Skip to content

CIO VERDICT — BANPU

1. The call

  • Rating: HOLD (bias to add on weakness below ฿4.80)
  • 2-year target price: ฿5.60 (vs current ฿5.35, implies +4.7% capital + ~12% cumulative dividends ≈ +17% total return)
  • Suggested position size: 1.0% of book (starter, tactical)
  • Time horizon: 24 months
  • Why: Specialists triangulate to a tight base-case fair value (Analyst ฿6.30, Quant mean-reversion ฿6.61, Trader base ฿5.00–6.00) — call it ฿6.00 — but the data integrity is Grade C with the shareholders/filings pages literally 404 during a live conflicted merger. That combination forbids conviction. The stock is at chart support with a fat (cyclical) dividend, but it's also a structurally challenged coal name with ESG-mandate cap and a related-party take-private in mid-flight. Expected value modestly positive; risk-adjusted edge thin. The honest answer is a small position, not a thesis bet.

2. How the specialists voted

Persona Verdict 2y target Key insight
Auditor Grade C / High confidence real n/a Prices clean; SET shareholders & filings 404; no financials extractable. Hard cap on conviction.
Trader PASS / small tactical long at support Base ฿5.00–6.00; bear ฿3.50–4.20; bull ฿8.50–9.50 Broken multi-year trend; distribution candles; ฿5.30 triple-tested floor.
Analyst HOLD, add on weakness ฿6.30 mid (range ฿5.50–7.50) Cyclical coal cash cow + USD optionality (BKV) + ESG multiple cap.
PI Refuse to opine (data destroyed by 404s) qualitative Cannot ID directors, family stakes, or insider behaviour. BPP merger mechanics opaque.
Prosecutor Grade C — elevated, structural qualitative Live conflicted RPT (BPP merger); 5–9% tail-risk discount warranted.
Forensic U — Unrated (cannot grade) qualitative 10–15% discount for merger opacity; 3–5% for filings-page blackout.
Quant model median ฿4.07 (GBM) – ฿6.61 (MR) low confidence 90% CI 24m: ~฿1.5–฿12. Noise > signal.
Geopolitical Neutral with negative skew adj −2% (range −10% to +8%) ESG headwind largely priced; merger + US gas the swing factors.

3. Where they converge (the well-supported view)

  1. The price is in a structural downtrend — Trader's chart read, Quant's −6.3% CAGR, Analyst's "round-trip" narrative, and Geopolitical's "stranded-asset partially priced" all agree.
  2. Base-case fair value clusters at ฿5.50–6.60. Analyst ฿6.30, Quant MR ฿6.61, Trader base midpoint ฿5.50. That's a tight ~20% band — unusual given the data gaps and the source for it.
  3. The BPP merger (Oct 2025) is the single dominant 24-month corporate event — and its terms/completion are absent from the package. Auditor flagged it, PI refused to opine on it, Prosecutor put it as Priority 1 enforcement risk, Forensic warrants a 10–15% discount on it.
  4. ESG/mandate exclusion is a multiple cap, not a bear thesis. Analyst, Geopolitical and Quant all observe the structural de-rating is already in the −80% drawdown.
  5. No analyst sees a screaming buy at ฿5.35. Trader says PASS, Analyst says HOLD, Quant says noise dominates, Geopolitical neutral.

4. Where they diverge (the real questions)

Q1: Is the drift mean-reverting (cyclical) or trending (structural decline)? - Quant GBM: trending → ฿4.07 in 24m. - Quant MR / Analyst: cyclical → ฿6.30–6.61. - My call: Split the difference, weighted toward MR. The −80% drawdown already reflects a decade of de-rating; coal isn't going to zero in 24 months (China + India + Asian utilities still buying); BKV provides USD/gas optionality that didn't exist in 2016. Weight 55% MR / 35% GBM-geom / 10% bull.

Q2: How heavy is the BPP merger discount? - Forensic: 10–15% until terms clarified. - Prosecutor: 2–3% expected drag (probability-weighted litigation). - Geopolitical: ±5% on terms. - My call: 7% net discount embedded in target. Three independent sources confirmed the announcement; absence of completion news 8 months later is the worry. Most likely it closes — but minority-sweetener risk to BANPU is real.

Q3: Tactical entry — at support or wait for break? - Trader: PASS or 0.5R tactical with stop ฿5.18. - Analyst: add on weakness below ฿4.50. - My call: Trader is right tactically. ฿5.30 is the line. If it breaks on volume, the next visible level is materially lower and that's where Analyst's "cyclical entry" thesis triggers.

Q4: Does the data-integrity gate force HOLD? - Auditor C grade + Forensic U + Prosecutor C + PI's refusal = yes. Even if everything else said BUY, I cannot size aggressively into a name where the shareholders page is 404 during a related-party merger.

5. The 2-year forecast — my synthesized number

Anchor: weighted blend of model outputs, less RPT discount, plus geopolitical skew.

  • Bull case (P=20%): ฿8.00. Requires: BPP merger closes accretively, Newcastle coal back to US$140+/t, BKV re-rates on Henry Hub >US$3.50, THB weak. Analyst bull ฿9.00+ haircut for ESG cap.
  • Base case (P=50%): ฿5.80. Requires: BPP closes on neutral terms by YE26, coal flat at US$110–120/t, BKV holds, status-quo Thai politics. ~9x mid-cycle EPS + 6% yield.
  • Bear case (P=30%): ฿3.80. Requires: coal to US$90/t OR merger blocked/repriced, OR ฿5.30 breaks and momentum unwinds. Quant bear realistic.

Expected value: 0.20×฿8.00 + 0.50×฿5.80 + 0.30×฿3.80 = ฿1.60 + ฿2.90 + ฿1.14 = ฿5.64

Plus est. 24m cumulative dividends ~฿0.65 (assuming ฿0.30–0.35/yr base-case payout) → total EV ≈ ฿6.29 vs ฿5.35 spot = +17.6% total return over 24m, ~8.5% IRR.

6. Position-sizing math

  • Edge (total return EV vs spot): +17.6% over 2y ≈ 8.5% IRR
  • THB risk-free ~2.0% → excess return ≈ 6.5%
  • Vol (Quant ann.): 38.6%
  • Vol-adjusted edge (Sharpe-equiv): 6.5% / 38.6% = 0.17 — thin
  • Kelly fraction (full): edge/var = 0.065 / 0.149 ≈ 4.4%
  • Quarter-Kelly (standard family-office prudence): 1.1%
  • Data-integrity haircut: ×0.75 (Auditor C, filings 404) → ~0.8%
  • Round to: 1.0% of book starter; max 2.5% if data gaps resolve favourably

Rationale for not going to zero: dividend yield + chart support + USD optionality justify holding something. Rationale for not going higher: thin Sharpe, opaque RPT, no conviction in drift sign.

7. Risk gates

  • Tail-risk gate: Prosecutor C (not D/F) and Forensic U (unrated, not D/F). Gate does not force HOLD/SELL by itself — but combined with data-integrity, it caps me at small. If BPP merger terms surface as visibly minority-unfriendly (e.g., EGM <90% non-interested support), Prosecutor upgrades to D and I'm out.
  • Data-integrity gate: Auditor Grade C with shareholders + filings pages 404 during a live conflicted merger. This is the binding constraint. It caps position at ≤2.5% and forbids any high-conviction BUY regardless of model output. It also forced me to discount every persona that needed financials (Forensic, parts of Analyst's quantitative claims). I gave full weight only to Trader (price-only), Quant (price-only summary stats), Geopolitical (structural), and Prosecutor (event-driven).

8. What I don't know

  1. BPP merger swap ratio, EGM result, and closing status. Without this, NAV impact is a ±10% guess. Resolved by: SEC Thailand 247-4 filing, IFA report, EGM vote tally.
  2. BANPU's net debt, parent-only vs consolidated leverage, and BKV financing structure. USD 1.5bn US capex announcement against unknown balance sheet is the dilution/refinancing risk. Resolved by: Fresh SET 56-1 pull, BKV NYSE 10-K, audited 2025 consolidated accounts.
  3. Controlling shareholder identity, free float, and pledge status. SET shareholders page 404. Without it I'm trading blind to who controls the merger vote. Resolved by: Fresh SET shareholders pull or SEC Form 246-2 disclosures.

9. Triggers to re-rate

Buy more (scale to 2.5%) if: - BPP merger closes on terms market reads as fair (BPP minority approval ≥95% non-interested) AND BANPU re-rates above ฿6.20 on volume; OR - ฿4.50–4.80 reached on capitulation flush (Analyst's cyclical entry) AND fresh SET filings confirm net debt manageable (<3.0x EBITDA); OR - Newcastle thermal coal breaks above US$140/t AND Henry Hub above US$3.50/MMBtu.

Sell down (to 0.5%) if: - Daily close < ฿5.18 on volume > 100M (Trader stop-line breached); OR - BPP EGM defers, sweetens against BANPU minorities, or SEC raises fairness objections; OR - Coal breaks US$95/t Newcastle and looks structural; OR - People's Party probability in 2027 polling rises through 25%.

Get out completely if: - Prosecutor grade upgrades to D (insider trades surface in 29-Oct-25 window, or fairness opinion withheld); OR - Forensic statements appear and reveal goodwill impairments/ARO under-provisioning that move net debt/EBITDA >5x; OR - BANPU equity raise announced to fund US$1.5bn US gas capex (dilution at trough cycle is the classic Banpu pattern Analyst flagged); OR - ฿4.50 breaks on the downside without a buyer showing up — that opens Quant's GBM path to ฿2–3 and the bear case becomes the base case.

Final word to the family: This is not a name I want to defend in front of you in 18 months. It's a small, optionality-driven position — chart support + dividend + USD asset hedge + merger simplification — sized for the fact that we're working from a half-broken data package on a company in the middle of a related-party take-private. If the SET filings page comes back online and the merger closes cleanly, I'll have a real opinion. Until then, 1% is the discipline.