TNB: State-controlled steel mill with stretched EV/EBITDA and limited liquidity
Intrinsic value VND 9,676 vs market VND 10,100 — implied downside -4.2%
Business Overview
Công ty Cổ phần Thép Nhà Bè - VNSTEEL (TNB) is a steel producer listed on UPCOM with 29,428,877 issued shares. The company operates in the domestic steel/metals value chain (ICB: Kim loại) and serves construction and industrial customers. Reported revenues rose to VND 1,902.9 bn in 2025 (from VND 1,391.4 bn in 2024), reflecting a cyclical recovery in steel demand. TNB remains majority-owned by Tổng Công ty Thép Việt Nam (84.77%), which makes corporate actions and dividend policy effectively controlled by the state shareholder.
Investment Thesis
TNB’s recovery in top-line volumes is visible: revenue grew to VND 1,902.9 bn in 2025 with a reported Revenue YoY of 0.3685 (36.9%). However, profitability remains thin: EBIT margin of 0.44% and net profit margin of 0.26% translate into ROE of 0.0191 (1.9%) and ROA of 0.0082 (0.8%). Valuation is inconsistent with operating metrics — the stock trades at EV/EBITDA 22.0743 versus a sector EV/EBITDA of 9.14 and our fair EV/EBITDA input of 10.47, which compresses implied upside.
Liquidity and investor access are material constraints: average daily volume over two weeks is 22.0 shares and foreign ownership room is 0.0%, limiting re-rating potential from international flows. Ownership concentration (84.77% state ownership) lowers free float and increases the influence of SOE directives on capital allocation, dividends and restructuring. Earnings quality is middling (50.4/100), and our valuation model yields an intrinsic value of VND 9,676 per share with very_low model confidence, leaving an implied downside of -4.2% from the match price of VND 10,100. Given the narrow downside/upside band relative to execution and liquidity risks, the risk-adjusted case for adding exposure is weak.
Valuation Commentary
EV/EBITDA mid-cycle: we apply a mid-cycle EBITDA to a calibrated fair EV/EBITDA multiple and subtract net debt to derive equity value per share.
- Mid-cycle EBITDA input from the model (mid_cycle_ebitda) and our own median EBITDA history
- Fair EV/EBITDA multiple of 10.47 (source: own_history) vs sector EV/EBITDA 9.14
- Net debt figure included in the EV calculation as provided in model inputs
- Sanity calibration (isotonic) increased the reported intrinsic value from raw_intrinsic_value VND 5,380.9 to VND 9,676 but flagged illiquid trading
The model produces an intrinsic value of VND 9,676 per share implying -4.2% vs the current market price; model confidence is very_low. Given the high EV/EBITDA multiple implied by current market pricing (22.1x) versus our fair multiple (10.47x) and the illiquidity flag, we have low conviction in the precision of the target — the downside/upside is narrow relative to execution and market access risks.
Bull vs Bear
- Revenue recovered to VND 1,902.9 bn in 2025 (from VND 1,391.4 bn in 2024), indicating demand momentum in the cycle
- P/B of 0.8842 implies balance-sheet support relative to book (BVPS VND 11,422.8426), offering some capital value floor
- State ownership can enable access to preferential procurement or inter-company contracts via Tổng Công ty Thép Việt Nam (84.77% holder), stabilizing volumes in downturns
- EV/EBITDA at 22.0743 is well above sector median (9.14), indicating the market prices better-than-reported profitability that is not yet realized
- Margins are very thin: EBIT margin 0.44% and net profit margin 0.26%, and ROE is only 1.91%, limiting earnings upside
- Liquidity is minimal (avg_volume_2w 22.0) and foreign_room 0.0% prevents foreign buying — re-rating opportunities are constrained
- Concentrated state ownership (84.77%) and zero foreign room reduce free-float — minority shareholders face execution and governance risk
Sector Context
The Vietnamese steel sector is cyclical and sensitive to construction and infrastructure investment. VAS accounting practices and state ownership in the sector can obscure comparability; investors should adjust for related-party transactions and SOE directives when analysing earnings. SBV macro policy (credit growth quotas) can throttle working-capital financing for steelmakers during policy tightening, and banks' exposure to SOEs means restructuring or VAMC solutions may be used for distressed leverage. Peer valuation dispersion is wide: sector median upside is 5.6% but top peers show >40% implied upside while several small caps exhibit double-digit downside and very_low model confidence, underscoring idiosyncratic risk in the subsector.
Risk Factors
- Execution and margin risk: EBIT margin 0.44% and net profit margin 0.26% leave little buffer for raw-material cost or demand shocks
- Liquidity and marketability: avg_volume_2w 22.0 shares and UPCOM listing reduce tradability; illiquid flag in model inputs increases price impact risk
- Concentrated ownership: Tổng Công ty Thép Việt Nam holds 84.77%, limiting minority influence and increasing policy/execution risk from SOE mandates
- Valuation mismatch: market EV/EBITDA 22.0743 vs fair multiple 10.47 — a heavy premium that could reverse if EBITDA disappoints
- Foreign access: foreign_room 0.0% prevents incremental foreign demand and may cap potential rerating
- Earnings quality: score 50.4/100 suggests moderate quality — monitor working-capital accruals and one-off items closely
Catalysts
- Sustained EBITDA improvement or margin recovery that narrows the gap with the implied EV/EBITDA multiple
- Any state-led restructuring or asset-transfer announcements from the majority shareholder that improve free-float or governance
- Improvement in trading liquidity or re-listing/upgrade that increases investor access (changes in foreign_room would be material)
Forensic Assessment
No Beneish M-Score or explicit forensic flags were provided (mscore null). Red flags array is empty. Earnings quality is moderate at 50.4/100, which does not signal acute manipulation but warrants scrutiny of accruals and related-party transactions given large state ownership. In short: no formal forensic alarms, but quality is only middling and governance concentration is a practical concern.
Track Record
The model track record spans 12 years with a hit_rate of 0.45454545454545453 (45.5%), indicating roughly coin-flip historical directional accuracy. Average historical upside for the model was 132.895% (132.9%), but that distribution is skewed and past performance has limited predictive value for this illiquid, state-controlled UPCOM name. Given the very_low confidence on the current valuation, historical hit rate provides limited comfort.
Written by a language model on 2026-08-10 from this page’s own model outputs and financial statements, and may quote figures from that date. Descriptive analysis, not investment advice — no buy, sell or hold recommendation is given or implied.