Cầu Đuống (CDG): deeply depressed price vs mid-cycle EV/EBITDA implies 26.5% upside but liquidity and model confidence are limiting factors
Intrinsic value VND 3,035 vs market VND 2,400 — implied upside 26.5% (model confidence: low).
Business Overview
Công ty Cổ phần Cầu Đuống operates in construction and building materials under ICB3 'Xây dựng và Vật liệu' and is listed on UPCOM. The company reported revenues of VND 25.5 bn in 2023, VND 23.7 bn in 2024 and VND 27.8 bn in 2025, showing recovery in 2025. Reported net profit swung from VND 2.1 bn in 2023 and VND 1.3 bn in 2024 to VND 10.1 bn in 2025. Total assets stood at VND 55.1 bn (2023), VND 52.7 bn (2024) and VND 58.3 bn (2025).
Investment Thesis
1) Valuation gap vs trading price: Our mid-cycle EV/EBITDA model produces an intrinsic price of VND 3,035 per share versus the current UPCOM match price of VND 2,400, implying 26.5% upside. The model uses a fair EV/EBITDA of 4.0 (own-history) compared with a sector median EV/EBITDA of 9.85, and incorporates modest reported net debt. 2) Profitability profile: Recent margins are healthy — net profit margin of 36.2% and gross margin of 31.4% — and returns are solid with ROE at 21.0% and ROA at 18.1%. EPS is VND 2,904 and BVPS is VND 14,766, and the company trades at P/E 0.8x and P/B 0.16x using latest reported multiples. 3) Practical constraints to conviction: Model confidence is explicitly low, the stock is illiquid (avg 2-week volume reported as 0.0 and UPCOM quotes show 1y high/low both at VND 2,400), and foreign room is 0.0%. These factors increase execution risk and market-impact costs for larger allocations. 4) Ownership and strategic angle: A large strategic shareholder, Tổng công ty Đầu tư Phát triển Hạ tầng Đô thị UDIC, holds 42.6839%, which supports control stability but concentrates governance and may limit free float and liquidity. Taken together, the numerical upside (26.5%) is noteworthy but offset by low model confidence, illiquidity and concentrated ownership, meaning upside is not risk-free.
Valuation Commentary
Mid-cycle EV/EBITDA model calibrated to company history with isotonic calibration to cap illiquid upside.
- Fair EV/EBITDA = 4.0 (own_history) versus sector EV/EBITDA = 9.85
- Mid-cycle EBITDA anchor used in model (mid_cycle_ebitda input)
- Net debt is modest (model input net_debt = 213,241,073), reducing enterprise value haircut
- Sanity calibration reduced raw intrinsic (raw_intrinsic_value VND 5,641.6) to final VND 3,035 because of illiquidity and isotonic calibration
The VND 3,035 intrinsic price implies 26.5% upside to the current VND 2,400. Confidence in this valuation is low due to illiquidity (zero average 2-week volume) and model calibration that capped a higher raw intrinsic value. Treat the target as a valuation reference rather than a high-conviction entry trigger.
Bull vs Bear
- Valuation gap: model intrinsic VND 3,035 vs market VND 2,400 implies 26.5% upside.
- High reported profitability: Net profit margin 36.2%, gross margin 31.4%, EBIT margin 16.6% are above typical small-cap construction peers.
- Strong ROE/ROA: ROE 21.0% and ROA 18.1% indicate efficient capital use relative to the sector.
- Liquidity and execution risk: average volume 2-week = 0.0 and 1y high/low both at VND 2,400 — market cannot easily transact without price impact.
- Low model confidence: valuation flagged as 'low' and 'illiquid' with isotonic calibration reducing raw intrinsic value from VND 5,641.6 to VND 3,035.
- Concentrated ownership: UDIC holds 42.6839%, limiting free float and potentially reducing catalyst frequency for re-rating.
Sector Context
The Vietnamese construction sector is sensitive to public capex cycles, SBV credit quotas for developers, and VAS accounting conventions that can compress or defer recognized margins relative to cash flows. Peer median upside in our coverage is 9.6%, and sector EV/EBITDA median is 9.85 — CDG's fair EV/EBITDA of 4.0 is well below the sector median, reflecting either structural discounts for small, illiquid UPCOM names or genuinely lower capital intensity. State ownership norms and SOE payout/mandates also matter: a large state-related shareholder (UDIC) controls 42.7%, which aligns with common SOE governance patterns in construction but may also limit market float. Compare CDG to small-cap peers where EV/EBITDA multiples can vary widely; top peers in our sample show upside of 30–39% but also low-to-medium confidence levels.
Risk Factors
- Market liquidity: avg_volume_2w = 0.0 and 1y high/low both VND 2,400 — inability to exit positions without slippage.
- Model and calibration risk: valuation confidence labelled 'low' and calibration reduced a higher raw intrinsic value (raw_intrinsic_value VND 5,641.6) to VND 3,035.
- Shareholder concentration: UDIC owns 42.6839%, which can limit free float and create single-party governance risk.
- Sector cyclicality: construction revenues are sensitive to public/private project timing; revenue swung from VND 23.7 bn (2024) to VND 27.8 bn (2025).
- Distribution constraints: foreign_room = 0.0% restricts demand from overseas investors.
- Small absolute scale: reported totals (revenues and profits in the tens of billions VND) mean single projects or receivable turn changes can materially swing results.
Catalysts
- Improvement in liquidity or transfer to a mainboard/listing venue that increases visibility and trading activity.
- A repeat of 2025 earnings momentum or a material contract win that lifts revenue above the VND 27.8 bn level.
- Corporate actions from the controlling shareholder (UDIC) such as a divestment or strategic partnership that increases free float.
- Re-rating if sector EV/EBITDA compression reverses and small-cap construction multiples expand toward the sector median of 9.85.
Forensic Assessment
No Beneish M-Score supplied (mscore = null) and no red flags or positive signals reported in the forensic inputs. Earnings quality is moderate-high at 70.0/100, which reduces immediate forensic concerns. Given the absence of M-Score data and lack of forensic red flags, the primary concerns are market structure (illiquidity) and concentrated ownership rather than evidence of earnings manipulation.
Track Record
Model historical track record spans 10 years with a hit rate of 44.4% (model directional calls matched next-year direction in fewer than half the years). The average realized upside in past calls is high (avg_upside_pct 282.4%) but driven by a small number of large outliers; the hit rate suggests modest predictive reliability. Combine this with current low model confidence to treat the present valuation as indicative but not definitive.
Written by a language model on 2026-08-11 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.