Gary Rieschel

1. Snapshot and testable edge thesis

Rieschel's defensible philosophy is a system for investing where markets grow faster than their information, legal and managerial institutions. Map how value is migrating; put real authority in locally trusted partners; specialize by sector; reconstruct operating truth in the field; require material opportunities to reach collective review; and let company proof and realizable liquidity—not available fund capital or paper marks—set the pace. [researcher inference · moderate] 2005 writing · 2009 diligence profile · 2016 process account

This is an institution-building thesis, not a claim that Rieschel personally selected Qiming's winners. It predicts locally embedded decision rights, sector-specific gates, consistent use on ordinary and failed deals, and better outlier-removed vintage outcomes after the SoftBank process correction. Full partnership, pipeline and cash-flow records could falsify every part. testable profile

2. Core philosophy

Map value migration before selecting a company

The 2005 statement argues that standards, manufacturing density, design tools, technical labor, market scale and risk capital can shift profit from assembly toward design, brand and distribution. [investor-stated · strong] It is a market-formation map, not a security-selection algorithm: company capture, price, governance and exit must still be proved. writing analysis

Embed authority rather than export a brand

Rieschel moved to Shanghai, chose Duane Kuang as co-founder and used an independent Chinese identity rather than a foreign franchise label. He later argued that an investor should live locally for roughly the holding period. [documented behavior · strong] Fudan trust note · Homeward note

Make proof sector-specific

His 2013 remarks distinguish Internet products that may be underwritten before revenue from healthcare companies that need discovery or efficacy evidence and cleantech companies that need customers or a credible service model. [investor-stated · strong] The public deal files are too incomplete to show consistent application. Thoughtful China note

Reconstruct ground truth

The 2009 record describes site visits, mystery rentals, ownership checks, nearly a year of market mapping and receipt-level financial reconstruction. [documented behavior · moderate] These practices reduce dependence on weak records; one selected article cannot establish their portfolio-wide use or causal effect. NYT/Qiming profile

Match capital and liquidity to evidence

SoftBank's 1999 vehicle was reportedly deployed in six months, marked at 3x and ultimately below 1x. The outcome is [retrospective witness · moderate] [single-source] [disputed]; Rieschel's pacing lesson and Qiming's later emphasis on steadier deployment and repeated review are [investor-stated · strong]. SoftBank failure lab

3. Sourcing

The visible sourcing system combines local partner networks, sector reputation, founder and portfolio referrals, university/community ties and systematic market mapping. Gan & Lee entered through a school-parent connection when the healthcare team could not obtain a meeting; that example proves access, not selection skill. Qiming's current page says portfolio referrals increasingly matter, but publishes no channel conversion or performance cohort. [investor-stated · moderate] Gan & Lee/Xiaomi lab · Qiming about

An implementable sourcing ledger should record: introducer and economic interest; local relationship owner; cold-route comparison; sector map position; why the company is visible now; conflicts; prior passes; and whether a material opportunity reached partnership review. The ByteDance account makes escalation part of sourcing quality, not only IC quality. [researcher inference · strong] ByteDance lab

4. Picking

Team

Rieschel repeatedly emphasizes trust, curiosity, generosity, directness and willingness to expose bad news. Equal senior economics and partner succession make those traits institutional inputs rather than founder charisma tests. [investor-stated · moderate] Candor and generosity are plausible signals, but no named accepted/rejected cohort validates predictive power. Fudan trust note · Hoover note

Product and technical proof

The evidence threshold should match sector physics: behavior and retention may lead revenue in Internet products; healthcare needs scientific, clinical and regulatory proof; industrial/climate systems need repeatable yield, counterparties, capex and scale-up evidence. [investor-stated · strong] sector thresholds · LanzaTech lab

Market and ecosystem

Require a reinforcing system of demand, talent, infrastructure, standards and capital—but identify the scarce complement and who controls it. A large country, category or technical wave is not a moat. [researcher inference · strong] The 2005 testimony's correct directional confidence coexisted with insufficient sensitivity to VIE enforceability, capital reversal and exit closure. writings playbook

Data, governance and truth

When accounting, ownership or customer records are unreliable, create independent evidence: observe sites, reconcile receipts, verify legal control, sample customers and identify stop authority. Walk away when ownership cannot be established. [documented behavior · moderate] Diligence intensity cannot cure a prohibited transaction, a non-transferable license or absent liquidity. 2009 profile · Treasury boundary

Price, ownership and exit

Company success, market capitalization, beneficial ownership and fund return are different variables. Underwrite security, cost, dilution, reserves, realization path and vehicle beneficiary. The reviewed public record supports no Rieschel personal IRR and no complete Qiming fund return. [researcher inference · strong] economic-contribution view

Decision card

Before approval, record:

1. the value layer moving and the company's evidence of capture;
2. the local decision owner, sector specialist and independent references;
3. proof thresholds and kill conditions by technical/commercial/regulatory risk;
4. ownership, price, reserve need and three-year exit-closure case;
5. conflicts, dissent, earlier passes and escalation path;
6. the next evidence that changes the follow-on decision; and
7. which claim is investor-stated, documented, inferred or still unavailable.

[researcher inference · strong] This card operationalizes the corpus; it is not a recovered Qiming template.

5. Deal and portfolio mechanics

Use staged checks when a proof point materially changes technical, market, regulatory or financing risk. Preserve reserves through long evidence cycles, but do not infer Qiming's historic reserve policy from LanzaTech and WeRide follow-ons. Separate entry decision, board work, follow-on, exit and later narration. [researcher inference · strong] portfolio forensics

At portfolio level, track stage, sector, geography, partner, sourcing channel, check, ownership, reserve, write-off, realized proceeds and public/unrealized value on a consistent vintage basis. Qiming's current 580+ investments, 210+ exits and 80+ unicorns are promotional live counts whose definitions and dated denominator are not public. The 33 selected Qiming lineages here are at most 5.7% research-selection coverage, never a hit-rate sample. [researcher inference · strong] denominator

Pareto analysis needs three views. Gan & Lee, Tigermed and Xiaomi lead the visible economic story; Qiming's local institution and Tigermed matter historically; SoftBank pacing and ByteDance matter most for decision learning. No public cash-flow schedule supports a literal 80/20 split. [researcher inference · strong] Pareto views

6. Vehicle and incentive fit

SoftBank Venture Capital/Mobius, SBCVC, SAIF, Ceyuan, Ignition, Qiming's China USD and RMB funds, guidance capital, U.S. healthcare vehicles and any Rieschel personal commitments are distinct beneficiaries and governance systems. Never pool their companies or economics. [researcher inference · strong] vehicle scopes

Qiming's reported equal managing-partner cash and carry should reduce deal-hoarding and internal status incentives, while weekly collective review should distribute judgment. [investor-stated · moderate] Taomee and Jasper filings confirm shared authority in named structures, not the full partnership or labor performed. Equal economics can also enable groupthink, free riding or slow decisions; ByteDance shows a pre-IC escalation breach. vehicle economics

7. Team attribution and work with founders

Company creation belongs first to founders and employees. Deal credit must be decomposed into sourced, sponsored, diligenced, approved, negotiated, board work, recruited, follow-on, exit influence and later narrator; absent evidence remains unknown. Shared IC or beneficial-ownership filings do not prove individual labor. [researcher inference · strong] people map

Public boundaries are unusually clear in several cases: Hans Tung led Qiming's Xiaomi investment; Rieschel credits Kuang's long company support; Nisa Leung originated Gan & Lee; Rieschel held documented Tigermed and LanzaTech boards. Those facts strengthen a team-franchise thesis and weaken a personal-deal legend. [documented behavior · strong] profile authority map

Rieschel calls venture a service industry. The test is observable founder benefit—recruiting, customer introductions, governance, regulatory navigation, financing or difficult truth—not board tenure or a testimonial. The public record does not estimate causal board value. [researcher inference · strong] Hoover note · board evidence boundary

8. Views on the craft

The corpus treats venture as long-duration apprenticeship, localized judgment and organizational service. Pattern recognition must be paired with curiosity because a remembered pattern can be false in a new market. Partner generosity matters because equal economics and succession depend on sharing opportunity and credit. [researcher inference · moderate] Hoover note

Macro knowledge helps define where to look, not what to buy. Rieschel's talks repeatedly separate China innovation from simplistic copying while later acknowledging that legal and liquidity regimes can overwhelm company quality. The craft therefore needs both structural patience and the humility to stop when title, price or exit cannot be underwritten. [researcher inference · strong] talks synthesis · forecast scorecard

9. Signature bets and decision process

Tigermed

The public rationale was a domestic CRO able to perform trials to global analytical standards; Rieschel served on the board and [self-reported] [single-source] that the company returned Fund I. The outcome is [retrospective witness · moderate]; entry terms, realization and causal board work are unavailable. Tigermed lab

LanzaTech

Qiming led an $18 million 2010 round around industrial-waste-gas fermentation; Rieschel joined as Qiming's nominee and remained through the public transition. [contemporaneous record · strong] Fourteen years of governance is not proof of return. LanzaTech lab

Gan & Lee and Xiaomi

These are Fund II power-law cases with divided credit. Leung originated Gan & Lee; Kuang and Tung receive Xiaomi work credit; issuer/firm records support bounded ownership and chronology. Rieschel's reported $1 billion Gan & Lee return is [self-reported] [single-source] and unaudited; Xiaomi proceeds are unknown. [researcher inference · strong] combined lab

SoftBank 1999 and ByteDance

The SoftBank loss shows capital pacing, paper-mark and reserve risk; ByteDance shows a material opportunity bypassing collective review. They are more useful for process design than any success anecdote because both expose a rule and its failure. [researcher inference · strong] SoftBank lab · ByteDance lab

10. Mental models and maxims

Six models survive the evidence threshold:

1. follow value migration, then prove company capture;
2. embed local authority for the full risk horizon;
3. match proof thresholds to sector physics;
4. reconstruct operating truth when institutions are weak;
5. match capital pace to proof and realizable liquidity; and
6. align senior economics, then force material opportunities into collective review.

Each model has provenance, applications, counterexamples, falsifiers and a dated 2026 translation in mental-models.md. “Curiosity and generosity predict investment outcomes” remains a candidate, not a promoted model, because no named outcome cohort tests it.

11. Evolution and contradictions

[researcher inference · strong] The changes show learning, but later clarity cannot be projected backward into the original decisions. writings chronology · spoken change ledger · forecast tensions

12. Failures and limits

• SoftBank/Mobius: rapid deployment, survival triage and a reported sub-1x outcome. Process lesson supported; complete fund economics and numbering disputed. [retrospective witness · moderate] [single-source] [disputed] failure lab
• ByteDance: passed twice outside partnership review. Process failure supported; rejection rationale and counterfactual value unavailable. [investor-stated · moderate] [single-source] miss lab
• China liquidity/regulatory turn: Rieschel says Qiming missed the reversal. Diagnosis is retrospective; write-down and process response are not public. [retrospective witness · moderate] Hoover note
• Inktomi and ordinary companies: contemporary reporting records the miss; Vancl, Kaixin, Touch Media, LP Amina and others prevent a winner-only narrative, but cannot be labeled write-offs without evidence. [researcher inference · moderate] adverse record
• Legal/adverse boundary: targeted U.S. searches found no proceeding naming Rieschel/Qiming. That is not clearance, and non-U.S., private or sealed matters remain outside the record. [researcher inference · insufficient public record] search boundary

Failure and adverse record shows why the philosophy must be tested on full cohorts rather than successful retrospectives.

13. Historical operating environment

SoftBank's U.S. venture operation invested through the 1990s Internet adoption and 2000 crash, when public marks could accelerate private deployment before distributions existed. Qiming formed while China's local venture, accounting, legal and reference infrastructure was immature but engineering talent, domestic demand and capital were expanding. Its 2005–12 opportunity set cannot be evaluated using today's institutions or restrictions. historical context

By the mid-2010s, competition, valuations and RMB liquidity changed entry conditions; later VIE, listing, state-policy, data and geopolitical restrictions changed exit and control. Treasury's outbound program, effective 2025-01-02 for specified China/Hong Kong/Macau semiconductor, quantum and AI transactions, is a current transfer constraint—not evidence that a 2005 deal violated a later rule. Treasury program

14. Modern VC translation (as of 2026-08-22)

What transfers

• Map the full technical, talent, customer, regulatory, capital and exit system. [researcher inference · strong] Model 1
• Put meaningful economics and authority with locally embedded specialists. [researcher inference · strong] Model 2
• Define stage gates by sector physics, not a universal revenue template. [researcher inference · strong] Model 3
• Build independent operating evidence where records or incentives are weak. [researcher inference · strong] Model 4
• Match check size and deployment pace to the next binding proof. [researcher inference · strong] Model 5
• Escalate material passes and record dissent before the outcome is known. [researcher inference · strong] Model 6

What changed

Cross-border advanced-technology investing now faces outbound controls, data/security rules, fragmented standards, uncertain listing routes and higher political tail risk. This translation treats model access, evaluation, inference cost, data rights and security as explicit AI evidence gates; a large technical market remains insufficient without legal title and realizable economics. [researcher inference · strong] Treasury current boundary · modern model translations

A 2026 implementation

Run one pre-IC memo with five ledgers: value-layer capture; local authority and conflicts; sector evidence/kill gates; security/capital/liquidity; and pass/escalation/dissent. At follow-on, compare dated predictions with actual proof. At portfolio review, publish outlier-removed realized/unrealized results and a pass/loss denominator. [researcher inference · strong] six-model implementation · denominator boundary

Misuse guards

Do not use “local context” to excuse opacity, “patient capital” to excuse missing milestones, “service” to claim founder value, “team” to erase responsibility, or a correct macro theme to ignore price and exit. Do not retroactively apply current restrictions to historic decisions. [researcher inference · strong] model failure ledgers · historical boundary

15. Comparison with completed peers

The closest comparison set is Peter Brooke for cross-border institutional translation, Jim Breyer for U.S.–China networks and vehicle attribution, and Yigal Erlich for incentive-led ecosystem design. The cells compare disclosed doctrine and boundaries, not returns or rank. [researcher inference · moderate]

Underlying controls were reopened in the peer packages: Brooke's Computer History Museum oral history, Breyer's Facebook filing and team-attributed Forbes account, and the OECD's Yozma structure. The shared lesson is that institution building, access and company outcomes do not by themselves establish personal alpha. [researcher inference · strong]

16. The strongest case against greatness

The available record may describe a senior sponsor who correctly recognized China's rise, recruited exceptional local partners and narrated a favorable selected record. The economic leaders are partner- and founder-attributed; ordinary outcomes are mostly invisible; three reported return claims are unaudited; and Qiming's full cash flows, losses, passes, ownership and partner labor are not public. SoftBank's failed fast-deployment vehicle and Qiming's ByteDance bypass also show that articulated rules were not always operating controls. [researcher inference · strong] profile skeptical case

The thesis earns respect because it predicts records that could disprove it. Until those records are available, the strongest conclusion is institution-building skill with unproved personal investment alpha—not a quantified Midas ranking. [researcher inference · strong] retrieval ledger

17. Analyst study guide, glossary, and go deeper

Recommended order

1. Read the 2005 authored statement note before later success stories.
2. Read the SoftBank and ByteDance labs before the winners.
3. Compare Tigermed, LanzaTech, Gan & Lee and Xiaomi for personal/team/vehicle boundaries.
4. Use talks.md to distinguish repeated doctrine from one-off remarks.
5. Apply the mental models to a current opportunity, including the exceptions.
6. Finish with the adversarial review and retrieval ledger.

Ranked primary sources

1. 2005 USCC written statement — best decision-date intellectual anchor because Rieschel states the China value-chain, talent, demand, capital and IP-risk thesis before Qiming's results; several estimates still lack a source appendix. [contemporaneous record · strong]
2. 2005 USCC hearing transcript, printed pp. 43–54 and 56–62 — best contemporaneous cross-examination of the thesis and product/IP rule; it is policy testimony, not an investment memo. [contemporaneous record · strong]
3. Xiaomi prospectus, “Major Pre-IPO Investors,” printed pp. 193–199 and LanzaTech 2024 proxy, printed pp. 9 and 17 — best issuer controls for bounded ownership and governance; neither filing supplies Qiming's entry cost or realized return. [contemporaneous record · strong]
4. 2025 Hoover official transcript — richest direct-voice account of formation, partner design, misses and liquidity; its deal and return memories are retrospective, selected and unaudited. [retrospective witness · moderate]
5. Taomee SEC 13G and Jasper SEC 13G — best legal-entity controls for shared voting/dispositive authority; authority is not labor, sourcing or personal economics. [contemporaneous record · strong]

Ranked secondary and edited sources

1. 2016 AVCJ/Qiming profile — highest-density formation, IC, incentive, pacing and failure account; direct answers are valuable, but the SoftBank outcome remains single-source and vehicle numbering disputed. [retrospective witness · moderate]
2. Forbes 2000 profile — strongest near-contemporaneous adverse control on SoftBank deployment, survival triage and Rieschel's operating style; it predates the final fund outcome. [contemporaneous record · moderate]
3. 2009 New York Times profile mirrored by Qiming — best field-diligence and partner-credit account from Qiming's early operating period; selected cases cannot prove portfolio-wide use or returns. [documented behavior · moderate]
4. 2016 AVCJ succession report, PDF pp. 10–13 — best edited account of shared economics, internal promotion and Rieschel's transition; legal agreements and exact carry allocations remain private. [retrospective witness · moderate]
5. Qiming/China Entrepreneur on Gan & Lee — strongest partner-attributed chronology for Nisa Leung's work and the long listing path; firm-hosted retrospective evidence does not prove cash-flow economics. [retrospective witness · moderate]

Glossary

• Decision-date freeze: only evidence available when the decision occurred.
• Documented behavior: an action supported by contemporaneous or primary records.
• Investor-stated: the investor's claim, not independently verified fact.
• Researcher inference: a synthesis that must expose assumptions and falsifiers.
• Vehicle: the legal/economic beneficiary whose security and cash flows matter.
• Paper mark: an unrealized valuation, not a distribution.
• DPI / TVPI / IRR: distributions, total value and time-weighted fund metrics; none is publicly complete here.
• Selection coverage: the share of a public company denominator researched; not a hit rate or economic coverage measure.
• VIE: a contractual control structure whose enforceability and regulatory treatment can be material.

Reusable questions

• Where is value moving, and what proves this company captures it?
• Who is locally accountable, economically aligned and independently referenced?
• What sector-specific evidence unlocks the next check, and what kills the deal?
• Which entity owns the security, who did the work, and who receives the economics?
• What survives a three-year exit closure and an outlier-removed portfolio test?
• Which material opportunities never reached the partnership, and why?
• What private record would most likely falsify the story being told?

Pareto 80/20 — what drove the record

The public record supports a power-law interpretation, but not a literal 80/20 calculation. Qiming says it has backed more than 580 companies, yet it does not publish paid-in capital, write-offs, dilution, distributions, fees, carry, or company-level cash flows. Rieschel's 2025 retrospective [self-reported] identifies Tigermed as sufficient to return Fund I and [self-reported] says Gan & Lee returned $1 billion to Qiming, while crediting Duane Kuang's years of work with Xiaomi; those are unusually useful attribution signals, but they remain unaudited, gross/net basis and timing unspecified. Hoover, transcript, answer beginning “The deals are the deals” Pareto judgment: [investor-stated · moderate] for directional contribution; [insufficient public record] for a numeric split.

Economic-contribution view

Dominant-outlier test. Fund II: removing Gan & Lee's [self-reported] $1 billion leaves Xiaomi as a prospectus-documented ownership position, but no cash-flow schedule establishes Xiaomi's realized contribution or whether the remaining portfolio diversified the vehicle economically. Removing both Gan & Lee and Xiaomi leaves no public cash-flow schedule from which to rank Fund II's remainder. Fund I: Tigermed's separate [self-reported] whole-fund claim cannot be used to test Fund II; removing Tigermed leaves no reconstructable Fund I performance. Evidence label: [researcher inference · moderate] that a few outliers drove visible value; [insufficient public record] for concentration ratios, personal hit rate, DPI, TVPI, IRR or loss ratio.

Historical-importance view

Decision-learning-value view

Evidence boundary and searched denominator

Scope, entities and inclusion rules

This review searched the period 1995–2026 and kept five economically distinct scopes separate:

1. SoftBank Venture Capital and its Mobius Venture Capital successor, including the 1995–2000 U.S. funds and the seven companies named in SoftBank's 2002 official account;
2. SoftBank-sponsored China vehicles—SoftBank China Venture Capital, SoftBank Asia Infrastructure Fund/SAIF and Ceyuan—as institution-building evidence, not as Rieschel-personal company returns;
3. Qiming's China USD/RMB franchise from Fund I through current vehicles, for which the official live claim is 580+ companies, 210+ IPO/M&A/other exits and 80+ unicorns;
4. Qiming U.S. Healthcare Fund II and public filings that name Rieschel as a shared control person; and
5. personal fund commitments to Ceyuan and Ignition, which are not Qiming assets and for which no public cash flows were located.

The company ledger includes every name in a deduplicated 43-lineage material-evidence union: eight named SoftBank/Mobius lineages (the seven in SoftBank's official account plus Buy.com in contemporaneous reporting), one SAIF outcome named by Rieschel (Shanda), Qiming's 18-company 2011 milestone roster, thirteen additional flagships named in its 2022/current profile, WeRide, Jasper Therapeutics and the documented ByteDance miss. ByteDance is kept in the anti-portfolio rather than the 42-lineage invested denominator. Follow-ons, aliases and mergers count once: GeoCities/Yahoo, Concentric/XO, Dianping/Meituan, and eHi's financing-to-privatization lineage are each single lineages. The press claim that Rieschel bought stakes “like Yahoo” is folded into the GeoCities/Yahoo lineage because the reviewed record cannot separate a Mobius venture investment in Yahoo from SoftBank parent's strategic Yahoo stake.

Search variants included Gary Rieschel, Gary E. Rieschel, Gary Edward Rieschel, SoftBank Venture Capital, SOFTBANK Technology Ventures, SOFTBANK Capital Partners, Mobius Venture Capital, SB China Holdings, SoftBank China Venture Capital, SBCVC, SoftBank Asia Infrastructure Fund, SAIF Partners, Ceyuan, Qiming, Qiming Venture Partners, Qiming Ventures, Qiming Weichuang, and 启明创投. Routes included Qiming's team/about/portfolio/news archives, SoftBank annual reports, SEC EDGAR, HKEX/SZSE issuer materials, U.S. Treasury rules, contemporaneous Forbes/NYT reporting, AVCJ profiles and the Hoover transcript. Commercial databases were discovery aids only.

Denominator reconciliation

Researched means the named lineage was checked for chronology, vehicle, team or outcome evidence. Sufficiently reconstructed means a primary/contemporaneous record supports at least the decision/outcome boundary; it never means complete economics.

Coverage fraction: not computable. The public 580+ Qiming denominator is a live cumulative firm claim, not a dated deal schedule; the official portfolio page omits entry terms and may mix active, exited and follow-on positions. The 33-company Qiming invested subset is 33 / 580+ = at most 5.7% by company count, but that is a research-selection ratio, not economic coverage. The eight named SoftBank lineages cannot be reconciled to Rieschel's unspecified twelve. Survivorship bias is severe because official biographies and milestone releases foreground IPOs, M&A and unicorns while write-offs and ordinary deals remain unobserved.

Historical market, regulatory and base-rate context

U.S. internet venture, 1995–2001

When Rieschel and Son formed the U.S. venture operation, Yahoo and Excite were young, Google did not yet exist and Facebook was years away. The investable thesis was mass internet adoption, but capital velocity soon outran underwriting: Forbes described 1,500 monthly inbound plans, a 99% rejection rate, and public-market damage that repriced two contemporaneous private rounds from requested $40 million valuations to $15 million. AVCJ/Qiming, opening paragraphs; Forbes, paras 29–38 These are contemporary opportunity-set facts, not a base rate for venture success.

The crash changed the relevant decision from “capture market share” to cash survival. Rieschel's team used weekly red/yellow/green reports with revenue, burn, cash and runway; the contemporaneous account quotes imminent layoffs, sales or closures. Forbes, paras 29–32 A [single-source] later profile says the 1999 vintage went from a 3x paper mark to below 1x; the vehicle identity is [disputed] against contemporary numbering. AVCJ/Qiming, paras 20–24 Causal lesson: [investor-stated · strong] that pacing and marked value were unsafe; no public vintage-matched benchmark permits a claim that the loss was better or worse than peers.

China venture formation, 1999–2012

Rieschel's 2025 account says early China VC lacked the accounting, legal, reference and market-data infrastructure familiar in Silicon Valley. Consumer internet investments frequently used variable-interest-entity structures because foreign ownership of internet licenses was restricted; he describes their long-term legal viability as uncertain and says four government bodies issued conflicting fund guidance. Hoover transcript, historical answer on accounting, VIEs and ministries Evidence label: [investor-stated · strong] for his contemporary underwriting problem; [investor-stated · moderate] for market-wide generalization.

By 2009/2011, contemporaneous reporting described unreliable financials and limited market data, while Qiming answered with site visits, cash-receipt reconstruction and niche-leadership filters. The same report cited 138 venture-backed Chinese IPOs raising an estimated $21 billion in the prior year, a reminder that the exit window and underwriting weakness coexisted. NYT/Qiming, paras 24–31 Qiming Fund III's 2011 release reported more than 50 investments, half early and half expansion stage, with $0.5 million–$30 million company commitments. Qiming Fund III announcement, paras 16–18 No defensible contemporaneous loss or return base rate for comparable China funds was located; one is not manufactured here.

Liquidity and regulatory regime shift

Rieschel says Qiming and peers missed the tightening that followed China's 2015–2016 market break; he describes 2018–2022/23 listing approvals and weak liquidity as trapping both foreign and RMB capital. Hoover transcript, market-regulation answer This is a 2025 retrospective, not contemporaneous proof of each regulator's intent. Evidence label: [investor-stated · moderate] for the portfolio-level diagnosis; [insufficient public record] for the affected Qiming NAV or distributions.

The modern U.S. comparison is not dimensionally identical to Qiming's early China period. Treasury's outbound-investment rules, effective 2 January 2025, prohibit or require notification for certain U.S.-person transactions involving China/Hong Kong/Macau entities in semiconductors, quantum technologies and AI. U.S. Treasury, “Outbound Investment Security Program,” program regulations Modern comparison: [researcher inference · strong] that a U.S.-linked China fund today faces a formal national-security compliance layer absent from the 2005 Qiming formation story; this says nothing about whether a specific Qiming transaction is covered.

Vehicle economics and incentives

People, capital and influence map

Decision-rights rule. Unless a filing or direct account says otherwise, sourced, sponsored, diligenced, approved, negotiated, board work, recruited, follow-on, exit influence, and later narrator remain unknown. Shared GP/IC power is not sole authorship; later narration is not ex-ante evidence.

Comprehensive scoped deal ledger

Legacy SoftBank/Venture Capital/Mobius company ledger

Sponsored China fund outcome

Qiming 2011 milestone-roster ledger

The 18 rows below reconcile every company named in Qiming's Fund III closing announcement. Unless stronger evidence is cited, the announcement establishes only firm portfolio membership as of May 2011; entry date, vehicle, security, ownership, dilution, board role, follow-ons, exit proceeds and Rieschel-personal attribution are not publicly established. Qiming Fund III announcement, company roster

Later flagship and U.S.-healthcare ledger

Ex-ante forensic cases and decision labs

Case 1: Tigermed, 2007–2012 — a foundational Fund I deal

Decision-date freeze. Rieschel dates Qiming's investment to 2007/08 and says about 95% of Tigermed's trials then served generic drugs in China; the firm saw a need for a domestic CRO able to perform clinical trials to global analytical standards. Hoover transcript, Tigermed answers The public record does not disclose the exact date, check, security, valuation, ownership, competing bids, IC dissent, milestone schedule, reserves or kill criteria. The 2012 IPO and later trial-mix reversal were not knowable at entry. Rationale label: [investor-stated · moderate]; it is retrospective, not an investment memo.

What could be known. China lacked a comparable local CRO platform in Rieschel's account; clinical quality, regulator acceptance, customer trust, trained staff and cash runway were obvious underwriting questions. Those risk categories are analyst reconstruction. Qiming's surviving 2008 Fund II announcement confirms healthcare focus and an early/selected-growth mandate, but not Tigermed's specific terms. Qiming Fund II announcement, paras 17–27

Reader exercise — answer before reading the governance and outcome record:

• Would you invest then? Decide whether the absence of a global-standard domestic CRO is validated demand or an adoption/regulatory trap.
• What would kill the deal? Predefine trial-quality failures, sponsor churn, regulator rejection, inability to recruit clinical staff, unreliable site data and runway thresholds.
• What ownership/terms fit the risk? Specify staged financing, information/board rights, compliance covenants and reserves without using the later IPO value.
• What evidence unlocks the next check? Require repeat multinational sponsors, audited trial-quality metrics, regulator acceptance, margin progression and staff retention.

Governance and milestones. Rieschel served on the board and later described a two-hour CSRC examination covering business, gross margin and risks before the listing. Qiming and Ignition were the only institutional investors identified in Qiming's IPO release. Hoover transcript, board-exam answer; Qiming IPO release Recruiting, follow-on authorization and exit influence remain unknown.

Outcome revealed. Tigermed issued 13.4 million shares at RMB37.88 in August 2012, raising RMB508 million; its first-day close implied about RMB2.7 billion market capitalization. Qiming calls it its first domestic IPO. Qiming IPO release Rieschel later [self-reported] that the company returned the entire Fund I and said that within five years its trial mix moved from 95% generics to 95% novel products. Hoover transcript These are reported results, not audited cash flows or proof Qiming caused the industry shift.

Actual decision and result: Qiming invested, Rieschel served on the board and the company listed in 2012; the [self-reported] whole-fund return is not economically reconstructed.

Case 2: LanzaTech, 2010–present — governance through a long commercialization cycle

Decision-date freeze. In July 2010 Qiming led an $18 million round to commercialize gas-fermentation technology; contemporaneous reporting said LanzaTech had operated a New Zealand pilot since 2008, had agreements involving Baosteel and the Chinese Academy of Sciences, targeted a pre-commercial plant in 2011, and aimed eventually at a 200-million-liter commercial facility. NBR, 13 Jul 2010 The record does not provide Qiming's check, valuation, security, ownership, technical-diligence file, dissent, downside cases or tranche milestones.

Ex-ante thesis and risks. Rieschel publicly framed the opportunity as producing fuel from industrial waste gas without food/land competition while reducing emissions. NBR, Rieschel statement That is a contemporaneous thesis. Scale-up yield, feedstock contracts, capex, permitting, commodity prices, project finance, partner concentration and time-to-plant are analyst-created risk prompts. Rationale label: [investor-stated · strong]; mechanism support: [contemporaneous record · moderate] because commercial scale was not yet proven.

Reader exercise — answer before reading the governance and outcome record:

• Would you invest then? Underwrite 2010 pilot evidence and counterparties, not today's climate narrative or listing.
• What would kill the deal? Set biological-yield, feedstock, capex, permitting, offtake, partner and runway thresholds.
• What ownership/terms fit the risk? Choose milestone tranches, board/information rights, project-finance protections and follow-on reserves.
• What evidence unlocks the next check? Require stable pilot runs, independently verified economics, binding industrial/offtake contracts and a financed demonstration plant.

Governance and follow-ons. Rieschel joined the board in 2010 as Qiming's nominee and remained through the 2023 business combination; in 2024 he chaired compensation. LanzaTech 2024 proxy, printed pp.9 and 17 / PDF pp.17 and 25 A 2014 $60 million round added Mitsui and other investors; the announcement records continued Qiming participation only indirectly and does not disclose fund ownership. GlobeNewswire, 26 Mar 2014

Outcome revealed. LanzaTech became public through a business combination in 2023 and Qiming's nominated director remained. A listing does not establish a Qiming realization, multiple or profit. The fourteen-year board duration is strong governance evidence and weak return evidence. Lesson label: [documented behavior · strong] for patient governance; [researcher inference · moderate] for the need to reserve/syndicate long-cycle technical risk.

Actual decision and result: Qiming led the round and appointed Rieschel; a public transition followed after a long syndication/commercialization cycle, but the venture return remains unknown.

Case 3: Gan & Lee and Xiaomi — the Fund II power-law pair with divided credit

Decision-date freeze. Qiming invested in Gan & Lee in 2010, before the STAR Market and while the firm expected a long wait for a profitable-company main-board listing. Leung describes Gan & Lee as her first Qiming biopharma deal. Qiming/China Entrepreneur, paras 20–29 For Xiaomi, the reviewed evidence establishes Qiming II ownership and early-investor status, but not the entry memo, price, check, terms or dissent. Xiaomi prospectus, printed pp.193–199 / PDF pp.7–13; Qiming/AFR profile, opening

Alternatives and attribution. Rieschel says a school-parent connection opened the Gan & Lee founder meeting after Qiming's healthcare partner could not obtain one; this is network access, not proof of selection skill by Rieschel. He credits Lei Jun for Xiaomi's execution and Duane Kuang for advising the company over years. Hoover transcript, Fund II answer Attribution label: [investor-stated · strong] that Leung/team deserve Gan & Lee credit and Kuang/team deserve Xiaomi credit; [insufficient public record] for each IC vote and negotiation.

Reader exercise — answer before reading the outcome record:

• Would you invest then? For Gan & Lee, test product quality, clinical/regulatory path, reimbursement and listing patience; for Xiaomi, test founder/product/brand/channel economics available at entry.
• What would kill the deal? Define clinical/product failures, governance opacity, unit-economics weakness and inability to finance through a long exit wait.
• What ownership/terms fit the risk? Set rights and reserves separately for regulated biopharma and consumer hardware; do not use later market caps as entry inputs.
• What evidence unlocks the next check? Require regulatory/product milestones and repeat demand, plus a credible local listing and liquidity path.

Outcome revealed. Gan & Lee listed in June 2020 after two prior IPO attempts; Qiming's retrospective calls it the largest pre-IPO institutional investor and Rieschel [self-reported] $1 billion returned to Qiming. Qiming/China Entrepreneur, paras 20–29; Hoover transcript Xiaomi's 2018 prospectus shows the three Qiming II entities at about 3.98% pre-offering and about 3.72% after the offering, but supplies no realized proceeds. Xiaomi prospectus, shareholding table and “Major Pre-IPO Investors,” PDF pp.7–13

Actual decision and result: the shared Qiming II vehicle held both outliers; available evidence assigns the active partner credit to Leung and Kuang, not solely to Rieschel.

Case 4: SoftBank/Mobius 1999 vintage — the full failure lab

Decision-date freeze. The reviewed record has no contemporaneous 1999 IC memo, price schedule, reserve plan or full opportunity set. Internet adoption was the stated theme, but the failed fund's complete portfolio, underwriting memos, ownership, write-offs and LP cash flows remain unavailable. The exercise therefore tests whether a sound secular theme can justify six-month deployment without importing the later paper mark or crash.

Reader exercise — answer before reading the deployment and outcome record:

• Would you commit or deploy then? Separate belief in the internet from price, pacing and reserves.
• What would kill the deal/fund pace? Set valuation, cohort-loss, burn/runway, follow-on-capacity and public-window triggers.
• What ownership/terms fit the risk? Preserve downside rights and reserves; do not let a later paper mark justify accelerating uncalled capital.
• What evidence unlocks the next check? Require operating milestones, cash-duration stress tests and realized liquidity—not comparable-company marks.

Original decision and warning signs revealed. A [single-source] 2016 AVCJ profile says the 1999 vintage was fully deployed within six months, marked at 3x a year later and ultimately failed to return 1x; which contemporaneously named vehicle it describes is [disputed]. Rieschel acknowledged that permitting deployment so quickly was an error. AVCJ/Qiming, paras 20–24 Forbes reported that by July 2000 Rieschel's unit had deployed at record speed into wireless, supply-chain, internet telephony, graphics and financial services; contemporary warning signs included an 80% SoftBank stock fall, survival triage, layoffs, cash deadlines and collapsing private valuations. Forbes, paras 11–38

Capital and governance response. The team circulated weekly red/yellow/green cash reports, chose which of more than 100 companies to stop supporting and laid off partners. NYT/Qiming, paras 37–41 The reviewed record does not show company-specific reserve decisions, board actions, sale/closure dates or whether outside LPs approved pacing.

Outcome and process change. A [single-source] retrospective says the fund failed to return capital; Mobius later stopped new investing. At Qiming, Rieschel adopted steadier deployment, at least two formal discussions per deal and equal senior economics, explicitly linking caution to the SoftBank experience. AVCJ/Qiming, paras 23–24 and 35–47 Lesson label: [investor-stated · strong] for pacing/process change; [researcher inference · moderate] that equal compensation reduces but cannot eliminate rush-to-deploy incentives.

Numbering conflict. AVCJ calls the failed 1999 vehicle “Fund IV”; Forbes in 2000 calls the July 1999 $660 million vehicle Fund V, while SoftBank's 2002 table lists Technology Ventures V at $606 million and Capital Partners at $1.45 billion. Forbes, para 15; SoftBank, printed p.28 / PDF p.30 The conflict may reflect parallel/twin vehicles or journalistic numbering; this review does not silently resolve it.

Actual decision and result: a [single-source] retrospective says capital was deployed in six months, the mark rose to 3x and the fund then returned less than 1x; the vehicle identity remains [disputed]. Company-level economics remain unavailable.

Case 5: ByteDance — the twice-passed anti-portfolio lab

Decision-date freeze. No contemporaneous ByteDance pitch or rejection file is public. The available account provides no dates, product stage, valuation, check, proposed ownership, competitive map, IC memo, rejection rationale or evidence other partners would have invested. Later scale must not be inserted into the original information set; without the original packet, the investment question must remain unresolved.

Reader exercise — answer before reading the pass and later outcome:

• Would you invest then? Write the decision using only a retrieved contemporary pitch; if it cannot be retrieved, state insufficient public record.
• What would kill the deal? Predefine retention, content-supply, moderation/regulatory, distribution, founder and financing risks.
• What ownership/terms fit the risk? Choose an option-sized first check and follow-on milestones without using the later ByteDance outcome.
• What evidence unlocks the next check? Require cohort retention, creator supply, engagement quality, cost-efficient distribution and regulatory resilience.

Pass and process failure revealed. Rieschel said one Qiming partner passed on ByteDance twice without showing it to partners. Hoover transcript, hiring/process answer For a team-oriented system, the documented defect was not merely a wrong forecast; it was that a potentially material opportunity never reached collective review. Rieschel said the partner left the firm, framing two undisclosed passes as carrying a cost. It is unknown whether the departure was immediate, solely caused by ByteDance, voluntary or a formal policy change. Attribution label: [investor-stated · strong] for pass/escalation failure; [insufficient public record] for causality of departure and foregone return.

Counterfactual boundary. Opportunity cost cannot be calculated. Qiming might have declined in IC, received too little allocation, failed to follow on or sold early. A famous missed company is evidence about review architecture, not proof that the rejecting investor destroyed the company's later valuation in fund profits.

Actual decision and result: Qiming passed twice before partnership review; the company later became an obvious anti-portfolio outlier, but the fund's counterfactual economics are unknowable.

Portfolio-level forensics

The public record does not support a Rieschel-personal hit rate or personal IRR. It supports three narrower claims: he helped design and govern multiple vehicles; he held documented boards in Tigermed/LanzaTech and other early Qiming companies; and he publicly assigns many company-specific wins to partners and founders.

Failure, miss, criticism and adverse record

SoftBank/Mobius rapid deployment and crash

The original thesis was broad internet adoption; the warning signs were falling public/private valuations, rising burn/runway stress and a six-month investment pace. The team shifted to weekly survival reporting, cut companies and partners, and a [single-source] retrospective says the 1999 vintage finished below 1x. The documented process response at Qiming was slower deployment, repeated IC discussion and equal senior economics. Forbes, paras 29–38; AVCJ/Qiming, paras 20–24 and 35–47 Attribution remains shared with Son/Fisher/partners and the market regime; fund numbering is [disputed], and cash-flow details remain unavailable.

ByteDance and the missing rejection denominator

The twice-passed ByteDance opportunity is a full process miss, and Rieschel says the responsible partner left. No other named Qiming pass was found, so one spectacular anecdote cannot establish the false-negative rate or prove the corrected process catches later outliers. Hoover transcript

China liquidity/regulatory miss

Rieschel says Qiming and the wider VC market missed the regulatory/liquidity reversal after the 2015–2016 break. Capital response, write-downs, reserve changes and specific portfolio damage are not public. Hoover transcript, answer on CSRC/liquidity Postmortem status: [investor-stated · moderate] for the diagnosis; [insufficient public record] for financial consequences and subsequent process change.

Adverse, litigation and regulatory search boundary

Targeted searches across SEC/EDGAR, DOJ, FTC, OFAC, congressional/public-report indexes and general web/court-result queries found no enforcement matter directly naming Rieschel or Qiming in the reviewed routes. This is not a clearance opinion: Chinese dockets, private arbitration, LP disputes and sealed matters are not comprehensively searchable. A U.S. government-hosted 2022 security study mentions a reported Qiming public-markets vehicle only in a generalized discussion of Chinese VC/PE influence and IP risk; it is not a finding of misconduct by Qiming or Rieschel. Government-hosted study, printed p.52 n.64

The investee record contains ordinary company risk and competitive loss—Kaixin's social-network thesis, Vancl's unresolved outcome, LanzaTech's long commercialization cycle—but the searched evidence does not permit labeling any of those a Qiming write-off. Absence of public loss data is a denominator defect, not evidence of a clean record.

Uncertainty and retrieval ledger

The defensible conclusion is deliberately narrower than the legend: Rieschel was a consequential vehicle builder, governor and board investor across U.S. internet and China venture formation. The visible Qiming economics appear power-law driven, but the best public evidence allocates company credit across Nisa Leung, Duane Kuang, local teams and founders; it does not support a Gary Rieschel personal return series.

Coverage and evidence base

The accessible authored corpus contains one qualifying priority work, and this page does not inflate it with reporter-written profiles or interview transcripts. The sole item is Rieschel's seven-page prepared statement for the U.S.-China Economic and Security Review Commission on 2005-04-21; it was accessible in full and is deeply analyzed below and in its exact-schema note. Prepared statement, PDF pp. 1–7

Named routes searched were Qiming's English/Chinese Insights and Newsroom, LinkedIn Pulse, Medium, Forbes author pages, PERC research, 25iq, Google Books/Scholar patterns, Wayback discovery, public shareholder/LP communications, SEC/USCC testimony, library and oral-history catalogs, and personal-domain/name variants through 2026-08-22. No Rieschel-authored book, recurring blog, annual letter, shareholder letter, public LP letter or essay series was found. The five accessible edited texts plus one inaccessible EY panel text remain valuable direct-voice evidence but are not presented as writings; the full source map records their authorship and access boundaries.

Source-by-source analysis

2005 — China's Global Technology Competitiveness

Authorship, context and access. Rieschel submitted the seven-page statement as executive managing director of Mobius Venture Capital to a public commission hearing on China's high-technology development. It is a near-decision-date source: he had already helped establish multiple China venture institutions and would form Qiming that year, but none of Qiming's later signature results was available. Full official text and a full hearing transcript survive, allowing the prepared argument to be separated from oral delivery and later success narration. Prepared statement, title/PDF p. 1 · official transcript, printed pp. 43–54/PDF pp. 55–66

Central argument. [investor-stated · strong] China was not merely the next low-cost production base. Its manufacturing density, technical education, huge consumer market, standards access, product-cycle compression, design tools, overseas returnees and growing risk capital formed a system capable of moving from assembly to design, then brand and global innovation. Rieschel's conclusion was categorical: these elements would make China a major innovative force. Prepared statement, PDF pp. 2–7

Reasoning chain. The paper proceeds layer by layer:

1. Proprietary, vertically integrated systems had favored scale, but standards-based markets increasingly rewarded nimble firms able to innovate around a common core. Prepared statement, PDF pp. 1–2
2. Outsourced manufacturing created supplier capability; economics then pulled suppliers toward higher-margin design, while faster product cycles pushed brands to rely more heavily on that supply network. Prepared statement, “China's Competitiveness in Outsourcing,” PDF pp. 2–3
3. Design capability plus a large home market let Chinese companies move toward their own brands and channels. Quanta, Flextronics, Haier, Huawei and Lenovo illustrate different steps in the proposed migration. Prepared statement, PDF pp. 3–4
4. Universities, foreign R&D facilities and returnees enlarged the technical and managerial pool, while English study and global standards reduced interface costs. Prepared statement, PDF pp. 4–5
5. Venture capital, equity incentives, market liquidity and reform supplied the final catalyst, directing talent and capital toward risky new companies. Prepared statement, “The Final Catalyst,” PDF pp. 5–6

Evidence and examples. [investor-stated · moderate] The examples are concrete enough to make the mechanism falsifiable: reported manufacturing-versus-design margins; Quanta's designer count; Flextronics' design/manufacturing role; Haier's brand ascent; Huawei's domestic base; Lenovo's IBM PC acquisition; foreign R&D hiring; engineering graduates; and China-focused capital formation. But the paper gives no references, data definitions or adverse company sample. These are evidence of Rieschel's 2005 information set, not independently verified measurements. Prepared statement, PDF pp. 3–6

The hearing supplies a useful contemporary disagreement rather than hindsight. Accel partner Carl Everett said China had generated little advanced R&D and was then strongest at cost-down and volume production. [contemporaneous record · strong] [researcher inference · moderate] The two accounts could coexist under different definitions: Rieschel described a system likely to become innovative, while Everett asked where original advanced breakthroughs already existed; a rigorous 2005 IC would have separated those propositions. Official transcript, Everett statement and Rieschel panel discussion, printed pp. 43–54/PDF pp. 55–66

Assumptions. The thesis depends on at least seven contestable premises: supplier competence travels upward into differentiated design; standards remain open enough to access; Chinese consumer scale rewards domestic leaders rather than forcing price competition; financial/accounting reform continues; IP enforcement strengthens as domestic owners gain assets; foreign and domestic capital remain available; and liquidity venues remain credible. [researcher inference · strong] The paper acknowledges several of these risks but turns the conclusion into a “guarantee,” leaving no sensitivity case for reform reversal, VIE ambiguity, capital controls, geopolitical bifurcation or exit closure. Prepared statement, PDF pp. 6–7

Decision rules and operational implications. Three supportable rules emerge, with strict attribution boundaries:

• [investor-stated · strong] Underwrite the direction of value migration, not the current low-cost layer. Ask whether the company is gaining design, software, brand or channel power as its ecosystem standardizes. Prepared statement, PDF pp. 2–4
• [investor-stated · moderate] Treat market scale, talent, infrastructure and risk capital as complements. Capital without technical depth and a home-market learning loop is not the same opportunity. Prepared statement, innovation recap and risk-capital section, PDF pp. 4–6
• [researcher inference · moderate] Convert every macro thesis into company-level proof: evidence of proprietary design, time-to-market advantage, customer willingness to pay, margin progression, talent retention, governance enforceability and an executable exit. The prepared statement itself does not provide this checklist. Prepared statement, PDF pp. 2–7

Relationship to actual decisions. [researcher inference · moderate] Qiming's later sector organization and early-stage China focus are consistent with the ecosystem thesis, but sequence is not causation. Duane Kuang, Nisa Leung, JP Gan, William Hu, Hans Tung and other partners supplied local sourcing, sector judgment and company work; no source shows this paper was Qiming's IC blueprint. The clearest contrary attribution is Xiaomi: Hans Tung led the investment, and Rieschel later credited Kuang's long advisory work. AVCJ profile, sections on sector teams and shared IC · Forbes on Hans Tung, opening Q&A · Hoover transcript, answer beginning “fund two wound up being quite special”

Biographical and intellectual-development evidence. [researcher inference · strong] The Japan comparison shows that Rieschel carried an operator's value-chain framework into investing: identify how standards, architecture, cycle time and distribution change where profit pools reside. His testimony dates this perspective before Qiming and connects it to work across Japan, China, India, Europe and the United States. It does not establish that biology training drove the method or that he alone identified China. Prepared statement, PDF pp. 1–2

Evolution and contradictions. Later Qiming-era remarks preserve the long-horizon, local-team and ecosystem emphasis while moderating the capital thesis. In 2016 Rieschel pointed to SoftBank's fast-deployed 1999 fund—reported 3x on paper but below 1x after the crash—as the reason Qiming should be more conservative. [investor-stated · moderate] [single-source] [disputed] Capital changes from catalyst to hazard when price, pace and exit capacity detach from company progress. AVCJ profile, opening section

The legal thesis also became more explicit. In 2025 he described VIE ambiguity, conflicting regulators and unreliable early accounting/reference infrastructure, and argued that liquidity—not geopolitics alone—now constrained venture. [investor-stated · moderate] This adds texture missing from the 2005 reform trajectory, but it is retrospective and should not be projected backward as though the original paper fully priced those risks. Hoover transcript, answers beginning “some of the challenges” and “the single biggest issue is the liquidity side”

Cross-source synthesis

There is only one authored writing, so no honest cross-writing recurrence test is possible. A cross-record comparison can distinguish what remained stable from what later voice added:

Intellectual chronology, evolution, and contradictions

• 1989–93 operator comparison: Rieschel observed Japan's vertically integrated technology model and later used it as the contrast for a standards-based, faster-cycle world. [investor-stated · strong] Statement, PDF pp. 1–2
• 1995–2000 scale experiment: SoftBank pursued Internet concentration and exceptionally rapid capital deployment. Contemporary coverage records Rieschel's excitement and emotional involvement as vulnerabilities. [contemporaneous record · moderate] Forbes, profile entry “Gary Rieschel, Softbank Venture Capital”
• 2005 opportunity system: the writing articulates a broad China flywheel—manufacturing, design, talent, consumer scale, standards, capital and reform. It is a thesis of market formation, not a deal checklist. [investor-stated · strong] Statement, PDF pp. 2–7
• 2006–16 organizational response: Qiming chose a local brand, sector teams, equal partner incentives, formal shared IC and steady deployment. [documented behavior · moderate] The public record shows consistency with 2005 but not a causal memo-to-design chain. AVCJ profile, formation/team sections
• 2016–25 risk correction: capital velocity, exit management, VIE ambiguity, liquidity and geopolitics occupy more of the framework. The stable core is local capability and ecosystem depth; the revised boundary is that capital abundance and nominal reform do not guarantee realizable returns. [researcher inference · strong] AVCJ profile, closing warning · Hoover transcript, answer beginning “the single biggest issue is the liquidity side”

The unresolved contradiction is confidence calibration. “It will not fail in China” and “guarantees” are stronger than the paper's own premises permit. The later public record shows real category wins and institution building, but also fast-deployment loss, ByteDance passes, Vancl restructuring, legal ambiguity and liquidity traps. Those adverse cases do not refute China's innovative scale; they refute treating correct macro direction as sufficient investment selection. [researcher inference · strong] Prepared statement, risk-capital and conclusion sections, PDF pp. 5–7 · AVCJ profile, opening SoftBank-fund postmortem · AVCJ Fund IV/Vancl report, PDF pp. 8–9 · Hoover transcript, answers on ByteDance and liquidity · U.S. Treasury, “Outbound Investment Security Program”

Decision playbook and modern translation

Rule 1 — Follow the value layer, then prove capture

• Historical rule: [investor-stated · strong] Seek firms moving from low-margin execution into design, software, standards influence, brand or distribution as product cycles shorten. Statement, PDF pp. 2–4
• IC diagnostics: [researcher inference · moderate] Which layer captures incremental gross profit? What proprietary capability survives supplier substitution? Is time-to-market improving? Does the customer pay for differentiation? Does the company own the brand/channel or remain interchangeable? Prepared statement, PDF pp. 2–4
• Kill criteria: [researcher inference · moderate] No measurable margin/value migration; dependence on one buyer or subsidy; standardization erases differentiation; working-capital needs grow faster than pricing power. Prepared statement, PDF pp. 2–4
• Failure mode: [researcher inference · moderate] Mistaking ecosystem growth for company-specific capture. Prepared statement, PDF pp. 2–4

Rule 2 — Underwrite complements, not isolated metrics

• Historical rule: [investor-stated · strong] Market, talent, infrastructure, standards and risk capital must reinforce one another. Statement, innovation recap, PDF pp. 4–6
• IC diagnostics: [researcher inference · moderate] Which complement is genuinely scarce? Who controls it? Can the company survive if foreign capital, cross-border data, a listing venue or a government-created demand channel disappears? Prepared statement, PDF pp. 4–7
• Kill criteria: [researcher inference · moderate] The thesis requires every complement to improve simultaneously; legal entity or license cannot be diligenced; exit depends on a single policy window. Prepared statement, PDF pp. 4–7
• Failure mode: [researcher inference · moderate] Using a country-level growth statistic as a company moat. Prepared statement, PDF pp. 4–7

Rule 3 — Match capital pace to proof pace

• Historical rule: this is an evolution, not stated in the 2005 paper. [researcher inference · strong] Risk capital is catalytic only while deployment and price remain linked to operating evidence; the SoftBank postmortem shows paper marks can vanish before distributions. Statement, PDF pp. 5–6 · AVCJ profile, opening SoftBank-fund postmortem
• IC diagnostics: [researcher inference · moderate] What milestone reduces the next uncertainty? How much capital and time reach it? Which valuation assumption requires an open IPO market? What does the position look like if the exit window closes for three years? AVCJ profile, sections on the failed SoftBank vintage and Qiming's investment pace
• Kill criteria: [researcher inference · moderate] Round size is driven by peer pace; valuation needs perfect follow-on liquidity; no reserve or downside plan; paper mark is treated as DPI. AVCJ profile, sections on the failed SoftBank vintage and Qiming's investment pace
• Failure mode: [researcher inference · moderate] Confusing capital availability with evidence of company quality. Prepared statement, PDF pp. 5–6 · AVCJ profile, opening SoftBank-fund postmortem

Modern VC translation (as of 2026-08-22)

What transfers. [researcher inference · strong] Value-chain migration, complementary ecosystems, local decision authority and capital-to-milestone discipline remain useful. A modern analyst can apply them to AI infrastructure/applications, industrial automation, biotech services and climate supply chains by tracking who captures margin and data, which standards open or close markets, and whether deployment produces proprietary learning. Prepared statement, PDF pp. 2–7 · AVCJ profile, Qiming formation and pacing sections

What does not transfer unchanged. The 2005 assumption of increasingly frictionless U.S. risk capital into Chinese advanced technology is no longer a safe baseline. Treasury's final outbound-investment rule, effective 2025-01-02, prohibits or requires notification for certain U.S.-person transactions involving China/Hong Kong/Macau semiconductors, quantum technologies and AI. [contemporaneous record · strong] U.S. Treasury, “Outbound Investment Security Program” and “Program Regulations”

Observable applicability signals. [researcher inference · strong] The thesis is live when supplier density shortens iteration time, technical talent moves into startups, customers demand world-standard products, unit margins shift toward proprietary layers, legal ownership and data paths are enforceable, and several independent exit routes exist. It is weak when state procurement is the only customer proof, the VIE/license chain is opaque, outbound rules cover the technology, or paper valuations lack realizations. Prepared statement, PDF pp. 2–7 · Hoover transcript, answers on VIE structure and liquidity · U.S. Treasury, “Outbound Investment Security Program”

Most likely misuse. A modern analyst may quote the correct long-run direction—China innovates, AI diffuses, healthcare demand rises—and skip entry price, ownership, partner attribution, export controls, liquidity, or the company denominator. The SoftBank fund history demonstrates why thematic correctness and fund returns are different propositions. [researcher inference · strong] AVCJ profile, opening section

Grounded current example. Rieschel's 2025 account treats low capital supply amid a larger Chinese entrepreneurial system as an attractive valuation setup, while also describing severe liquidity and regulatory constraints. [investor-stated · provisional] That is a testable forward thesis, not a verified return: it predicts favorable entry prices but still requires lawful sector screening, company-quality evidence and multiple realizable exits. Hoover transcript, answers beginning “the single biggest issue is the liquidity side” and “the money's down to invest in China”

Best writings and unresolved record

Only one qualifying writing survives the searched public record, so a “top five” would be fabricated.

1. China's Global Technology Competitiveness (2005) — the best and only full authored text because it freezes the thesis before Qiming's outcome history, lays out a causal system rather than a slogan, and is preserved by an official archive with a hearing transcript and a contemporaneous dissenting witness. Prepared statement

What could change this synthesis:

• a public Qiming annual/LP letter or internal memo would show whether the macro thesis became a company/portfolio process;
• a stable copy of the 2008 EY IPO panel may add governance and post-listing rules, but it remains spoken/editorial rather than authored;
• public archive or library evidence of a book, essay series or collected correspondence would expand the denominator;
• source appendices for the 2005 statistics could distinguish rigorous underwriting premises from persuasive hearing rhetoric;
• rejected investments, loss memos and contemporaneous IC documents would test whether value-chain analysis actually drove selection rather than later narration.

Coverage and evidence base

Rieschel's public-speaking record is most useful as a longitudinal test: it moves from a pre-Qiming China thesis and IP-risk rule, through Qiming-era selection and firm-design claims, to later trust, failure, policy-miss and liquidity reflections. USCC, printed pp. 43–50 and 61–62, Thoughtful China 00:19:17–00:24:12, USC 00:28:34–00:42:05, and Hoover 00:18:40–00:32:10 — researcher inference / strong. The strongest conclusion is institutional rather than heroic: his durable edge claim is building a local, sector-specialized, equal-economics partnership that can outlast him. 2015 formation account, Fudan section 01, and Hoover 00:06:02–00:07:21 — researcher inference / strong.

The denominator is 29 distinct appearance/source families from 2005–25. Mirrors, reposts and multiple cuts of one event count once. It comprises the nine priority families mapped in sources.md plus 20 additional families found through podcast, YouTube, conference, university, Chinese-language and decade sweeps. Public access yielded 19 captioned videos, one official hearing transcript, and four direct/edited interview or live-note texts; five families remain audio/video metadata without a usable public transcript or caption track. The twelve evidence-rich families below have exact-schema notes: eight full/near-full and four partial. All nine of nine mapped priorities are represented; six are full and three retain an explicit excerpt/summary/no-transcript limit — researcher inference / strong.

Evidence labels are two-axis. Attribution is investor-stated, contemporaneous record, documented behavior, retrospective witness, or researcher inference. Support is strong, moderate, provisional, or insufficient public record. A label does not convert a self-reported return into audited performance.

Appearance denominator and access ledger

“Inventory only” means the route was opened and, where present, captions were recovered, but it was not promoted to a separate note because the appearance was redundant, short, policy-heavy, or lower in decision evidence. It remains in the denominator.

The Asia Society conference program mapped in sources.md establishes a moderator role but contains no substantive remarks, so it is a metadata control rather than a thirtieth appearance — researcher inference / strong.

Source-by-source guide to the twelve analyzed families

Recurring stories and frameworks

Build the institution before scaling the picks

The repeated design is equal managing-partner economics, sector specialization, local leadership and succession. The 2015 interview reports that Qiming equalized managing-partner compensation, invested early and organized around sector expertise; the 2025 retrospective says this was intentional preparation for a firm lasting at least thirty years. 2015 formation account and Hoover 00:06:02–00:07:21, 00:12:56–00:13:22 — retrospective witness / moderate. The causal claim remains unproven because partnership agreements and carry schedules are private, but its consistency across a decade strengthens the institutional reading. 2015 formation account and Hoover 00:06:02–00:07:21 — researcher inference / moderate.

The negative template is his SoftBank fund recollection: very fast deployment, too little deal dissent and one partner with thirty-five boards. Qiming's stated lesson was that partners can remain collegial while strongly opposing a deal. USC 00:39:28–00:42:05 — retrospective witness / moderate.

People are tested through behavior, not adjectives

The founder screen becomes more observable over time:

• Directness: a pitch should become a substantive dialogue; the founder should answer clearly and have researched the investor. USC 00:22:40–00:24:33 — investor-stated / strong.
• Disclosure: missing a plan is normal; concealing the miss until asked exactly the right question is disqualifying over a seven-to-eight-year relationship. Thoughtful China 00:21:51–00:22:56 — investor-stated / strong.
• Generosity: an employee option pool is a founder-level revealed-preference test; partner generosity can be known only after working together. Thoughtful China 00:23:36–00:24:12 and Hoover 00:11:46–00:12:38 — investor-stated / strong.
• Service: he describes venture as support for entrepreneurs, not the origin of the value. His comment that VCs become “full of themselves” limits personal-attribution claims. Hoover 00:10:41–00:12:06 — investor-stated / strong.

Proof thresholds should follow sector risk

Rieschel described different proof thresholds rather than one stage rule: Internet could be pre-revenue; healthcare should pass discovery or show efficacy; capital-intensive cleantech needed traction or a service path. Thoughtful China 00:19:17–00:20:18 — investor-stated / strong. Reading the 2019 admission that Qiming's 2006–07 semiconductor investments were too early as negative evidence supports a broader timing rule: a correct long-run theme may still be uninvestable before its ecosystem is ready. Fudan section 04 — researcher inference / moderate.

Innovation is a system, not an invention count

The 2005 model combined market demand, technical people and risk capital. The 2015 lecture stated the ingredients as problem + people + demand + capital and added leapfrogging: weak physical retail or enterprise legacy systems can accelerate a new platform. USCC, printed pp. 43–50 and PKU 00:07:20–00:10:22 — investor-stated / moderate.

By 2025, Rieschel described the mechanism this way: China may not originate every invention, but fast adoption, a large market and dense manufacturing let firms iterate to best-in-class products. Government can set incentives and standards, yet private firms diffuse technology. Hoover 00:33:17–00:43:22 and 00:47:16–00:48:04 — investor-stated / moderate.

Trust is both an opportunity and an exit variable

In 2013, founder truthfulness is the trust test. In 2019, missing consumer/institutional trust becomes a market-selection question. In 2025, legal predictability and functioning capital markets become the system-level requirement. Thoughtful China 00:21:51–00:24:12, Fudan section 05, and Hoover 00:55:29–00:55:44 — investor-stated / strong. Reading these as one progression—from people diligence, through product adoption, to DPI—is synthesis: missing trust appears to raise transaction costs at every layer. Thoughtful China 00:21:51–00:24:12, Fudan section 05, and Hoover 00:55:29–00:55:44 — researcher inference / moderate.

What changed across twenty years

Contradictions, tensions, and forecast scorecard

• Unlimited capital versus missing liquidity. The 2005 thesis used incoming venture capital as a catalytic advantage; 2025 says exit liquidity is the primary problem. Both can be true because capital raised and cash returned are different. The earlier language was still overconfident. USCC, printed p. 50 and Hoover 00:29:06–00:32:10 — researcher inference / moderate.
• Under-ten-percent failures versus zombie mortality. The 2012 loss claim probably counted only then-recognized failures; the 2020 warning says financing had delayed death. Without a stable denominator, neither figure should appear as a track record. China Money 00:00:56–00:01:16 and USC 00:30:04–00:31:24 — researcher inference / strong.
• China-expert humility versus macro certainty. In 2020 he says there are no China or America experts, only partial views; his 2005 and 2015 forecasts often sound certain. Scoring the forecasts rather than resolving the tension in his favor is an analytical choice. Peggy Liu 00:08:04–00:09:26, USCC, printed pp. 43–50, and PKU 00:11:05–00:17:16 — researcher inference / strong.
• Central control cannot pick winners versus government scale. Across 2005 and 2025 he distinguishes setting incentives from selecting commercial winners. The apparent contradiction dissolves only if direction, subsidy, adoption and firm-level innovation remain separate analytical steps. USCC, printed p. 43 and Hoover 00:21:45–00:24:17 — researcher inference / moderate.
• 2012 top-quartile return prediction — unresolved. Rieschel said a two-times return had historically been top quartile and expected all Qiming funds to fall well within that group. The corpus has no vintage-matched LP cash-flow schedule, independent benchmark or definition of the intended return metric. Validation requires every fund then referenced to rank top quartile on a fixed, vintage-matched benchmark; one fund below that benchmark falsifies “all our funds.” The separate 2.0× marker cannot be scored until its gross/net and realized/unrealized basis is known. China Money 00:04:41–00:04:53 — researcher inference / insufficient public record.
• 2015 personalized-medicine forecast — unresolved and weakly falsifiable as spoken. Rieschel predicted China could lead personalized medicine but supplied neither horizon nor comparative outcome metric. Portfolio examples do not validate national leadership. A fair prospective test needs a fixed date and a peer-country measure of approved products, clinical adoption or patient outcomes; failure to lead on the chosen measure would falsify the claim. PKU 00:15:01–00:16:32 — researcher inference / insufficient public record.
• 2015 autonomous-city forecast — unresolved and weakly falsifiable as spoken. He proposed that autonomous vehicles could reorganize cities but gave no horizon, city cohort or land-use threshold. Validation requires a prospectively fixed Chinese-city cohort and measurable mobility-driven land-use or infrastructure change; no such change by the chosen date would falsify the forecast. PKU 00:16:35–00:17:16 — researcher inference / insufficient public record.
• M&A optimism versus weak exits — unresolved and weakly falsifiable as spoken. In 2013 he expected time-to-market pressure to improve acquisition behavior; in 2016 and 2025 he still described poor liquidity and unreliable listings. He gave no horizon or baseline, and the corpus contains no comparable exit-series test. Under a fixed post-2013 window and definition, no increase in strategic M&A's share of China venture exits after controlling for total exit volume would falsify the operationalized forecast. Thoughtful China 00:15:50–00:16:36, Forbes “biggest disappointment”, and Hoover 00:29:06–00:32:10 — researcher inference / insufficient public record.
• IP enforcement forecast — unresolved and weakly falsifiable as spoken. In 2005 he expected local founders to press for stronger enforcement and changed product selection meanwhile, but gave no horizon or outcome threshold. No comparable enforcement series was recovered. Under a fixed period and definition, no improvement in founder enforcement outcomes—such as adjudication, damages or piracy-loss recovery—would falsify the operationalized claim; absent that test it is not “correct.” USCC, printed pp. 50 and 61–62 — researcher inference / insufficient public record.
• Scarcity as a good vintage. The 2025 claim that capital scarcity could produce an attractive vintage is explicitly forward-looking. No result exists as of the recording; an exit/DPI test should be set before later outcomes are known. Hoover 00:48:04–00:50:35 — investor-stated / insufficient public record.

Biography and institution signals

• Pre-Qiming: the 2005 hearing identifies Rieschel with Mobius and captures his China thesis before Qiming. The 2025 chronology adds SoftBank China Venture Capital, SAIF and Ceyuan, but it is a later recollection with transcript errors. USCC witness statement and Hoover 00:01:03–00:06:18 — contemporaneous record / strong for the 2005 role; retrospective witness / moderate for the fuller sequence.
• Qiming formation: Duane Kuang is the indispensable co-founder/partner in the 2015, 2019 and 2025 accounts. Equal economics, sector teams and local succession are institution signals; they prevent a lone-founder interpretation. 2015 interview formation section and Fudan section 01 — retrospective witness / moderate.
• Japan and cross-cultural formation: the Peggy Liu conversation ties Japan (1988–93), his wife Yuka and Masayoshi Son to an authenticity rule: work across cultures without pretending to become the other culture. 00:00:54–00:06:56 — retrospective witness / moderate.
• Seattle transition: the 2016 interview was conducted as he moved from Shanghai to Seattle, saying a remote investor should not make China deals without a holding-period-length local commitment. He framed the move as succession plus a China-informed U.S. healthcare strategy. Forbes “Why leave China now?” — investor-stated / strong.

Deal, miss, and attribution signals

Contemporaneous evidence versus retrospective memory

The 2005 hearing, 2012 China Money excerpt, 2013 panel, 2015 lecture/interview and 2016 Forbes Q&A are the principal contemporaneous anchors. They show what Rieschel was willing to say before later policy and liquidity outcomes: China capability optimism, copy-risk avoidance, top-quartile confidence, sector proof thresholds, early-stage institutional design and an initial liquidity warning. USCC, printed pp. 43–62, China Money 00:00:24–00:04:53, Thoughtful China 00:12:12–00:24:12, PKU 00:00:22–00:17:16, and Forbes Q&A — contemporaneous record / strong.

The 2019–25 accounts contain partner credit, the SoftBank failure, semiconductor timing, Didi/Kuaidi and ByteDance passes, spectacular company-return claims, and the missed state-control inflection. Fudan section 01 and section 04, USC 00:25:01–00:42:05, and Hoover 00:06:02–00:32:10 — retrospective witness / moderate. They are more useful for claimed lessons than for original causation because memory selection, survivor bias and firm storytelling remain active risks — researcher inference / strong. The synthesis therefore preserves both layers:

1. Use the early source to define the forecast or rule. USCC hearing and China Money excerpt — researcher inference / strong.
2. Use the later source to identify the claimed outcome or lesson. USC 00:25:01–00:42:05 and Hoover 00:06:02–00:32:10 — researcher inference / strong.
3. Require filings, cash flows, IC records or independent deal attribution before converting either into a verified track record. China Money's prospective return claim, Hoover's firm-return claims, and Forbes's Xiaomi attribution — researcher inference / strong.

Criticism, failure, litigation, and adverse evidence

The most material adverse evidence in the recovered spoken corpus is self-supplied; that makes it useful against hagiography but insufficient as an independently verified loss record. USC 00:28:58–00:42:05, Fudan section 04, and Hoover 00:12:27–00:20:11 — researcher inference / strong.

• a SoftBank fund that, in his telling, would not return capital despite early paper gains. USC 00:39:28–00:42:05 — retrospective witness / provisional;
• substantial losses from investing too early. USC 00:28:58–00:29:03 — retrospective witness / moderate;
• Qiming's early semiconductor investments that were not very successful. Fudan section 04 — retrospective witness / moderate;
• Didi/Kuaidi and ByteDance passes. Forbes and Hoover 00:12:27–00:12:56 — retrospective witness / moderate;
• acknowledged failure to anticipate the 2015–16 control turn and later listing/liquidity break. Hoover 00:18:40–00:20:11 — retrospective witness / moderate.

Exact-name searches combining Gary Rieschel/Qiming with lawsuit, sued, litigation, complaint, SEC enforcement, investigation, criticism, failure and underperformance produced no responsive personal or firm enforcement proceeding. Results were ordinary offering/adviser records, public-company director filings, irrelevant same-name/property records and secondary criticism. This is not legal clearance: Chinese court databases, subscription dockets, historical entity aliases and private disputes remain outside the search. No adverse spoken allegation from another first-hand witness was found; therefore the record is strongest on his admissions and weakest on independent challenge. Adverse-search boundary — researcher inference / moderate.

Ranked listening and reading path

1. USC video conversation, 2020 — the best teaching source. It combines a founder screen, stage/fund-life design, Tigermed team attribution, “too early” losses, zombie-company caution and the full SoftBank deployment/board-capacity failure — researcher inference / strong.
2. Hoover / China Considered, 2025 — the best complete retrospective and the only priority appearance with both official transcript and manual captions. Read return claims skeptically; use it for origin, partner, succession, policy-miss and liquidity models — researcher inference / strong.
3. USCC hearing, 2005 — the indispensable ex-ante anchor. It lets later China capability and IP claims be scored without hindsight leakage — researcher inference / strong.
4. Thoughtful China panel, 2013 — the highest density of reusable decision rules: sector proof thresholds, voluntary disclosure, employee-pool generosity and track-record denominator skepticism — researcher inference / strong.
5. PKU lecture, 2015 — the clearest contemporaneous statement of leapfrogging, “driving in fog,” consumer-quality transition and manufacturing/product iteration — researcher inference / strong.

Remaining gaps and research boundary

• No publisher transcript was found for Inspired by Example 2024, Sinica 2020 or China Money 2022. The playable audio routes are preserved; no synthetic quotation or inferred takeaway was created — researcher inference / strong.
• The public 2012 China Money video is a 5:02 excerpt, not the complete episode. Its fund claims require the full recording and LP cash-flow evidence — contemporaneous record / strong.
• The 2017 Silicon Dragon keynote video has no recoverable public caption track. It remains inventory only — researcher inference / strong.
• The Qiming/Fudan live-note and Hoover transcript normalize or misspell Duane Kuang, Nisa Leung, SAIF, Ceyuan, company names and technical terms. Short quotations were limited to clear audio/caption agreement; names were normalized only when independent context was strong — researcher inference / strong.
• No Computer History Museum, Stanford/Bancroft or other named oral-history collection was found. University, library/finding-aid, Wayback and decade searches yielded event appearances rather than a dedicated oral history. Source inventory and search boundary — researcher inference / insufficient public record.
• Most return evidence remains firm- or Rieschel-reported. No LP cash-flow schedule, investment-committee memo, complete loss denominator or partner-by-deal allocation was public in this pass. China Money return prediction and Hoover company-return claims — researcher inference / insufficient public record.
• No outreach, paywall bypass, private-database access or private recording request was attempted — researcher inference / strong.

Model 1 — follow value migration, then prove company capture

• Rule — Map how standards, cycle time, talent, infrastructure and distribution move profit pools; then require company-specific evidence of durable capture. [researcher inference · strong] 2005 writing note
• Evidence status — Strong as a pre-Qiming market thesis and moderate as an implemented selection edge because original deal memos, a complete portfolio and matched passes are unavailable. [researcher inference · moderate]
• Provenance and origin — Rieschel's 2005 statement describes manufacturing-to-design-to-brand migration after his Japan and cross-border operating experience. The hearing preserves contemporary disagreement over whether China already produced advanced innovation. [investor-stated · strong] writing analysis · USCC talk note
• Mechanism and reasoning — Ecosystem mapping identifies where capability and bargaining power may accumulate; company proof prevents a correct country or category forecast from becoming an indiscriminate investment. [researcher inference · strong]
• Historical and vehicle context — The thesis predates Qiming. Qiming's sector teams and China funds were shared vehicles; later outcomes cannot be credited to Rieschel or treated as proof that the paper was the IC blueprint. [documented behavior · strong] Fund II announcement
• When to use — Supply-chain shifts, new technical standards, industrial upgrading, healthcare infrastructure, climate platforms and AI layers where value can move across hardware, model, application, workflow and distribution. [researcher inference · strong]
• How to apply — Draw the value chain at entry and downside; identify the scarce complement; test proprietary capability, time-to-market, margin, customer willingness to pay, legal title and exit; state what would show that value bypasses the company. [researcher inference · strong]
• Examples — Three bounded cases, including an exception. [researcher inference · moderate]
  • Retrospective application: Tigermed as domestic clinical-research infrastructure; the whole-fund return and industry-causality claims remain unaudited. Tigermed lab
  • Team-attributed application: Xiaomi captured consumer hardware, brand and channel value, but Hans Tung/Duane Kuang and Qiming II—not Rieschel alone—own the public investment attribution. Gan & Lee/Xiaomi lab
  • Exception: correct China innovation direction did not prevent early semiconductor disappointments or later liquidity constraints. PKU note · Hoover note
• Evidence — A decision-date official statement, an official hearing, contemporaneous 2009 diligence reporting and later issuer/firm records establish the thesis and bounded examples; none supplies a controlled selection study. [contemporaneous record · strong] NYT/Qiming
• Limits, failure modes, and exception ledger — A growing ecosystem can reward suppliers, incumbents or consumers rather than the startup; policy, price, dilution and exit can absorb value; broad theses invite confirmation bias; selected winners conceal false positives. [researcher inference · strong]
• Evolution and contradictions — Capital changes from an almost unqualified 2005 catalyst to a pacing and liquidity risk after SoftBank and China's exit reversal. The model survives only if capital and reform are treated as contingent complements. [researcher inference · strong] writings evolution
• Modern VC translation (as of 2026-08-22) — In AI, distinguish model, compute, data, application, workflow and distribution value; prove retention, inference economics, data rights and buyer control instead of treating model adoption as company capture. [researcher inference · strong]

Model 2 — embed local authority for the full risk horizon

• Rule — Do not invest cross-border from periodic visits: place trusted local partners with real economics and decision authority close to founders, customers and regulators for the expected holding period. [investor-stated · strong] Homeward note
• Evidence status — Strong for stated doctrine, Rieschel's Shanghai move, Kuang's co-founder role and named GP/IC structures; insufficient for authority and economics across every Qiming vehicle. [documented behavior · strong]
• Provenance and origin — Japan/Asia operating experience preceded the doctrine; Qiming then adopted an independent local identity and Kuang partnership, and the 2019 Fudan account calls partner choice foundational. [retrospective witness · moderate] profile chronology · Fudan trust note
• Mechanism and reasoning — Embedded partners shorten feedback, improve references and interpretation, build founder trust and see institutional or policy changes earlier. Real economics prevent local roles from becoming ceremonial. [researcher inference · strong]
• Historical and vehicle context — SBCVC, SAIF, Ceyuan, Qiming China and Qiming U.S. used different leaders and beneficiaries. Institution sponsorship is not company-selection credit, and one GP filing is not firmwide governance. [contemporaneous record · strong] vehicle map
• When to use — Markets where ownership, customer truth, regulation, talent and distribution are relationship- or locality-dependent and where the holding period exceeds a short visit cycle. [researcher inference · strong]
• How to apply — Name the local decision owner, economics, veto, conflicts, reference network, succession path and escalation rule; test decisions made against headquarters' initial view; require an independent legal/entity map. [researcher inference · strong]
• Examples — Three bounded cases. [researcher inference · strong]
  • Documented application: Rieschel moved to Shanghai and co-founded a locally branded Qiming with Kuang. AVCJ formation account
  • Team-attributed application: Leung built the healthcare practice and originated Gan & Lee; Kuang/Tung carried Xiaomi work. Gan & Lee source · Hans Tung source
  • Boundary case: Seattle-based U.S. healthcare activity used a separate local vehicle and leadership initiative; the record does not prove that Rieschel could remotely recreate the original China edge. Jasper 13G · AVCJ succession account
• Evidence — Official biography/filings, AVCJ succession reporting and direct Fudan/Homeward remarks corroborate formation and authority from different routes; partnership agreements and decision logs remain private. [contemporaneous record · strong] AVCJ succession PDF
• Limits, failure modes, and exception ledger — Local partners can share local consensus errors, relationship bias or policy exposure; headquarters can retain hidden vetoes; equal title may mask unequal economics; embeddedness cannot overcome prohibited transactions or closed exits. [researcher inference · strong]
• Evolution and contradictions — Rieschel evolves from foreign sponsor across SBCVC/SAIF/Ceyuan to Qiming co-founder and then a Seattle-based successor/sponsor. The contradiction is that later remote U.S.–China work sits beside his warning against remote investing. [researcher inference · moderate]
• Modern VC translation (as of 2026-08-22) — For any cross-border AI, biotech or climate mandate, publish a country-by-country decision-rights and beneficiary map; treat local references as evidence inputs, not a substitute for independent security, legal and customer checks. [researcher inference · strong]

Model 3 — match proof thresholds to sector physics

• Rule — Ask for the next binding proof appropriate to the sector: product behavior for Internet, scientific/clinical/regulatory evidence for healthcare, and repeatable technical economics plus customers for industrial or climate systems. [investor-stated · strong] Thoughtful China note
• Evidence status — Strong as a direct 2013 rule and moderate as a career-wide practice because only selected deals expose decision-date evidence and terms. [investor-stated · strong]
• Provenance and origin — Qiming's specialist teams were operating by the Fund II period; Rieschel later made their different evidence thresholds explicit in a filmed panel. [documented behavior · strong] Fund II · 2013 panel note
• Mechanism and reasoning — Generic stage metrics hide different failure chains. Sector-specific gates buy the information that most changes survival, valuation and the next financing decision. [researcher inference · strong]
• Historical and vehicle context — Qiming I–III mixed technology/consumer, healthcare and cleantech across early and expansion stages. The public record does not reveal reserves, approval thresholds or whether later funds kept the same definitions. [contemporaneous record · strong] Fund III
• When to use — Portfolios spanning software, fintech, biotech, devices, healthcare services, hardware, climate and other long-validation technical categories. [researcher inference · strong]
• How to apply — Map risks in dependency order; for each, state evidence, independent verifier, owner, cost, date, kill condition and what the next check purchases; separate technical proof from customer, regulatory and financing proof. [researcher inference · strong]
• Examples — Three bounded cases, including an exception. [researcher inference · strong]
  • Retrospective application: Tigermed required trial quality, sponsor trust, clinical staff and regulator acceptance; the exact entry gates are unavailable. Tigermed lab
  • Contemporaneous application: LanzaTech entered with pilot operation, named industrial counterparties and a proposed pre-commercial plant; scale-up, yield, capex and offtake remained binding. LanzaTech lab
  • Exception: an Internet product may justify pre-revenue investment, but ByteDance proves that no threshold works if the opportunity never reaches review. ByteDance lab
• Evidence — Direct video provides the rule; Qiming fund announcements and company/issuer records corroborate sector structure and bounded cases. No private milestone budget or rejected cohort was found. [contemporaneous record · strong] LanzaTech proxy
• Limits, failure modes, and exception ledger — Gates can reward easy proxies, underfund platform work or create financing cliffs; regulation may change after proof; one sector can contain multiple business models; strong science cannot cure price or exit. [researcher inference · strong]
• Evolution and contradictions — The 2005 ecosystem thesis becomes explicit sector evidence in 2013 and later liquidity/legal gating. The contradiction is that public Qiming outcome reporting still emphasizes listings and unicorns rather than proof histories. [researcher inference · strong]
• Modern VC translation (as of 2026-08-22) — For AI, separate model capability, evaluation validity, data rights, inference cost, security, workflow adoption, buyer authority and renewal; do not use a benchmark score as a universal sector gate. [researcher inference · strong]

Model 4 — reconstruct operating truth when institutions are weak

• Rule — When accounting, ownership, customer or market records cannot be trusted, build an independent fact pattern in the field and walk away if legal control or operating reality remains opaque. [documented behavior · strong] 2009 NYT/Qiming
• Evidence status — Moderate to strong for several contemporaneously reported practices; insufficient for portfolio-wide consistency or causal outcome improvement because the source is a selected case family. [documented behavior · moderate]
• Provenance and origin — Rieschel's 2005 hearing identified weak IP enforcement; the 2009 profile records Qiming's site, receipt and ownership verification; the 2025 account adds unreliable early accounting/legal/reference infrastructure. [retrospective witness · strong] USCC note · Hoover note
• Mechanism and reasoning — Independent observation and reconciliation reduce founder-reporting, intermediary and database error; explicit walk-away conditions prevent more diligence from rationalizing uncertain title. [researcher inference · strong]
• Historical and vehicle context — Early China venture lacked familiar audit, market-data and enforcement systems, while later VIE, state and exit risks changed the perimeter. The method belongs to Qiming teams, not Rieschel alone. [researcher inference · strong]
• When to use — New geographies, offline-to-online services, channel-heavy businesses, marketplaces, regulated services, complex ownership, disputed KPIs and AI products whose logs or evaluations are founder-controlled. [researcher inference · strong]
• How to apply — Triangulate legal title, receipts/bank movement, customer existence/use, physical operations, employee/partner references and regulator/license status; record sampling limits and a hard unresolved-title stop. [researcher inference · strong]
• Examples — Three bounded cases. [documented behavior · moderate]
  • Documented application: eHi mystery rentals and site visits tested service reality before investment. NYT/Qiming, diligence section
  • Documented application: Kaixin followed nearly a year of social-network mapping rather than a single pitch. NYT/Qiming
  • Documented rejection boundary: cloudy ownership or groups of relatives on payroll drove separate walk-aways; receipt reconstruction concerned an investment that proceeded. The unnamed companies prevent independent outcome scoring. NYT/Qiming
• Evidence — A contemporaneous New York Times profile mirrored by Qiming anchors all three practices; the official hearing and later direct Rieschel Hoover transcript corroborate the institutional problem, not the cases' returns. [contemporaneous record · moderate]
• Limits, failure modes, and exception ledger — The selected anecdotes may be unusual; intensive diligence is costly and can delay; samples can be gamed; title can remain unenforceable; verified operations can still have poor economics, policy exposure or no exit. [researcher inference · strong]
• Evolution and contradictions — The method shifts from copy/IP and record verification to a broader recognition that even valid companies can be trapped by legal and liquidity regimes. It cannot support a claim that diligence solved China risk. [researcher inference · strong]
• Modern VC translation (as of 2026-08-22) — Reproduce AI usage from raw customer-side evidence, independently evaluate outputs and costs, verify data/model rights, trace revenue to bank/contract records, and map security/legal stop authority. [researcher inference · strong]

Model 5 — match capital pace to proof and realizable liquidity

• Rule — Deploy only as quickly as operating proof, reserves and plausible realization capacity justify; never let available capital or a paper mark set the pace. [investor-stated · strong] SoftBank failure lab
• Evidence status — Strong as a direct failure-derived rule; moderate as a validated Qiming performance edge because complete matched-vintage cash flows and reserve records are unavailable. [investor-stated · strong]
• Provenance and origin — The 1999 SoftBank vehicle was reportedly deployed in six months, marked at 3x and ultimately below 1x; 2000 reporting captured survival triage, and later accounts link Qiming's caution to that loss. The fund outcome is [retrospective witness · moderate] [single-source] [disputed]; the contemporary triage is [contemporaneous record · moderate]. Forbes 2000 · AVCJ/Qiming
• Mechanism and reasoning — Slower, milestone-linked deployment preserves option value and follow-on capacity, limits correlated vintage/price risk and prevents unrealized comparable marks from being mistaken for distributable return. [researcher inference · strong]
• Historical and vehicle context — SoftBank's fund numbering and final cash flows remain disputed; Qiming's sequential fund sizes and formal review are documented, but reserves, recycling, fees/carry and net returns are not. [contemporaneous record · strong] vehicle economics
• When to use — Hot markets, large funds, long technical cycles, policy-dependent exits, cross-border securities and any round whose private mark is anchored to a fragile public comparable. [researcher inference · strong]
• How to apply — Predefine the next proof, cost, time and kill condition; stress a three-year closed exit window; budget company runway and fund reserves separately; report cost, realized value, unrealized value and net LP metrics distinctly. [researcher inference · strong]
• Examples — Three bounded cases. [researcher inference · strong]
  • Failure: SoftBank's accelerated 1999 deployment and below-1x endpoint. SoftBank lab
  • Long-cycle application: LanzaTech required successive technical/commercial financings and fourteen years of governance; return remains unknown. LanzaTech lab
  • Liquidity exception: Gan & Lee waited through multiple listing attempts; patient duration may be necessary, but a reported $1 billion distribution still needs cash-flow verification. Gan & Lee source
• Evidence — Contemporaneous Forbes evidence, independent AVCJ retrospective, Qiming fund releases and issuer/company records triangulate pace and examples; no audited fund ledger closes the causal test. [contemporaneous record · strong]
• Limits, failure modes, and exception ledger — Speed can be rational for scarce allocation or low-cost option checks; slow funds can still overpay; milestones can be gamed; exit closure can exceed fund life; reserves can double down on weak assets. [researcher inference · strong]
• Evolution and contradictions — Capital moves from 2005 catalyst to post-SoftBank hazard and, by 2025, liquidity becomes a binding China constraint. The correct macro direction never cancels security timing. [researcher inference · strong] Hoover note
• Modern VC translation (as of 2026-08-22) — For AI and advanced technology, size rounds against evaluation, deployment, unit-cost, security, legal and renewal gates; show an ex-top-one fund case and forbid model/vendor marks from substituting for distributions. [researcher inference · strong]

Model 6 — align senior economics, then force material opportunities into collective review

• Rule — Reduce partner deal-hoarding with shared senior economics and require every potentially material opportunity, dissent and repeat pass to reach a specialist-informed partnership review. [researcher inference · strong] AVCJ process account · ByteDance lab
• Evidence status — Strong for reported equal economics, weekly meetings, repeated formal discussion and named shared IC authority; moderate for efficacy because ByteDance demonstrates a breach and no decision cohort exists. [documented behavior · strong]
• Provenance and origin — Rieschel links Qiming's measured, team-oriented structure to lessons from SoftBank. Taomee's 13G documents four-person IC authority, while the 2025 account records a partner passing ByteDance twice without escalation. [retrospective witness · strong] Taomee 13G · Hoover note
• Mechanism and reasoning — Shared economics lowers incentives to own or rush deals; sector specialists improve questioning; mandatory escalation limits one-person false negatives; written dissent makes process failure auditable. [researcher inference · strong]
• Historical and vehicle context — Equal cash/carry is reported for managing partners and one 2020 filing shows equal ownership of Qiming Corporate GP VI. Neither fact establishes every vintage, vote threshold, work allocation or beneficiary. [contemporaneous record · strong] HKEX GP VI filing
• When to use — Multi-partner funds, specialist franchises, geographically distributed teams and categories where a single outlier can dominate returns or a pass can disappear before IC. [researcher inference · strong]
• How to apply — Define materiality and automatic escalation; log every pass and re-pitch; assign specialist and red-team reviewers; record votes/dissent/conflicts; audit outcomes and false negatives by partner/channel; separate shared economics from contribution review. [researcher inference · strong]
• Examples — Three bounded cases, including the decisive exception. [documented behavior · strong]
  • Documented structure: Qiming's five managing partners reportedly met weekly and required at least two formal discussions per deal. AVCJ/Qiming
  • Shared authority: Kuang, Rieschel, JP Gan and Robert Headley comprised the relevant Taomee IC; the filing proves authority, not labor or vote. Taomee 13G
  • Exception/failure: ByteDance was passed twice by one partner without reaching colleagues; later scale cannot be converted into foregone Qiming return. ByteDance lab
• Evidence — AVCJ direct reporting, SEC/HKEX filings and the direct Rieschel Hoover transcript corroborate structure and failure from separate routes; the LPA, IC minutes and full pass ledger are missing. [contemporaneous record · strong]
• Limits, failure modes, and exception ledger — Shared economics can enable free riding; collective review can create groupthink, slow scarce deals or diffuse accountability; escalation floods IC unless materiality is explicit; firing a partner does not prove a corrected system. [researcher inference · strong]
• Evolution and contradictions — SoftBank's pace and individual excitement give way to equal economics and repeated review; ByteDance shows the formal system failed before review. The model is credible only if bypasses are measured and corrected. [researcher inference · strong]
• Modern VC translation (as of 2026-08-22) — Maintain a signed opportunity ledger with automatic second review for repeated passes, unusually large potential markets, partner conflicts and out-of-thesis technical shifts; publish false-negative and override audits to the partnership. [researcher inference · strong]

Evidence gaps and candidate models

[researcher inference · insufficient public record] These are retrieval hypotheses, not additional hidden models. Each promoted model above contains the required twelve components, at least three bounded applications or exceptions and multiple independent evidence routes; none is represented as a universal Rieschel algorithm.

What the evidence supports

The strongest defensible conclusion is institutional, not heroic. Rieschel articulated a value-chain and China-capability thesis before Qiming's outcomes, acknowledged a major SoftBank pacing failure, moved to Shanghai, chose a locally experienced co-founder, accepted shared senior incentives and authority, helped organize specialist teams, and held documented board roles. Those records support a plausible edge in institutional architecture for low-information cross-border markets. They do not establish a personal IRR, hit rate or causal share of Qiming's company outcomes. [researcher inference · moderate] profile edge · 2005 statement · AVCJ process account

The intellectual record has one qualifying authored work rather than an inflated essay corpus. The 2005 statement maps manufacturing, design, brand, talent, demand and risk capital before Qiming existed. Twelve deep spoken notes then test the doctrine over twenty years: sector proof thresholds, local presence, shared economics, founder candor, SoftBank loss, semiconductor timing, ByteDance escalation failure and the later China liquidity constraint. [documented behavior · strong] writings coverage · talks coverage

The investment evidence is bounded but useful. The 43-lineage material union contains 42 invested lineages and the ByteDance miss. Tigermed and LanzaTech support specific Rieschel governance roles; Gan & Lee and Xiaomi support Qiming vehicle outcomes while allocating partner credit to Nisa Leung, Duane Kuang and Hans Tung; SoftBank and ByteDance provide adverse process labs. The best visible return figures remain [self-reported] and [single-source]. [researcher inference · moderate] denominator · five decision labs

What could falsify the thesis

The favorable thesis predicts that private records would show local partners with real decision rights and economics; sector-specific proof gates applied before outcomes; material opportunities escalated into shared review; slower deployment and better reserve discipline after SoftBank; and ordinary, outlier-removed Qiming vintages comparing favorably with appropriate peers after price, ownership and liquidity are controlled. [researcher inference · strong] profile falsifiers

It would be weakened or falsified by partnership records showing ceremonial local roles or materially unequal authority; IC files showing that public rules were absent or selectively remembered; a full pass/loss ledger showing no escalation change after ByteDance; same-basis cash flows showing no improvement over SoftBank; or partner-work records showing that Kuang, Leung, Tung, Gan, Hu and others supplied the decisive judgment while Rieschel mainly supplied capital, access or narration. [researcher inference · strong] philosophy thesis

The simpler alternative remains plausible: a correct China macro call, favorable 2005–16 ecosystem beta, U.S. institutional capital and exceptional local partners produced a selected set of visible outliers. Qiming's public 580+ company universe lacks a dated full schedule, losses, write-offs, opportunity set and audited fund cash flows. That alternative explains the record without repeatable Rieschel personal alpha. [researcher inference · moderate] skeptical profile

Attribution and denominator audit

Five entity scopes remain separate: SoftBank Venture Capital/Mobius; sponsored China institutions SBCVC/SAIF/Ceyuan; Qiming China USD/RMB and guidance vehicles; Qiming U.S. healthcare vehicles; and personal commitments. One company appearing in several sources is one lineage, while separate exposures to a parent and acquired company are not silently pooled. The corrected union is 42 invested lineages plus one miss: eight SoftBank/Mobius, one SAIF, eighteen Qiming 2011 roster companies and fifteen later Qiming/U.S. lineages, plus ByteDance. [researcher inference · strong] scope and rules

Qiming's live claims—580+ investments, 210+ exits and 80+ unicorns—are promotional cumulative counts. The reviewed set includes 33 selected Qiming invested lineages, eight sufficiently reconstructed for bounded chronology/ownership/governance/outcome, and 572+ without sufficient reconstruction. The 33/580+ figure is at most 5.7% research-selection coverage, not economic coverage or a hit rate. Jasper is the sole evidenced Qiming U.S. Healthcare company lineage; LanzaTech remains a cross-border Qiming case and is not recast as a U.S.-fund investment. [researcher inference · strong] corrected denominator table

Attribution is decomposed into sourced, sponsored, diligenced, approved, negotiated, board work, recruited, follow-on, exit influence and later narrator. The public record supports Rieschel board roles at Tigermed and LanzaTech; Leung's Gan & Lee role; Tung/Kuang's Xiaomi work; and shared GP/IC authority in named structures. Shared control does not prove labor, and founder/company value is not investor-created value. [researcher inference · strong] people map

Transfer limits for modern VCs

The portable core is procedural: map the value layer; embed real local authority; set proof by sector physics; independently reconstruct operating truth; match capital to proof and realizable liquidity; and audit whether collective review catches material dissent and passes. [researcher inference · strong] The six mental models turn those ideas into dated field guides with provenance, mechanisms, examples, exceptions and falsifiers.

The least portable elements are China-era sector labels, relationship prestige, one partner's retrospective memory, and historical access to cross-border capital. Current advanced-technology investing may face outbound controls, data/security rules, VIE/entity risk, fragmented standards and uncertain listing routes. Treasury's program is a current constraint, not evidence that historic investments violated later rules. [researcher inference · moderate] modern translation

Do not use “local context” to excuse opacity, “patient capital” to excuse missing milestones, “service” to claim founder value, “team” to erase responsibility, or a correct macro forecast to ignore price, dilution and exit. Modern AI applications in the package are explicitly researcher-created translations rather than claims about Rieschel's historical practice. [researcher inference · strong] model guardrails · modern translation

Unresolved evidence

The decisive missing objects are Qiming vintage cash-flow schedules; a complete dated 580+ company/write-off/pass ledger; partnership agreements, IC minutes, rejection logs and reserve policies; the twelve unnamed SoftBank investments in Rieschel's biography; the identity and final cash flows of the failed 1999 vehicle; entry and realization records for Tigermed, Gan & Lee, Xiaomi and LanzaTech; the ByteDance rejection files; partner work and carry allocation; board interventions and founder testimony; current ADV/vehicle records; and non-U.S. legal dockets. retrieval ledger

The direct-voice frontier is also explicit. No authored book, blog, annual letter or public LP-letter series was found. The 2012 China Money source is an excerpt; three later podcasts and one keynote remain untranscribed; the 2008 EY panel body and current ADV routes remain unavailable. No outreach, private-database access or paywall bypass occurred. sources boundary · talks gaps

Corrections made during adversarial review

Central-claim citation audit trail

Ordinary-claim citation audit trail

Independent deal and team cross-checks

Mechanical reconciliation and citation QA

Twenty-point acceptance matrix

Reader conclusion

Gary Rieschel belongs in the canon as a consequential cross-border venture institution builder and reflective governor, not as a publicly proven personal-return champion. The reusable lesson is to build the information and decision system before admiring the outliers: local authority, sector proof, ground-truth diligence, paced capital and auditable collective review. The record is strong enough to teach that system and weak enough to keep the personal-alpha claim open. [researcher inference · moderate] profile edge · denominator · models

Direct writings

• ★★★ | Written statement before the U.S.-China Economic and Security Review Commission | https://www.uscc.gov/sites/default/files/4.21-22.05rieschel_gary_wrts.pdf | 2005-04-21 | authored testimony, 7 PDF pages; source type: primary; reliability A | Verification: full, PDF pp. 1–7. Why: earliest complete ex-ante thesis on China's manufacturing-to-design shift, market scale, talent, venture capital and IP risk. Limit: many quantitative premises are uncited contemporary estimates.

Edited interviews, event summaries and press text

• ★★★ | Profile: Qiming Venture Partners' Gary Rieschel | https://www.qimingvc.com/cn/node/3424 | 2016-04-20 | AVCJ profile/Q&A mirrored by Qiming; source type: primary interview plus secondary framing; reliability A/B | Verification: full, headings from opening through “An evolving market.” Why: dot-com failure, Qiming formation, sector structure, IC, equal economics and pacing. WRITINGS boundary: reporter-authored profile, not counted as a Rieschel writing. Limit: retrospective and firm/LP-facing; one fund-size figure appears as a site transcription error.
• ★★★ | Q&A: Qiming Venture Partners' Gary Rieschel | https://www.qimingvc.com/en/news/qa-qiming-venture-partners-gary-rieschel | 2017-11-13/17 | AVCJ direct Q&A mirrored by Qiming; source type: primary interview; reliability A | Verification: full Q&A. Why: SoftBank, SBCVC, SAIF, Ceyuan, Qiming formation, USD/RMB investor incentives, team durability and capital-market structure. WRITINGS boundary: interview text, not an authored essay. Limit: retrospective; financial claims lack LP schedules.
• ★★ | Founder of Qiming Suggests on Entrepreneurship and Funding | https://www.qimingvc.com/en/news/founder-qiming-suggests-entrepreneurship-and-funding | 2012-04-28 | Fudan event summary; source type: edited/translated remarks; reliability B | Verification: full page, sections “Entrepreneurship” and “Financing.” Why: discipline, openness, timing, macro fit, risk honesty, financing preparation and concise pitch rules. WRITINGS boundary: event summary, reserved for TALKS/press corroboration. Limit: summary, not transcript; “Gray” errors signal editing/translation noise.
• ★★ | Think Globally, Act Globally | https://www.qimingvc.com/en/news/think-globally-act-globally | 2015-07-14 | RMI profile with direct quotations; source type: primary remarks/organizational profile; reliability A/B | Verification: full page, sections on China strategy and long-term partnerships. Why: long-horizon partner selection, practical economic rationale, environmental formation and cross-country transfer. WRITINGS boundary: RMI-authored profile, reserved for press evidence. Limit: philanthropic/promotional context; not a portfolio memo.
• ★★ | The future opportunity and challenge is building trust | https://www.qimingvc.com/cn/news/%E5%90%AF%E6%98%8E%E8%A7%82%E7%82%B9-gary-rieschel%EF%BC%9A%E6%9C%AA%E6%9D%A5%E7%9A%84%E6%9C%BA%E4%BC%9A%E5%92%8C%E6%8C%91%E6%88%98%EF%BC%8C%E5%B0%B1%E6%98%AF%E5%BB%BA%E7%AB%8B%E4%BF%A1%E4%BB%BB | 2019-10-29 | selected Chinese live-note from Fudan forum; source type: edited remarks; reliability B | Verification: accessible Chinese body, sections 01 onward. Why: partner trust, institutionalization and healthcare change. WRITINGS boundary: live-note reserved for TALKS. Limit: selected shorthand and translation boundary; numerical claims are not audited.
• ★★ | Perspectives on global IPO markets — post-IPO value creation panel | https://capitalaberto.com.br/wp-content/uploads/2012/06/Global_IPO_trends_report_2009.pdf | 2009 report on 2008-10 panel | edited panel transcript in EY report; source type: primary remarks; reliability A/B | Verification: indexed result exposed relevant Rieschel passages, but direct PDF fetch returned an internal error. Why: post-IPO governance and investor-relations advice. Limit: not content-analyzed; lawful live route must be retried or archived.
• ★ | Gary Rieschel-authored blog, book, annual letter or public LP letter | no qualifying work found | through 2026-08-22 | negative archive result | Verification: searched Qiming Insights/Newsroom, LinkedIn Pulse, Medium, Forbes author pages, PERC, 25iq, Google Books/Scholar patterns and Wayback discovery queries. Why: prevents interviews from being mistaken for a prolific authored corpus. Limit: private LP communications may exist but are not legitimately public.

Talks, hearings, podcasts and video

• ★★★ | China's High Technology Development hearing transcript | https://www.uscc.gov/sites/default/files/transcripts/4.21-22.05HT.pdf | 2005-04-21/22 | official transcript, 352 PDF pages; source type: primary public proceeding; reliability A | Verification: Rieschel statement and Q&A at printed pp. 43–54/PDF pp. 55–66. Why: contemporaneous direct voice, including IP and competitive-risk questions. Limit: OCR artifacts; some quantitative claims repeat the written statement.
• ★★★ | Venture Capital in China with Gary Rieschel | https://www.hoover.org/research/venture-capital-china-gary-rieschel | recorded 2025-04-15 | full official podcast transcript; source type: primary retrospective interview; reliability A | Verification: full transcript; claims located by the host question and the opening words of Rieschel's answer. Why: Qiming origin, VIE/legal ambiguity, partner economics, hiring, misses, fund claims, government, liquidity and geopolitics. Limit: retrospective and auto-transcript errors, including names.
• ★★★ | Homeward Bound: Gary Rieschel Talks About China VC | https://www.forbes.com/sites/russellflannery/2016/08/01/homeward-bound-gary-rieschel-talks-about-china-vc-huawei-and-colliding-garbage-trucks/ | 2016-08-01/02 | full Forbes direct Q&A; source type: primary interview plus editorial framing; reliability A/B | Verification: full original article opened and analyzed in its note. Why: local-presence rule, succession, cross-border healthcare, stock-market reform and liquidity warning. Limit: retrospective interview; company and fund claims remain unaudited.
• ★★★ | Gary Rieschel on China Money Podcast | https://www.youtube.com/watch?v=_bgWIrbJzik | 2012-09-17 | public podcast excerpt; source type: primary appearance; reliability A/B | Verification: 5:02 video and auto-captions analyzed in its note. Why: early Qiming performance claim, slowing-growth test, exit trust and fund prediction. Limit: excerpt only; “top quartile” and under-10% failure claims are self-predictions without stable denominators.
• ★★ | Video: Conversation with Gary Rieschel | https://www.youtube.com/watch?v=-2oD_JXfrVE | 2020-05-22 | USC event video; source type: primary appearance; reliability A | Verification: full 48:55 video and captions analyzed in its note. Why: founder screen, innovation, fund-life design, Tigermed team attribution and SoftBank failure. Limit: retrospective claims and auto-caption noise.
• ★★ | Inspired by Example — Gary Rieschel | https://www.buzzsprout.com/2250911/episodes/15278748-episode-15-gary-rieschel | 2024-06-19 | playable podcast and show notes; source type: primary appearance metadata; reliability B | Verification: 33:47 audio route and show notes analyzed in its note. Why: leadership across the U.S., Japan and China and a stable original episode route. Limit: no public transcript or chapters; no substantive quotation or inferred audio takeaway used.
• ★★ | Peggy Liu / JUCCCE conversation | https://www.youtube.com/watch?v=1jyfcKRf8NE | recorded 2020 | full video conversation; source type: primary appearance; reliability A/B | Verification: 40:24 video and captions analyzed in its note. Why: cross-cultural formation, epistemic humility, resilience and network leverage. Limit: retrospective biography and auto-caption noise; limited deal evidence.
• ★★ | China investors should get in early | https://www.qimingvc.com/en/news/china-investors-should-get-early-says-xiaomi-venture-capitalist-gary-rieschel | 2015-11-06 | long interview/profile mirrored by Qiming; source type: primary interview plus secondary framing; reliability A/B | Verification: full live page. Why: Qiming formation, early-stage/sector operating design, China macro selection and diligence. Limit: headline over-associates Rieschel with Xiaomi; Hans Tung led the deal.
• ★★ | Gary Rieschel: The Masteries of Entrepreneurial Financings | https://www.qimingvc.com/en/news/gary-rieschel-masteries-entrepreneurial-financings | 2012-05-04 | Chinese Fudan event write-up; source type: translated primary remarks; reliability B | Verification: full Chinese body accessible in search/open. Why: Chinese-language version/provenance control for the English summary. Limit: likely same underlying talk; count as a mirror, not an independent work.
• ★ | Seeking Truth Through Facts conference program | https://asiasociety.org/sites/default/files/inline-files/Future%20of%20U.S.%20%26%20China%20Conference%20-%20Seeking%20Truth%20Through%20Facts%20Program%20%5BFINAL%5D.pdf | 2025 | program PDF; source type: primary event metadata; reliability A | Verification: indexed but browser PDF fetch returned internal error. Why: current Asia Society role and moderator route. Limit: no remarks or transcript.
• ★ | China Has a Great Investment Story | https://www.youtube.com/watch?v=C-wPSWFBnRk | 2013-05-07 | SuperReturnTV interview; source type: primary appearance; reliability A/B | Verification: full 9:29 video and captions. Why: early China growth, IP and technology-transfer context. Limit: short promotional clip and caption noise; inventory only.
• ★★★ | VC Investing in China | https://www.youtube.com/watch?v=osdUt-jgq0A | 2013-10-29 | Thoughtful China panel; source type: primary appearance; reliability A | Verification: full 28:42 video/captions and deep note. Why: sector proof thresholds, founder disclosure, employee-pool test and track-record skepticism. Limit: auto-caption name/number errors.
• ★ | China Does Not Need to Be Afraid of Apple | https://www.youtube.com/watch?v=Svf_YZwSAx8 | 2014-09-14 | Silicon Dragon interview; source type: primary appearance; reliability A/B | Verification: full 10:08 video and captions. Why: contemporary China consumer and technology competition. Limit: short, partially redundant and inventory only.
• ★★★ | China—An Emerging Innovation Power | https://www.youtube.com/watch?v=ZgspwAx42i4 | 2015-05-25 | PKU HSBC lecture; source type: primary appearance; reliability A | Verification: full 19:17 video/captions and deep note. Why: leapfrogging, ambiguity, JGE-to-brand shift and dated healthcare/autonomy forecasts. Limit: unsourced figures and caption errors.
• ★ | Silicon Dragon SF keynote | https://www.youtube.com/watch?v=RwOpvSN7odU | 2017-10-13 | conference keynote; source type: primary appearance; reliability B | Verification: 22:37 video opened. Why: decade-sweep conference route. Limit: no public caption track recovered; inventory only.
• ★ | Made in 2025 | https://www.youtube.com/watch?v=xSzi94P5lHk | 2018-04-18 | CGTN interview; source type: primary appearance; reliability A/B | Verification: full 5:04 video and captions. Why: dated China industrial-policy view. Limit: short broadcaster clip; inventory only.
• ★ | China Allows Tech Companies to Get to Critical Mass | https://www.youtube.com/watch?v=--K4XJWLxNs | 2019-11-21 | Bloomberg interview; source type: primary appearance; reliability A/B | Verification: full 6:33 video and captions. Why: China scaling and market context. Limit: short and macro-heavy; inventory only.
• ★ | Gary Rieschel on VC, tech, and the U.S.–China relationship | https://www.sinicapodcast.com/p/gary-rieschel-of-qiming-venture-partners-2bf | 2020-01-02 | Sinica podcast; source type: primary appearance; reliability A/B | Verification: 1:00:57 playable audio and show notes opened. Why: full-length decoupling and technology retrospective route. Limit: no publisher transcript/captions; inventory only.
• ★ | China may have 12% GDP medical sector | https://www.youtube.com/watch?v=RzHrKtGvcC8 | 2020-03-24 | China Admissions interview; source type: primary appearance; reliability A/B | Verification: full 10:16 video and captions. Why: dated China-healthcare thesis. Limit: forecast/self-report and caption noise; inventory only.
• ★ | Bloomberg: China's healthcare opportunity | https://www.youtube.com/watch?v=fIGwgV-lwqQ | 2020-04-13 | Bloomberg interview; source type: primary appearance; reliability A/B | Verification: full 9:42 video and captions. Why: healthcare, stage and first-investor claims. Limit: short and unaudited firm statistics; inventory only.
• ★ | TikTok, WeChat and U.S.–China decoupling | https://www.youtube.com/watch?v=NUX2oe_SjNs | 2020-08-17 | NCUSCR event; Rieschel moderator; source type: primary event; reliability A | Verification: full 1:15:23 video and captions. Why: public-policy moderator role and adversarial context. Limit: guests supply most substantive evidence; inventory only.
• ★ | Silicon Dragon Online with Gary Rieschel | https://www.youtube.com/watch?v=IvM_dM7JUM8 | 2020-09-30 | long video interview; source type: primary appearance; reliability A/B | Verification: full 59:06 video and captions. Why: pandemic-era China venture and technology route. Limit: auto-captions and overlap; inventory only.
• ★ | Silicon Dragon Online with Gary Rieschel | https://www.youtube.com/watch?v=GsAfgPvRunc | 2020-11-06 | long video interview; source type: primary appearance; reliability A/B | Verification: full 58:02 video and captions. Why: capital-markets and RMB-risk-capital update. Limit: auto-captions and thematic overlap; inventory only.
• ★ | Looking back at a pandemic year | https://www.explorebigsky.com/2020-looking-back-at-a-pandemic-year/38078 | 2021-04-02 | embedded interview/edit; source type: primary remarks plus framing; reliability B | Verification: full page and Rieschel excerpt opened. Why: dated pandemic/economic forecast. Limit: no transcript/caption recovered; local-policy focus and inventory only.
• ★ | U.S.–China Tech, Reset? | https://www.youtube.com/watch?v=AH_SvCigks0 | 2022-03-05 | Silicon Dragon panel; source type: primary appearance; reliability A/B | Verification: full 1:12:17 video and captions. Why: post-crackdown/decoupling phase. Limit: multi-speaker captions and macro emphasis; inventory only.
• ★ | Gary Rieschel discusses U.S.–China competition and venture investment | https://podcasts.apple.com/de/podcast/podcast-gary-rieschel-discuss-us-china-competition/id1204174064?i=1000559110941 | 2022-04-29 | China Money Podcast; source type: primary appearance metadata/audio; reliability B | Verification: playable 1h17m audio and summary opened. Why: later counterpart to the 2012 episode. Limit: no transcript/captions; inventory only.
• ★ | The U.S. and China Innovation Race | https://www.youtube.com/watch?v=Lr_oVwyaNX4 | 2024-03-08 | Asia Society panel; source type: primary appearance; reliability A/B | Verification: full 50:28 video and captions. Why: current innovation/geopolitics phase. Limit: multi-speaker and overlaps Hoover; inventory only.
• ★ | China as a Tech Superpower | https://www.youtube.com/watch?v=z5K5Ykg2_5g | 2025-08-06 | Asia Society discussion; source type: primary appearance; reliability A/B | Verification: full 1:00:47 video and captions. Why: post-Hoover capital-market and technology update. Limit: multi-speaker and caption noise; inventory only.
• ★ | China 2026: What to Watch — closing remarks | https://www.youtube.com/watch?v=LFRejQpzp88 | 2025-12-16 | Asia Society closing remarks; source type: primary appearance; reliability A/B | Verification: full 22:43 video and captions. Why: current public-role and outlook endpoint. Limit: ceremonial and forecast-heavy; inventory only.

Biography, institutional history and contemporary press

• ★★★ | Gary Rieschel official biography | https://www.qimingvc.com/en/team/gary-rieschel | current; accessed 2026-08-22 | firm biography; source type: primary self-description; reliability A for role/C for promotional claims | Verification: full. Why: education, operating roles, institution-formation claims, geography and present firm scale. Limit: capital raised is not AUM; “twelve companies over $1B” is unaudited and grammatically ambiguous.
• ★★★ | Directors, senior management and staff — Gary Edward Rieschel | https://www.hkexnews.hk/listedco/listconews/gem/2002/1019/e_8088pro-20000405chap016.pdf | 2000 prospectus, hosted 2002 | public-company filing; source type: primary regulatory; reliability A | Verification: full PDF downloaded from the live route; Rieschel biography on printed p. 90. Why: dates/titles for Intel, Sequent, Cisco, nCUBE and SoftBank plus education. Limit: director biography, not employer records.
• ★★★ | Transition time | https://www.avcj.com/digital_assets/AVCJ161108.pdf | 2016-11-08 | AVCJ magazine cover story, PDF pp. 10–13; source type: strong secondary plus direct interview; reliability A/B | Verification: full relevant article. Why: compensation transition, shared economics, promotion, key-person status, Seattle move and succession. Limit: no partnership agreements or carry percentages.
• ★★★ | Who is Qiming Venture Partners? | https://www.thewirechina.com/wp-content/uploads/2022/04/Who-is-Qiming-Venture-Partners.pdf | 2022-04-10 | Wire China firm map, 6 PDF pages; source type: strong secondary; reliability B | Verification: full. Why: LPs, vehicles, team, portfolio scale, policy alignment and U.S. institutional exposure. Limit: PitchBook/media compilation; selected deals and no complete cash-flow ledger.
• ★★★ | Twenty hot venture capitalists | https://www.forbes.com/asap/2000/0529/098_print.html | 2000-05-29 | contemporaneous Forbes profile; source type: strong secondary; reliability B | Verification: full Rieschel profile entry under “Gary Rieschel, Softbank Venture Capital.” Why: formation, pace, board load, Inktomi miss, excitement/attachment weaknesses. Limit: compressed magazine treatment and round-level anecdotes.
• ★★★ | Seeding China's Start-Up Scene, With a Nod to Silicon Valley | https://www.qimingvc.com/en/news/seeding-chinas-start-scene-nod-silicon-valley | 2009-09-10 | New York Times profile mirrored by Qiming; source type: strong secondary/direct interview; reliability A/B | Verification: full article, sections on eHi/Kaixin diligence, the opaque customer-services company and SoftBank triage. Why: eHi/Kaixin/anonymous-company diligence, walk-away conditions, six Rieschel boards, co-founder attribution and SoftBank triage. WRITINGS boundary: reporter-authored press, not a Rieschel writing. Limit: promotional mirror; early fund gains are reported and subsequent crash figures lack a cash-flow schedule.
• ★★★ | Damn the Torpedoes | https://www.forbes.com/forbes/2000/0724/6603116a.html | 2000-07-24 | contemporaneous Forbes profile; source type: strong secondary with direct quotes; reliability B | Verification: full article. Why: SoftBank vehicle size, deployment speed, incentives and self-reported early fund performance before bust. Limit: performance was mark/sale-based and Rieschel-reported, not an audited net cash-flow schedule.
• ★★ | China's Venture Capitalist | https://www.gluckman.com/Venture%20Capitalist.html | c. 2011 | republished profile; source type: secondary; reliability B/C | Verification: live page body surfaced via search, direct open returned internal error. Why: local diligence, audit skepticism, career narrative and Qiming's early scene. Limit: republication provenance/date need reconciliation before quotation.
• ★★ | Gary Rieschel — UC San Diego China Forum | https://china.ucsd.edu/policy/china-forum/participants/gary-rieschel.html | current; accessed 2026-08-22 | university program biography; source type: institutional metadata; reliability B | Verification: full. Why: current public roles, education and scale claim. Limit: current page says over $11B AUM versus Qiming's $9.5B capital-raised language; metrics/dates differ.
• ★★ | Gary Rieschel — CNAS | https://www.cnas.org/people/gary-rieschel | current; accessed 2026-08-22 | think-tank biography; source type: institutional metadata; reliability B | Verification: full biography. Why: current policy roles, healthcare/cleantech focus and joint-venture boards. Limit: self-supplied/promotional biography; no underlying appointment documents.
• ★★ | Gary Rieschel — PERC | https://www.perc.org/people/gary-rieschel/ | current; accessed 2026-08-22 | nonprofit board biography; source type: institutional metadata; reliability B | Verification: full. Why: SoftBank-board tenure and environmental affiliation. Limit: stale “over $2B” firm scale signals page age.
• ★★ | American-style Venture Capital in China | https://www.thewirechina.com/wp-content/uploads/2022/07/Sebastian-Mallaby-on-American-style-Venture-Capital-in-China.pdf | 2022-07 | Sebastian Mallaby interview, 8 PDF pages; source type: strong secondary; reliability B | Verification: full, especially PDF p. 3. Why: peer context for locally led China franchises and declining need for U.S. control. Limit: synthetic interview about the industry, not deal-level evidence.
• ★ | Think Globally, Act Globally — Medium mirror | https://medium.com/solutions-journal-summer-2015/think-globally-act-globally-411305488561 | 2015-06-18 | RMI/solutions-journal mirror; source type: duplicate secondary; reliability B | Verification: full page found/opened. Why: provenance and dead-link backup. Limit: same underlying work as Qiming/RMI page; do not count independently.

Funds, deals, filings and attribution controls

• ★★★ | Qiming Venture Partners III Form D | https://www.sec.gov/Archives/edgar/data/1522075/000152207511000002/xslFormDX01/primary_doc.xml | 2011-05-26 | SEC exempt-offering notice; source type: primary regulatory; reliability A | Verification: full; issuer and signature rows. Why: Cayman vehicle and Duane Kuang GP-signatory evidence. Limit: Form D does not report fund performance, IC or carry.
• ★★★ | Qiming Fund VI corporate ownership in a listing draft | https://www1.hkexnews.hk/listedco/listconews/sehk/2026/0102/11974288/sehk25121700479.pdf | draft filed 2026-01-02; facts through 2025 | HKEX draft prospectus; source type: primary regulatory; reliability A | Verification: PDF p. 19. Why: Cayman structures, GP voting power, Rieschel/Kuang 30%+ shareholder threshold and dated AUM. Limit: draft status; one vehicle family, not firmwide economics.
• ★★★ | In Their Own Words: Hans Tung on Xiaomi | https://www.forbes.com/sites/russellflannery/2014/12/11/in-their-own-words-investor-hans-tung-on-how-xiaomis-lei-jun-leads/ | 2014-12-11 | direct investor Q&A; source type: primary witness/secondary publication; reliability A/B | Verification: full. Why: explicit statement that Tung led Qiming's Xiaomi investment and how he knew Lei Jun. Limit: retrospective and does not disclose full IC/ownership chain.
• ★★ | New Midas Lister Hans Tung | https://www.forbes.com/sites/laurahe/2013/05/08/new-midas-lister-hans-tung-on-the-next-big-idea-in-china-after-xiaomi/ | 2013-05-08 | Forbes profile/interview; source type: strong secondary; reliability B | Verification: full. Why: contemporaneous Xiaomi, Tung board roles and Qiming consumer-deal attribution. Limit: valuation/revenue figures are reported and not Qiming-return data.
• ★★ | GGV hires early Xiaomi backer Hans Tung | https://techcrunch.com/2013/10/21/hans-tung-ggv/ | 2013-10-21 | TechCrunch report; source type: secondary; reliability B | Verification: full article opened. Why: independent corroboration of Tung's Xiaomi attribution and departure from Qiming. Limit: short personnel report.
• ★★ | Qiming announces $320M Fund II | https://www.qimingvc.com/en/news/qiming-venture-partners-announce-320m-fund-ii | 2008 | firm announcement; source type: primary firm record; reliability A/C | Verification: full announcement, sections naming fund size, mandate, team and portfolio examples. Why: fund size, LP statement, sector/stage mandate and named contemporaneous team/company examples. Limit: promotional; no terms or performance.
• ★★★ | Qiming About Us — current firm denominator | https://www.qimingvc.com/en/about | current; accessed 2026-08-22 | firm portfolio/strategy page; source type: primary firm claim; reliability A for what Qiming currently claims/C for performance | Verification: full, sections “Why Choose Us,” “Exceptional Investment Performances” and “What Do We Invest.” Why: live claims of 11 USD funds, 7 RMB funds, $9.5B capital raised, 580+ companies, 210+ exits and 80+ unicorns; supplies the current denominator and its promotional framing. Limit: cumulative live totals are undated, selected and unaudited; “other” exits, write-offs, follow-ons and company-level cash flows are not disclosed.
• ★★ | AVCJ Fund IV and Vancl adverse control | https://www.avcj.com/digital_assets/8467/AVCJdec09_2014_supp.pdf | 2014-12-09 | AVCJ report, PDF pp. 8–9; source type: strong secondary; reliability B | Verification: full relevant article. Why: $500M Fund IV, valuation pressure, Vancl restructuring/failure narrative and JP Gan attribution. Limit: firm-level and selected examples; no complete denominator.
• ★★ | AVCJ YPX thesis retrospective | https://www.avcj.com/digital_assets/9515/29AVCJ01262016.pdf | 2016-01-26 | AVCJ report, PDF p. 2; source type: strong secondary/direct witness; reliability B | Verification: full relevant section. Why: Hans Tung's original offline-growth thesis and e-commerce disconfirmation. Limit: one former partner's retrospective; outcome economics absent.
• ★★ | Qiming Investment Adviser Public Disclosure | https://adviserinfo.sec.gov/firm/summary/160844 | current; accessed 2026-08-22 | SEC/IAPD summary; source type: primary regulatory; reliability A | Verification: route opened; dynamic body was not rendered, search index showed active exempt-reporting-adviser status from 2012 and disclosure category. Why: regulatory identity and ADV route. Limit: latest ADV PDF and disclosure answers require direct retrieval.
• ★★ | Qiming Weichiang Form ADV | https://reports.adviserinfo.sec.gov/reports/ADV/309395/PDF/309395.pdf | current filing route | SEC/IAPD filing; source type: primary regulatory; reliability A | Verification: route attempted; browser returned internal error. Why: management-entity registration and control research. Limit: not content-analyzed; verification requires lawful direct retrieval.
• ★ | Signal investor profile | https://signal.nfx.com/investors/gary-rieschel | current; accessed 2026-08-22 | aggregator portfolio; source type: database lead; reliability C | Verification: only search-index text was accessible in this pass. Why: U.S. investment leads such as Madaket and LanzaTech. Limit: incomplete and conflates personal/firm roles; verify each deal independently before citation.

Skeptical, regulatory and modern-context controls

• ★★★ | U.S. Treasury Outbound Investment Security Program | https://home.treasury.gov/policy-issues/international/outbound-investment-program | final rule effective 2025-01-02 | official rule/program page; source type: primary regulatory; reliability A | Verification: full page opened. Why: modern transfer boundary for U.S. investment in China/Hong Kong/Macau semiconductors, quantum and AI. Limit: later than most Qiming investments; do not retroactively treat current restrictions as 2005 knowledge.
• ★★ | China's private-sector promotion law adoption notice | https://www.npc.gov.cn/npc/c2/c30834/202504/t20250430_445088.html | adopted 2025-04-30; effective 2025-05-20 | official NPC notice; source type: primary legal; reliability A | Verification: lawful official route attempted; browser returned internal error. Why: contemporary counterweight to “crackdown” and state-support claims. Limit: Chinese text/body not analyzed here; require direct text before substantive use.
• ★ | Gary Rieschel/Qiming lawsuit or enforcement proceeding | no responsive proceeding found in opened search results | searched 2026-08-22 | negative search result | Verification: queried exact-name criticism, underperformance, lawsuit, sued and litigation; ran official-domain combinations for SEC enforcement, DOJ, FTC and OFAC, plus Congress/GovInfo security-report indexes. Results were ordinary Form D/ADV/issuer ownership records, an unrelated government-hosted risk study, irrelevant names or Qiming's notice about third-party impersonation—not a proceeding against Rieschel or Qiming. Why: records the mandatory skeptical search. Limit: not a legal clearance; Congress direct-page access was blocked, and Chinese court databases, subscription dockets, private arbitration and historic entity aliases remain outside this pass.

Coverage assessment

Inventory. This map contains 65 annotated routes: 61 distinct substantive works, two duplicate mirrors and two negative-search records. The route breakdown is 1 qualifying direct writing, 7 edited-text/negative-authorship routes, 29 spoken routes, 13 biography/context routes, 12 fund/deal/filing routes and 3 skeptical/modern controls. The Fudan Chinese/English and RMI/Qiming/Medium pairs are explicit mirrors, not independent works; the authored-corpus and lawsuit/enforcement rows record negative searches rather than works. The spoken map reconciles all 29 appearance families in talks.md; the Asia Society program is metadata for a role rather than a thirtieth appearance. Skeptical annotations are attached to the underlying AVCJ SoftBank-fund, AVCJ Vancl, Wire China and SEC rows rather than duplicated here. The sole accessible priority writing—the 2005 prepared testimony—was deeply processed into notes/w-uscc-china-technology-2005.md and writings.md. Reporter-authored profiles, Q&As, transcripts and event summaries remain interview/talk/press evidence, not an inflated authored corpus. The EY panel PDF and any undiscovered private/archival writing remain outside the analyzed denominator.

Searches completed. Name variants Gary Rieschel, Gary Edward Rieschel, common misspellings Gary Reischel/Gary Reichel, plus SoftBank Venture Capital, Mobius, Qiming, 启明创投 and partner names were searched across Qiming's English/Chinese newsroom, USCC, SEC EDGAR/IAPD, HKEX, AVCJ, Forbes, TechCrunch, Hoover, USC, Fudan, Asia Society, PERC, CNAS, LinkedIn, Medium, 25iq, Google Books/Scholar patterns, YouTube/podcast indexes and Wayback discovery. Decade sweeps covered 1995–2005, 2006–15, 2016–25. The mandatory criticism, underperformance, lawsuit, sued, litigation and SEC combinations were run, followed by Rieschel/Qiming searches restricted to DOJ, FTC, OFAC, SEC and Congress/GovInfo domains.

Direct-voice and archive boundary. The official 2005 testimony and transcript are the strongest primary archive. No public annual letters, shareholder letters, LP letters, book, personal blog or authored essay series was found. Qiming's newsroom largely mirrors journalism or event summaries. Wayback, public library/finding-aid, Computer History Museum, Stanford/Bancroft oral-history and 25iq routes yielded no named primary Rieschel collection in this pass. No outreach, archive request or paywall bypass was attempted.

Access barriers. The EY IPO PDF, IAPD Form ADV, China NPC notice and Asia Society program produced browser errors or metadata-only access; the exact route and missing body are preserved above. TALKS recovered full captions for USC 2020 and Peggy Liu 2020, a 5:02 excerpt for China Money 2012, and stable audio/show notes but no transcript for Inspired by Example 2024. Sinica 2020, China Money 2022 and the 2017 Silicon Dragon keynote also remain untranscribed.

Portfolio-denominator scope. The reconciled analysis keeps these entities and vehicles distinct: SoftBank Venture Capital funds; Mobius continuation; personal capital; SoftBank China Venture Capital; SAIF; Ceyuan; Ignition; Qiming USD Funds I–IX; Qiming RMB funds; Qiming U.S. healthcare vehicles; managing-director/co-investment vehicles; and the Chinese government guidance fund mentioned in 2025. Qiming's current selected portfolio and Wire China's 400+ company claim are not an auditable denominator.

Historical-context routes. Use contemporaneous 2000 Forbes coverage, 2005 USCC evidence, 2008 Fund II announcement, 2014/2016 AVCJ market data, official VIE/export-control rules and company filings. Later recollection must not leak backward into ex-ante cases.

Completed-peer comparisons. Jim Breyer is the closest completed cross-border institutional comparator for U.S./China networks and team/vehicle attribution; Yigal Erlich is useful for institution formation and public/private incentive design but is a different country/vehicle; Georges Doriot supplies a historically distant company-building comparison. Neil Shen, Kathy Xu and Masayoshi Son would be closer subject-matter peers but are not completed canon packages and must not be used as completed-peer comparisons.

Highest-value retrieval leads. Obtain the 2008 EY panel PDF through a stable archive; download current Qiming ADV filings; acquire or lawfully transcribe the three audio-only podcasts and 2017 keynote; seek public LP annual-report fund marks/cash flows; reconstruct the failed SoftBank 1999 vehicle and each Qiming vintage; and obtain decision/realization records for Tigermed, Gan & Lee, Xiaomi, LanzaTech and the ByteDance miss. These gaps remain explicit rather than being silently resolved.

Source notes