barrels, someone must have the capital, the technical capability and the operating record. Luisa's Orinoco point gives a concrete test: greenfield development needs upfront infrastructure, and infrastructure needs a balance sheet, not enthusiasm.

Next time a headline of this size lands, I plan to check whether those three slots can be filled. If they can't, it goes on a watchlist rather than into a position.

### "Government backing — doesn't that make it safer?"

Bob's line made me laugh: having Uncle Sam sit at the table as your business partner is not necessarily the thing you want.

His reasoning: the intent is to reduce risk, but the arrangement introduces new uncertainty — an unprecedented structure, legal exposure, political controversy, public scrutiny. De-risking and up-risking at once. Which is why Chevron, Exxon and Conoco, companies that have gone in and out of Venezuela for a century, would not want this. Not timidity — they know what success looks like and what risk looks like.

The reasoning travels. When a thesis rests mainly on official backing, that backing is itself a variable. Backers change, get challenged, have terms. Policy support and operating capability are different species: one lets you sleep tonight, the other keeps you around in three years.

The test I would write down: **remove the policy-support line from the thesis — does the rest still stand?** If it does, policy is a bonus. If it doesn't, policy is the foundation, and this particular foundation gets voted on.

![Two stacks of blocks: on the left, pulling out the bottom policy block topples the three blocks above it; on the right, pulling out the top policy block leaves the three below still standing.](/figures/policy-as-foundation-or-bonus-en.svg)

### "I invest over years; policy only runs to the next election"

The sharpest detail in the episode is a timestamp written into the documents: 12:01 p.m., 20 January 2029.

Bob's read is blunt: even if the structure survives legal review, a Democratic president can end it with a signature. A Republican successor would likely want to preserve the broader relationship, but might not spend political capital defending this particular deal. Either way the odds look stacked against it. Meanwhile, field development runs twenty years, and heavy oil runs slower still.

![Two bars of very different lengths on the same timeline: the contract validity ends in early 2029, while the field development cycle runs twenty years and extends further with a dashed line.](/figures/contract-clock-vs-field-clock-en.svg)

Listening to that, I thought about the "policy beneficiary" names in my own portfolio. The mismatch between investment horizons and political ones is structural; it does not go away when a different party wins. Since it can't be removed, it can at least be made explicit: how long is this thesis valid for, and which leg fails first if the government changes?

Luisa supplies the other half. What long-term capital needs is rule of law, government capacity, a working electricity system, a legitimate government — slow to build, and once built, nobody dismantles them on day one. Bob agreed: what he wants is a government still standing after 2028 and on speaking terms with both parties in Washington. That is the real guarantee that commitments get honoured.

Same idea in a portfolio: **the durability of a moat depends on whether it rests on institutions or on relationships.** Relationship-based advantages restart when the relationship changes hands. Institution-based ones survive a transition. When I look at a company's long-run advantage, I try to sort it into one of those two piles — and anything that won't go into the institutions pile doesn't get a long valuation horizon from me.

## Worth a look

- Columbia Energy Exchange, episode of 8 September 2026 (Center on Global Energy Policy, Columbia University; energypolicy.columbia.edu)
- Bob McNally, *Crude Volatility: The History and the Future of Boom-Bust Oil Prices* (2017) — the starting point for the logic behind oil's booms and busts
- Venezuela's 1975 nationalization law and the recently amended hydrocarbons law: the 25-year cap, and the tax and royalty brackets for greenfield investment, are all in the text
- The IEA's recent revisions to the peak-demand narrative, and the upstream underinvestment discussion around it
- The century of U.S.–Venezuela oil history Bob sketched: Chevron and Exxon arriving in the 1920s, Juan Pérez Alfonzo pushing fifty-fifty profit sharing, studying the Texas Railroad Commission's quota system while in exile, and going home to help found OPEC. Any one of those threads teaches more than a week of headlines.

## One thing to take with you

One idea: **tell apart what rests on institutions and what rests on relationships.**

Institutions are what survives a change of personnel — rules, processes, terms written down and enforced. Relationships are attached to a particular person: smooth while they're there, back to zero when they leave. Both move things forward; their durability differs by an order of magnitude. What the two guests kept circling back to is one sentence: shortcuts run on relationships, the slow road builds institutions, and the slow road has no substitute.

![On a single timeline with three handover points, the relationship line above breaks and restarts at each point, while the institution line below runs continuously through all of them.](/figures/institution-vs-relationship-en.svg)

Here's something I've tried, offered rather than assigned. Take a sheet of paper and write down five things you currently rely on to keep working — at work, at home, nothing to do with investing. Maybe "a colleague shields me from noise," "a group chat reminds me of deadlines," "one teacher keeps my kid on top of homework." Then mark each one: **I** or **R**. If that person left tomorrow, would it still run? Yes means institution. No means relationship.

You'll see the shape of your own list. The first time I did it, four of my five were R — and I'd assumed all four were handled. Nothing needs fixing immediately. Pick the one you'd least like to lose, and think about the single step that would move it toward I.