Sông Đà 10 (SDT): deep-value construction name with sizable state control and illiquidity constraints
Intrinsic value VND 5,059 vs market VND 4,000 — implied upside 26.5% (model confidence: low).
Business Overview
Công ty Cổ phần Sông Đà 10 (SDT) is a construction group listed on UPCOM operating in construction and building materials (ICB: Xây dựng và Vật liệu). The company executes civil and infrastructure projects and is majority-controlled by the state-owned Tổng Công ty Sông Đà (62.27%), which shapes strategic project access and cash-flow patterns. Revenue has been volatile: VND 702.7 bn in 2023, VND 1,022.1 bn in 2024 and VND 775.5 bn in 2025, reflecting project timing and backlog realization.
Investment Thesis
SDT screens as an undervalued, asset-rich construction name on simple multiples: P/B of 0.2 and P/E of 3.0 based on latest reported ratios. Our mid-cycle EV/EBITDA valuation implies an intrinsic price of VND 5,059 (using a fair EV/EBITDA of 6.67 versus sector median 9.85), producing a 26.5% upside to the current match price of VND 4,000. Earnings recovery is visible: net profit swung to VND 66.2 bn in 2025 after losses in 2023-24, and margins are positive (net profit margin ~8.1%, gross margin ~10.0%).
Offsetting the valuation attraction are execution, liquidity and governance constraints. The stock trades on UPCOM with low liquidity (avg volume two weeks 23,263 shares) and a flagged illiquid upside cap; foreign ownership room is closed (0.0%), limiting demand from foreign investors. Financial leverage is material (Debt/Equity 1.3) and ROE is modest at 7.8% compared with construction peers where sector EV/EBITDA is higher. Model confidence is low (calibrated), so the 26.5% implied upside requires accepting both execution risk and marketability risk.
Valuation Commentary
Valuation uses an EV/EBITDA mid-cycle approach: mid-cycle EBITDA multiplied by a fair EV/EBITDA multiple, less net debt, converted to a per-share intrinsic value and calibrated isotonic to historical model outputs.
- Mid-cycle EBITDA used: VND 133.2 bn (model input).
- Fair EV/EBITDA multiple applied: 6.67 (own-history calibration) vs sector EV/EBITDA 9.85.
- Net debt deducted: VND 62.5 bn.
- Sanity calibration: isotonic recalibration produced raw intrinsic value uplift but final is capped for illiquidity.
The model implies VND 5,059 per share (26.5% upside) but confidence is low due to UPCOM illiquidity and calibration limits. The valuation gap versus sector EV/EBITDA suggests re-rating potential if EBITDA stabilizes and liquidity improves; conversely, upside may not realise promptly given trading constraints and state ownership concentration.
Bull vs Bear
- Cheap on balance-sheet multiples: P/B 0.2 and P/E 3.0 imply substantial asset backing per share.
- EBITDA mid-cycle and fair EV/EBITDA 6.67 produce intrinsic VND 5,059 — 26.5% above market price VND 4,000.
- Profitability turnaround: net profit recovered to VND 66.2 bn in 2025 after losses in 2023-24.
- Low reported EV/EBITDA (3.9) vs sector median (9.85) leaves room for valuation multiple expansion if execution and reporting persist.
- Illiquid listing: avg volume two weeks 23,263 shares and model sanity flags ('illiquid', 'illiquid_upside_capped') constrain price discovery and realisation of intrinsic value.
- High ownership concentration: Tổng Công ty Sông Đà holds 62.27%, reducing free float and potential for minority-friendly corporate actions.
- Operational volatility: revenue declined YoY -24.4% recently and assets have fallen (total assets VND 2,021.1 bn in 2025 vs VND 2,288.5 bn in 2023), indicating project timing and collection risk.
- Balance-sheet strain: Debt/Equity 1.3 increases refinancing and interest-rate sensitivity in a credit-constrained environment.
Sector Context
The Vietnamese construction sector faces cyclical project flows, SBV-directed credit growth quotas and periodic SOE procurement that drive revenue spikes and lulls across contractors. VAS accounting can make working-capital and receivable dynamics opaque (contract revenue recognition and progress billing practices). Many listed contractors trade at higher EV/EBITDA multiples (sector median EV/EBITDA 9.85) than SDT, reflecting differing project quality, backlog visibility and foreign investor interest. For banks and large contractors, legacy VAMC bonds and land-use-rights exposure in real estate counterparties remain cross-sector risks that can affect subcontractor payments and liquidity.
Risk Factors
- Illiquidity risk: UPCOM listing, low two-week average volume (23,263 shares) and model flags limit ability to monetise the intrinsic premium.
- Concentrated ownership: state majority holder (62.27%) reduces free float and could prioritise SOE policy objectives over minority returns.
- Execution & backlog risk: volatile revenue (2025 revenue VND 775.5 bn; revenue YoY -24.4%) implies sensitivity to project awards and collections.
- Leverage and refinancing: Debt/Equity 1.3 and net debt VND 62.5 bn expose the company to interest-rate and credit-access pressures.
- Valuation calibration uncertainty: model confidence is low and intrinsic value was isotonic-calibrated and capped for illiquidity.
- No foreign room: foreign_room 0.0% limits demand from international funds, constraining re-rating potential.
- Sector/regulatory risk: SBV credit policies and SOE payout/assignment mandates can affect sector cashflows and contract timing.
Catalysts
- Contract wins or backlog rehiring that stabilise revenue growth and restore multi-year EBITDA.
- Improved liquidity or transfer to a more liquid market segment — would materially increase probability of multiple expansion.
- Clear reduction in net debt or visible deleveraging that improves Debt/Equity from current 1.3 level.
- Operational transparency improvements (better disclosure of contract margins and receivables) that raise investor confidence.
Forensic Assessment
No Beneish M-Score or forensic red flags are provided (mscore null and no red_flags). Earnings quality scores 70/100, indicating acceptable but not pristine earnings visibility. Given VAS accounting nuances in construction and the company's state ownership, focus remains on receivable recognition, progress-billing practices and related-party transactions; however, no explicit forensic alerts are present in the dataset.
Track Record
Model track record spans 12 years (2015–2026) with a hit rate of 45.5% — a mediocre historical directional accuracy. Average historic upside for calls has been large (avg_upside_pct 172.9%), but the hit rate indicates substantial variability; treat model outputs as informative inputs rather than definitive signals.
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.