The protein system
The eight members feed roughly 237 million people.[1] Population-weighted protein supply across the six FAOSTAT covers, 97.3 per cent of the population, is 79.2 grams per person per day, of which 30.6 grams is animal-source.[1] The spread is wide, from Malaysia at 93.3 grams total and 54.65 animal-source to Cambodia at 64.4 and 21.89.[1]
Aquatic animals carry the load: the eight members landed 5,255,030 tonnes of wild aquatic animals and farmed 2,695,261 tonnes in 2023, about 34 kilogrammes live weight per person, alongside 1,853,032 tonnes of farmed aquatic plants that are not protein.[2] That distinction matters most in the Philippines, where 68 per cent of the farmed tonnage is seaweed.[2]
On staples the region sits close to balance and slipping. Rice production of 36.0 million tonnes against consumption of 37.0 million in MY2025/26 gives self-sufficiency of 97.2 per cent.[4] The aggregate conceals a distribution problem: Myanmar exports 2.3 million tonnes and Cambodia 3.4 million, yet Comtrade shows 41 per cent of Myanmar's 2025 rice shipments going to China, 24 per cent to the EU27 and 7 per cent to Madagascar against 4.4 per cent to the Philippines,[44] and 6.85 million tonnes of Cambodian output crosses the border as raw paddy, which USDA attributes to Vietnam.[33][5] The deficit members buy from India, Vietnam and Thailand, none of them inside the boundary. Nutritional outcomes lag the aggregates: roughly a quarter of children under five are stunted, from 46.7 per cent in Timor-Leste to 21.2 per cent in Malaysia.[7]
Durable domestic capacity: 50
Quantitative score 55, adjusted down 5 points
The indicators put the pillar at 55; we deduct 5, taking it to 50. In aggregate the region produces most of its own protein, and the mainland endowment is real. The weakness is durability and distribution rather than volume.
The Philippines, 49 per cent of the regional population,[1] is the problem case. Rice self-sufficiency of 71.7 per cent in 2024 was its lowest since 1988, on record imports of 4.74 million tonnes.[9] It imports 37 per cent of its meat, including 44 per cent of pork and 62 per cent of beef.[4] Fisheries output fell for a third consecutive year to 3,956,913 tonnes in 2025, 7.1 per cent below the 2023 level, with marine municipal landings down 6.7 per cent.[10] Malaysia's rice self-sufficiency has fallen to 52.9 per cent from 62.1 per cent in 2020, with beef import dependency at 83.5 per cent.[17][18] Singapore's farms supply 25 per cent of protein consumed[22] and Brunei's rice crop about 8.5 per cent of its demand.[26]
We deduct 5 points for conflict degrading nominal capacity. Myanmar is 23 per cent of the region's population and its second-largest producer, at 1.62 million tonnes of capture fisheries, 1.20 million tonnes of aquaculture and 2.3 million tonnes of rice exports,[2][4] yet WFP projects 12.4 million of its people facing acute hunger in 2026 against 3.6 million already displaced and 4 million projected by the year end,[29][30] with FAO GIEWS putting the acutely food-insecure count nearer 8.5 million,[25] and IFPRI's survey of 2,734 producers found productivity down 8 per cent in the 2025 monsoon and profit per acre down 40 per cent to a six-year low.[31] Output that neither reaches people inside the same borders nor sustains the farmers producing it should not count at face value.
Resource headroom: 46
Quantitative score 51, adjusted down 5 points
The indicators put the pillar at 51; we deduct 5, taking it to 46. Water is abundant on the mainland and land is not. Renewable internal freshwater runs from 4,203 cubic metres per person per year in the Philippines to 25,188 in Laos, every member except Singapore far above the 1,700 stress threshold.[6] Arable land is the opposite: 22.93 million hectares for 237 million people, about 0.097 hectares each, roughly half the world average.[3]
What binds is the composition of the protein mix and the state of the wild base. The region imports 6.71 million tonnes of corn and 10.54 million tonnes of wheat a year alongside its soybean meal, with about 3.3 million tonnes of the Philippine wheat going to feed,[4] so each additional kilogramme of poultry, pork or farmed fish is grown on imported land. The wild base that supplies the cheapest protein has meanwhile passed its peak: South China Sea stocks are down 70 to 95 per cent since the 1950s, with catch rates down 66 to 75 per cent over 20 years;[39] Malaysian marine capture stocks fell in five of the last six years;[19] and Philippine aquaculture fell 1.9 per cent in 2025 with commercial capture down 2.2 per cent.[10]
Our 5-point deduction reflects that per-capita freshwater statistics record Mekong water as domestic when its timing is not sovereign. The Mekong River Commission's Council Study projected total fish biomass down 35 to 40 per cent by 2020 and 40 to 80 per cent by 2040 under basin hydropower,[41] and its sediment modelling puts the load reaching the delta 67 per cent lower under the 2020 development scenario and 97 per cent lower under the 2040 scenario, leaving 3 per cent of the 2007 load.[46] WWF records an 88 per cent population collapse in Tonle Sap fish in a lake producing about 2 per cent of the global inland catch.[40] A resource base whose replenishment schedule is set upstream does not deliver the headroom its water statistics imply.
The freshwater abundance that remains within the region's own control is genuinely being mobilised rather than sitting idle. Land equipped for irrigation as a share of arable land rose from 20.4 to 25.5 per cent across the seven members with a FAOSTAT series between 2000 and 2024, led by Malaysia's rise to 52.0 per cent and the Philippines' rise from 27.0 to 38.6 per cent.[60] Cambodia is the exception: its equipped share fell from 7.3 to 6.5 per cent as arable land grew 11.4 per cent faster than water infrastructure did.[60] Agricultural credit functions at scale only in Cambodia, whose orientation index of 0.842 in 2024 sits close to the 1.00 parity value and has risen steadily since 2004; the Philippines and Malaysia sit further below parity and essentially flat, Timor-Leste's index is near zero throughout, and Myanmar, Brunei and Laos carry no FAOSTAT agricultural credit series of any kind.[61] Neither irrigation nor credit converts into more land: regional arable land grew 11.5 per cent in absolute terms between 2000 and 2024 but fell 20.3 per cent per person as population grew 40.0 per cent, and every member's arable land per capita declined.[60][62] Malaysia's National Agrofood Policy shows the ceiling directly. It targeted 75 per cent rice self-sufficiency by 2025 against an actual 52.9 per cent, a shortfall its Agriculture Ministry attributes in part to irrigation upgrades covering only a small share of what padi land needs.[63][64] The Philippines shows the same ceiling from the productivity side: the Rice Competitiveness Enhancement Fund put PHP 60 billion into farm mechanisation, seed and extension between 2019 and 2025, and a 2026 evaluation found genuine yield gains among participating farmers, yet national rice self-sufficiency fell over the same period to its 37-year low.[65][66][9]
Import exposure: 57
Quantitative score 55, adjusted up 2 points
The indicators put the pillar at 55; we add 2, taking it to 57. Removing Thailand flips the staple trade position: the seven members USDA covers, all but Timor-Leste, were net importers of about 0.67 million tonnes of milled rice in MY2025/26 and are forecast at a net 1.90 million tonnes in 2026/27.[4] Exposure is concentrated and specific: the Philippines is the world's largest rice importer at 5.6 million tonnes forecast for MY2026/27;[4] Malaysia's food import bill reached RM 93.8 billion in 2024 on an import dependency ratio of 57.9 per cent;[18] Singapore imports more than 90 per cent of its food, including 96 per cent of pork;[22][4] and Brunei imports 97.6 per cent of its rice on USDA figures, against a ministerial self-sufficiency claim implying 91.5 per cent.[4][26]
We add a small positive adjustment for geography, which remains genuinely favourable. The region straddles rather than sits behind the chokepoints that define exposure elsewhere: Singapore and Malaysia's west-coast ports sit on the Strait of Malacca, through which some 29 per cent of global maritime oil flows pass, while the Philippines faces the open Pacific and Myanmar the Bay of Bengal.[43] Vietnam and Thailand remain immediate neighbours even though both are scored outside this entity, so supply lines stay short and multiply routed.
We limit the uplift to 2 because intra-regional supply has twice been withdrawn at short notice: Malaysia halted exports of 3.6 million live chickens a month from 1 June 2022 while supplying roughly a third of Singapore's chicken,[48] and the Philippines suspended its own rice imports for four months, under Executive Order 93 from 1 September 2025 and then Executive Order 102 to 31 December, after traders cut palay buying prices to as low as PHP 8 to 10 per kilogramme against a production cost of PHP 12 to 14.[13][47] Neighbours legislate for their own farmers first. Where concentration is measured directly against customs data it does not always fall where the region's own diplomacy would suggest: India, not Thailand or Vietnam, is now Singapore's largest rice supplier at 42 per cent of 2025 volume, with the two ASEAN suppliers together at 51 per cent;[44] India supplies 83 per cent of Malaysia's frozen beef imports and, on Khazanah Research Institute's reading, close to three-quarters of total beef supply;[44][18] and 60 per cent of Philippine fish meal comes from Malaysia, with Chile and Peru together under 5 per cent.[44]
Upstream dependence: 43
Quantitative score 38, adjusted up 5 points
The indicators put the pillar at 38; we add 5, taking it to 43. This is where the system binds. Feed protein is effectively 100 per cent imported,[4] and fertiliser is imported almost everywhere: the Philippines buys about 90 per cent of what it uses, on a market worth USD 3.6 billion in 2025.[15]
The 2026 Strait of Hormuz disruption showed what that costs. Urea rose above USD 850 a tonne in April 2026, 80 per cent above February, with the World Bank projecting the fertiliser index up more than 30 per cent across the year,[36] and FAO warns that a delay of even a few weeks forces farmers to reduce or abandon application, with consequences landing in the second half of 2026 and into 2027.[37] ESCAP puts regional oil up about 45 per cent and gas about 55 per cent since late February 2026.[38] Myanmar shows the farm-level transmission: urea up 23 per cent and input expenditure per acre up 63 per cent between 2023 and 2025, against farmgate paddy prices down 6 per cent.[31]
The upward adjustment reflects funded in-region capacity that did not exist five years ago and is substituting under live stress. PETRONAS's Sipitang plant in Sabah is rated at 3,850 tonnes a day of urea on 2,100 tonnes a day of ammonia, with further, unquantified urea capacity at Gurun and Bintulu,[20] and Brunei Fertilizer Industries converts 2,200 tonnes a day of ammonia into 3,900 tonnes a day of granular urea, producing over a million tonnes in 2023.[21] Those two plants alone carry roughly 2.6 million tonnes a year, comfortably above regional urea consumption, and Asia-Potash's Khammouane operation in Laos entered three-million-tonne capacity in December 2025.[34][35] Malaysia and Brunei supplied 137,900 of the Philippines' 395,000 tonnes of first-half 2025 urea imports,[14] which is why the 2026 Gulf shock reached Philippine farmers through price rather than physical shortage. We add 5 points for demonstrated substitution during the 2026 Gulf urea shock and no more: none of it substitutes for feed protein, which has no in-region source at scale.
Access and affordability: 53
Quantitative score 55, adjusted down 2 points
The indicators put the pillar at 55; we deduct 2, taking it to 53. Myanmar carries no Cost and Affordability of a Healthy Diet estimate in FAOSTAT's current release for any year 2017-2025, so its unaffordability component is scored qualitatively rather than dropped from the population weighting or carried on an unconfirmable figure: 5.4 per cent undernourishment and 26.7 per cent child stunting from FAO's own series, set against the World Food Programme's finding that 12.4 million people, close to a quarter of the country, are projected to face acute hunger in 2026, with 3.6 million people internally displaced and WFP's own delivery reaching only 1.5 million of the 12.4 million in need.[29][30] On the World Bank/FAO Cost and Affordability of a Healthy Diet measure, the members with a direct reading show unaffordability at 47.4 per cent in the Philippines and 51.9 per cent in Laos in 2024, against 1.6 per cent in Malaysia, among the lowest rates in Asia.[50] Weighted by population across the four members with any basis for a score, the unaffordability-equivalent component comes to 50. Undernourishment is comparatively reassuring at 3.8 per cent population-weighted across five members including Myanmar, though Timor-Leste's 18.7 per cent is the outlier.[8] Child stunting, the only indicator with full coverage, is 25.3 per cent population-weighted, above the roughly 22 per cent global average and ranging from 4.5 per cent in Singapore to 46.7 per cent in Timor-Leste.[7]
Price volatility and physical access are the weaker half. Food inflation ran at 5.20 per cent in the Philippines and 7.69 per cent in Cambodia in mid-2026,[51][53] while Laos is still working off a currency crisis that took food inflation to 52.69 per cent in May 2023 and the kip from about 9,000 to over 21,500 to the dollar between 2020 and 2025.[54] Myanmar's conflict, which is what puts a quarter of its population at acute hunger risk, and the Philippines' typhoon-driven crop losses in October and November 2025 both degrade physical access for the two members holding the largest share of the region's population.[29][56] Malaysia, Singapore and Brunei, with low and stable food inflation, are the diverging positive cases.[52][55]
We deduct 2 points because the unaffordability component still omits Cambodia, Singapore, Brunei and Timor-Leste. Timor-Leste's stunting and undernourishment, both the region's worst, point toward unaffordability above the level the scored members alone would suggest, so the component most likely still overstates the pillar even with Myanmar now priced on its own evidence. We do not raise the score for Timor-Leste's USD 18.75 billion Petroleum Fund, since that buffer belongs to crisis purchasing power under shock endurance and, at 0.6 per cent of the region's population, could not move the weighted figure materially.
Shock endurance: 45
Quantitative score 49, adjusted down 4 points
The indicators put the pillar at 49; we deduct 4, taking it to 45. Endurance rests on dietary breadth rather than on reserves or money. The ASEAN Plus Three Emergency Rice Reserve earmarks 787,000 tonnes, only 87,000 of them from ASEAN members, against regional consumption of roughly 101,000 tonnes a day, so the ASEAN tranche is under a day of cover.[42] The Philippine national buffer held 337,618 tonnes on 7 May 2026, 8.74 days of demand, below the nine-day statutory minimum under the Rice Tariffication Law, after procuring 13,127 tonnes against a January target of 313,315.[11][12] Singapore requires licensed rice importers to hold a stockpile set against their import volumes as a condition of the import licence, though the size of the multiple is not disclosed in either the government's own description of the scheme or its WTO notification.[24][49]
Fiscal depth sits in the wrong places. Singapore and Brunei, at USD 139,593 and USD 78,985 of GDP per head in PPP terms, hold 2.7 per cent of the population between them, while the five members holding 82 per cent of it sit between USD 4,110 and USD 10,748.[45] What does provide endurance is dietary breadth: rice, freshwater and marine fish, poultry, pork and pulses are all produced at scale inside the region, so substitution is routine, and the world rice market is amply supplied, with trade forecast at a record 63.1 million tonnes in 2026/27 and global ending stocks at 192.8 million tonnes.[5]
Crisis purchasing power, the financial capacity to buy through a disruption, is close to neutral against the region's other endurance indicators. Reserve cover is adequate across most of the region by IMF/World Bank standards: 7.33 months of imports in the Philippines and 7.65 in Cambodia in 2024.[57] Laos holds only 2.17 months on the last Lao value the World Bank series carries, for 2023, with the IMF putting 2024 cover nearer 1.7 months, against a public debt burden the World Bank puts at roughly 88 per cent of GDP in 2025, with the kip having fallen from about 9,000 to over 21,500 to the US dollar between August 2020 and October 2025 and a January 2025 survey finding a third of households cutting spending on food, health or education.[58] Myanmar has disclosed no reserve figure to the World Bank for any year after 2019, a blackout under military administration and sanctions that is itself informative about the state of its external buffers.[57] Singapore's reserve cover of 4.29 months understates its true capacity, since a trade-entrepot economy imports many multiples of what it consumes,[57] and Timor-Leste's USD 18.75 billion Petroleum Fund as of 31 May 2026, against a population of 1.4 million, is a genuine buffer that the GDP-per-capita figure above does not show, though it is too small a population share to move the region.[59] The indicators put this pillar at 49: sound reserves in the Philippines and Cambodia are weighed against Laos's currency crisis and Myanmar's disclosure blackout, so financial firepower remains concentrated in members carrying a few per cent of the population rather than in the two largest.
We deduct 4 points because reserve and fiscal indicators cannot capture a member already inside the shock. Myanmar's 12.4 million people projected to face acute hunger in 2026, 1 million of them at emergency levels, and 400,000 acutely malnourished young children and mothers[29][30] sit alongside a farm economy running down its own capital, with rice prices up 5.2 per cent year on year in June 2026 and fuel and fertiliser well above year-earlier levels,[31][32] so a quarter of the region enters the next shock with less absorptive capacity than it had a year ago. The deduction would stand on FAO's lower count of about 8.5 million as readily as on WFP's.[25] No further deduction applies for biological shocks: Philippine African swine fever cases fell from 98 active barangays at end-2025 to 27 by 6 February 2026,[16] and pork output is forecast up 1.5 per cent to 990,000 tonnes,[4] which reads as recovery rather than permanent loss.
Trajectory
Three variables matter most here. First, feed protein, where nothing in the pipeline helps: until part of the 5.26 million tonnes of soybean meal is originated regionally or displaced by lower-conversion protein, upstream stays in the 40s.[4] Second, the Philippine buffer, where a reserve at 8.74 days against a nine-day floor is a policy failure with a visible remedy, and the four-month 2025 import suspension shows a government willing to trade consumer exposure for farmgate prices.[11][13][47] Third, fertiliser: the Malaysian, Bruneian and Laotian plants are the clearest improving structural fact here, and a Laos potash ramp to five million tonnes would turn the region from potash importer into exporter.[34][35]
Working the other way, the wild capture base is in structural decline across the South China Sea, the Mekong and Philippine municipal waters,[39][40][46][10] and the substitution it forces runs towards feed-dependent farmed protein, deepening the upstream exposure. The South China Sea figures date from 2017 and no comparable regional assessment has been published since, so the direction is better established than the current level.[39] Singapore's recalibration, retiring 30-by-30 for revised 2035 targets,[23] is an honest read of what land-scarce production can deliver. Myanmar's trajectory is a function of the conflict, where we see no basis for assuming improvement.
Country notes
Philippines. Half the region's population and the entity's centre of gravity. Rice self-sufficiency of 71.7 per cent in 2024 was the lowest since 1988,[9] imports are forecast at 5.6 million tonnes in MY2026/27 and 37 per cent of meat is imported.[4] Fisheries have fallen three years running to 3,956,913 tonnes, 7.1 per cent below 2023,[10] arable land is 0.049 hectares per person,[3] and the rice buffer sits below its statutory floor.[11] Protein supply of 77.5 grams,[1] improving ASF containment[16] and Pacific-facing ports are the positives.
Malaysia. The wealthiest large member and the most exposed on protein composition. Animal protein of 54.65 grams a day is the region's highest[1] and eggs and duck meat exceed demand, yet beef is 83.5 per cent imported with India supplying 83 per cent of the frozen beef it buys, rice self-sufficiency is 52.9 per cent and fish self-sufficiency has fallen to 90.2 per cent.[17][18][19][44] It is also the region's most important upstream asset, and the source of 60 per cent of Philippine fish meal.[20][14][44]
Myanmar. Nominal capacity that neither reaches people nor sustains itself: 2.82 million tonnes of aquatic animals[2] and 2.3 million tonnes of rice exports[4] on strong endowments,[6][3] set against 12.4 million projected to face acute hunger in 2026 on WFP's count and about 8.5 million on FAO's, 3.6 million displaced[29][25] and farm profits at a six-year low.[31] The binding constraints are conflict and input costs rather than resources.
Cambodia. Rice-rich and protein-thin. The highest arable land per person at 0.236 hectares,[3] but 6.85 million tonnes of the surplus leaves as raw paddy across the border, so milling and value capture sit outside.[33] Total protein of 64.4 grams and animal protein of 21.89 grams are the region's lowest,[1] its capture fisheries depend on the depleted Tonle Sap flood pulse,[40] and it imports 310,000 tonnes of soybean meal a year.[4]
Laos. Land-locked, water-rich and the most improved on upstream. Freshwater of 25,188 cubic metres per person is the region's highest[6] and protein supply of 81.7 grams second only to Malaysia's,[1] against stunting of 32.8 per cent[7] and rising rice imports.[4] Asia-Potash's three-million-tonne capacity puts potash inside regional shipping distance,[34][35] while GDP per head of USD 8,873 limits its ability to buy through a shock.[45]
Singapore. The most import-dependent member and the best prepared. Over 90 per cent of food imported, from more than 180 countries and regions, with local farms supplying 25 per cent of protein consumed[22] and revised 2035 targets replacing 30 by 30.[23] Rice importers must hold a stockpile against their import volumes as a licence condition,[24][49] and GDP per head of USD 139,593 buys diversification no other member can afford,[45] though at 106 cubic metres of freshwater per person and 600 hectares of arable land wealth cannot buy a resource base.[6][3] Its rice supply now leans on India at 42 per cent of volume rather than on its ASEAN neighbours.[44]
Brunei. Small, rich, and the holder of one genuinely regional asset. Self-sufficient in chicken meat and beyond it in eggs, giving animal protein of 46.3 grams a day as of 2009, the last year FAO published for it,[1] with domestic food output up 41 per cent by value between 2017 and 2024.[27] Rice self-sufficiency of about 8.5 per cent leaves most of the staple imported, 91.5 per cent on the ministerial figure and 97.6 per cent on USDA's.[26][4] Brunei Fertilizer Industries makes it one of only two nitrogen producers in the region.[21]
Timor-Leste. The weakest member on every nutritional measure: stunting of 46.7 per cent[7] and undernourishment of 18.7 per cent against 2.5 to 5.4 per cent elsewhere,[8] on protein supply of 65.4 grams.[1] About 60 per cent of cereals are imported,[28] aquatic production of 8,432 tonnes is negligible[2] and GDP per head of USD 4,110 is the region's lowest.[45] It is also absent from the USDA PSD database, so the regional staple aggregates exclude it.[4] The 2025 harvests came in above the five-year average, the only clearly favourable signal.[28]