CDR: Distressed construction smaller-cap with material forensic and liquidity constraints
Intrinsic value VND 5,376 vs market VND 4,400 — implied upside 22.2% (model confidence: low).
Business Overview
Công ty Cổ phần Xây dựng Cao su Đồng Nai (CDR) is a UPCoM-listed small-cap construction company operating in building and materials within Vietnam's construction sector (ICB: Xây dựng và Vật liệu). The company reported revenue of VND 109.0 bn in 2025 after VND 135.3 bn in 2024 and VND 115.6 bn in 2023, reflecting volatile top-line trends across the past three years. CDR's operations sit alongside state-linked corporate parents and strategic institutional holders: the largest shareholder is Công ty TNHH Một Thành Viên Tổng Công ty Cao Su Đồng Nai with ~29.0% ownership, followed by other institutional stakes (12.5%, 4.84%), leaving limited free float and no foreign ownership room (foreign_room 0.0%).
Investment Thesis
CDR's valuation case rests on a mid-cycle EV/EBITDA approach that produces an intrinsic value of VND 5,376/share versus the current UPCoM match price of VND 4,400 — an implied 22.2% upside. The company trades at a P/B of 0.37 and EV/EBITDA of 8.58, suggesting market pricing already discounts capital-intensity and leverage: net debt in the valuation model is VND 26,902,295,372 and the model flags the company as distressed due to a negative equity value floor adjustment (BVPS floor VND 12,044 with a 70% discount applied in calibration). Operationally, margins are thin (EBIT margin 3.72%; net margin 1.28%) and profitability metrics are muted — ROE 5.4% and ROA 1.7% — which limits upside absent clear improvement in earnings quality or cash conversion.
However, ownership concentration with a near-29.0% state-linked holder provides some strategic support and potential access to group contracts; revenue rebounded to VND 109.0 bn in 2025 after 2024's VND 135.3 bn, showing the business is not collapsing. The sector-level backdrop also matters: the median sector upside is ~9.6%, so CDR's 22.2% implied upside is above peers but the model confidence is low and several forensic and liquidity flags materially increase execution risk. Given the combination of modest upside, low model confidence, and elevated forensic concerns, the implied return does not sufficiently compensate for the risk and low liquidity.
Valuation Commentary
Mid-cycle EV/EBITDA: apply a fair EV/EBITDA multiple to a seven-year median (mid-cycle) EBITDA and subtract net debt to derive equity value per share.
- Mid-cycle EBITDA input: VND 2,624,536,935 (model mid_cycle_ebitda).
- Fair EV/EBITDA multiple used: 8.58 (fair_ev_ebitda).
- Net debt of VND 26,902,295,372 is deducted in the model.
- Model calibration: isotonic recalibration with a BVPS floor of VND 12,044 and a 70% discount due to distressed adjustment.
- Model confidence flagged 'low' with sanity flags: illiquid, low_earnings_quality, manipulation_risk.
The valuation produces intrinsic value VND 5,376 (upside 22.2%) but confidence is low because the model required distressed calibration and flags earnings quality and liquidity issues. The upside exceeds the sector median (9.6%) but is below our threshold for a high-conviction (>25%) conviction. Treat the result as indicative rather than definitive; execution and forensic risks could materially compress realized value.
Bull vs Bear
- Intrinsic value VND 5,376 implies 22.2% upside vs market VND 4,400, above sector median upside of 9.6%, leaving room for recovery if earnings normalize.
- Low valuation multiples: P/B 0.37 and EV/EBITDA 8.58 suggest the market already discounts downside, leaving upside from multiple re-rating if earnings quality and cash flows improve.
- Significant institutional ownership (largest holder ~29.0%) provides strategic stability and potential preferential access to group projects.
- Forensic red flags: Beneish M-Score -0.8538 places CDR in the 88th percentile among peers, and earnings quality is very low at 19.8/100 with cash conversion and receivables at 0/100, increasing manipulation and restatement risk.
- Illiquidity risk: average daily volume over 2 weeks is only 1,512 shares and the UPCoM listing plus zero foreign_room (0.0%) limit marketability and institutional demand.
- Balance-sheet strain and distressed model calibration: net debt is significant in the model (VND 26,902,295,372) and the model required a BVPS floor adjustment (BVPS VND 12,044) due to negative-equity-value concerns.
- Thin profitability and declining asset base: EBIT margin 3.72%, net margin 1.28%, ROE 5.4% with total assets shrinking from VND 94.4 bn in 2023 to VND 80.3 bn in 2025.
Sector Context
The construction & materials sector remains heterogeneous in Vietnam: large contractors with access to SOE-backed projects and healthy order books command higher multiples, while smaller UPCoM contractors like CDR face project concentration, late receivables, and working-capital pressure. Regulators (SBV) and state directives indirectly affect sector liquidity: SBV credit quotas and infrastructure fiscal flows can materially alter project pipelines. Accounting in Vietnam (VAS) allows provisions, revaluations, and related-party transactions that can obscure cash reality; forensic flags such as CDR's elevated Beneish M-Score and low earnings-quality score should be read in the VAS context. Peers in our universe show a median implied upside of ~9.6%; top peers can exceed 30% upside but often carry comparable confidence issues.
Risk Factors
- Aggressive accounting/manipulation risk: Beneish M-Score -0.8538 (above the -1.78 threshold) and earnings quality 19.8/100 signal elevated risk of aggressive revenue recognition or accrual management.
- Liquidity and marketability: avg_volume_2w 1,512 shares and UPCoM listing limit ability to trade positions; foreign_room 0.0% prevents foreign buying.
- High leverage relative to equity: model net debt is VND 26,902,295,372 and Debt/Equity 2.12 increases refinancing and covenant risk during downturns.
- Low cash conversion and receivables quality (both 0/100 in earnings-quality sub-metrics) threaten cash flow reliability and dividend capacity (dividend yield 0%).
- Concentrated ownership: top holder ~29.0% and several institutional owners reduce free float and could lead to correlated block trades or related-party contracting.
- Model and valuation uncertainty: model confidence rated 'low' with distressed calibration and isotonic recalibration changes the intrinsic estimate materially (raw_intrinsic_value VND 8,430.8 adjusted to VND 5,376).
Catalysts
- Improvement in cash conversion metrics or public evidence of clean audits that improve the earnings-quality score.
- A visible recovery in order intake or signed contracts that would lift mid-cycle EBITDA above the model input (VND 2.62 bn in mid-cycle EBITDA).
- Any change in major shareholder stance (e.g., a strategic sell-down or an articulated support plan) that alters liquidity and free float.
- Sector-wide re-rating from improved public investment or SBV credit loosening that benefits smaller contractors.
Forensic Assessment
Forensic indicators are the principal concern. The Beneish M-Score of -0.8538 exceeds the -1.78 threshold used to flag manipulation likelihood and sits in the 88th percentile versus Vietnamese peers, while the year-on-year deterioration (+1.39) suggests an adverse trend. Earnings quality is very low at 19.8/100 with cash-conversion and receivables sub-scores at 0.0/100 — these are strong red flags for revenue recognition and cash-realization reliability. The input summary also references an Altman Z-Score near 2.01 (cautionary zone), reinforcing balance-sheet stress. Positive counterpoints include a Piotroski F-Score of 5/9, indicating not all fundamentals are failing, but overall forensic signals require caution and reduce confidence in reported earnings and model outputs.
Track Record
Model track record spans 10 years with a hit rate of 66.7% (i.e., two-thirds of prior yearly directional calls matched subsequent price direction). The historical average upside for past calls is high (avg_upside_pct 119.2%), but that figure is skewed by outliers; a 66.7% hit rate is reasonable but not exceptional. Given the current 'low' model confidence and the forensic flags specific to CDR, past model performance should be treated with caution and not taken as a guarantee of forward accuracy.
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.