KOSDAQIT & Software477850

MakinaRocks

₩22,900▲ 0.44%2026-10-02 close
Market Cap
₩403.9B
Turnover
₩6.8B
Volume
300K
Shares out.
17.8M
PER
—
PBR
6.9×
EPS
—
Dividend Yield
—

PER, PBR and dividend yield are calculated from the latest confirmed results (EPS, BPS, dividend per share) and the current share price · Prices as of the 2026-10-02 close

01

Report overview

Orders Racing Ahead; Break-Even Set for 2027

Q1 2026 revenue surged 167.6% YoY powered by its closed-network industrial AI OS 'Runway', yet sustained operating losses and an IPO valuation anchored to 2028 earnings estimates make the 2027 break-even target the critical re-rating inflection point.

  1. 1

    Runway holds a near-monopolistic position as an enterprise AI OS for closed-network industrial environments, creating structural barriers that cloud-centric competitors cannot overcome

  2. 2

    Q1 2026 revenue of KRW 3.0B (+167.6% YoY per DART), extending a three-year top-line growth run: KRW 5.2B (2023) → KRW 8.3B (2024) → KRW 11.5B (2025) per recent reports

  3. 3

    Q1 2026 new orders of KRW 7.5B (~2.8x YoY); secured annualized backlog of KRW 13.1B as of March 31 covers more than half the KRW 22.5B full-year target

  4. 4

    Blue-chip clients include Samsung, Hyundai, SK, LG, and key defense agencies (ADD, Joint Chiefs of Staff), with strategic investors such as SKT, Hyundai Motor, and Naver reinforcing the commercial ecosystem

  5. 5

    IPO pricing based on 2028 estimated earnings embeds significant forward expectations; the 2027 break-even target is the key re-rating trigger while lock-up expiry overhang remains a recurring supply risk

02

Business structure

MakinaRocks was co-founded in 2017 by CEO Yoon Seong-ho (MIT physics, formerly Samsung Electronics and SK Telecom) and is an industry-specialized AI company targeting demanding AI deployment environments across manufacturing, defense, energy, semiconductor, and battery sectors.

Its flagship product 'Runway' is an enterprise AI operating system covering the full AI lifecycle—from data collection, storage, and processing to model development, deployment, and management—with integrated MLOps, LLMOps, RAG, on-premise/on-device AI, and multi-agent AI capabilities.

The platform's defining competitive edge is its ability to deliver cloud-comparable AI operations within air-gapped closed networks and factory server environments, addressing structural constraints that cloud-centric competitors cannot overcome in settings where data exfiltration is prohibited and security requirements are extreme.

According to recent reports, approximately 68% of 2025 revenue originated from the top-5 manufacturing conglomerates and the defense sector, with clients including Samsung Electronics, Samsung Display, Hyundai Heavy Industries, SK On, Hyundai Mobis, the Agency for Defense Development, Joint Chiefs of Staff, and the ROK Navy First Fleet.

The revenue mix is currently divided between annual and monthly license fees and project-based solution sales; the company's strategic priority is shifting toward a higher proportion of recurring Runway license revenue to improve earnings predictability and margin profiles over time.

Beyond Runway, the company supplies 'DrawX', an AI application automating drawing-based cost estimation workflows, and has deployed over 6,000 industry-specific AI models to date.

Domestically, comparable players include Nota, Superb AI, and OnePredict—all listing or preparing to list—yet MakinaRocks maintains a meaningful head start through its closed-network specialization and blue-chip reference roster.

A strategic investor roster spanning SK Telecom, Hyundai Motor, Naver, LG Technology Ventures, Hanwha, GS, and POSCO collectively functions as a structural commercial network asset, facilitating customer introductions and ecosystem expansion across key domestic industries.

03

Recent trends

Per DART disclosures, Q1 2026 revenue reached KRW 3.0 billion, a 167.6% YoY increase, while the operating loss narrowed to -KRW 2.4 billion (+24.6% improvement YoY) and the net loss similarly improved to -KRW 2.4 billion (+30.0% improvement YoY), signaling that loss leverage is beginning to moderate as the revenue base expands.

According to recent reports, full-year 2025 revenue was approximately KRW 11.5 billion with an operating loss of approximately KRW 8.0 billion, continuing a three-year top-line growth trajectory from KRW 5.2B (2023) to KRW 8.3B (2024) to KRW 11.5B (2025).

On the order front, per recent reports, the company recorded KRW 20.5B in new orders in 2025, followed by KRW 7.5B in Q1 2026 alone—approximately 2.8x the year-ago figure of ~KRW 2.7B—with the secured annualized revenue backlog standing at KRW 13.1B as of March 31, 2026.

However, meeting the full-year KRW 22.5B target requires additional H2 order wins and contract execution beyond the current backlog.

On the stock price front, MakinaRocks debuted on KOSDAQ on May 20, 2026, reaching KRW 60,000—the 400% 'fourfold' ceiling—on its first trading day, with public offering subscription deposits reaching approximately KRW 13.9 trillion, the highest of any 2026 IPO.

The physical AI sector premium compressed rapidly thereafter: shares declined to approximately KRW 41,050 by May 27 and stand at KRW 26,450 as of June 5, representing a ~56% drawdown from the IPO-day high.

Nonetheless, the current price still carries a ~76% premium over the IPO offering price of KRW 15,000, indicating that meaningful forward growth expectations remain embedded in the market valuation.

04

Outlook

The medium-term core scenario centers on scaling recurring Runway license revenue and reaching break-even in 2027.

Based on company-provided guidance, internal targets stand at KRW 22.5B for 2026 and KRW 37.3B for 2027—thresholds at which management believes BEP is achievable—and Q1 order acceleration has delivered an early positive signal toward this path.

Of the approximately KRW 36.4B raised through the IPO, roughly 77% has been earmarked for Runway platform enhancement, proprietary foundation model development, and agentic AI R&D, while the remaining ~23% funds overseas market expansion centered on Japan.

In Japan, the local subsidiary established in April 2025 has already secured four customers in automotive and industrial machinery within less than a year; in Europe, a partnership with KUKA subsidiary Device Insight provides an initial commercial foothold.

The structural acceleration of AI adoption in manufacturing and defense—sectors where closed-network, on-premise deployments are often non-negotiable—supports the medium-to-long-term demand thesis for Runway.

The approaching pipeline of physical AI peer IPOs (Superb AI, InEase, OnePredict) could sustain sector-level interest but may also dilute investor attention and intensify competitive dynamics over time.

The pace of H2 2026 order wins and the rate at which recurring Runway license revenue displaces project-based solutions will be the key determinants of 2027 BEP path visibility.

05

Bull factors

Structural scarcity of closed-network AI OS

Runway is effectively the only enterprise AI OS supporting closed-network environments in security-critical manufacturing and defense settings, occupying a space that cloud-centric peers structurally cannot penetrate.

The inability of data-exfiltration-restricted factory environments and defense-classified networks to adopt public cloud solutions creates a near-moat effect for the company.

An already-validated reference base spanning Samsung, Hyundai, LG, SK, and major defense agencies generates powerful commercial leverage in new customer acquisition and creates high switching costs that reinforce customer retention.

Steep order growth and robust forward pipeline

According to recent reports, new orders of KRW 20.5B in 2025 were followed by KRW 7.5B in Q1 2026 alone (~2.8x YoY), accelerating the leading indicators of future revenue growth.

The secured backlog of KRW 13.1B as of March 31, 2026 already covers a significant portion of the full-year KRW 22.5B target, providing early revenue visibility.

Order growth consistently outpacing revenue recognition creates a structural pipeline advantage that supports confidence in revenue growth continuity across subsequent quarters.

Defense AI expansion and strategic investor ecosystem

Selection as one of the '2025 Defense Innovation Companies 100' (the sole entrant in the AI platform category) and confirmed engagements with the Agency for Defense Development and Joint Chiefs of Staff reflect a privileged position in the high-security defense AI market.

Recent industry analysis indicates that defense AI companies command a valuation premium multiple times that of commercial peers, with predictable government procurement cycles providing structural revenue stability.

The strategic investor network—spanning SK Telecom, Hyundai Motor, Naver, and POSCO—acts as a commercial 'super-reference' ecosystem, generating incremental domestic and overseas business opportunities that pure-play competitors lack.

06

Bear factors

Persistent losses and accumulated deficit burden

Per DART disclosures, the Q1 2026 operating loss of -KRW 2.4B shows an improving trend, yet achieving 2027 break-even requires concurrent revenue growth and cost efficiency gains over the next 1-2 years.

According to recent reports, the accumulated deficit as of 2025 stood at approximately KRW 65.3B (~6x annual revenue), posing a meaningful financial strain, and sustained operating cash outflows could exhaust IPO proceeds and eventually necessitate additional capital raises.

The IPO pricing methodology reliant on 2028 estimated earnings introduces inherent execution sensitivity: any delay in reaching projected profitability would likely precipitate downward valuation pressure.

Elevated valuation and lock-up overhang

At the current price of KRW 26,450 and a market capitalization of approximately KRW 500 billion, the stock trades at a PSR exceeding 40x on 2025 revenue per recent reports, reflecting a substantial forward growth premium already priced in.

According to recent reports, 38.5% of total shares were freely tradeable immediately following listing, and upcoming lock-up expiry events are expected to generate recurring supply overhang that could weigh on the stock price at each release date.

Despite a ~56% correction from the IPO-day high of KRW 60,000, the possibility of further valuation compression toward more normalized levels cannot be ruled out.

Customer concentration and project-based revenue volatility

According to recent reports, approximately 68% of 2025 revenue was concentrated in the top-5 manufacturing conglomerates and the defense sector, meaning contract terminations or execution delays at a single large client could have a disproportionate impact on quarterly results.

The still-high proportion of project-based solution revenue creates significant quarter-to-quarter variability in revenue recognition timing, introducing material uncertainty around annual target achievement.

If the strategic transition toward recurring Runway license revenue proceeds more slowly than planned, the company's recurring earnings base could remain limited and the profitability improvement timeline could be meaningfully deferred.

07

Risk factors

Delayed Profitability Risk

Achieving the 2027 break-even target requires steep revenue ramps to KRW 22.5B in 2026 and KRW 37.3B in 2027, contingent on new order wins and contract execution proceeding as planned in H2 2026.

Macroeconomic headwinds or capital expenditure deferrals by major clients could delay order recognition and push the break-even timeline beyond 2028. Any such delay in the profitability inflection point would directly add downward pressure to a currently elevated share price.

Intensifying Technology Competition

Global technology giants including NVIDIA, Microsoft, and Amazon are aggressively expanding into industrial AI and edge AI, creating conditions for increasing competition in closed-network specialized solutions over the medium to long term.

Domestically, a pipeline of physical AI IPO candidates—including Superb AI, InEase, and OnePredict—broadens investor choices while intensifying competitive dynamics within the sector.

If large-scale incumbents develop or acquire closed-network AI OS capabilities comparable to Runway through organic development or M&A, the technology differentiation premium embedded in MakinaRocks' valuation could erode.

Macro and Regulatory Risk

A manufacturing sector downturn or broad-based corporate IT and AI investment contraction could directly weaken new order momentum and revenue growth trajectories.

Defense AI contracts are inherently exposed to government budget allocation changes and defense procurement process delays, limiting the predictability of public-sector revenue streams.

In overseas markets, particularly Japan and Europe, tightening AI regulations (e.g., the EU AI Act) and evolving data sovereignty requirements may constrain market penetration speed and introduce additional compliance costs.

08

Overall view

마키나락스는 폐쇄망 산업 AI OS 분야에서 실질적인 기술 자산과 검증된 대형 고객 레퍼런스를 보유하고 있으며, DART 공시 기준 2026년 1분기 매출의 167.6% 급증과 가파른 수주 파이프라인은 중기 성장 잠재력의 실체를 보여준다.

그러나 2026년 1분기에도 영업손실 -24억원이 지속되고, 최근 보도에 따르면 누적 결손금이 653억원(매출의 약 6배)에 달하는 등 수익화까지 상당한 거리가 남아 있다.

현재 주가(26,450원) 기준 2025년 매출 대비 PSR이 40배를 상회하며, IPO 공모가 산정 자체가 2028년 추정 실적에 기반하고 있어 밸류에이션 부담이 여전히 상당하다. 보호예수 해제 일정에 따른 반복적 수급 압박과 대형 고객 집중 리스크 역시 단기 주가 변동성을 심화시키는 요인이다.

성장 방향성과 기술 차별성은 긍정적이나, 현 밸류에이션이 실적 집행 위험을 충분히 반영하지 않은 상태로 판단된다. 2026년 하반기 신규 수주의 가시화와 2027년 BEP 달성 경로가 구체화되는 시점이 보다 실질적인 재평가의 출발점이 될 것이다.

09

Sources

  1. Korea Exchange (KRX) — Prices · Market Cap · Volume
  2. FSS electronic disclosure system (DART) — Financial statements · Dividend filings
Show 9 more articles and sources
  1. valueline.co.kr
  2. v.daum.net
  3. smedaily.co.kr
  4. unicornfactory.co.kr
  5. thelec.kr
  6. m.ekn.kr
  7. v.daum.net
  8. thevc.kr
  9. kitpa.org

Report written 2026-06-05 · Data as of 2026-06-04

This content is AI analysis of market data and web search results, provided for information only. It is not a solicitation or recommendation to invest. Investment decisions and their consequences are the investor's own responsibility. Data may be delayed or contain errors.