Bahrain diversified early and ran out of room

1. Where things stand

The debt, and which number actually matters. Per the World Bank, government debt was 142.5% of GDP in 2025 and is projected at 138% by end-2026, and is not expected to stabilise over the medium term.

The stock is alarming and the flow is what bites. Debt service consumes 33% of total government revenue. A third of everything Bahrain collects is committed before a single service is delivered, a single salary paid or a single project funded.

That ratio is the mechanism by which Bahrain has less room than states with worse-looking debt levels but lighter service burdens. Two countries can carry identical debt-to-GDP and be in completely different positions depending on the coupon, the maturity profile and the revenue base underneath it. Bahrain is on the wrong side of all three.

The reform package, and the honest way to describe it. Per the register, in December 2025 the Cabinet approved a local corporate tax, a 20% cut in government administrative spending, higher fuel and gas prices for business users, and larger transfers from state-owned companies. These are politically expensive measures and the state took them.

They raise 1.1% of GDP against a deficit of 10.5% of GDP, about one tenth of the gap.

Both halves of that sentence have to travel together, and almost nobody carries both. "Bahrain is reforming" is true. "Bahrain's reforms cannot close the deficit" is also true. Quoting either alone misrepresents the position, and the two together are the whole story.

The production shock. Oil output fell by about three quarters in March–April 2026 on the Hormuz closure and infrastructure damage, the sharpest production shock recorded anywhere in this GCC series.

Growth forecasts do not agree with that narrative. The IMF projects +3.3% real GDP for 2026 while the same period is described as the economy weakening sharply, and no source reconciles the two. We carry both because there is no arbiter, and we would caution anyone citing the 3.3% as settled.

The lifeline, and its shape. Bahrain's only significant oilfield, Abu Safah, is shared with Saudi Arabia and operated by Aramco. Financially, a $10bn fiscal-balance programme came from Saudi Arabia, the UAE and Kuwait in 2018, and a separate $7.5bn GCC development fund allocation dates from 2011.

Those are two different programmes, seven years apart, with different structures and different sponsors. They are not a running facility and should not be added together into a single headline figure, a point we return to below, because the habit of treating them as standing support is the foundation of the largest risk here.

The bright spot. Aluminium Bahrain is one of the world's largest single-site smelters at roughly 750,000 tonnes a year, and reported H1 2026 profit of BD 140.2m (US$372.8m), up 228% from BD 42.7m.

And a qualifier that matters more than it appears: while hydrocarbons are a fifth of GDP, crude is still 48.2% of total exports. The domestic economy is diversified. The external account is not.

2. The ten emerging risks

Ranked by how much damage the event would do. Bahrain's register is unusual in this series because it is not ten independent exposures. Most of these are routes to a single question that nobody has answered in writing, and the first risk is that question.

2.1 The debt is called sustainable, and the assumption holding it up is unwritten

Bahrain's debt path is routinely described as sustainable. No source found describes a current support facility. The sustainability case quietly assumes that regional sponsors step in again, and nobody has written that assumption down.

Start with what the past programmes are not. The 2018 fiscal-balance package and the 2011 development allocation were two distinct interventions, seven years apart, with different structures and different sponsor groups. Treating them as evidence of a standing arrangement is the error this entry exists to name, and it is a common one.

Now ask what the sustainability case actually rests on. A state carrying 142.5% debt-to-GDP against a $60 breakeven gap needs something load-bearing underneath that description. It is not the primary balance, which the reform package closes only a tenth of. It is not the discretionary budget, which debt service has already reduced to a fraction of headline. It is not asset sales at any scale that has been announced.

What remains is an expectation. The market believes Bahrain's neighbours will not let it fail, and that belief is doing the work that fiscal arithmetic cannot.

That expectation may well be correct, and we want to be fair to it. Bahrain's strategic position gives its neighbours strong reasons to support it, and they have done so twice within living memory. The base rate is good.

But an assumption nobody states is an assumption nobody stress-tests. Nobody has published what triggers support, what conditions would attach, how large it would be, or what happens if it is slower than needed. And the sponsors themselves are under documented fiscal pressure now in a way they were not in 2018, which is the tenth risk on this list.

2.2 A second production shock arrives with nowhere to absorb it

Output fell by about three quarters in two months. With a $60 breakeven gap and a third of revenue pre-committed to debt service, a repetition or a persistence has nowhere to go except borrowing.

The mechanism is simple and unforgiving, and it is worth walking through because it explains why Bahrain's position deteriorates faster than its headline numbers suggest.

Most states meet a revenue shock by cutting discretionary spending. That is what a discretionary budget is for. Bahrain's discretionary budget is what remains after 33% of receipts have gone to debt service, and that remainder is simply too small to absorb a shock of this size. There is nothing to cut that would make a material difference.

So the adjustment falls on borrowing, into a stock already at 142.5% of GDP. And each iteration makes the next one harder, because the additional borrowing enlarges the debt service line, which shrinks the discretionary budget further, which makes the following shock even harder to absorb.

This is a ratchet rather than a cycle. Cycles return to where they started. Ratchets do not.

It is also the fastest route to the risk above. A second shock does not gradually erode Bahrain's position. It accelerates the arrival of the financing moment at which the unwritten assumption finally gets tested.

2.3 Debt fails to stabilise and the 138% projection is missed

Debt is projected at 138% of GDP by end-2026, and the World Bank does not expect it to stabilise over the medium term. A missed projection would confirm the trajectory that the projection is meant to reassure about.

The projected fall from 142.5% is doing a great deal of work in how Bahrain is currently assessed. It is the number that supports the argument that the position is being actively managed, and it appears in almost every summary of the country's finances.

It depends on a deficit path that the reform package addresses only marginally.

Notice also that two World Bank statements sit uncomfortably beside each other here. The near-term projection shows improvement. The medium-term assessment says debt is not expected to stabilise. Both are published, and in practice only the first gets quoted.

A missed projection would be read as the first hard evidence that consolidation is not working, and it would arrive well before any of the larger structural risks on this list resolve. That timing is why it ranks third: it is the earliest observable signal that the rest of this article is right.

2.4 Reform succeeds, is read as a turning point, and the gap remains

The package closes roughly a tenth of the gap. Stabilising the position on revenue alone would need measures around ten times larger. The risk is not that reform fails. It is that reform succeeds and is mispriced.

Closing a tenth of a gap is not failure. It is arithmetic, and the measures behind it are real and politically costly. Bahrain has done something difficult.

But a package sized at a tenth of the problem produces exactly the evidence a market wants to see: direction of travel, demonstrated political will, delivered measures with dates attached. Meanwhile the fiscal position is substantially where it was.

That combination is more dangerous than visible failure, and this is the counterintuitive point worth sitting with. Visible failure forces the larger conversation with sponsors, with creditors with the public. Visible partial success postpones it, because spreads tighten and rating commentary warms at precisely the moment when the arithmetic says the case for harder measures should be getting stronger.

The consequence is delay, and delay is expensive here in a way it is not elsewhere. Every quarter that the larger package and the sponsor conversation are deferred is another quarter of debt service accruing against a gap that has not moved.

2.5 The corporate tax underdelivers in its first years

A local corporate tax is the centrepiece of the reform package and is new to Bahrain. First-year collection from a tax with no administrative history rarely matches its forecast.

The measure is sound in principle and is exactly the kind of structural revenue reform the position requires. It broadens the base rather than raising an existing rate, which is the harder and better choice.

It is also the component with the least precedent. A state introducing corporate taxation has to build the assessment machinery, the collection systems, the appeals process and the compliance culture at the same time as it books the revenue. Every jurisdiction that has done this has taken time to reach forecast yield, and the shortfall in the early years is normally material rather than marginal.

Underdelivery here would not be a policy failure so much as an implementation lag, and lags of one to two years are ordinary. The problem is that the package is already only a tenth of the gap. If its largest single component delivers late, the effective contribution in 2026 is smaller than the 1.1% of GDP currently being counted, which makes the fourth risk above arrive sooner.

No source found breaks the 1.1% down by measure, so the corporate tax's share of it cannot be established. This is our read of a common implementation pattern rather than a Bahrain-specific finding.

2.6 Alba was struck, and is running at about 30%

CORRECTION, 2026-08-20. The first version of this article said no source examined Alba's input exposure, and treated the smelter as Bahrain's one clearly-working asset. That was wrong. Alba was physically attacked on 28 March 2026. The vault's aluminium note now carries the event and this section has been rewritten around it. The original reasoning about input fragility was sound; the conclusion that nothing had happened was not.

Alba was struck on 28 March 2026, injuring two employees, and is running at roughly 30% of its 1.62 Mt capacity. EGA's Al Taweelah in the UAE was hit in the same attack and faces about 12 months to restore full production.

That reframes the H1 result entirely. The 228% profit rise describes a period that closed before the strike, and quoting it as evidence of Alba's condition is now a category error rather than merely incomplete.

The mechanism that makes this severe is specific to smelting. Pot lines run at 13–15 MWh per tonne, and if the current stops for a few hours the bath freezes into rock, destroying the cells. That converts a short interruption into a year-long outage, and it is why restart costs run to roughly a thousand times those of restarting a coal-fired unit.

The comparison that matters sits one country away. Qatalum was warned by QatarEnergy of a fuel shortage and executed a controlled 40% shutdown with no physical damage. Its pots were preserved and it may return to full output by Q4 2026. Alba and EGA were attacked without notice.

For a continuous-process asset, notice period is a material variable in its own right, separate from damage severity. That is the transferable finding here, and it applies to any physical asset in the vault, not only smelters.

Bahrain's exposure is also larger than the single plant. Alba alone is 2.16% of world aluminium capacity, and 8.03% of world smelting sits inside the Strait of Hormuz, against imports of alumina and carbon anodes that must arrive by sea through the same strait, at roughly 3 tonnes of cargo per tonne of metal on the round trip.

2.7 The external account transmits what the domestic economy has hedged

Hydrocarbons are a fifth of GDP but 48.2% of total exports. Bahrain earns its foreign currency the same way its undiversified neighbours do, so a production shock transmits regardless of GDP composition.

This is the cleanest illustration in this series of why the word "diversified" is meaningless without a denominator, and it explains something that otherwise looks contradictory.

On output, Bahrain is the most diversified economy in the Gulf. On external earnings, it is barely diversified at all. Both statements describe the same country in the same year.

The distinction matters because the currency peg, the import bill and the external debt are all settled in foreign exchange rather than in GDP. A domestic services economy of banking, retail and logistics generates dinars. It does not generate the dollars that a hydrocarbon export does. That is why the production shock reached the fiscal position as directly as it did in a state where oil is only a fifth of output.

The risk is that Bahrain continues to be assessed on the GDP figure, which is the more flattering number and the less relevant one for the exposure that actually binds.

2.8 The Abu Safah arrangement is revisited

Bahrain's only significant oilfield is shared with Saudi Arabia and operated by Aramco. If the allocation, terms or operating arrangement were ever revisited, Bahrain would be negotiating over the asset that produces its revenue, from a position it does not control.

No other GCC producer is in this position, and it is worth stating plainly what it means. A national hydrocarbon base that is jointly held and operated by another state's national oil company is a physical dependency, not merely a financial one. Bahrain does not set the production level, does not control the operating decisions, and does not own the operator.

The timing is the only reason to raise it now. Saudi Arabia is under its own documented fiscal pressure, and arrangements that are comfortable when a sponsor is flush tend to be examined when it is not. That is a general observation about how such arrangements behave over time, not a specific claim about this one.

This is entirely our read. No source found suggests any revision is contemplated, and it ranks eighth for exactly that reason. We include it because a dependency of this structure does not stop being a dependency merely because it is currently stable, and a register that omitted it would be describing a country with more control over its own revenue than Bahrain actually has.

2.9 A rating action narrows market access

Bahrain funds its deficit in markets while carrying the region's weakest fiscal metrics. A downgrade would raise borrowing costs on a debt stock where a third of revenue already goes to service it.

The arithmetic here is unusually direct. At 142.5% of GDP with 33% of revenue committed to service, the sensitivity of Bahrain's budget to its own cost of borrowing is the highest in the GCC by a wide margin. A modest widening in spreads is a material fiscal event rather than a market inconvenience.

The channel to a downgrade runs through the earlier risks rather than independently. A missed debt projection, an underdelivering reform package, or a second production shock would each supply the evidence an agency would act on. That is why it sits ninth: it is a consequence of the risks above rather than a separate cause.

The reason to list it separately is that it is the point at which the other risks stop being analytical and start costing money. It is also, unlike most items here, something that arrives on a specific date with a public announcement attached.

No rating action is currently signalled, and this is our inference from the fiscal position rather than from any agency commentary.

2.10 The sponsors' own fiscal pressure removes the option

The support assumption requires Saudi Arabia, the UAE and Kuwait to have both the will and the capacity to act. All three now carry documented fiscal pressure of their own, in a way they did not in 2018.

This is the risk that modifies the first one, and the two should be read together.

Look at where the 2018 sponsors are today. Saudi Arabia is running a contested breakeven well above prevailing prices and has already cut its own capital programme by more than half. Kuwait has just restored debt issuance after eight years without a funding instrument, and is debating whether to open its sovereign fund. The UAE is the most comfortable of the three, and it is managing a property correction and a multi-gigawatt compute build-out at the same time.

None of that means support would be refused. The event here is subtler and more likely: a package that is smaller, slower, or more conditional than the 2018 one, arriving at a Bahrain whose gap is larger than it was then. Conditionality in particular changes the picture, because conditions attached to support constrain exactly the fiscal choices Bahrain would otherwise use to respond.

This is our read, and no source found addresses sponsor capacity in relation to Bahrain. It ranks last because it modifies the first risk rather than standing alone. We include it because the first risk cannot honestly be assessed without it. The question is not only whether the sponsors would act, but what they can afford to do when asked.

3. What this means

Bahrain is diversified without reserves. Kuwait is undiversified with them. Bahrain is the more exposed of the two, because the debt stock consumed the fiscal space diversification was meant to buy. Diversification reduces variance. Only a balance sheet creates capacity. That is the most transferable lesson in this series and it is our inference rather than a published finding.

Almost every risk here terminates in the same place. The production shock, the reform mispricing, the Alba exposure and the rating channel all end at borrowing, and borrowing ends at the support question. This is not a register of ten independent exposures. It is a set of routes to one unwritten assumption.

The domestic and external accounts tell opposite stories. A fifth of GDP against 48.2% of exports is the cleanest example in this series of why a diversification statistic is meaningless without its denominator.

What to watch. Three things would move this assessment materially: any public description of a current support facility, the first-year yield of the corporate tax against its forecast, and whether debt reaches the projected 138%. The first would resolve most of this article; the absence of it is the article's central point.