Genemap Insights How it works Try Genemap Sign in
πŸ‡°πŸ‡· Country deep-dive Β· Production systems Β· ~12 min read Β· 11 May 2026

The Korean Hanwoo system, properly modelled.

Hanwoo is Korea's domestic beef breed β€” small, intensely marbled, fed to 30 months on a barley-and-rice-bran ration, sold through a grading system where one marbling tier above neighbour can be the difference between an unprofitable kill and a record price. The bioeconomic shape is unlike anywhere in the Western beef world. Here's how the engine models it correctly.

3.4M
Hanwoo on hand Β· Korea 2025

A small national herd by global standards β€” Korea imports more beef than it produces β€” but the per-head economics are 2–4Γ— any Western beef system. A 1++ A graded Hanwoo carcass clears β‚©20m+ at wholesale; a 3-grade carcass clears β‚©9m. The genetics that move animals from 3 to 1+ to 1++ to 1++ A are the platform's load-bearing economic axis.

The grade economics, in numbers.

The Korean Beef Quality Grading System runs on two orthogonal axes: a Yield Grade (A / B / C, based on carcass yield) and a Quality Grade (1++ / 1+ / 1 / 2 / 3, based on marbling, meat colour, fat colour, texture and maturity). The combination is what the consumer-facing market actually pays for. Marbling drives the quality axis almost entirely; the platform's bioeconomic anchor treats the MARB EBV (or KAPE equivalent) as the single most economically valuable trait in Korea.

GradeMarbling scoreAverage wholesale (β‚©/kg cwt)Share of national kill
1++ A8–9 (extreme)β‚©28,900~12%
1++8–9β‚©25,700~20%
1+6–7β‚©22,300~26%
14–5β‚©19,100~22%
22–3β‚©15,400~14%
31β‚©12,200~6%

The premium between adjacent grades is not linear, and the premium for 1++ A over 1++ (same marbling score, just better yield grade) is itself ~12% β€” that's a structural feature of the grid most US/AU/EU pricing tables don't have an analogue for. Producers chase the top tier hard. So does the platform's bioeconomic anchor when the producer's market is Korea.

What Korean producers actually select on.

KAPE (Korea Animal Products Quality Evaluation Service) publishes the breed evaluation for Hanwoo. The trait set is similar to BREEDPLAN or BIF in structure but with weight shifted toward carcass quality and away from growth. The platform reads it natively (Tier 1 ingestor) and re-weights for the producer's specific operation.

The 30-month finish window.

Korean Hanwoo are typically slaughtered at 28–32 months, ~720 kg liveweight, after a long intensive finish on barley, rice bran and concentrate. That's roughly twice the time-on-feed of a US Choice-Prime finish and ~40% longer than even a long-fed Wagyu programme. The feed cost is correspondingly high β€” the breakeven per head is driven primarily by getting the 1++ premium, secondarily by feed-conversion efficiency over that long window.

This shapes the engine's selection objective. Growth-trait weights are softer than US systems (because the producer is paying to grow the animal for an extra year either way), but marbling-trait weights are 3–5Γ— heavier. The production-system modifier "hanwoo_intensive" in the platform automatically applies this re-weighting; any KAPE DGV report or imported Hanwoo bull catalogue feeds straight into a Korean-shape rank page.

Trait notation, side by side.

The platform's translator handles the breed-society semantic differences. The mapping for the Korean audience reading reports across countries:

Canonical (Genemap)AU / NZ (BREEDPLAN)US (BIF / IGS)Korea (KAPE)
Marbling scoreIMF (EBV)MARB (EPD)MS / 근내지방도 (DGV)
Carcass weightCWT (EBV)CW (EPD)도체쀑 (DGV)
Eye muscle areaEMA (EBV)REA (EPD)등심단면적 (DGV)
BackfatRIB (EBV)FAT (EPD)λ“±μ§€λ°©λ‘κ»˜ (DGV)
Growth rate (yearling)400WT (EBV)YW (EPD)μΌλ‹Ήμ¦μ²΄λŸ‰ (DGV)
Calving easeCE-DIR (EBV)CED (EPD)λΆ„λ§Œλ‚œμ΄λ„ (DGV)
Cow longevity / stay-ability(not natively published)STAY (EPD)λ²ˆμ‹μˆ˜λͺ… (DGV)

What that looks like on a real-shape Hanwoo operation.

The numbers below are derived from a real-shape 280-head Hanwoo cow-calf-finish operation in Gyeongsangbuk-do, finishing all progeny on a barley-and-rice-bran ration to 30 months, with a typical 1++ achievement rate of ~38% and a 1++ A rate of ~14%.

TraitIndustry-default weightProducer-fit weightWhy the shift
MS (marbling)β‚©68,500/scoreβ‚©89,200/scoreThis producer's 1++ β†’ 1++ A conversion rate is 36%, materially above national average.
CW (carcass weight)β‚©2,150/kgβ‚©2,080/kgSlight softening β€” producer is consistently above the 720 kg target so marginal CW is less valuable.
EMA (eye muscle)β‚©9,400/cmΒ²β‚©11,700/cmΒ²Better-yielding carcasses compound the 1++ A premium for this producer.
BF (backfat)βˆ’β‚©4,200/mmβˆ’β‚©6,800/mmThis producer's grid penalty for excess backfat is steeper than national average.
WG (growth rate)β‚©140/g/dayβ‚©115/g/dayThe producer's feed-conversion efficiency is below average β€” slower-growing animals can still hit 1++ economically here.
STAY (cow longevity)β‚©68/dayβ‚©68/dayIndustry default carries through.

The pattern is consistent with what the closed-loop calibration loop is designed to find: this producer's calibration multipliers reflect their specific grid achievement, their specific feed-conversion economics, and their specific cow-cycle structure. Marbling is even more valuable here than on the national average; growth is slightly less valuable; backfat is more penalised. A KAPE-published Hanwoo bull catalogue feeds through this re-weighting on the rank page and produces a meaningfully different top-10 list than the same catalogue would on a national-average operation.

The national grading system pays for marbling. The producer's specific grid pays for marbling even more. The engine reflects both.

What's wired natively for Korea.

As of May 2026, the platform reads the following Korean sources natively:

What this means in code.

The production-system modifier for Korean Hanwoo lives alongside the other 12 archetypes. Simplified shape:

// core/js/production-system-modifier.js β€” Hanwoo intensive overlay
SYSTEM_MODIFIERS['hanwoo_intensive'] = {
  // Hanwoo MS economic weight multiplier β€” 3.5Γ— the temperate-system default.
  // Driven by the 1++/1++A grade premium dwarfing any Western marbling premium.
  marbling: 3.5,

  // Eye muscle area compounds with marbling on the yield-grade axis
  ema: 1.4,

  // Growth softens β€” the 30-month finish window means marginal growth
  // matters less than getting the carcass quality right
  yw_400: 0.55,

  // Backfat penalty steeper β€” Korean grid penalises excess fat heavily
  rib_backfat: 1.8,

  // Cow size penalty stays at default β€” extensive land cost is the constraint
  mcw: 1.0,

  // Heat tolerance ingestible if KAPE publishes it (R&D pilot 2026)
  heat_tolerance: 1.0,

  // Documented in: insights/korean-hanwoo.html, methodology.html
};

This single overlay tells the closed-form bioeconomic derivation that this producer's marbling is worth 3.5Γ— what it's worth on the industry-default temperate template. The producer-fit Ξ²Μ‚ multipliers from the closed-loop calibration sit on top, further sharpening to the producer's actual grid achievement.

Where the platform can go further for Korean producers.

References cited inline: Lee et al. (2014) Journal of Animal Science 92: 3683–3692 β€” Hanwoo genetic parameters; Park et al. (2020) Asian-Australasian Journal of Animal Sciences 33: 1083–1089 β€” KAPE evaluation methodology; Wagyu and Hanwoo comparative work in Yoon et al. (2018) Animal Production Science 58: 1538–1545. Full Genemap bibliography at research.html. KAPE_Hanwoo native ingestor and DGV semantic mapping in core/js/catalogue-parsers.js (registered in core/js/eval-system-translator.js); the hanwoo_intensive_30mo and hanwoo_intensive production-system modifiers in core/js/production-system-modifier.js.

Acknowledgement: the engine's Korean DGV semantics were built against publicly-available KAPE data and benefit from substantial published work by the National Institute of Animal Science (NIAS) and Seoul National University quantitative-genetics groups. The platform welcomes academic collaboration with Korean teams; reach the engineering team via for-researchers.html.