Skip to content

CIO VERDICT — THANI (Ratchthani Leasing PCL)

1. The call

  • Rating: HOLD (with a small starter long permitted for risk-tolerant sub-books)
  • 2-year target price: ฿1.80 (vs current ฿1.69, implies +6.5% price / ~+17% total return incl. assumed ~5% dividend yield)
  • Suggested position size: 0.5% of book (cap; trader-style satellite only)
  • Time horizon: 18–24 months
  • Why: The bottom-up triangulation is genuinely tight (Trader base ฿1.75, Analyst FV ฿1.65–2.00, Quant median ฿1.80) — but it triangulates to almost nothing. Expected total return barely clears the risk-free rate. The Auditor graded the data C with Medium confidence it's real, and critically every fundamental input — NPL, NIM, coverage, dividend, shareholder list — is on the DO-NOT-FABRICATE list. The Analyst, Forensic and Prosecutor reports are all conditional/incomplete because the filings page 404'd. You cannot underwrite a high-conviction call on a leasing company without seeing its loan book. Range trade only, on the tape.

2. How the specialists voted

Persona Verdict 2y target Key insight
Auditor C / Medium n/a Trust the 60d tape; financials/shareholders/filings are absent (404s + JS shells)
Trader Small swing LONG ฿1.75 base / ฿2.40 bull / ฿1.20 bear Clean ฿1.58 floor; base-building; 1:2 R/R; size <1%
Analyst HOLD ฿1.65–2.00 (mid ~฿1.82) ~10% upside + 5–6% yield; truck cycle hasn't turned; TIDLOR ate the mindshare
PI Inconclusive (black box) n/a No board/shareholder data; May 8–11 volume spike unexplained — possible block
Prosecutor B– (low-moderate) qualitative No enforcement found; RPT structure with TCAP is the latent land mine; -5–10% tail discount
Forensic INCOMPLETE → C provisional qualitative Cannot grade — no statements. Sector-conditional unverified-state discount 30–55%
Quant Low confidence Median ฿1.80, 90% CI [฿0.75–฿3.70] Kurtosis 16 → fat tails; basing regime real but only 12m of evidence vs 6y downtrend
Geopolitical HEADWIND (mild-moderate) -15% to +3% adj OCPB HP-rate-cap extension + EV-truck residual-value risk; both under-priced

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

  1. The bottom-up targets triangulate around ฿1.75–฿1.82. Trader (฿1.75), Analyst (฿1.82 mid), Quant (฿1.80 median) all land within ~4% of each other. That is unusually tight — and tells you the upside is real but small.
  2. No fraud, no enforcement, no smoking gun. Prosecutor (B–), Forensic (no distress signature in tape), Auditor (orderly OHLC) all agree this is not an accounting blow-up risk. It is a credit-cycle problem.
  3. The 60-day tape is clean and tradeable. ฿1.58 floor defended 4×, ฿1.76 ceiling rejected on volume. Range-bound base, low realised vol (~14% ann.).
  4. This is a structurally challenged name, not a hidden gem. −77% max drawdown, −8% 10y CAGR, flat-to-decade-ago in nominal terms. TIDLOR has been promoted to SET50; THANI got bypassed.
  5. TCAP parentage = funding moat + RPT/governance overhang. Every persona flagged both sides of this coin.

4. Where they diverge (the real questions)

Q1: Is the 12-month basing regime real, or is it a pause inside a 6-year downtrend? - Trader leans constructive (higher lows, base-building, accumulation candles). - Quant is explicitly agnostic — 12m sample is statistically insufficient to reject the prior trend. - My call: Trader is overweighting recent tape; Quant is right. Until we see a Stage-3 NPL inflection in actual filings, the prior is continuation. Side with Quant.

Q2: How big is the unverified-state discount? - Forensic says 30–55% to optimistic-case FV (sector-conditional). - Prosecutor says 5–10%. - Analyst implicitly applies ~0% (uses inferred run-rate). - My call: Forensic's 30–55% is appropriate if you're sizing a core position. For a 0.5% trading position on tape-only evidence, Prosecutor's 5–10% is the right gauge. The discount scales with position size and time horizon.

Q3: How much does the geopolitical/regulatory headwind matter? - Geopolitical: -15% to +3% over 24m, skewed negative; EV-truck residual risk under-priced. - Analyst: acknowledges EV residual risk as "slow burn" #4 risk. - My call: Geopolitical is right that bottom-up analysts under-weight collateral obsolescence. Adjust base case down by ~5%.

Q4: What was the May 8–11 volume event? - PI calls it a possible controlling-shareholder block trade or related-party reduction. - Prosecutor flags it as a textbook ramp-then-dump silhouette. - Trader read it as distribution into strength. - My call: All three are saying the same thing — there was distribution by someone informed. This is a red flag against the bullish base-building narrative. Whoever sold on 2026-05-11 (31.7M shares) knew something we don't.

5. The 2-year forecast — my synthesized number

  • Bull case (P=20%): ฿2.20. Requires Thai commercial-vehicle NPL inflection (2 consecutive quarters of declining Stage-3), BoT cuts in H2 2026, and either a TCAP corporate action or sector re-rating. +30% price + ~12% cumulative dividends = ~42% total return.
  • Base case (P=55%): ฿1.75. Range-bound continuation, NPLs stabilise but don't improve, dividend intact at ~฿0.08–0.10. +4% price + ~11% dividends = ~15% total return.
  • Bear case (P=25%): ฿1.20. Truck NPL re-acceleration, OCPB extends HP rate cap to commercial vehicles, dividend cut, OR the May 8–11 selling resumes. -29% price + ~5% partial dividend = ~-24% total return.

Expected value (price only): 0.20×฿2.20 + 0.55×฿1.75 + 0.25×฿1.20 = ฿0.44 + ฿0.96 + ฿0.30 = ฿1.70. Expected total return incl. dividends: ~0.20×42% + 0.55×15% + 0.25×(-24%) = +8.4% + 8.3% − 6.0% = +10.7% over 2 years (~5.2% annualised).

That is barely above the Thai 2y risk-free. The skew is symmetric-to-slightly-negative. The market is roughly right at ฿1.69.

6. Position-sizing math

  • Edge (annualised expected excess return vs 2y Thai govvies ~2.5%): ~5.2% − 2.5% = +2.7%
  • Vol-adjusted edge: 2.7% / 31% = 0.087 (Sharpe-equivalent)
  • Kelly fraction (full): edge / variance ≈ 2.7% / 9.6% = ~28% — but this is a toy number on a fat-tailed, low-liquidity, data-incomplete name. Discard.
  • Practical Kelly (× 0.25 fractional Kelly): 7%
  • Hard constraints that override Kelly:
  • Data-integrity gate (Auditor C, Medium): cap at 2%
  • Forensic INCOMPLETE: cap at 1%
  • ADV ฿8–12M/day, can't fill more than ฿2–3M without being the market: cap at ~1% of a ฿15bn book
  • Quant low confidence + kurtosis 16: cap at 0.5–1%
  • Final position size: 0.5% of book (~฿75M) — but given ADV constraints, even this requires ฿2M/day VWAP work over ~6 weeks. Realistically start at 0.25% (~฿37M) and only build to 0.5% if the tape holds ฿1.60.

7. Risk gates

  • Tail-risk gate: Prosecutor B–, Forensic provisional C. Neither D/F. Gate NOT triggered — bottom-up not overridden. But Forensic is provisional because they couldn't read the statements, not because the statements are clean. Treat as one notch worse than face value.
  • Data-integrity gate: Auditor C / Medium. This is the binding constraint. Per protocol, I may NOT issue a high-conviction BUY/SELL, must cap position size, must lean HOLD. This is exactly what I've done. Without verified shareholders, filings, or financials, any larger position is gambling, not investing.

8. What I don't know

  1. The actual loan book quality. NPL ratio, Stage-2/Stage-3 migration, coverage ratio, credit cost trajectory. Resolves with: the 56-1 One Report FY2025 and Q1-2026 quarterly. Without this, the entire investment thesis is sector-inferred, not company-verified.
  2. Who sold 31.7M shares on 2026-05-11? Block trade by TCAP? Director sale? AMC liquidation? Resolves with: SET Form 246-2 / Form 59 filings for May 2026, and SETTrade time-and-sales for 2026-05-07 to 05-12. This single fact could flip the bull/bear narrative.
  3. Is the dividend intact? Trader/Analyst assume ฿0.08–0.10 DPS; Auditor flagged dividend history as absent. Resolves with: the FY2025 dividend declaration and AGM minutes. A dividend cut would gap the stock down 15%+ immediately.

9. Triggers to re-rate

  • Buy more (upgrade to BUY, scale to 2–3% of book) if:
  • 56-1 One Report verifies NIM stable >4%, Stage-3 declining 2 consecutive quarters, dividend maintained ≥฿0.08, AND
  • Daily close above ฿1.76 on >20M shares with no fade, AND

  • May 8–11 volume event is explained by a benign cause (e.g., index rebal, not insider/parent sell-down).

  • Sell down (cut to 0%) if:

  • Close below ฿1.58 on any volume (Trader's line; base thesis broken), OR
  • Dividend cut announced, OR
  • ADV collapses below 2M shares for 2+ weeks (can't exit), OR

  • OCPB signals HP rate-cap extension to commercial vehicles.

  • Get out completely (and short-list for SELL) if:

  • Stage-3 ratio jumps >150bps QoQ, OR
  • Auditor change at next AGM, OR
  • TCAP announces stake reduction or forced placing, OR
  • Form 59 reveals director-sales cluster >฿5M in 30-day window around the May spike.

Family-office bottom line: This is not a stock — it's a placeholder. The numbers triangulate to "fine," but every fundamental input is on the do-not-fabricate list. We are paid to have conviction, and we don't have it here. Take a 0.5% rental at ฿1.65–1.68 with stop at ฿1.58 if the trading desk wants the optionality; otherwise sit out and wait for the 56-1 One Report. The asymmetry isn't here yet.

(~1,950 words)