18. Four worlds in 2050: scenarios for Vietnam's protein supply

In one paragraph. Chapter 15 found two external uncertainties that decide most for alternative protein in Vietnam: whether imported feed protein comes under chronic stress, and whether Asia's fermentation industry stays an open market or falls under Chinese price dominance. Crossing them gives four worlds for 2050. In a Regional workshop, imports stay cheap and the region trades openly, so Vietnam specialises in contract fermentation, functional aquafeed ingredients and food ingredients. In a Comfortable price-taker world, cheap soy and cheap Chinese products leave little room for domestic production. In a Security build-out, dear and unreliable imports and regional co-operation make Vietnam's residues, aquafeed buyers and, later, clean power the base of a domestic feed-protein industry. In a Squeezed importer world, imports are stressed and Chinese producers capture the regional market, so efficiency and local niches carry the load. Vietnam's own choices decide where it lands inside each world: the same external conditions can produce a hub or a price-taker.

#18.1 The two axes

#18.2 The four worlds

#A. Regional workshop (benign imports, open market)

• Picture in 2050. Imported soy and maize stay cheap on average. Several Asian countries run fermentation plants and trade under shared or mutually recognised approvals. Vietnam, with a working new-food route and open pilot plants, hosts contract fermentation for regional brands, sells functional feed ingredients for shrimp and pangasius and makes textured and fungal food ingredients. High-value precision-fermentation ingredients arrive after first-generation patents lapse in the mid-2030s ¹⁴ generalLow. Bulk microbial feed protein stays small because soybean meal is cheap.
• Protein balance. Soybean-meal need of about 7.0 to 10.4 Mt, depending almost entirely on the efficiency programme; microbial protein about 1 to 2% of high-protein feed protein, mostly as functional ingredients estimate.
• Active hubs. Ho Chi Minh City with Dong Nai; Tay Ninh (former Tay Ninh and Long An) sugar hub; the Mekong river belt (Y. Targets and hubs).
• Grows from: 2035 Scenario 3, regulated regional hub (Outlook).
• Signposts. An ASEAN reliance or mutual-recognition mechanism for novel foods; Thai, Indian or Korean plants exporting across the region; fishmeal below USD 1,800 per t for a year; China's soybean imports below 100 Mt a year.

#B. Comfortable price-taker (benign imports, Chinese price dominance)

• Picture in 2050. Imported protein stays cheap and Chinese scale sets the price of every fermentation product. Vietnam imports soybean meal and Chinese yeast, bacterial and textured proteins. Domestic alternative protein stays at chay processing, tofu and a few feed-additive lines. Feed mills gain from cheap inputs, but the country stays exposed to any later shock.
• Protein balance. Soybean-meal need of about 10.4 to 11.7 Mt (trend to high-income demand); microbial protein 1% or less of high-protein feed protein, largely imported estimate.
• Active hubs. Ho Chi Minh City with Dong Nai, at small scale.
• Grows from: 2035 Scenario 1, import price-taker.
• Signposts. Imports of yeasts and protein preparations from China growing above 20% a year for three years; three or more Chinese single-cell or mycoprotein plants above 20 kt a year by 2030; non-Chinese Asian pilot plants underused or closing.

#C. Security build-out (chronic stress, open market)

• Picture in 2050. Repeated fishmeal shortages, more frequent crop losses in exporting countries and shipping scares make imported protein dear and unreliable. Asian governments co-operate on approvals and invest in domestic protein. Vietnam builds residue-based feed protein on cassava, rice straw and brewery side streams, becomes a regional source of aquafeed functional ingredients and demonstrates protein from renewable power on the Mekong shrimp coast (Ca Mau and Vinh Long) from the early 2030s, scaling it after about 2040.
• Protein balance. Soybean-meal need of about 7.0 to 8.8 Mt, with efficiency and alternative protein adding up; microbial protein 4 to 8% of high-protein feed protein, a stretch above any analogue today (Protein balance 2050) estimate.
• Active hubs. All six candidate hubs, including the Mekong shrimp coast.
• Grows from: 2035 Scenario 2, feed-first domestic build-out, then Scenario 3.
• Signposts. Fishmeal above USD 2,500 per t for 12 months; two extreme El Niño events within a decade; joint maize or soybean losses above 10% in two major exporters in one year; war-risk insurance suspended on a main shipping lane.

#D. Squeezed importer (chronic stress, Chinese price dominance)

• Picture in 2050. Imported protein is dear and unreliable, and Chinese producers, with scale and state backing, capture most of the regional market for microbial and fermented protein. Vietnam depends on China for both soy-competing and fishmeal-competing protein. Domestic production survives where local feedstock, freshness or species-specific function give an edge, or where the state treats capacity as strategic.
• Protein balance. Soybean-meal need of about 7.0 to 10.4 Mt, with efficiency carrying most of the reduction; domestic microbial protein 1 to 3% of high-protein feed protein, more imported estimate.
• Active hubs. The Mekong river belt; Tay Ninh selling sugar; Ho Chi Minh City with Dong Nai at small scale.
• Grows from: 2035 Scenario 1 under stress, or Scenario 2 with state support.
• Signposts. The chronic-stress signposts of Scenario C together with the Chinese-dominance signposts of Scenario B; an export-dumping episode for microbial protein.

Full descriptions: scenarios_2050.csv. All 16 signposts, with thresholds, where to watch and their September 2026 status: signposts_2050.csv and X. Drivers and signals.

#18.3 Where the signposts point in September 2026

The evidence points both ways on both axes, which is why we assign no probabilities.

• Import stress: short-term stress, long-term baseline benign. The 2026 anchoveta season landed about a quarter of a quota that was already the lowest in a decade, and world fishmeal output fell 26% year on year in January to April 2026 ^15,16 generalMedium signal. NOAA gives over 90% odds of a very strong El Niño in winter 2026 to 2027 ⁵ generalHigh projection (published probability). The fishmeal-to-soy price ratio reached 6.2 in August 2026 ¹⁷ generalHigh. Yet the OECD-FAO baseline to 2035 expects ample soy ⁴ generalMedium projection.

• Fermentation market: Chinese scale growing, but not alone. China added large yeast-protein and mycoprotein capacity in 2025 and 2026 ^10,11 VN-adjacentMedium signal, while its flagship gas-fermentation plant halted ¹⁸ VN-adjacentMedium. Thailand, India and Korea now fund public pilot plants and food-tech plans (2026 to 2030) ^12,19,20,21 VN-adjacentMedium signal. US clearances went to Chinese and non-Chinese precision-fermentation firms alike in 2024 to 2025 ^22,23 generalMedium.

#18.4 Wildcards that jump between worlds

Some wildcards would move Vietnam from one world to another within a year (X. Drivers and signals, wildcards.csv).

Sources: ^5,8,10,24,25 generalLow.

#18.5 Stress-testing the sixteen plays

We scored each play from Plays in each world: 2 where it thrives, 1 where it holds with conditions, 0 where it struggles. The scores are our judgement from the evidence in chapters 15 to 17 VN-directLow estimate.

Classes: no-regret (strong in all four); robust (at least viable in all four); bet (weak in at least one world but strong in another: needs signposts before scaling); option (weak in at least one world and strong in none: cheap to hold as research). Reasons for each score: play_robustness.csv.

Readings.

• Public infrastructure is the safest investment. Feed-trial capacity, a protein-quality laboratory and an open cost and price series pay off in every world, because they serve imports and domestic products alike and carry the efficiency lever that chapter 17 found strongest.
• Most business plays are robust but not strong everywhere. Seven of the ten hold in every world with conditions; none thrives in the Comfortable price-taker world.
• Two edition 1.0 plays are bets on one axis each. Bulk fishmeal replacement (T3) needs chronic import stress; high-value precision fermentation (T7) needs an open regional market. Each should scale only when its signposts trip (Robust moves).
• The Security build-out world is the only one where everything works. Readiness matters more than prediction: plants, sites and finance prepared in the late 2020s can scale quickly if that world arrives, and cost little if it does not.

#18.6 What this means

• Investors: underwrite plays that hold in all four worlds and thrive in two (T2, T4, T5, T6, T9). Hold bets on T3 and T7 as options with named triggers: for T3, fishmeal above USD 2,500 per t for 12 months and a pilot cost within 1.2 times fishmeal-protein parity; for T7, a new-food route, an ASEAN reliance mechanism or regional capacity outside China, and an anchor customer (Robust moves).
• Policy makers: the no-regret plays are all public goods and all cheap. Inside the same external world, a new-food route and open pilot capacity decide whether Vietnam gets World A's outcomes or ends up a price-taker, as in World B.
• Startups: species-specific function, local feedstocks and freshness are the edges that survive Chinese price dominance.
• Manufacturers: worlds C and D are the ones that hurt feed mills. Formulation flexibility, multi-origin sourcing and qualified alternative ingredients are the hedges.
• Research and international bodies: tracking the 16 signposts yearly, and publishing them, is a low-cost service to every actor.

Related: Outlook (2035 scenarios), Drivers to 2050, Robust moves, X. Drivers and signals.