Deal Origination System

Systematic infrastructure for succession deal sourcing in Hong Kong & Taiwan
The same architecture that runs the best trading desks — applied to deal origination
The Data Play Nobody Is Running P0

The broker ecosystem in HK and Taiwan is not thin. It is hiding in plain sight inside government databases that nobody in succession PE is querying programmatically.

🇹🇼

Taiwan GCIS — The Entire SME Map, Free and Queryable

data.gcis.nat.gov.tw · findbiz.nat.gov.tw · Open API · Monthly CSV dumps · 1.63M companies

What exists: Taiwan's MOEA publishes the entire company registration database through GCIS (Government Commerce and Industry Services). Every registered company in Taiwan — name, address, representative (負責人), paid-in capital, industry code, registration date, and current status. Updated monthly. Downloadable as CSV. Free API access with a simple email application (1-2 days approval).

What you can do with it that nobody is doing:

Registration date filterCompanies registered 1970-1995 = founded during the "economic miracle." Their founders are now 65-85. Filter by industry code (manufacturing: C-codes). This gives you a target list of aging manufacturing companies by sector and region.
Capital size filterPaid-in capital TWD 10M-500M = SME sweet spot. Too small below, too large above. Cross-reference with industry code to find precision machinery companies with TWD 50M+ capital registered in the early 1980s.
Representative name + cross-referenceThe representative (負責人) field gives you the founder's name. Cross-reference with TWSE filings, industrial association membership lists, and trade show exhibitor directories. When the same name appears as representative of a 40-year-old company AND in a family business executive education program at NTU — that is a live succession signal.
Status change monitoringGCIS publishes dissolution and modification lists. A company changing its representative after 30+ years of the same name is a succession event — either it already happened or it's about to. Monitor these changes monthly.
findbiz.nat.gov.twThe front-end search portal. Search by company name, address, representative name, or director name. Cross-domain search links to factory registrations. Combine with GCIS API for bulk queries.

The build: A Python script that pulls GCIS data, filters for manufacturing companies registered pre-1995 with capital >TWD 50M, cross-references representative names against age-proxy signals (registration date, no representative change in 20+ years), and outputs a ranked target list by sector and region. This is a weekend project. Nobody has built it because nobody in Taiwan's PE market is thinking about succession as a systematic sourcing problem.

🇭🇰

Hong Kong ICRIS — Director Data at Scale

icris.cr.gov.hk · Companies Registry · 1.4M+ registered companies · Director index searchable

What exists: Hong Kong's Companies Registry maintains ICRIS (Integrated Companies Registry Information System). Over 1.4 million registered companies. Searchable by company name, registration number, or director name. Key data: incorporation date, current status, registered office, director details (name, appointment date), company secretary, share capital.

Post-2023 privacy change: The December 2023 ICRIS overhaul restricted director residential addresses and full ID numbers to "specified persons" (regulated entities). Director names, appointment dates, and company affiliations remain publicly searchable. This actually matters less than it sounds — you don't need addresses, you need patterns.

What you can do with it:

Director longevityA director who has held the same position since incorporation in 1985 is very likely the founder. Companies with single long-tenured directors and no recent board changes are prime succession targets.
Incorporation date filterSame logic as Taiwan. Companies incorporated 1970-1995, still active, private limited company — that is your universe. HK has roughly 1.27M private companies on the register.
Annual return gapsA company that has filed annual returns consistently for 30 years and then misses one — something changed. The owner may be ill, distracted, or checked out. This is a signal.
Company secretary changesWhen a company switches from a named individual secretary to a corporate secretarial firm (like Tricor, Vistra, or Boardroom), it often signals professionalization before a sale or estate planning. Monitor these changes.
OpenCorporates bulk dataOpenCorporates indexes HK Companies Registry data and offers API access. For bulk analysis (filtering 1.4M companies), this is more practical than ICRIS's web interface. Combine with industry classification from Business Registration data.

The limitation: HK's Companies Registry does not classify companies by industry. Business Registration Certificates (from Inland Revenue Department, not Companies Registry) carry the business nature code. Getting both datasets to talk to each other requires matching on company name or UBI (Unique Business Identifier, introduced Dec 2023). Not trivial, but doable.

Why this matters: Every fund in the world sources deals through brokers, platforms, and relationships. Nobody in HK or Taiwan succession PE is building programmatic sourcing from government data. This is the equivalent of on-chain analysis for crypto — the data is public, freely available, and nobody is reading it systematically. The first fund that builds this pipeline owns the information advantage.
The Architecture
"The best trading desks don't win on speed. They win on processing and routing. PINECONE needs the same architecture for succession deals."

Event-driven trading desks layer four modules — market monitoring, alert systems, flow tracking, and fundamental research — into a single decision framework. The same architecture applies to deal origination. The signals are different:

Trading Desk AnalogyPINECONE DOI EquivalentWhat It Catches
On-Chain Monitoring
wallet flows, exchange deposits
Platform Monitoring
listing aggregation, broker scraping
New listings on SMERGERS, DealStream, Kumo, Axial, BizBuySell, broker websites. The "exchange flow" of deal origination.
Twitter Alert System
keyword tracking, priority routing
Ecosystem Signal Tracking
advisor alerts, event monitoring, news
Broker newsletters, MAPECT announcements, law firm deal alerts, industrial association event calendars, obituaries/retirements in trade press, M&A league table changes.
Exchange Flow Monitoring
inflows/outflows, anomaly detection
Relationship Pipeline CRM
broker touchpoints, referral tracking
Track every broker, CPA, lawyer, banker interaction. Score relationships by deal quality. Measure referral conversion. Alert on stale relationships.
Event-Driven Research
fundamental analysis, scenario planning
Target Intelligence Research
company analysis, owner profiling
Deep research on targets that pass initial screen. Owner background, family dynamics, financial forensics, durability audit. The "trade thesis" equivalent.
The key insight: The edge is not speed. It is processing and routing. You will not hear about every succession deal first. The ones you catch, you process faster and route to a decision more efficiently than anyone else in the market. That is the entire game.
Module 1: Platform Monitoring P0

Passive deal aggregation. Set it. Forget it. This catches 5-10% of real deals — the ones that surface publicly. The other 90% come from Modules 2 and 3.

Platforms to Monitor
🌐

Kumo — Deal Aggregator

HKTW 100K+ listings · AI-powered · $49/mo starter

What it is: Aggregates business-for-sale listings from thousands of brokers and dozens of marketplaces globally. AI cleans unstructured data, deduplicates, and generates concise summaries. Covers North America, Europe, and Asia.

How to use it: Create saved searches with PINECONE's criteria: geography (HK, Taiwan), revenue floor ($1.5M+), industry filters (manufacturing, services), keyword filters ("retirement", "succession", "family business", "owner retiring"). Set daily email alerts. Export to CRM weekly.

Limitations for Asia: Coverage is strongest in US/Europe. HK and Taiwan listings are thinner. Best used as a passive sweep layer, not primary sourcing.

Action: Sign up, create 3-4 saved searches, set daily alerts. Review weekly. Low effort, catches outliers.

🔵

SMERGERS — Asia SME Marketplace

HKTW 131 HK listings · 21 TW brokers · Free buyer account

What it is: The most relevant online marketplace for HK/TW SME deals. 110K+ member network across 170 countries. Self-described as "the first online investment bank for SMEs." Lists both businesses for sale and brokers/advisors.

HK coverage: 131 active listings (businesses for sale + investment). 35 businesses actively for sale. 23 active brokers. Mix of SFC-regulated entities, clinics, F&B, tech companies, licensed companies (money service operators, trust companies).

TW coverage: 21 active brokers. Listings thinner — Taiwan SME owners don't typically list online. More useful for identifying active brokers than finding deals directly.

How to use it: Create buyer profile with clear succession mandate. Browse HK/TW listings weekly. More importantly: use the broker directory to identify and contact active intermediaries. Message 5-10 brokers with your buy criteria. SMERGERS itself is the fishing pond — the brokers are the real fish.

🟡

DealStream — HK Focused Listings

HK Hundreds of HK listings · Stronger for licensed entities · Free browse

What it is: Another online marketplace, stronger in HK specifically. Listings skew toward regulated/licensed businesses (SFC brokers, asset management cos, TCSP trusts, money lenders). Also has F&B, tech, manufacturing.

How to use it: Browse HK listings filtered by revenue range and industry. Cross-reference with SMERGERS to avoid duplicates. DealStream often has listings SMERGERS doesn't and vice versa.

Limitation: Many "shell company" listings — entities being sold primarily for their licenses, not operating businesses. Filter carefully for actual operating companies with revenue.

🟣

Additional Platforms to Monitor

Axial · BizBuySell · Dealsuite · SourceCo · Inven
AxialUS-focused but has international deal flow. PE and lower middle market. Less relevant for HK/TW but worth monitoring.
BizBuySellLargest US marketplace. Minimal Asia coverage. Skip for now.
DealsuiteEurope-focused. Strong cross-border M&A matchmaking. Could catch HK/European deals. Worth registering as a buyer.
SourceCoAI-driven off-market sourcing. 200M+ SMB database. Better for proactive outreach than passive monitoring. Consider for Phase 2 when you want to do direct outreach to TW manufacturing targets.
InvenAI M&A search. 23M companies, 430M contacts, 93% matching accuracy. Enterprise tool. Consider when scaling.
Cyndx FinderAI deal search in 7 languages including Chinese. Could be useful for Taiwan.
Reality: The best deals never appear on platforms. This layer catches 5-10% of what's available. Set it and forget it. Your real edge is Modules 2 and 3 — and the GCIS/ICRIS data play described below.
Module 2: Ecosystem Signal Tracking P0

The deals that matter never appear on platforms. They surface through signals in the ecosystem — an obituary in the trade press, a CPA asking about estate planning, a founder who stopped attending industry events. This module catches those signals.

Signal Sources by Market
🇭🇰

Hong Kong — Signal Sources

Big 4 newsletters · HSBC Commercial alerts · HKICPA · Chamber events · Trade press
SourceSignal TypeAction
PwC HK Deals NewsletterMarket trends, deal announcements, sector reportsSubscribe. Read monthly. Note sectors with increasing activity.
Mazars / PKF / Baker Tilly HKMid-tier audit firms see aging owner financials firsthandMeet M&A advisory partners. Present buy mandate. They see succession signals before anyone: declining owner engagement, estate planning questions, children refusing to take over.
HSBC Commercial BankingRM network covers thousands of HK SMEsBuild relationship with 2-3 senior RMs who cover SME clients. They know which owners are thinking about exit.
HKTDC / HKGCC eventsTrade Development Council + General Chamber of Commerce host industry eventsAttend quarterly. Meet founders face-to-face. The 65+ year old at the networking table who's there alone (no team) is your target.
HKICPAInstitute of CPAs. Members are the most trusted advisors to SME owners.Join SME-focused working groups. Present at events on "succession planning options." Position as buyer, not seller.
South China Morning Post / HKEJBusiness obituaries, retirement announcements, company anniversaries (40th, 50th — founder is 65+)Set Google Alerts for: "Hong Kong" + ("retirement" OR "succession" OR "family business" OR "next generation" OR "stepping down")
Private bank wealth advisorsUBS, HSBC Private Banking, JP Morgan PB cover HNW foundersHarder to access. Attend private banking conferences. Position as solution for "my client who wants to sell."
🇹🇼

Taiwan — Signal Sources

MAPECT · Industrial associations · Big 4 audit teams · Local CPAs · MOEA data
SourceSignal TypeAction
MAPECT eventsTaiwan M&A and PE Council. Annual awards + conferences. Convenes M&A community.Join as member. Attend every event. Sponsor award category. This is where Taiwan's M&A deals get discussed.
PwC Taiwan Deals AdvisoryIntegrated M&A: legal, tax, corporate finance, DD. Audit and tax client base = sourcing goldmine.Meet deals advisory / corporate finance team. Present succession buy mandate with clear criteria. Their audit partners see aging owners firsthand.
TAMI / TEEMA / TCIAIndustrial associations: Machinery, Electronics, Chemicals. Trade shows, member directories, founder networks.Join 2-3 associations in target sectors. Attend trade shows (TIMTOS, SEMICON Taiwan, TAITRONICS). Founders attend these — meet them where they are.
Local CPA firms (會計師事務所)Taiwan's most trusted SME advisors. Equivalent of Japan's zeirishi. They know which owners are thinking about exit.Build referral network with 10-15 CPA firms in Taipei, Taichung, Kaohsiung. Offer referral fee (1-2% of deal value) for qualified succession introductions.
MOEA SME White PaperAnnual data on SME demographics, sectors, regional distribution.Read annually. Identify sectors with highest concentration of aging founders. Use as targeting intelligence.
Lee and Li / LCS & PartnersTop Taiwan law firms. Their M&A teams see deal flow constantly.Engage as legal counsel. Their client referral network is the most valuable in the country. Being their client = seeing their deal flow.
CDIB Capital deal overflowThe only PE fund targeting succession. They pass on deals that are too small or don't fit.Build relationship with Steven Wu (Head of Buyout). Understand where they pass on deals. Their rejects = your targets.
NTU/NCCU business school eventsExecutive education programs attract aging founders seeking succession guidance.Guest lecture on "succession options." Attend family business programs. Founders self-select into these programs.
Taiwan CPA referral network — the killer move: In Japan, NGTG built their pipeline through M&A advisors and regional banks, paying advisory fees. In Taiwan, CPAs play this role. Build a structured referral program: fixed fee for introduction that passes first screen, success fee on close. Start with 10 CPAs, scale to 50. This is your proprietary sourcing channel that nobody else has built.
Alert Configuration

For each signal source: tracked keywords, priority level, and routing.

ChannelKeywords / TriggersPriorityRoute To
Google Alerts"Hong Kong" + succession / retirement / family business / selling business / "next generation"HIGHSlack #deal-signals
Google Alerts"Taiwan" + 接班 (succession) / 傳承 (inheritance) / 退休 (retirement) / 出售 (for sale)HIGHSlack #deal-signals
SMERGERS / DealStreamNew listing in HK or TW with revenue > $1MMEDIUMSlack #platform-listings
Kumo daily digestAsia filter, manufacturing + services, succession keywordsMEDIUMEmail → weekly review
MAPECT website / socialsEvent announcements, deal awards, member newsLOWSlack #ecosystem-intel
Trade press (SCMP, HKEJ, 工商時報)Founder obituaries, retirement, company milestones, management changesHIGHSlack #deal-signals
Module 3: Relationship Pipeline CRM P0

Every broker, CPA, lawyer, banker, and advisor is a node in your system. Track: last contact date, deal quality of referrals, conversion rate, relationship health. Stale relationships (>30 days no contact) are dead relationships. The CRM enforces discipline.

CRM Architecture (Notion or HubSpot)
1

Contact Database

Every intermediary, advisor, and potential referral source. Fields: name, firm, role, market (HK/TW), sector expertise, last contact date, relationship score (1-5), deal referral count, conversion rate.

2

Interaction Log

Every meeting, call, email, event. Auto-calculate "days since last contact" and flag relationships going stale (>30 days = yellow, >60 days = red). Track liveliness of relationships the way a trading desk tracks liveliness of positions.

3

Deal Pipeline

Every deal that enters the funnel. Stages: Signal → First Screen → Owner Meeting → LOI → Due Diligence → Close. Track source attribution back to which contact/channel originated the deal.

4

Attribution Analytics

Which channels produce the most deals? Which contacts have the highest conversion? Where should you invest more relationship time? This is your edge framework. Not all sources are equal. Rank them. Double down on what converts.

Initial Contact Database: HK
📋

HK Broker & Advisor Target List (28 contacts)

Tier 1 (meet first) · Tier 2 (month 2-3) · Tier 3 (ongoing)
TierName / FirmWhyFirst Contact Method
TIER 1
Week 1-4
PwC HK — Deals AdvisoryBig 4 see aging owners in audit client base. Integrated M&A advisory.Request intro meeting. Present succession buy mandate.
Mazars HKMid-tier audit. More SME exposure than Big 4. Less saturated.Meet M&A advisory partner directly.
PKF HKMid-tier audit and advisory. Strong SME practice.Same approach as Mazars.
Ivory CapitalBoutique IB in HK. Mid-market M&A advisory.Direct approach. Position as permanent-hold buyer.
HSBC Commercial Banking — 2-3 Senior RMsLargest SME lending book in HK. RMs know who's thinking about exit.Warm intro via HSBC contact. Or cold approach RM team covering your target sectors.
Baker Tilly HKMid-tier audit. Corporate advisory practice.Meet advisory team.
TIER 2
Month 2-3
Armor CapitalBoutique M&A advisory. Cross-border.Direct approach.
HKGCCHK General Chamber of Commerce. Events, committees.Join. Attend SME committee events.
Synergy Business Brokers (Asia)Award-winning business broker with HK presence.Register as buyer. Submit acquisition criteria.
SMERGERS HK brokers (top 5)23 active brokers on platform. Message top 5 by deal count.Message through platform with clear buy criteria.
DealStream HK brokersActive deal listers. Cross-reference with SMERGERS.Register as buyer. Contact active listers.
TIER 3
Ongoing
Private bank wealth advisorsUBS, JP Morgan PB teams cover HNW founders.Conference networking. Harder to access cold.
Family office networksSome HK family offices sell operating businesses.HKVCA events, family office conferences.
Insolvency practitionersDeloitte, KPMG restructuring teams. Not succession per se, but distressed exits by aging owners.Niche. Monitor for opportunities.
Initial Contact Database: Taiwan
📋

TW Broker & Advisor Target List (30+ contacts)

Tier 1 (meet first) · Tier 2 (month 2-3) · Tier 3 (build over time)
TierName / FirmWhyFirst Contact Method
TIER 1
Week 1-4
Lee and Li, Attorneys-at-LawTaiwan's oldest and largest law firm. Wrote the BMAA. #1 Bloomberg M&A league tables. Their client referral network is the most valuable in the country.Engage as legal counsel. Request meeting with M&A practice group.
LCS & PartnersAdvised on the Jintex/CDIB succession deal. Top M&A law firm.Engage as alternative/complementary counsel. Discuss succession deal flow.
PwC Taiwan — Deals AdvisoryBig 4 with integrated M&A advisory. Audit client base = sourcing goldmine.Meet deals advisory team. Present succession buy mandate with clear financial criteria.
MAPECTTaiwan M&A and PE Council. Industry body. Annual awards. Convenes the community.Join as member. Attend next event.
CDIB Capital GroupOnly PE fund targeting succession. Did Jintex. Their rejects = your targets.Request meeting with Steven Wu (Head of Buyout). Position as potential co-investor/partner.
TIER 2
Month 2-3
Deloitte / KPMG / EY TaiwanBig 4 audit and M&A advisory. Similar approach to PwC.Meet deals advisory teams. Present buy mandate.
ONEtoONE Corporate FinanceGlobal mid-market M&A. China/Asia coverage.Direct approach. Register buy mandate.
YCP GroupM&A target sourcing consulting with Taiwan expertise. Cross-border specialist.Engage as sourcing consultant for Taiwan mandate.
Winkler PartnersBoutique law firm. Strong cross-border M&A practice.Meet M&A team. Alternative to Lee and Li / LCS.
Mega Financial / First Financial / Taiwan Business BankDomestic banks with deep SME lending relationships.Meet commercial banking teams covering SME segments.
TIER 3
Build over 3-6 months
Local CPA firms (10-15 in target cities)The killer sourcing channel. CPAs are the most trusted SME advisors. Structured referral program.Identify CPAs in Taipei, Taichung, Kaohsiung through MAPECT, law firm referrals, TWSE company filings. Offer structured referral fee.
TAMI / TEEMA / TCIAIndustrial associations. Founders attend trade shows and events.Join 2-3 in target sectors. Attend events. Meet founders directly.
NTU / NCCU business schoolsFamily business executive education programs.Guest lecture on succession. Attend events.
SMERGERS TW brokers (top 5)21 active brokers on platform.Message through platform with buy criteria.
Module 4: Target Intelligence Research P1

Once a signal passes the first screen, deep research kicks in. Three layers: (1) Public intelligence — filings, industry reports, customer landscape. (2) Proprietary intelligence — owner profiling from advisors, CPAs, contacts. (3) Pattern recognition — what to systematize so next time you move faster.

1

Company Intelligence Gathering

Public filings (TWSE, IRD annual return, Companies Registry), industry reports, customer reviews, competitor landscape, supply chain position, key contracts. The "general knowledge" layer.

2

Owner Profiling

Age, family situation, children's involvement (or lack thereof), health signals, estate planning status, emotional attachment to business, prior succession conversations. The "on-chain" equivalent — proprietary intelligence from advisors, CPAs, industry contacts.

3

First Screen Checklist Application

HK: Revenue > HK$10M? Margin > 5%? Succession stated reason? 10+ years operating? No China dependency >50%? Not a shell entity? → If 5/6 pass, proceed to Owner Meeting.
TW: Revenue > TWD 50M? Margin > 5%? Succession stated reason? 10+ years operating? Manufacturing facility acceptable? No China corridor >50%? DIR approval feasible? → If 6/7 pass, proceed.

4

Owner Read (Five Archetypes)

Apply the five owner archetypes from the playbook. Is this owner the Reluctant Patriarch? The Legacy Builder? The Pragmatic Founder? Each archetype requires different messaging, timeline, and deal structure. The scenario planning here is the same discipline applied to fundamental research — map the outcomes before you invest the time.

Operational Workflow: Signal → Close
S

Signal Detected

Platform listing, broker referral, ecosystem signal, direct outreach response, event meeting. Logged in CRM with source attribution.

1

First Screen (24-48 hrs)

Apply checklist. Does it meet minimum criteria? If yes, move to research. If no, log reason and archive. Target: 80% screened out here.

2

Target Intelligence (1-2 weeks)

Company research + owner profiling. Build initial investment thesis. Identify key unknowns that can only be resolved by meeting the owner.

3

Owner Meeting

First meeting is NOT a pitch. It's a listening session. Understand the owner's story, their fears, their hopes for the business and employees. Apply the Owner Read framework. Target: 50% screened out after first meeting.

4

Investment Memo + LOI (2-4 weeks)

Full investment memo following playbook template. Valuation. Structure. LOI submission. This is where the deal gets real.

5

Due Diligence + Close (2-6 months)

Financial, legal, operational DD. Taiwan: add DIR approval, manufacturing assessment, China subsidiary review. HK: add offshore structure review, lease analysis, regulatory compliance.

The funnel math: If you want 2-3 deals per year, you need roughly: 200+ signals detected → 40 that pass first screen → 15-20 that get target intelligence → 8-10 owner meetings → 3-4 LOIs → 2-3 closes. This means your DOI system needs to generate 4+ qualified signals per week. That's why all four modules matter — no single channel produces enough volume alone.
Recommended Tech Stack
FunctionToolCost
CRM & PipelineNotion (current) or HubSpot Free CRM. Notion is fine for 0-50 active contacts. HubSpot when you scale past that.$0-50/mo
Deal AggregationKumo (primary) + manual monitoring of SMERGERS, DealStream$49-99/mo
Alert SystemGoogle Alerts (free) + Slack channel routing + Zapier for platform alerts$0-20/mo
ResearchCompanies Registry (HK), TWSE filings (TW), MOEA data, industry reportsVaries
CommunicationSlack (internal routing) + Email (broker comms) + WeChat/LINE (Asia contacts)$0
Document ManagementGoogle Drive or Notion for memos, NDAs, deal documents$0-12/mo

Total infrastructure cost: ~$70-180/month. This isn't a capital problem — it's a discipline problem. The system only works if you actually check the alerts, follow up with brokers, and log every interaction.

90-Day Build Plan
WeekActionDeliverable
1-2Set up Kumo + SMERGERS + DealStream buyer accounts. Configure saved searches and alerts. Set up Google Alerts for HK/TW succession keywords. Create Notion CRM database.Module 1 live. Alert system active.
2-4Begin Tier 1 outreach: HK (PwC, Mazars, PKF, Ivory Capital, HSBC RMs). TW (Lee and Li, PwC Taiwan, MAPECT).5-8 Tier 1 meetings scheduled.
4-6Complete Tier 1 meetings. Log all contacts in CRM. Begin Tier 2 outreach. Join MAPECT. Register with Dealsuite and Axial as buyer.Contact database at 20+ entries. First deal signals incoming.
6-8Review first pipeline. Apply First Screen Checklist to any signals. Begin target intelligence on anything that passes screen. Start building Taiwan CPA referral network (identify first 5 CPAs).First target intelligence reports. CPA referral program designed.
8-10Continue Tier 2/3 outreach. Attend first MAPECT event. Attend one HK industry event. Follow up on all stale relationships (>30 days).Contact database at 30+. Pipeline has 3-5 signals in research.
10-12Review system performance. Which channels are producing? Where to double down? First owner meeting if pipeline is mature enough. Refine alert configuration based on signal quality.DOI system v1 operational. Attribution data available. First owner meeting possible.