Skip to content

CIO VERDICT — MIDA (Mida Assets PCL)

1. The call

  • Rating: AVOID / SELL (do not initiate; exit if held)
  • 2-year target price: ฿0.28 (vs last print ฿0.41, implies −32%)
  • Suggested position size: 0% of book (hard zero — not even a speculative tag)
  • Time horizon: 24 months, with a 6-month surveillance-risk window inside that
  • Why: Seven specialists and one auditor have handed me a file in which the only things I actually know are (i) the stock has destroyed 57% of capital over a decade with an 82.5% max drawdown, (ii) it has just ripped +70% in six weeks on volume bursts 30× normal with zero verifiable news, and (iii) I have no income statement, no balance sheet, no shareholder register, no filings, and no auditor identity. The Auditor graded the data C with two SET pages returning 404. The Forensic Accountant graded provisional D. The Prosecutor flagged the tape as a textbook SET Market Surveillance pattern. Under our tail-risk and data-integrity gates, that is a no-touch. The fact that the chart looks "interesting" is irrelevant when I cannot underwrite a single fundamental input.

2. How the specialists voted

Persona Verdict 2y target Key insight
Auditor Data grade C, Medium-High real n/a 2 SET pages are 404s; zero fundamentals extractable; news is ticker-collision with Malaysian agency
Trader Small spec long, tight stop ฿0.30–0.40 base / ฿0.70–0.85 bull / ฿0.15–0.20 bear Real reversal pattern but treat as orphan; max 0.5–1% only for micro-cap specialists
Analyst HOLD (avoid) ฿0.25–0.40 anchor Refuses to publish point forecasts; data package insufficient
PI Inconclusive — refuses to fabricate n/a Names, board, shareholders not in package; "someone is accumulating with no story"
Prosecutor Grade C, lean cautious Apply 15–25% legal-tail discount Tape consistent with SEA §244 pattern; would open a surveillance file
Forensic Grade D (provisional, incomplete) n/a Cannot compute Beneish/Altman/Sloan — refuses; pledge-liquidation hypothesis live
Quant Model median lower ฿0.33 at 24m, 90% CI ฿0.12–฿0.95 Mean-reversion + GBM both point down; current = R4 speculative regime
Geopolitical Neutral with negative skew −15% to +10% 20% probability of SEC/SET surveillance action causing 20%+ forced drawdown

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

Seven of eight reports converge on the same operational conclusion despite different lenses:

  1. The data is not investable. Auditor (C), Analyst (refuses to rate), PI (refuses to profile), Prosecutor (cannot certify clean), Forensic (provisional D) all explicitly say they will not fabricate to fill gaps. That is unanimous.
  2. The June 2026 ramp is anomalous and unexplained. Trader, Analyst, PI, Prosecutor, Forensic, Quant and Geopolitical all flag the +70% / 27.7M-share spike on zero verifiable news as suspicious. Nobody has a benign explanation.
  3. The 10-year tape is a value destroyer. −57% cumulative, −82.5% max DD, decade of underperformance. No one disputes this.
  4. Liquidity is broken. Multiple zero-volume sessions, sub-1M median ADV, ฿0.01 tick = 2.5% spread. Exit risk is real for any meaningful size.

4. Where they diverge (the real questions)

Disagreement #1 — Trader (small spec long) vs everyone else (avoid). The Trader sees a clean breakout pattern off capitulation lows and would play it small with a hard stop at ฿0.27. The other six personas don't dispute the chart — they dispute trading something this opaque at all. My call: side with the six. The Trader's setup is real but it is a tape-reader's trade, not a family-office position. We are a ฿15bn book; we have no edge in scalping ฿250k-turnover micro-caps against a buyer who, per the PI's read, "knows something we don't." The Trader himself caps it at 0.5–1% and says "if you don't already trade Thai micro-caps, don't touch this." We don't.

Disagreement #2 — Quant's wide CI vs Geopolitical's directional skew. Quant says ฿0.12–฿0.95 at 24m (Gaussian; real tails wider). Geopolitical says −15% to +10% with surveillance tail. My call: weight Geopolitical higher in the near term (6m), Quant's mean-reversion higher in the medium term (12–24m). Both point down or sideways. Neither supports the Trader's bull case as a base.

Disagreement #3 — Is the June spike accumulation or distribution-pump? Trader leans "informed accumulation." Prosecutor and Forensic lean "manipulation or forced cover." PI says unknowable. My call: irrelevant. In both cases the asymmetry is bad for us — if it's a pump, we're the exit liquidity; if it's informed accumulation by insiders, we have no information edge and they do. There is no scenario in which we are the smart money here.

5. The 2-year forecast — my synthesized number

I weight Quant's mean-reversion model (฿0.28–฿0.33), Analyst's anchor (฿0.25–฿0.40), Trader's bear (฿0.15–฿0.20), and apply Geopolitical's 20% surveillance-tail and Prosecutor's 15–25% legal-tail haircut.

  • Bull (P=15%): ฿0.65. Requires the June catalyst to be a real, disclosed, value-accretive event (asset sale, gold-loan recap, dividend resumption) AND surveillance does not flag AND property cycle troughs. Triple-conditional.
  • Base (P=55%): ฿0.27. Mean-reversion to the 2020–early-2026 trading range as the unexplained ramp fades. Roughly Quant's median.
  • Bear (P=30%): ฿0.15. Surveillance flag / cash-balance / trading halt; OR forced-pledge liquidation; OR distribution rally completes and stock breaks ฿0.27 support back toward the multi-year low near ฿0.05–฿0.15.

Expected value = 0.15 × 0.65 + 0.55 × 0.27 + 0.30 × 0.15 = ฿0.29

Vs spot ฿0.41 → expected return −29% over 24 months. Skew is negative and tails are fat to the downside.

6. Position-sizing math

  • Edge: EV is negative 29% over 24m → there is no long edge to size. A short is theoretically interesting but ruled out on borrow availability, illiquidity, and squeeze risk in a name that prints +20% days.
  • Vol-adjusted edge: −29% / 60% ann. vol × √2 = −0.34 (Sharpe-like). Negative.
  • Kelly fraction: Kelly = edge / variance. With negative edge, Kelly is negative → optimal long position is zero (or short, which we cannot execute).
  • Practical position size: 0%. Even ignoring the negative edge, the data-integrity gate (Auditor C, Forensic D) caps any high-conviction position; the tail-risk gate (Prosecutor C with active surveillance hypothesis) forces HOLD/SELL.
  • For comparison: if I forced a long against my will at the Trader's framing, Kelly × 0.25 capped at micro-cap illiquidity ceiling (ADV/100) would still come out at <10 basis points of book. That is not a position; it is rounding error. Skip it.

7. Risk gates

  • Tail-risk gate: TRIGGERED. Prosecutor grade C with active "File #1: market manipulation" hypothesis and 35% probability of SET surveillance flag in 6 months. Forensic grade provisional D. Per our framework, D/F on forensic forces HOLD/SELL regardless of trader/analyst views. This alone ends the debate.
  • Data-integrity gate: TRIGGERED. Auditor grade C with explicit DO-NOT-FABRICATE list covering every fundamental line item. Two SET pages returned 404. No income statement, no shareholder register, no filings. Per our framework, this caps conviction and prohibits a high-conviction BUY/SELL. Combined with the tail-risk gate, the only permissible outputs are AVOID, HOLD-trim, or exit. I have chosen AVOID/exit.

Any specialist who did publish numeric targets (Trader, Quant) was working off price data only — which the Auditor cleared as real. I have used those numbers. Any specialist who used invented fundamentals — none did; all refused — would have been discounted. The discipline shown by Analyst, PI, and Forensic in refusing to fabricate is exactly why I trust their "do not underwrite" verdict.

8. What I don't know (and what would change my mind)

  1. What caused the June 2026 ramp. Resolved by: SET filings 2026-05-25 → 2026-06-22 (Form 246-2, asset acquisition/disposal, Form 59 insider trades). If the catalyst is a clean disclosed event (e.g., land monetisation at a verified premium), the bull case probability rises materially.
  2. Whether MIDA is solvent and undiluted. Resolved by: latest 56-1 One Report, audited FY25 financials, share count history. If balance sheet shows net cash, clean equity history, and identifiable book value > ฿0.41, the avoidance call weakens.
  3. Who the controlling shareholders are and whether shares are pledged. Resolved by: SET shareholder list + SEC 59-2 pledge filings. Forced-pledge-liquidation is the Forensic's #1 bear scenario; ruling it in or out shifts the bear-case probability by ~15 points either way.

9. Triggers to re-rate

  • Re-rate to HOLD / small starter (up to 0.25% book) if: (a) SET filings retrieved show a benign, disclosed catalyst for the June spike; (b) audited FY25 financials show positive CFO and no going-concern; (c) shareholder register obtained shows no pledged-share overhang; (d) no SET surveillance designation within 6 months.
  • Re-rate to BUY (up to 0.75% book, Kelly × 0.25 with micro-cap liquidity cap) if: all four above PLUS price holds ฿0.35 on declining volatility AND clean dividend resumption confirmed.
  • Confirm AVOID / actively recommend exit if held if: (a) SET imposes C-sign, cash-balance, or trading halt; (b) any rights issue, PP, or warrant announcement; (c) auditor change or qualified opinion; (d) price closes below ฿0.27 on >5M volume (Trader's own stop).
  • No short at any trigger — borrow, illiquidity, and tail-squeeze risk make it uninvestable from the short side too.

Family-facing one-liner: We are not buying a stock when seven of our own specialists tell us they cannot see the financials, the regulator is likely already looking at the tape, and the expected return is negative 29%. There are 500 better names on the SET. Pass.

(Word count: ~1,640)