2026-07-12 · 2026-07 / week-2
Hong Pu Prices Ex-Dividend Failure, Not the First Buyback
Hong Pu Prices Ex-Dividend Failure, Not the First Buyback
Opportunity Ranking
| Rank | Idea | Discovery Lane | Why It May Be Best Now | Evidence Freshness | Catalyst Window | Near-Term >5% Move Case | Asymmetry | Main Reason to Reject |
|---|---|---|---|---|---|---|---|---|
| 1 | Hong Pu Real Estate 2536.TW long |
Taiwan low/mid-cap property / first treasury buyback / ex-dividend failure | The stock sits near a 52-week low and about 0.55x book after failing to fill a TWD 2.00 ex-dividend gap, while the company has just launched its first treasury buyback in roughly 30 years as a listed company. | High. Buyback decision dated 2026-07-08, buyback window began 2026-07-09, quote checked 2026-07-09/10 from Taiwan market data. | 2026-07-09 to 2026-09-06 buyback window, 2026-07-13 dividend record date, 2026-07-30 cash dividend payment. | A move from TWD 19.55 toward the TWD 20.80 limit-up reference, TWD 21.00 pre-dividend area, or TWD 22.00 book-discount repair is plausible if the corporate bid turns the failed ex-dividend trade into a shareholder-return tape. Evidence quality: medium-high. | Best mix of fresh catalyst, valuation gap, and surprise. | Property-cycle risk is real, and the buyback is only 1.8% of shares. |
| 2 | Shinpung 002870.KS long |
Korea low-priced KOSPI / on-market buyback | Fresh KRW 1.0bn buyback decision for 1,081,081 shares, with the program running from 2026-07-10 to 2026-10-08. | High. Korean articles and KRX/KIND screen dated 2026-07-09. | 2026-07-10 to 2026-10-08. | The stock already traded up to KRW 1,130 on 2026-07-10; further >5% is possible if the 108,108-share daily order limit absorbs liquidity. | Moderate. | No cancellation commitment, business quality is weak, and the first-day move may have pulled forward the clean setup. |
| 3 | LECIP Holdings 7213.T long |
Japan local small-cap / sub-JPY 800 / buyback execution update | It fits the Japan search preference: local Japanese small-cap, sub-JPY 800, and an active buyback status update dated 2026-07-01. | Medium. July 1 execution update is fresh, but the main authorization is from May. | Buyback mandate through 2027-03-23, next status update likely early August. | A thin JPY 500s stock can move >5% on execution proof, but the program is small and the catalyst is slow. | Lower. | Already screened as a runner-up in recent Desk runs; no materially new thesis. |
| 4 | Hong Kong small-cap special dividend screen | Hong Kong local-market / dividend event | HKEX entitlement screens showed upcoming July and August dividends, including small-cap names. | Medium. HKEX entitlement table checked during the run. | Mostly July to August ex-date windows. | Some dividend-driven >5% moves possible, but clean long asymmetry was weaker than Hong Pu. | Lower. | No better low/mid-cap long with fresh, size-relevant evidence surfaced in the research window. |
| 5 | Singapore low/mid-cap capital-return screen | Singapore local-market / SGX buyback and dividend scan | SGX screens remain useful, but this run did not surface a fresher non-duplicate long than Hong Pu. | Medium. Search run during 2026-07-12. | No superior dated catalyst found. | Not underwritten. | Low. | No candidate beat Hong Pu on freshness, valuation wedge, and immediate catalyst. |
Selected opportunity: Hong Pu Real Estate Development 2536.TW long.
Why this one now: the market is treating Hong Pu like a property stock that failed its ex-dividend test. The filing tape now says something narrower: a TWD 6.506bn market-cap company at roughly 0.55x book has authorized a 6,000-lot treasury buyback for the first time in its public-market life, immediately after the stock touched a 52-week low.
Why it can jump more than 5% soon: the current reference price is TWD 19.55. A return to TWD 20.80 is roughly 6.4%; a return to the TWD 21.00 area seen before the ex-dividend drop is about 7.4%; a move to TWD 22.00 is about 12.5%. The trigger is not an earnings forecast. It is the July 9 to September 6 buyback window, plus the July 30 dividend cash payment.
What should surprise the reader: the buyback looks small at 1.8% of shares, but it is large relative to the immediate narrative. Hong Pu is not buying after a normal dip. It is buying after an ex-dividend failure, near a 52-week low, below 0.6x book, after posting a 20-year high in first-half cumulative revenue.
Search and Duplicate Control
This run was explicitly scoped to low/mid-cap long ideas in Japan, Korea, Hong Kong, Taiwan, and Singapore. Searches used local-language or market-native terms, including 自己株式取得, 800円以下, 자사주 취득 결정, 주식소각, 股份回購, 特別股息, 庫藏股, 董事會決議買回, and SGX share buyback.
Recent primary Desk topics were excluded: Dida 02559.HK, Shinyoung 001720.KRX, Mgame 058630.KQ, Taroko 1432.TW, Kona I 052400.KQ, Seohee Construction 035890.KQ, Pungkang 093380.KQ, NCS&A 9709.T, Merchant Bankers 3121.T, Miraial 4238.T, Mitani Sekisai 5273.T, Leader Electronics 6867.T, Space Shower SKIYAKI 4838.T, Tein 7217.T, Koh Brothers Eco Engineering 5HV.SI, Valuetronics BN2.SI, Swancor 3708.TW, and K Auction 102370.KQ.
Japan compliance note: LECIP 7213.T was the compliant Japan candidate because it is a local small-cap and trades below JPY 800. It was rejected because the July 1 update only showed partial execution of an older, small buyback program. Hong Pu won because the evidence was fresher, the price was closer to a local stress point, and the buyback represented a sharper narrative break.
The Setup
Hong Pu is a Taiwanese residential and commercial property developer. The stock closed at TWD 19.55 on the latest Taiwan quote captured from Cnyes, up 3.17% on the day, with a TWD 18.85 to TWD 29.70 52-week range, TWD 6.506bn market value, TWD 35.24 book value per share, 5.90x trailing P/E, 0.55x P/B, and a TWD 2.00 annual dividend shown by Cnyes. Goodinfo lists the 2026 ex-dividend date as 2026-07-07, the dividend record date as 2026-07-13, and the cash payment date as 2026-07-30.
The local tape matters. Cnyes reported that Hong Pu went ex-dividend for TWD 2.00 on 2026-07-07, with an ex-dividend reference near TWD 19.60, then slipped into an unfilled ex-dividend trade. On 2026-07-08, the board approved a treasury-share buyback for 6,000 thousand shares from 2026-07-09 to 2026-09-06, with a buyback range of TWD 13.5 to TWD 31.5, and said purchases would continue even below the lower bound. Cnyes described this as Hong Pu's first treasury buyback in more than 30 years as a listed company.
The Mispricing
The weak version of the market story is simple: Taiwanese builders are cyclical, project revenue is lumpy, June revenue collapsed month over month, and a small 1.8% buyback does not fix property-cycle risk.
That argument is not wrong. It is incomplete.
The market is pricing the failed dividend fill and June revenue air pocket as if they define the company. The filing path says the opposite. Hong Pu's first-half 2026 cumulative revenue was TWD 4.991bn, up 35.7% year over year, and Goodinfo marks it as a 20-year high for the comparable period. The company is also buying back stock just below TWD 20 while Cnyes shows book value per share at TWD 35.24.
The disagreement is not "cheap property stock goes up." It is sharper: price says the ex-dividend failure is the relevant signal; management's first buyback says the near-book discount and tape damage became actionable.
Price
| Item | Level / Data | Source and Timestamp |
|---|---|---|
| Last captured price | TWD 19.55 | Cnyes Taiwan quote page, market closed, checked 2026-07-12; page shows latest quote and 2026-07-09 quote fields |
| Same quote on Yahoo Singapore mirror | TWD 20.20 | Yahoo Finance Singapore 2536.TW, checked 2026-07-12; use with caution because Taiwan-market mirrors can lag or differ |
| Market cap | TWD 6.506bn | Cnyes quote page |
| Book value per share | TWD 35.24 | Cnyes quote page |
| P/B | 0.55x | Cnyes quote page |
| Trailing P/E | 5.90x | Cnyes quote page |
| 52-week range | TWD 18.85 to TWD 29.70 | Cnyes quote page |
| Annual cash dividend | TWD 2.00 | Cnyes and Goodinfo dividend calendar |
| Buyback size | 6,000 thousand shares, about 1.8% of equity | Cnyes report on board decision |
| Buyback range | TWD 13.5 to TWD 31.5, with purchases continuing below TWD 13.5 | Cnyes report on board decision |
| Buyback window | 2026-07-09 to 2026-09-06 | Yahoo Taiwan / Cnyes disclosure coverage |
| June 2026 revenue | TWD 15.3mn, down 97.1% YoY | Goodinfo monthly revenue table |
| 1H 2026 cumulative revenue | TWD 4.991bn, up 35.7% YoY | Goodinfo monthly revenue table |
Positioning
The direct positioning evidence is the corporate bid. It is not perfect, but it is filing-linked and time-bound.
Hong Pu has 333mn shares outstanding on the Cnyes page. A 6mn-share buyback is only 1.8% of shares, so the thesis does not rest on a huge denominator change. It rests on timing and signal quality: the first buyback in about 30 years follows an ex-dividend failure, near a 52-week low, with the buyback ceiling set far above the market price.
Missing data: no reliable live short-interest, borrow-cost, options-chain, foreign ownership flow, or fund-flow dataset was available in this run. Taiwan property small/mid-caps are not ideal for clean derivatives positioning. The best available positioning evidence is the disclosed corporate buyer, the low turnover shown on quote pages, and the price response around the ex-dividend date.
Catalyst
The first catalyst is already live: the buyback window began on 2026-07-09 and runs to 2026-09-06. The immediate test is whether the company is a real buyer, not a press-release buyer.
The second catalyst is the dividend cash date. Goodinfo lists the TWD 2.00 cash dividend payment for 2026-07-30. That matters because an ex-dividend failure can flip if cash returns to shareholders while the company is also in the market buying stock.
The third catalyst is proof that June's revenue collapse was a timing gap, not a demand crack. Goodinfo shows May 2026 revenue of TWD 4.047bn and June revenue of only TWD 15.3mn, while cumulative first-half revenue still rose 35.7% year over year. A property developer can show violent monthly revenue timing. The market may be treating June as signal rather than project-recognition noise.
Payoff Map
This is a common-stock long with a defined invalidation point, not a clean merger-spread trade. The expected value can be estimated from price targets, but it should not be treated as a precise model because buyback execution speed and project revenue recognition are not disclosed daily.
At TWD 19.55:
- Top case: the stock refills the ex-dividend gap, re-rates toward TWD 23.50, and still trades at only 0.67x book.
- Base case: it returns to TWD 21.80 to TWD 22.00 as the buyback proves real and the July 30 cash payment resets the tape.
- Bottom case: it breaks the TWD 18.85 52-week low because the buyback is too small, property sentiment worsens, or July and August revenue confirm a deeper issue.
Probability-weighted target using TWD 23.50, TWD 21.80, and TWD 17.50 is TWD 21.34, or about +9.2% before dividends and trading costs.
Price Target and Probability Map
| Scenario | Probability | Target / Level | Return / Payoff | Time Horizon | Conditions Required | Evidence Quality |
|---|---|---|---|---|---|---|
| Top Case | 25% | TWD 23.50 | +20.2% before dividend cash | 1 to 8 weeks | Buyback execution is visible in the tape, the stock clears TWD 21, and the market treats June revenue as timing noise rather than demand failure | Medium |
| Base Case | 50% | TWD 21.80 | +11.5% before dividend cash | 2 to 6 weeks | Corporate bid absorbs post-dividend selling and the July 30 cash payment helps repair the failed ex-dividend trade | Medium-High |
| Bottom Case | 25% | TWD 17.50 | -10.5% | 1 to 6 weeks | Buyback is slow or symbolic, property sentiment weakens, or new monthly revenue confirms a real slowdown | Medium |
| Invalidation / Stop Condition | n/a | Sustained close below TWD 18.85 or disclosure that buyback execution is negligible | n/a | Immediate to September 2026 | The 52-week low breaks despite a live buyback, or the company does not materially execute inside the window | High |
Probability-weighted expected value: TWD 21.34 target, about +9.2% from TWD 19.55 before dividend cash, costs, and slippage.
Current market price / level: TWD 19.55 from Cnyes, market closed; Yahoo Finance Singapore mirror showed TWD 20.20, so the article uses the more conservative Cnyes local quote.
Timestamp: market data checked 2026-07-12 15:05 ICT; Taiwan market was closed for the weekend, so levels are last available closes.
Primary instrument: Hong Pu Real Estate Development common stock 2536.TW.
Alternative expressions considered: Taiwan property basket rejected because the thesis is company-specific; options rejected because no reliable liquid chain was available; waiting for buyback execution proof rejected because the first leg of rerating could occur before the next monthly status datapoint.
Confidence: Medium. The catalyst is fresh and visible; execution data is still missing.
What Would Prove This Wrong
The thesis fails if Hong Pu closes below TWD 18.85 while the buyback window is open. That would mean the corporate bid is not strong enough to defend even the recent low.
It also fails if management does not execute the buyback in meaningful size, if July or August revenue shows that May's recognition was a one-off without follow-through, or if Taiwan property financing conditions deteriorate enough that book value becomes a poor anchor.
The most fragile assumption is that the first buyback has information value. If it is symbolic capital-market theater, 1.8% of shares is not enough to force a rerating by itself.
Risk Audit
Strongest counterargument: Hong Pu is cheap because Taiwanese property revenue is project-timing driven and balance sheets carry inventory risk. A low P/B is not automatically a bargain if the market distrusts land values, future margins, or cash conversion.
Most fragile assumption: the buyback is a real signal rather than a small support operation after dividend disappointment.
What the market may already know: the first-half revenue number looks strong because May was unusually large. June then collapsed to TWD 15.3mn, so investors may be discounting the quality and timing of revenue, not ignoring it.
What could make the trade lose money even if the thesis is directionally right: the company may buy slowly, liquidity may be thin, and the stock can drift below the buyback range before monthly execution proof arrives.
Liquidity / execution risks: use limit orders only. The Cnyes page showed 1,355 lots traded in the captured session and seasonal average volume of 832 lots. That is tradable for small sizing, not for careless market orders.
Leverage risks: property developers carry cycle and financing risk. Goodinfo shows 26Q1 liabilities at 66.3% of assets and inventory at 71.2% of assets. That is the real balance-sheet risk.
Information reliability risks: the buyback data comes through local disclosure coverage and market-data pages rather than a freshly downloaded MOPS filing in this run. The numbers are consistent across Taiwanese financial media, but the next step would be direct MOPS document verification.
Invalidation trigger: sustained close below TWD 18.85, negligible buyback execution, or new revenue data confirming a structural slowdown.
Publish / revise / reject recommendation: publish as a medium-conviction Deep Dive, with explicit missing-positioning and execution-risk notes.
Best Trade Strategy
Direction: long Hong Pu common stock.
Preferred instrument: 2536.TW common shares.
Common-stock stance: acceptable only with limit orders and small sizing. The setup is a buyback-and-book-discount repair trade, not a high-liquidity index trade.
Options stance: avoid. No reliable liquid options chain was found in this run.
Take-profit / target: base target TWD 21.80; stretch target TWD 23.50 if buyback execution becomes visible and the stock clears the TWD 21 area.
Stop loss / invalidation: close below TWD 18.85, the captured 52-week low, or evidence that buyback execution is negligible.
Timeline: 1 to 8 weeks, anchored to the 2026-07-09 to 2026-09-06 buyback window and the 2026-07-30 dividend payment date.
Execution risks: thin turnover, weekend stale quote risk, Taiwan property sentiment, spread cost, and buyback execution uncertainty.
Do-not-trade conditions: do not enter after a one-session spike above TWD 22 without new execution evidence; do not enter if the company files weak buyback progress; do not enter if Taiwan property credit stress becomes the dominant market variable.
Monitoring checklist: daily close versus TWD 18.85, turnover versus 832-lot seasonal volume, MOPS buyback execution updates, July and August monthly revenue, and any company update on project recognition.
Live-price note: primary quote used is TWD 19.55 from Cnyes; market was closed when checked on 2026-07-12. A Yahoo mirror showed TWD 20.20, so sizing should use the next live Taiwan open rather than stale weekend prints.
Bottom Line
Hong Pu is not a clean secular compounder and not a simple yield trade. It is a short-window disagreement between a failed ex-dividend tape and a first-ever buyback. At TWD 19.55, the market prices the June revenue air pocket and property-cycle discount. The company is buying below 0.6x book after first-half revenue reached a 20-year high for the period. The trade is worth publishing because the catalyst is already live, the downside is monitorable, and the surprise is concrete: the first buyer after the failed dividend is not a newsletter reader. It is the issuer.
Research Quality Scorecard
| Criterion | Score | Evidence Note |
|---|---|---|
| Market disagreement | 4 | Clear tension between ex-dividend failure, low P/B, and first treasury buyback, though property-cycle risk keeps it below a 5. |
| Evidence base | 4 | Fresh Taiwanese market data and local buyback coverage; direct MOPS filing verification remains the main gap. |
| Positioning and flows | 3 | Corporate buyback is documented, but live short interest, foreign flow, and options data are missing. |
| Catalyst path | 4 | Buyback window and dividend payment date are specific and near-term. |
| Payoff architecture | 4 | Defined targets and stop; upside/downside skew is positive but not extreme because the buyback is only 1.8% of shares. |
| Invalidation discipline | 5 | TWD 18.85 low, buyback execution, and revenue follow-through are monitorable. |
| Differentiated insight | 4 | The non-obvious point is the first-buyback signal after an ex-dividend failure, not a generic cheap property call. |
| Client value | 4 | Useful as a capital-return tape watch even if no trade is taken. |
Total: 32 / 40.
Section 17 Quality Gate
| Gate | Answer |
|---|---|
| 1. Is the mispricing specific? | yes |
| 2. Is there evidence beyond narrative? | yes |
| 3. Is the positioning claim supported or clearly labeled as uncertain? | yes |
| 4. Is there a catalyst or plausible closing mechanism? | yes |
| 5. Is the downside case described honestly? | yes |
| 6. Is the strongest counterargument included? | yes |
| 7. Is the article useful even if the trade is not taken? | yes |
| 8. Are all factual claims sourced or marked as unverified? | yes |
| 9. Does the article avoid hype? | yes |
| 10. Does the headline match the actual evidence? | yes |
| 11. Does the article explain why this is the best opportunity right now? | yes |
| 12. Does the article explain why the selected asset can plausibly jump more than 5% soon? | yes |
| 13. Does the article identify what should surprise a sophisticated reader? | yes |
| 14. Does the article include top, base, and bottom targets with probabilities that add to 100%? | yes |
| 15. Does the main article file include its Research Quality Scorecard in a dedicated section? | yes |
| 16. Are all reader-facing tables kept as Markdown tables in the main article file? | yes |
| 17. If optional table images were explicitly requested, are they saved separately? | n/a |
| 18. If the task required an illustration prompt, is it included inline? | yes |
| 19. Does the file include a Best Trade Strategy section? | yes |
| 20. If technical signals are used, are they framed as timing inputs? | yes |
| 21. Unless scoped, were U.S., Japan, broader Asia, and Europe/UK screened? | n/a, user scoped to Japan, Korea, Hong Kong, Taiwan, and Singapore |
| 22. If Japan is in scope, did the screen prioritize local small/mid-cap and sub-JPY 800 names? | yes, LECIP was screened and rejected |
| 23. If live Substack finish was requested, was Substack updated? | n/a |
Sources
| Source | Use |
|---|---|
Cnyes Taiwan quote page for Hong Pu 2536 |
Latest captured quote, market cap, P/E, P/B, book value per share, 52-week range, turnover, dividend yield, and operating metrics. |
Goodinfo Hong Pu 2536 page |
Company basic data, dividend calendar, monthly revenue table, balance-sheet ratios, and historical dividend data. |
| TradingView / Cnyes article on Hong Pu first treasury buyback | Buyback decision, buyback size, price range, timing, first-buyback context, ex-dividend failure, and project pipeline context. |
| Yahoo Taiwan announcement on Hong Pu treasury-share board decision | Local disclosure coverage for the July 8 board resolution. |
Yahoo Finance Singapore mirror for 2536.TW |
Secondary quote cross-check, used with caution because mirror quotes can differ from local pages. |
| DigitalToday article on Shinpung buyback | Korea runner-up buyback size, period, purpose, and financial context. |
| Investing.com Korea article on Shinpung buyback | Korea runner-up daily order limit, prior close, treasury holdings, and buyback terms. |
| LECIP July 1 buyback status PDF | Japan compliant sub-JPY 800 runner-up buyback execution status. |
StockAnalysis page for LECIP 7213.T |
Japan runner-up quote, market cap, valuation, volume, and earnings date context. |
| HKEX entitlement table | Hong Kong dividend-event screen. |
AI Illustration Prompt
Create a realistic, high-value, high-end editorial cover image for The Mispricing Desk about Hong Pu Real Estate
2536.TW, a Taiwanese property developer whose stock failed to fill a TWD 2 ex-dividend gap just as the company launched its first treasury buyback in roughly 30 listed years. Composition: a rain-dark Taipei construction site at dusk, with a half-lit residential tower model on a broker's desk; beside it, a crisp treasury-share order ticket markedTWD 13.5-31.5presses down on a torn ex-dividend calendar page dated July 7. In the background, a muted TWSE order book glows near19.55, with0.55x booketched faintly into the glass. Mood: restrained, skeptical, analytical, no promotional real-estate glamour. Palette: concrete gray, deep teal, warm office tungsten, small red and green market lights. Style: realistic financial editorial photography suitable for The Economist, Barron's, or Bloomberg Markets. Include a subtle but clear watermark/text reading "The Mispricing Desk" on the lower-right edge of the desk blotter. No generic stock chart hero, no cartoon bulls, no hype, no AI slop.