2Six recommendations

1. Months 0–6: buy licences, not companies. Sign data-licensing pilots covering about 100,000 articles with two or three private publishers, for a few million dollars, to validate the cleaning pipeline and unit cost without triggering foreign-investment review.
2. Months 6–18: acquire private capacity and the business/management cluster. First choice is Business Perspectives (UA-02, 7.24) plus Virtus Interpress (UA-03, 6.56) as a package; second, the platform, team and content of DKS Center (UA-05) or Helvetica (UA-01) — explicitly without inheriting their titles or editorial boards.
3. For public and academy assets, abandon control and build jointly. Establish co-publishing and delegated-publishing relationships with the National Academy of Sciences (UA-11), Kyiv National University (UA-10) and Kyiv Polytechnic (UA-13). All three score 8 on discipline fit — the highest in the pool — and none can be acquired.
4. Treat journals by tier, not as a bundle. Take tier A (93 titles) as brands worth keeping and investing in; treat tier B (190) as the main co-publishing pool; use tier C (321) purely as licensed corpus; exclude tier D (1,236) entirely. Never buy "all the journals of a publisher".
5. Treat compliance and narrative as part of deal design, not post-closing PR. Sanctions and export controls, Ukrainian public-asset law, and FDI and exchange-control review must be handled up front; public positioning must be limited strictly to "improving the international dissemination and open access of Ukrainian research".
6. Build a two-way Ukraine-China scholarly channel over 36 months. One end is licensable, cleanable Ukrainian-language and English corpus; the other is European distribution and EU open-access compliance for Chinese English-language journals.

Dimension 1 — Would acquisition strengthen a Chinese publisher's global position?

3The benchmark: China supplies content but hosts little of it

China is the world's largest producer of research papers and one of the smallest hosts of them. Measured on a single-year, single-method basis for 2023:

• 812,907 journal articles by China-affiliated authors
• 41,765 articles published in China-hosted journals
• Domestic hosting rate: 5.14%
• Supply-demand ratio: 19.46x

In other words, for roughly every 19.5 papers a Chinese author publishes, only one can appear in a Chinese-hosted journal; the other 18.5 depend on foreign publishers for editing, production, indexing and sales. Independent Chinese official estimates put the same ratio at about 19.27x, closely corroborating this report's 19.46x.

4The only usable precedent: Science Press and EDP Sciences

On 28 June 2019, China Science Publishing & Media Ltd (Science Press, SSE: 601858) announced an agreement with four French learned societies — the French Physical Society, the French Chemical Society, the French Optical Society and SMAI — to acquire 100% of Edition Diffusion Press Sciences SA (EDP Sciences) at an enterprise value of EUR 12 million, cash-free and debt-free. EDP Sciences was founded by Marie Curie, Louis de Broglie and Jean Baptiste Perrin, and in 2018 published 75 journals, 58 in English and 17 in French.

Four transferable lessons and three limits:

Lessons. (1) What is bought is a seat, not earnings — the assets were international content, authors, markets, distribution channels and a platform on which to build a global editorial board. (2) Societies are the best sellers — the shareholders were four learned societies, not a commercial group. (3) Society assets in Europe can be sold outright — so the obstacle is price, narrative and governance, not law. (4) The scale is low tens of millions — this anchors the whole asset class well below trade and educational publishing.

Limits, which are exactly where Ukraine differs. EDP Sciences had a real brand and society relationships, so Ukrainian targets must be valued at a substantial discount to EDP's price per journal. EDP Sciences was a peacetime transaction, so Ukraine adds sanctions, exchange-control and FDI constraints. EDP's journals were internationally indexed while only about 11% of Ukrainian journals are, so buying into Ukraine buys content and capacity, not indexing seats.

5Three routes to global position

Route 1 — portfolio scale. Easy to buy, and the gain is diluted by quality. Citations per article across Ukrainian journals are 0.88; the largest single title by volume holds 90,263 articles and just 1,213 citations. The reputable titles are small: Eastern-European Journal of Enterprise Technologies (7,115 articles, 35,894 citations), SIGMA (2,449 / 18,853), Corporate Ownership and Control (3,058 / 19,469).

Route 2 — platforms and rules. This is where the largest gain lies and where acquisition works least. Ukraine's infrastructure layer — OUCI (the national citation index built on Crossref Cited-by and I4OC) and URIS/NRAT (the ministry's registration and full-text systems) — are public assets that cannot be acquired; only technical cooperation and standards work are available. OUCI scores highest in the whole pool on global-position gain (8/10), yet only 6/10 on tradability.

Route 3 — content and corpus. This route monetises most directly, and should not anchor valuation. The industry benchmarks reported publicly put large publisher–AI licensing in the tens of millions of dollars annually, with individual first-year deals below that. Ukrainian-language scholarly corpus is therefore reasonably valued in the low millions — not enough to support a brand-premium acquisition.

6Conclusion for Dimension 1

Acquisition can materially strengthen a Chinese publisher's global position, but only if four conditions hold: the buyer accepts that Ukraine can add volume and a European capacity interface but not prestige; the objective is reframed from buying publishers to acquiring capacity, relationships and data; the buyer accepts the cost of performing internationalisation itself; and reputational risk is treated as part of the transaction. The correct replication of the EDP Sciences precedent is low-price, dispersed and multi-form — licensing, delegation, joint venture and minority stakes — not a single premium controlling acquisition.

Dimension 2 — Can these targets fill disciplinary gaps?

7China's gap, by discipline (2023)

Three observations. The largest absolute gap is in medicine and the life sciences. The most severe relative gaps are in social science and business — a block long neglected by Chinese publishers, with gap ratios far higher than in the natural sciences. And there is a third, invisible gap: area studies and Global South knowledge, where the primary sources, local journals and indigenous scholarly discourse that Chinese area-studies programmes need are precisely what international publishers do not collect. Ukraine sits in that position.

8What Ukraine can actually fill

Business and social science — the strongest match. This is the only combination in the pool that is simultaneously a large Chinese gap, credible Ukrainian quality, and acquirable. Business Perspectives holds 12 journals with 9 DOAJ certifications, including Scopus-indexed titles with close to 20,000 cumulative citations; Virtus Interpress has a stable international authorship in corporate governance and risk. Both are private, both are in Sumy, both run the same model, and they can be bought as a package. This is this report's recommended first deal. Notably, business and management journals are over-represented in the top tier: 10 of the 30 highest-cited Ukrainian journals are in business, finance or management.

Engineering and agriculture — largest scale, quality needs screening. Kyiv Polytechnic holds 20 journals covering mechanical engineering, materials, electronics and energy, 8 of them DOAJ-certified — the best disciplinary fit in the pool — but it is a public university and cannot be acquired. Eastern-European Journal of Enterprise Technologies (7,115 articles, 35,894 citations) is among the few Ukrainian titles with real international reach. Honest caveat: Ukraine's engineering strength is in applied and traditional mechanical/materials work, not in semiconductors, advanced materials or AI.

Area studies and Soviet-legacy basic science — highest strategic value, most undervalued. SIGMA, Ukrainian Journal of Physics and Condensed Matter Physics carry genuine international reputations, representing the theoretical-physics and mathematical-physics depth of the Soviet tradition. Alongside them sits a body of Ukrainian- and Russian-language area-studies material that no international database indexes systematically.

9The language window

Three implications. Ukrainian is now a stable second language of Ukrainian scholarship (26,161 articles in 2024, about 34%), and Ukrainian-language scholarly text has almost no competing source worldwide. Russian-language output from Ukraine is disappearing, so the ten-year transition corpus is a time-limited collection opportunity. And because English remains the majority language, a buyer obtains Ukrainian and English text together, which is more valuable for multilingual training than a single-language corpus.

10What Ukraine cannot fill

Advanced materials, semiconductors and advanced manufacturing; top-tier medical journals (Ukraine can supply medical volume, not medical prestige); and high-end chemistry and chemical engineering. In these fields the gap requires European or American assets of the EDP Sciences type, or domestic build-out.

Dimension 3 — Are the targets actually for sale?

11The defining mismatch

Ukraine's academic publishing assets must be split into two classes, and the two behave in opposite ways.

The acquirable class: private commercial publishers. The journal system is highly fragmented — 63% of journals have no recorded publisher — but standing on top of that fragmentation is a group of wholly private limited companies. Their shareholders can decide to sell. There is no approval layer, no public-asset procedure, no membership vote. Publishing House Helvetica alone holds 139,409 DOIs.

The non-acquirable class: academy and public university assets. The highest-quality part of the system — the National Academy of Sciences, Kyiv National University and Kyiv Polytechnic — belongs to the state or to public institutions. Ukrainian law restricts the transfer of strategic and public assets to foreign buyers, and wartime conditions add political scrutiny.

In other words, the obstacle is not that nothing may be sold — it is that the good assets may not be sold, while the sellable assets are mediocre. That mismatch, not price, is the binding constraint. It dictates the deal form: use licensing and co-publishing to reach the good content, and use equity only to acquire capacity and corpus.

A second constraint must be handled up front: wartime capital and exchange controls, sanctions compliance, and Chinese outbound-investment (ODI) filing together push closing beyond 12 months. These can be partly ring-fenced by deal structure, but they cost time.

12Four classes of target

Class 1 — private commercial publishers: buyable, and worth buying.

These are worth buying not because the journals are good — most are mediocre — but because a purchase yields four things at once: a running publishing platform and team; a DOI-registered, structurally complete content stock; a local operating entity and bank account; and a network in the local scholarly community. What must be discounted is the brand: 12,642 articles at Zaslavsky carry roughly 2,000 citations. Valuation should rest on cleanable article count plus platform replacement cost, never on journal count times a per-title price.

Class 2 — public university publishing: legally not acquirable. Kyiv National University (53 journals, the largest cluster), Kyiv Polytechnic (20, engineering), Drohobych (11,038 articles), Ternopil (91,174 articles — the largest single content pool, but only 1,213 citations). These are the highest-quality assets and cannot be bought. Only three routes are available: delegated publishing services, co-publishing of English editions, and data licensing.

Class 3 — academy and professional associations: theoretically negotiable, practically very hard. The National Academy of Sciences (Ukrinformnauka, 48,075 DOIs, holding SIGMA and Ukrainian Journal of Physics) and the Ukrainian Assembly of Doctors of Science in Public Administration (48,196 DOIs, the second-largest registrant nationally). One undervalued opportunity: the academy's flagship publisher Naukova Dumka, which operated from 1922, ceased operations in 2024, leaving scholarly publishing assets in disposal. This is the only case in the pool of a high-reputation institutional asset actually being disposed of, and deserves top-priority diligence — though domestic Ukrainian parties may hold a right of first refusal.

Class 4 — platforms and infrastructure: not acquirable, but must be partnered with. OUCI, URIS and NRAT are public systems. The strategy is technical supply and standards cooperation: provide data conversion, DOI registration services, international distribution interfaces and long-term preservation, becoming part of the system without touching ownership.

13Five executable transaction forms

Dimension 4 — Value for Chinese AI companies' corpora

14Correcting a common misjudgement

The most common error is to confuse accessibility with licensability. Ukrainian content is exceptionally accessible: 1,338 of 1,840 journals (72.7%) are open access, covering 818,174 articles; 546 journals are DOAJ-certified; 393 (72%) use CC BY, the licence best suited to text and data mining and model training; 336 (61.5%) charge no article-processing fee; and 373 (68%) have preservation arrangements.

A principle stated explicitly: every corpus pathway discussed in this report rests on authorised licensing, open licences (CC BY) and statutory text-and-data-mining exceptions. It does not involve, and does not recommend, circumventing licences, technical protection measures or copyright. Being lawfully licensable is precisely what gives this corpus its commercial value.

15The corpus funnel

Four structural conclusions. The largest loss comes from licence clarity, not readability — the drop from 88.1% open access to 35.4% DOAJ-certified costs about half a million articles, because much content is free to read but carries no clear licence, or a licence unsuited to commercial training (CC BY-NC and CC BY-ND account for 125 titles). The second loss comes from machine-readability: most Ukrainian journals run on OJS, but structured full text is another matter, and PDF layout files are worth far less for training than JATS XML. The most insidious risk is the rights chain: with 63% of journals lacking a recorded publisher, there is no counterparty to sign with for 518,084 articles. Finally, exclusivity is the value watershed — if anyone can train on the same CC BY text, its marginal value to a buyer reduces to cleaning quality and time advantage.

16What the corpus is worth, and how much of valuation it should carry

Industry benchmarks put large publisher–AI licensing in the tens of millions of dollars annually, individual first-year deals below that, and the entire EDP Sciences asset — 75 journals, real brand, society relationships — at EUR 12 million in 2019. Against those anchors:

Corpus is the most tangible synergy of this transaction, but it is not the basis of valuation. The right use is as a negotiating instrument to lower consideration and bind long-term cooperation — not to justify an inflated price.

Supplementary dimensions: operability, valuation and deal structure

17Operability

The most underrated asset is the prevalence of OJS, because it means technical integration cost is far lower than expected — no production system needs rebuilding, only JATS XML conversion and platform interfacing. The corresponding risk is "buying a shell": because journal production depends on individuals rather than organisations, diligence must verify key editorial and production staff by name, or the buyer may acquire titles without capacity. Facilities in Kharkiv and Kherson have suffered damage and institutions continue to relocate; valuation should carry an explicit 15–30% discount for operational-continuity risk.

18Valuation

Academic journal assets have no active market, so valuation combines comparables with an asset approach, using three anchors.

Anchor 1 — EDP Sciences (the only high-quality comparable). EUR 12 million for 75 journals implies about EUR 160,000 per journal, or roughly EUR 20–40 per article. These cannot be applied directly to Ukraine for three reasons: EDP's journals were mostly internationally indexed while only ~11% of Ukrainian journals are; EDP had real brands and society relationships; and EDP was a peacetime transaction. This report applies a composite discount factor of 0.15–0.35.

Anchor 2 — price per article (best for corpus-type targets).

Anchor 3 — platform and capacity replacement cost. Rebuilding a production system serving tens of thousands of articles is estimated at 1–3 years and several million renminbi; this is the floor value of the platform in a Helvetica-scale target.

This report recommends capping the first deal at USD 10 million, because diligence costs are high, the exit path is unclear and there is no successful precedent; validating assumptions with a small transaction is far more rational than a single large controlling acquisition.

19Deal structure: four tools

Tool 1 — licence first, equity later. Phase one signs a 3–5 year data-licensing agreement with a right of first refusal; phase two executes the equity purchase after diligence and value validation. This avoids FDI review, validates corpus quality with real data, builds trust, and caps the downside at the licence fee if quality disappoints.

Tool 2 — asset purchase rather than share purchase. Acquire the platform, technology, employment contracts, content and data assets while explicitly excluding the original titles, editorial boards and editorial policy. This is the most effective tool against academic-integrity risk: inheriting the titles means inheriting the reputational exposure.

Tool 3 — joint venture with a local party. Establish a JV with a Ukrainian partner (or a university-affiliated entity), with the Chinese side taking a minority or equal stake and contributing technology, platform and capital, while the local side contributes content and relationships. This is the only lawful way to reach public university and academy assets, and it defuses the "foreign control of Ukrainian scholarship" narrative.

Tool 4 — staged consideration with milestone binding. Pay 40% at closing, 30% on delivery of converted, cleaned corpus, and 30% when target journals achieve DOAJ or Scopus status. This converts corpus quality and internationalisation from due-diligence assumptions into payment conditions.

Compliance checklist, to be handled before signing. China side: ODI filing, foreign-exchange registration, state-asset supervision (for state-owned buyers), listed-company disclosure. Ukraine side: wartime capital and exchange controls, antitrust filing, public-asset transfer procedure where applicable, sanctions screening. Third-country: US secondary-sanctions exposure — the single item most in need of prior legal opinion. Academic governance: editorial-independence guarantees, editorial-board retention, and a COPE compliance undertaking — the three terms Western scholarly communities watch most closely.

Targets: publisher ranking and the journal tier list

20The scoring model

Six weighted dimensions: global-position gain 22%, discipline fit 20%, tradability 20%, corpus value 16%, operability 12%, risk 10% (low risk scores high). Scores run 0–10.

The most important column is not the rank but the type. Seven of the top nine are private companies or partnerable infrastructure; the three highest-scoring targets on discipline fit (UA-11, UA-10 and UA-13, all 8/10) rank only 10th, 11th and 13th, for one reason: tradability of 2/10.

Note the Helvetica paradox. It holds the largest DOI stock in the country and scores 9/10 on tradability — the highest in the pool — yet ranks only 9th, because its global-position gain is 3 and its discipline fit is 4. This is the clearest proof that in Ukraine tradability and value are decoupled. Buy Helvetica for its pipeline, team and content, not for its titles.

21The journal tier list: which of the 1,840 are worth buying

The user's question also concerns the journals themselves. The answer is: never buy "all the journals of a publisher" — tier them and take only a minority.

Tier A — the 93 journals that carry the reputation

Tier A holds 176,000 articles and 440,134 citations: 49.6% of all citations from 5.1% of titles.

Science, technology and engineering

Mathematics and theoretical physics — small, prestigious, irreplaceable

Business, management and finance — China's largest gap, and the most tradeable

Tiers B, C and D

Tier B (190 journals) has verifiable impact (cumulative citations of at least 500 and at least 1.0 per article) and clear licensing, but has not yet built international visibility. Representative titles include Corporate Governance and Organizational Behavior Review (1,909 citations, 4.81 per article), Modern Economics, Science and Innovation, Biotechnologia Acta, Radioelectronic and Computer Systems, International Journal of Endocrinology (Ukraine), Medicni Perspektivi and Ukrainian Society. Their value is appreciation potential: they already have basic quality and citation traction and lack only English publication standards, JATS XML production, an international editorial board and a Scopus application — precisely the capabilities a Chinese publisher can supply at scale. Tier B is the main pool for co-publishing English editions.

Tier C (321 journals) holds the largest article count of any tier (352,541) but weak impact. It yields corpus, not seats. Principle: price per article, take no titles, inherit no editorial boards.

Tier D (1,236 journals) averages about 207 articles per title; 270 are DOAJ-certified but negligible in scale and impact. Per-title diligence would cost more than the asset is worth. Access their content, if at all, only through bulk CC BY licensing.

Three operating rules that follow. For tier A, prefer buying titles rather than bundles — the brand and editorial board are themselves assets, and where an A-tier journal sits inside a publisher target (such as Business Perspectives' nine DOAJ titles) it should anchor the consideration. For tiers C and D, prefer buying content rather than titles — the transaction documents should state explicitly that titles, editorial boards and editorial policy do not transfer. For tier B, what is bought is the right to renovate, so use staged consideration bound to milestones such as JATS XML delivery, maintenance of DOAJ certification and Scopus acceptance.

22Recommended combination

The five layers total an estimated USD 10–25 million — roughly one to two EDP Sciences transactions — but what is acquired is not a single brand but a sustainable European content-access system.

Risk matrix and compliance design

Sanctions and export controls (high). Ukraine is a war economy. A Chinese buyer deemed to facilitate sanctions evasion faces secondary sanctions exposure, determined by the target's historical counterparties, fund flows and customer base — several Ukrainian publishers have Russian-language or Russian-institutional histories that would transfer to the buyer. Required: OFAC/EU/Ukrainian sanctions screening, historical counterparty checks, an opinion from counsel competent in US sanctions law, and representations, warranties and indemnities in the agreement.

Public assets are non-transferable (high). The highest-quality parts of the system — the academy and public universities — cannot be sold to foreign buyers, and 63% of journals have no recorded owner. Required: establish ownership character in the first diligence phase, exclude public assets from equity acquisition, and pivot to delegation, co-publishing and licensing.

Academic integrity and predatory publishing (high). Ukraine is mid-way through an anti-predatory reform: MES Order No. 56 of 19 January 2026 rewrote the register rules, and Order No. 928 of 11 June 2026 set a 1,298-title Category B list, rejecting about 7% of applications, with all listed journals open access and 53% diamond. A Nature investigation into the "Tanu.pro" paper mill alleged about 1,500 compromised articles. Required: buy content and platform, not titles; sample-check self-citation, review times and citations; commit to COPE standards.

Public opinion and academic freedom (high). The core asset is the scholarly community's trust. There is an established reaction path: in September 2026, editors of a French journal resigned en masse after a change of ownership. Required: strict narrative discipline, contractual editorial-independence guarantees, retention of the existing editorial board, and preference for joint ventures and minority stakes over control.

A practical test: if a proposed deal contains no clause written specifically to address the four high risks above, the proposal is not yet mature.

Roadmap, KPIs and conclusion

Months 0–6 (USD 0.5–2.5m): licensing pilots and diligence. Three data-licensing agreements covering about 100,000 articles with an option over equity; build the PDF to JATS pipeline and measure unit cost; commission sanctions and ownership diligence on 20 candidates; sample-check journal quality against the tier list; open relationships with Kyiv National University, Kyiv Polytechnic and the academy. KPIs: 3 licences signed, 20 targets diligenced, unit cleaning cost measured, first equity target selected.

Months 6–18 (USD 5–15m): acquire capacity and platform. Complete the Business Perspectives plus Virtus Interpress package; acquire the platform, team and content of one or two private publishers without titles; sign co-publishing and delegated-publishing agreements with two or three public universities, starting from tier B; scale licensed corpus to 500,000 articles; dispose of tiers C and D in the acquired portfolio. KPIs: 1 equity deal closed, 3–5 delegated contracts, 500,000 licensable articles, JATS conversion completed.

Months 18–36 (USD 5–10m): embed into the system. Participate in OUCI/URIS as a technical service provider and international distribution node; reach a million-article licensable corpus; operate the two-way Ukraine-China channel; push tier A and B journals into Scopus and DOAJ. KPIs: 1m licensable articles, two-way channel operating, 10+ new Scopus titles, a replicable international publishing capability.

Stop conditions. Abandon or suspend if sanctions exposure cannot be ring-fenced; if ownership cannot be established; if systematic integrity problems cannot be isolated by buying content without titles; if key editorial staff refuse to stay; if measured corpus quality falls materially below the 35% XML assumption; or if the transaction becomes systematically framed as foreign control of Ukrainian scholarship.

In one sentence: Ukraine is not a market where buying one big brand changes the game. It offers a chance to buy, at relatively low cost, the capacity, relationships and structured data that others cannot replicate — and to plug them into one's own international platform. The real barrier is not money; it is the professional capability to design deal structures, ring-fence compliance and manage narrative.

Appendix A — Method, tier rules and limitations

Sources. All quantitative findings are reproducible from public data: OpenAlex (sources, works, publishers), Crossref (members), DOAJ, and the Ukrainian Ministry of Education and Science (Orders No. 56 and No. 928 of 2026), together with Open Science in Ukraine (Scopus/WoS lists) and the China Science Publishing & Media announcement of 28 June 2019.

Key definitions. Journals = OpenAlex sources records with the given country_code and type:journal. Articles = works aggregated by primary_location.source.id. Citations per article = total cited_by_count divided by total works_count. Supply-demand ratio = a country's authored output in the reference year divided by the capacity of journals hosted in that country in the same year. Domestic hosting rate = the reciprocal. All cross-country comparisons are year-aligned to 2023 and apply the same merge-artefact filter (sources above 200,000 articles excluded).

Six-dimension model. Global position 22%, discipline fit 20%, tradability 20%, corpus value 16%, operability 12%, risk 10% (low risk scores high). Composite = weighted sum, 0–10.

Journal tier rules.

Two caveats. Citations per article is not the whole of quality — it is heavily discipline-dependent (mathematical physics is structurally lower than life sciences), so tier A uses a disjunction of cumulative or per-article citations to avoid systematically underrating smaller fields. And DOAJ certification enters only the tier-B test, never tier A on its own: measured, 270 of the 546 DOAJ journals fall into tier D, because sound licensing is not the same as excellent scholarship.

Limitations. (1) This report does not assert that any target is actually for sale. All "tradability" scores are analytical inferences from ownership character and public information, not statements that a sale process exists. Financial and shareholding data for unlisted small publishers are generally not public. (2) The corpus funnel's last two levels (35% XML readiness, 45% exclusivity) are conservative assumptions, not measurements, and must be replaced by sampling in diligence; all other levels are measured. (3) OpenAlex publisher attribution is incomplete, so "no recorded publisher" does not mean "no sponsoring body". (4) Valuation ranges are analytical estimates; only the EUR 12 million EDP Sciences price is a disclosed comparable, and none of this constitutes investment advice. (5) Wartime capital controls, sanctions rules and Chinese outbound-investment policy may change rapidly and must be reconfirmed at the time of any transaction.