Navigating the Transition

Executive Summary

Two credible scenarios for oil demand now point in opposite directions after 2030. A strategy that requires knowing which is right is not a strategy.

Resource-dependent economies are told to diversify because their reserves will be stranded. The projections are less decisive than the advice. In the International Energy Agency's stated-policies scenario, oil demand peaks at about 102 million barrels a day around 2030 and then declines; in its current-policies scenario it rises to 113 million barrels a day by 2050. Both come from the same institution in the same publication.

This report treats that disagreement as the central fact rather than an inconvenience. When the direction is genuinely uncertain, the exposure that matters is not the expected path but the shortening and widening of the revenue distribution — and that is a balance-sheet problem with known treatments, not a forecasting problem.

Section 1 establishes that the exposure is not obsolescence. Section 2 states the mismatch. Sections 3 to 6 take the four treatments in turn — hedging, sovereign saving, liability reform and conversion — with the cases that show each working and failing. Section 7 sets out which options are open to which kind of country, and Sections 8 and 9 give recommendations and conclusions.

1. The Exposure Is Not Obsolescence

Reserves are not about to be worthless. The revenue they produce is becoming shorter and less predictable, which is what actually breaks budgets.

The IEA's stated-policies case has coal demand peaking before 2030 as Chinese declines outweigh growth in India and South-East Asia, and gas demand growing at nearly one per cent a year to 2035 before levelling off. Its current-policies case has oil still growing at mid-century. A finance ministry cannot build a twenty-year fiscal framework on the difference.

What it can build on is the shape of the risk. Three things are happening to resource revenue regardless of which path demand takes:

• The horizon is shortening. Even the optimistic case has a plateau rather than indefinite growth, which converts a perpetual income into a finite one with an uncertain end date.
• The variance is widening. Policy shifts, technology cost curves and geopolitics now move prices in both directions on short notice, and the polycrisis adds correlated shocks that arrive together.
• The buyers are changing their requirements. Methane rules, border carbon adjustments and procurement standards mean that market access increasingly depends on the emissions intensity of production, not only on price.

The third of those is the least discussed and the most concrete. The EU's carbon border mechanism entered its definitive phase in January 2026 covering cement, iron and steel, aluminium, fertilisers, electricity and hydrogen, so exporters of those goods now face a levy tied to the verified emissions embedded in production — a point developed in an earlier report. For a resource economy pursuing downstream processing, that turns the emissions intensity of its own power system into a determinant of whether the processed product can be sold, which is a new and binding constraint on the conversion strategy in Section 6.

None of this is obsolescence. All of it is a worse-behaved income stream — and the problem that creates depends entirely on what that income has been committed to.

2. The Mismatch

A treasurer would recognise this instantly: a volatile, depleting, foreign-currency asset funding rigid, growing, domestic-currency liabilities.

Table 1 — The two sides of a resource-dependent state's balance sheet

The framing is the author's, applied to the properties documented in Sections 1 and 3 to 6.

The final row is what makes this a polycrisis question rather than a commodity-cycle one. In an ordinary downturn a resource state runs a deficit and waits. When shocks arrive together — a price fall, a drought, a currency move, a debt repricing — the revenue and the obligations move in opposite directions at the same moment, and the buffer that would have absorbed one of them is absorbing all four.

Recognising the problem as a mismatch has one immediate advantage: mismatches have standard treatments. You can hedge the variance, lengthen the asset, shorten the liability, or change what the asset is. The next four sections take those in turn.

3. Treatment One: Hedge the Variance

The fastest of the four, the only one that works within a single budget year, and the one most often mistaken for a solution.

Mexico has for two decades run the largest sovereign commodity hedge in the market, buying put options that give it the right to sell crude at a set floor. Reported accounts put the programme at roughly US$1 billion a year in premiums, covering something like 200 to 300 million barrels; the government has since classified the details. In 2020, when prices collapsed, the programme paid out about US$2.4 billion.

Why, then, do so few producers do it? Three obstacles recur, and none is technical. The first is budget optics: a premium paid in a year when prices hold is recorded as money spent for nothing, which is difficult to defend in a legislature and easy to attack after the fact. The second is market depth — a programme covering hundreds of millions of barrels needs counterparties willing to take the other side, which exists for a few benchmark crudes and thinly or not at all for smaller producers and for most minerals. The third is basis risk: a hedge written against a benchmark protects the benchmark price, not the discount at which a particular grade actually sells.

Three properties make hedging attractive and limited in the same breath. It is fast — a decision taken in one budget cycle protects the next. It is honest about cost, since the premium is visible and paid up front, unlike the implicit cost of hoping. And it changes nothing structural: it converts a bad year into an average one and leaves the mismatch exactly where it was. A country that hedges and does nothing else has bought time at a price, which is worth doing only if the time is used.

4. Treatment Two: Lengthen the Asset

A sovereign fund converts a finite, volatile income into a perpetual, smooth one — but only if the withdrawal rule binds.

Norway's Government Pension Fund Global passed US$2 trillion, and the fiscal rule limits average transfers to the budget to the fund's expected real return, currently estimated at 3 per cent and reduced from 4 per cent in 2017. The design does two things at once: it insulates the budget from price swings, and it converts an exhaustible asset into an endowment whose real value is preserved for later generations.

Timor-Leste shows the same architecture without the discipline. Its Petroleum Fund, of the order of US$16 billion, finances more than 80 per cent of the state budget. The legislated Estimated Sustainable Income is also 3 per cent of total petroleum wealth — and withdrawals have exceeded it for years, financing deficits and drawing down the capital base. Analysts have warned of depletion between the late 2020s and the mid-2030s.

5. Treatment Three: Shorten the Liability

The only side of the balance sheet a government fully controls, and the one it is most reluctant to touch.

The clearest measure of liability rigidity is the fiscal breakeven price — the oil price at which the budget balances. It is a statement about spending, not geology, and it moves with what the state has committed to. IMF analysis reported in 2025 puts the regional median fiscal breakeven for Gulf exporters at about US$70 a barrel in 2025, falling to US$62 by 2030 as consolidation and diversification proceed. For Saudi Arabia the estimate depends on scope: roughly US$80–85 for the standard budget, about US$96 including the giga-projects, and above US$110 once Public Investment Fund outlays are counted.

Three liability reforms recur, and all of them are politically expensive:

• Fuel and electricity subsidies, which are pro-cyclical in the worst way: they cost most when prices are high and are hardest to remove when incomes are squeezed. Nigeria's 2023 removal, discussed in an earlier report, produced a rapid vehicle-conversion response precisely because it changed relative prices for everyone at once.
• Public wage bills, where resource booms have typically been converted into permanent employment, giving the state a liability that survives the revenue that created it.
• Off-budget commitments — state enterprise obligations, guarantees, giga-project pipelines — which do not appear in the headline breakeven until they are already due.

The sequencing point from Section 4 applies here with force. A state that cannot fund itself without exceeding its savings rule does not have a savings problem; it has an expenditure problem wearing a savings problem's clothes.

6. Treatment Four: Convert the Resource

The slowest treatment, the only one that changes what the country sells, and the one with the worst record when attempted without market power.

Conversion means using the resource endowment to build a different export. Indonesia's nickel policy is the most-cited recent case: an export ban on raw ore, reinstated from January 2020, pushed processing onshore. Nickel-related exports rose from about US$6 billion in 2013 to nearly US$30 billion by 2022, as stainless steel and battery materials replaced ore in the export mix, with roughly US$30 billion of associated downstream investment commitments.

The case deserves its prominence and also its caveats, which are usually omitted:

• It ran on market power. Indonesia could impose the ban because processors had nowhere else to go at that scale. A producer without that position imposing the same ban loses the revenue and gains nothing.
• The output skewed to stainless steel rather than the battery chain that motivated the policy, because that is what the available ore and the incentives suited.
• The processing is energy-intensive and largely coal-fired, which converts a mineral export into an emissions-intensive one — and, under border carbon rules, into a future market-access problem of the kind Section 1 described.

The more durable version of conversion is to sell the energy rather than the molecule: using a cheap domestic clean resource to make energy-intensive products for export, a proposition examined in an earlier report in this series. It is the same strategy — capture more of the value chain at home — with a compliance profile that improves rather than deteriorates over time.

7. Sequence, and Who Can Do What

The four treatments are not alternatives. They have an order, and the order is determined by which of them makes the others possible.

Table 2 — Which treatments are open to which kind of exporter

Archetypes are the author's, derived from the cases in Sections 3 to 6. Most countries sit between two of them.

7.1 How to tell whether it is working

Diversification programmes are announced more often than they are measured, and the headline indicators — a new industrial zone, a sovereign fund's nominal value, a target for non-oil GDP — can all rise while the underlying exposure does not change. Four indicators track the mismatch itself rather than the announcements:

• Non-resource revenue as a share of government spending, not of revenue. The relevant question is what proportion of the state's obligations could still be met if the resource revenue stopped, and measuring against spending answers it directly.
• The fiscal breakeven trend on a constant perimeter. A breakeven that falls only because the spending definition narrowed has not improved. The series matters more than the level, and the perimeter must not move.
• Real fund value per capita. A sovereign fund growing in nominal terms while population grows faster and inflation runs is shrinking as an endowment, which is precisely the variable the rule exists to protect.
• Export concentration. Standard concentration indices show whether conversion has genuinely broadened the export base or simply moved the same commodity one step down the processing chain — a distinction the nickel case makes vivid.

The third archetype deserves emphasis because it is the one being created now. A country whose export earnings come to depend on copper, nickel, lithium or cobalt has the same balance sheet as a petrostate: a volatile, foreign-currency, exhaustible asset against rigid domestic obligations. The demand outlook is friendlier, which makes the discipline harder to argue for and no less necessary. Transition minerals are not an exit from resource dependence; they are a renewal of it under better market conditions.

8. Recommendations

By actor, in the order the sequence implies.

8.1 Finance ministries

• Publish the fiscal breakeven on a consistent perimeter, including off-budget and state-enterprise commitments. A breakeven that excludes the giga-project pipeline understates the rigidity that will actually bite.
• Reform the liability first. Subsidy reform and wage-bill discipline are what make a savings rule survivable; without them the rule will be breached by governments with genuine needs and no alternative.
• Hedge deliberately and disclose the policy, if not the position. The premium is a budget line and should be argued for as insurance, with the expectation that most years it pays nothing.
• Write the withdrawal rule so that a breach is visible and costly — an independent estimate of sustainable income, published variance reports, and a requirement to legislate any excess explicitly.

8.2 Ministries of energy and industry

• Test conversion proposals for market power before committing. Export restrictions work where buyers have no alternative at scale and fail expensively where they do.
• Specify the energy source for processing at the outset. Coal-fired downstreaming converts a market-access problem in the future into the price of the strategy today.
• Prefer selling embedded clean energy — processed metals, fertiliser, fuels made with domestic renewable power — over selling raw molecules, where the resource endowment allows it.

8.3 Development finance and advisers

• Finance the transition of the fiscal system, not only of the energy system. Technical assistance on hedging, sovereign fund governance and subsidy reform addresses the binding constraint more directly than another renewable project.
• Support transition-mineral producers to install the rules early, while revenue is rising and the politics are easy, rather than after the first downturn.

9. Conclusions

Resource dependence is not defeated by predicting the transition correctly. It is managed by making the balance sheet survive being wrong.

The advice to diversify is sound and almost useless on its own, because it says nothing about sequence, cost or what to do in the meantime. Reframing the problem as an asset–liability mismatch makes the choices concrete. The asset is finite, volatile and foreign; the liabilities are open-ended, rigid and domestic; and in a polycrisis the two move in opposite directions at the same moment.

Four treatments follow, and the cases show each of them working and failing. Mexico's hedge converts a catastrophic year into an ordinary one for about a billion dollars a year, and changes nothing structural. Norway's fund turns an exhaustible income into a perpetual one because a 3 per cent rule is enforced; Timor-Leste's has the same rule and a fund financing over 80 per cent of the budget, drawn down beyond it for years, with a cliff now in view. Gulf breakevens show that the fragile variable is the spending perimeter rather than the reserve base. Indonesia's nickel ban moved US$6 billion of exports to nearly US$30 billion, on market power the next country may not have, and with coal-fired processing that buys a future compliance problem.

The sequence matters more than the menu. Liability reform is the precondition, because a savings rule binds only where the state can fund itself when it bites. Hedging buys the time in which the other treatments are executed. Conversion is the only one that ends the dependence, and it is the slowest — which is why it has to be started while the revenue that pays for it is still arriving.

And the newest instance of the problem is the one least likely to be treated: countries whose earnings are coming to depend on transition minerals hold the same balance sheet in a friendlier market. The friendlier market is what makes the discipline hard to argue for, and it is the only reason there is time to install it.

References

Every quantitative claim above is attributed inline. The principal sources are collected here.

• IEAWorld Energy Outlook 2025: Stated Policies Scenario and Current Policies Scenario — oil demand peaking at about 102 mb/d around 2030 in one case and rising to 113 mb/d by 2050 in the other; coal peaking before 2030 and gas growing to the mid-2030s in the stated-policies case. See also Carbon Brief's summary.
• UNCTADThe State of Commodity Dependence 2023 — about 53 per cent of member states commodity-dependent, over 80 per cent of least-developed and landlocked developing countries, around 60 per cent of small island developing states.
• Mexico's sovereign hedgeProgramme structure, scale and the 2020 payout as reported by World Oil and analysed by the Jain Family Institute; the IMF's comparison of hedging with self-insurance is here.
• Norges Bank Investment ManagementAbout the fund, and the Norwegian government's fiscal policy framework — the fund above US$2 trillion and the 3 per cent spending rule, reduced from 4 per cent in 2017.
• Timor-LesteEast Asia Forum and BenarNews on a Petroleum Fund of roughly US$16 billion financing more than 80 per cent of the budget, withdrawals above the legislated Estimated Sustainable Income, and depletion projected between the late 2020s and mid-2030s.
• IMF analysis of Gulf fiscal positionsRegional median fiscal breakeven of about US$70 a barrel in 2025 falling to US$62 by 2030, and Saudi estimates of roughly US$80–85, US$96 and above US$110 on widening spending perimeters, as reported by EFG International and World Oil; a sceptical reading of breakeven as a policy guide is at AGSI.
• Indonesia's nickel policyEast Asia Forum on exports rising from about US$6 billion in 2013 to nearly US$30 billion in 2022, and USITC on the ban's mechanics; a critical reassessment, including the skew to stainless steel, is here.
• H HeuristicsRelated reports in this series: convergence through crisis, on subsidy removal as a forced-replacement window; climate-resilient industrialization pathways, on selling embedded clean energy; and financing the synergy, on sovereign risk premia.